The Story of the Atomic Airplane
In 1980s, veterans of the Aircraft Nuclear Propulsion program got together and recorded nearly 13 hours of presentations and interviews describing the amazing effort to make a nuclear-powered airplane. The content is mostly focused on the Heat Transfer Reactor Experiments (HTREs) which can still be seen sitting out in the parking lot of the EBR-I museum in Idaho.
Dr. Jake Hecla, Stanton Postdoctoral Fellow at MIT, digitized this material with the help of John Webb. The people who saved this film and got it to Dr. Hecla are Lee Hite and David Carpenter. The late John Webb (W7NWH), Otto Woike, George Pomeroy, and Gunnar Thornton contributed to the development and preservation of this film.
This page contains a full transcript of the recordings with links to the specific timestamps in the films.
Parts 1-4
Parts 5-7
For a more ‘vivid’ view of the project, make sure to check out this 1960 film we got digitized showing the actual hardware these guys are talking about:
1 Intro
From 1946 through 51, the Fairchild Airplane Company conducted feasibility studies for nuclear propulsion for aircraft, or NEPA. Various cycles, power plants, and nuclear reactors were explored. Among the decisions was that the jet engine, still in its infancy, was to be used. In 1951, a decision was made that GEANP was to develop the direct cycle concept where the air from the compressor
was heated directly by the reactor fuel elements. Major elements of the ANP program are shown on the chart. During the 1951-53 early nuclear flight was an objective to determine operational feasibility of a nuclear-powered aircraft. The test aircraft was to be a B-36.
The power plant consisted of four modified J-47s designated the X-39. The energy was supplied by one R-1 reactor. The R-1 was water-moderated and had an aluminum structure. The fuel element consisted of a 310 stainless steel plus UO2 matrix, which was clad with stainless steel. The program was redirected in early 1953. Three heat transfer reactor experiments, HTREs, followed.
From 1952 through 61, the divided shield design was validated and various shield materials were tested in the tower shield facility at Oak Ridge. Also, 40 test flights of a B-36 with a small 2-megawatt thermal reactor aboard were conducted by Convair at Fort Worth. On January 31st of 1956, in the HTRE 1,
The first jet engine operated on nuclear power at the National Reactor Test Station in Idaho. The fuel material was nichrome plus UO2. Tests began in 1957 of the HTRE 2. It was similar in design to the HTRE 1, but with an 11-inch hole in the center of the core, where advanced fuel element and moderator designs were tested until 1961.
The HTRE 3 was built in a full-scale aircraft reactor configuration. It was tested using an air-cooled solid moderator, zirconium hydride, and nichrome with a few elements. Two X-39 engines were operated at full power. Nuclear starts were demonstrated. From 1956 through 59, the XMA-1A development program was to meet the Air Force Camel mission.
including supersonic sprint. New twin X-211 engines shared a reactor between them. A new solid moderator of yttrium hydride with capability of 2,300 degrees Fahrenheit was developed. However, the advanced fuel element, though structurally improved, did not yet have the capability of 1,800 degrees Fahrenheit turbine inlet temperature.
By 1959, development tests demonstrated that a BEO ceramic reactor would meet the 1,800 degree Fahrenheit turbine inlet temperature. The XNJ-140 power plant had one X-211 engine with a ceramic reactor. It provided 100 megawatts at cruise conditions. By March 1961, other competitive weapons systems such as missiles and nuclear submarines
had obviated the need for a nuclear-powered aircraft. Therefore, the A&P program was canceled. This is a model of a nuclear-powered jet engine, the XNJ-140. It is one-tenth scale. The actual engine was 37 feet in length. This is the compressor, the reactor with the shield assembly, the combustor,
the turbine, and the nozzle. By the end of 1960, 11 builds of the turbine machinery did operate in Cincinnati on chemical fuel. The XNJ-140 was scheduled to operate on nuclear power in 1962 at the National Reactor Test Station in Idaho.
This is a model of the Convair NX-2 aircraft. It was to fly in 1965. The canard configuration is well suited for a nuclear-powered aircraft. There are three XNJ-140 engines at the rear of the fuselage where they're accessible for remote handling and maintenance. 100 feet forward is the crew compartment, which is shielded.
Air inlets to the engines are located at the base of the wings. J-75s are mounted in pods and are used for takeoff only in the event of a loss of an engine. The hangar to house this aircraft and provide the remote maintenance was constructed in Idaho by April of 1961 when the Aircraft Nuclear Propulsion Program
was cancelled.
2.1 Background
The temperature reached 108 degrees in the afternoon in Mindanao. But lo and behold, we were issued new rifles, OD winter clothes, and OD blankets, as well as other equipment. Rumors were flying that the Japanese islands were next. Meanwhile, on the 6th of August, several people
were loading the bomb into the Enola Gay's bomb bay. One of these men was Gunner Thornton. Gunner, whom we have with us fortunately today, will you describe that short film? The first minor correction was on the 5th of August because the bomber took off quite early in the morning on the 6th of August and everything had to be ready naturally on the 5th.
So, basically, I was a member of the Los Alamos group out there, the so-called Project Alberta, which consisted of about 50 people from Los Alamos, which was a mixture of civilians and naval personnel and military personnel of different ranks and responsibilities, generally.
it was not a militaristic type of operation, and that Oppenheimer had insisted that this simply be a technical group. The Project Alberta Bunch was the one that actually brought the bomb. Their responsibility was to deliver the bomb to the Air Force, which was represented by the 509th Group, of which
Tibbets was the commander.
So the little piece of film that you saw was after we'd taken the assembled bomb out of the Quonset hut where we had done the assembly work and opened the rear doors to the backyard there, which was a backyard. This was actually the gun bomb to be used on Hiroshima. This backyard was shared then by another Quonset hut,
the Nagasaki bomb, the implosion bomb. So what you see is the bomb was wheeled out, and I think at that point we had it covered over with a piece of canvas, and so the guy in kind of ragged clothes, which we were apt to have at the time, kind of said, or acted like, get the show on the road.
pushed the rear end of the bomb around, started shoving it towards the gun carriage. That was me. So from then on I jumped on board the gun carriage, rode up to the loading pit, and generally once we got into the loading pit and in position, we handed it at that point over to the Air Force for most of the work. What were some of the events that led up to the atomic airplane?
Well, let's see. We could go back extremely ancient history, but maybe what we ought to do is go back to a book that was written by H.G. Wells. Let's see. H.G. Wells, in the year 1914, put out a book called The World Set Free. Now, this is a book in which he predicted the development of the atomic bomb.
And he had done this mostly on the basis of conversations with a physicist or chemist, Soddy, S-O-D-D-Y, who was a contemporary of the Curies. And he went into quite a bit of detail then about the story about the development of an atomic bomb. And Leo Szilard, who was the guy that got Einstein to write the famous letter to Roosevelt, which kicked off.
the atom bomb program, he had read this book. And I really didn't know its existence at the time. Much later, in the 1930s, there was a lot of talk about atom bombs. People didn't know how to release the energy yet, but there was a lot of talk about it. But Wells was talking in 1914. Well, although nothing was said,
At the time, he not only talked about atom bombs, he talked about atomic airplanes. Now, if you'll give me that book, I'll maybe read you a passage or two about what he said about atomic airplanes and atomic locomotives and atomic automobiles and all kinds of... Atomic roller skates? Well, we haven't quite got to the single-bladed roller skates yet, but let's see now.
Okay, in fact, in talking about atomic airplanes, for example, he says here, the new atomic airplane, bear in mind this is 1914, the new atomic airplane became indeed a mania. Every one of means was frantic to possess a thing so controllable, so secure, so free from the dust and danger of the road. And in France alone,
In the year 1943, Wells predicts, 30,000 of these new atomic airplanes were manufactured and licensed and soared humming softly into the sky. And with an equal speed, atomic engines of various types invaded industrialism. The railways paid enormous premiums for priority in the delivery of atomic traction engines.
Atomic smelting was embarked upon so eagerly as to lead to a number of disastrous explosions due to inexperienced handling of the new power, and the revolutionary cheapening of both materials and electricity made the entire reconstruction of domestic buildings a matter merely dependent on reorganization of methods, etc., etc. So he does a lot of talking about things here.
And the idea was that there was so much more energy in this atomic thing. The one thing that he's off by rather grossly is he figured that a ton of uranium was equal in energy. Let's see here now what he says. Okay, no, no, it was even less than that. He had a bottle of...
a pint of uranium oxide. And he figured that was equal in energy to 160 tons of coal. Now, if you work that out, he was really underestimating it. The real ratio is a pound of uranium-235 has, if you could release it, has the energy of about approximately 2 million pounds of coal. So you have a multiplication of about a factor of...
of two million, and that incidentally is one reason why the weight of the waste from the atomic energy is so low, because of the fact that there's a proportionate small amount of waste, and that's one reason why it's so silly that people are so worried about this, and why we can store the waste around on the sites. There is so little of it that from an engineering viewpoint, it just isn't a massive problem.
Not only that, but the waste from coal gets multiplied because you burn oxygen from the air that adds in the combustion products to the mass of waste. So the waste problem is really from conventional fuels is much worse than it is from nuclear. But you have to be careful, nevertheless. So the thing, though, that he gets at here is the
Everybody had their atomic airplanes and their atomic bombs, and they got to flying around and dropping bombs on each other's factories and countries and all that kind of stuff. And things kind of got out of control. And so after a while, various countries in the world thought, gosh, we just can't let this keep going like this. We've got to settle down and get this thing stabilized. And so they did.
And the net result of the development of the bomb in particular, but to some extent the nuclear airplane, but especially nuclear power and so forth, had the effect of stabilizing the world. That's why they say the world's set free. And what Wells says in here is that this would have happened eventually anyhow, because we would have kept fighting with conventional weapons and there would have been more people.
and it would have become intolerable that the world continue this way but the advent of the atomic bomb accelerated by several centuries the point at which the world people in the world realized that it was intolerable to just simply continue having wars so he says that was really was the big contribution and when you think about it that may be in fact probably is exactly what happened the fact that these bombs were used
We mentioned before already that Mr. Wells, H.G. Wells, wrote a book in which he mentioned nuclear aircraft and so forth. At the time, people knew there was energy locked up because they could tell from measurement of the energy of the particles that were released from uranium and so forth that there had to be a lot more energy available, that fire exceeded the kinetic energy from combustion or chemical processes.
You'd measure an alpha particle that had a huge amount of energy compared with some combustion product or chemical reaction product. So people did a lot of speculating. Wells did, and then afterwards. It would be very popular in the 1930s to speculate about nuclear airplanes, about nuclear ships, and so forth. And I was very much interested in all that kind of stuff.
but other people were also. And it started to appear in the newspaper. Newsweek magazine had started to give progress reports on the German atomic energy program back in the early 1940s. And those reports stopped at some time, but there were several issues in which they were saying something about that.
So during the war then, there were people that speculated about it. And in particular, in the NEPA final report, which came out in, what was it, 1951, and this was, what, NEPA 1830 or 1630, it goes into specifically the event.
the key events. It mentions the fact that people had speculated during the war. Some people who had gotten things like the K-25 plant going were still on the payroll. They were making the plant work better, but they had a little more time on their hands and they were able to start speculating. Some of the speculations involved not just the bomb work, but the
power producing reactors, nuclear airplanes, nuclear ships, and so forth. And in particular, there was a guy who worked for J.A. Jones Construction. And let me just read a little bit out of the NEPA final report. Here it says, in industry, two men were responsible to a large measure for the birth of the NEPA project.
And again, let me remind you, NEPA, Nuclear Energy for the Propulsion of Aircraft. These men were J. Carlton Ward, Jr., then president of the Fairchild Engine and Airplane Corporation, and Gordon Simmons, a young mechanical engineer, then in the employ of the J.A. Jones Construction Company, the construction company which erected the large K-25 gaseous diffusion plant at Oak Ridge.
Simmons then obviously had an insight. He was an insider as far as the Manhattan Project was concerned. And he probably knew a lot more about the bomb and a bunch of other things than people generally think. Because generally people working on the project weren't as ignorant of everything else. You hear these stories that they were compartmentalized and didn't know anything about what the other guys were working on. That was true to some extent, but the point is people kept their mouths shut.
And I'm sure that Simmons knew a great deal. He probably towards the end had more time on his hands because the K-25 plant was running and needed fixing and so forth. And it's not out of the question that, I just don't know what, if any part Fairchild had played in the Manhattan Project.
But I would not be at all surprised if at some point in time that Jay Carlton Ward, president of Fairchild, and Gordon Simmons had had some conversations, bootleg or otherwise, Fairchild was known as working on leading edge stuff. There was always at the front end of the innovative ideas as far as aircraft were concerned.
So Simmons had these ideas about making this nuclear airplane. So here's the key sentence, the very event. Simmons' thoughts were stimulated. Oh, wait a minute. Ward, it said, had been thinking about this. But it says that J. Cotton Ward's thoughts were stimulated by a letter dated 27 August 1945, which Gordon Simmons addressed.
to Sherman Fairchild, then chairman of the board of Fairchild Engine and Airplane Corporation, in which Mr. Simmons stated that he would like to be connected with a company that would be interested in applying nuclear energy to the propulsion of aircraft. Okay, so if you can say there's any point at which it started in the form of a piece of paper, it was that letter on August 27th.
1945, and I might point out that that was only ten days after the end of the war with Japan, so it is very likely that Gordon Simmons had composed this letter well ahead of time, but had waited until he could legally say something. Well, okay.
As a result, a conference was held in the Fairchild New York office in October 1945, attended by Simmons and others, at which it was decided that the Fairchild company was definitely interested in the sponsorship of such a project and would make representations to the military services to that effect. Now I'll read a little more here because it again brings in characters.
After subsequent conferences and further studies, consultation with Major General L.R. Groves, Bigger General Nichols of the Manhattan District, Rear Admiral Stevens of the Bureau of Aeronautics, and after evaluation of other proposals received from other companies such as Northrop, the Kellex Corporation, Gabriel Giannini Company, Army Air Forces, then part of the War Department, decided to sponsor.
a single unified project under which management, under the management of one industrial company with which all the companies in the recognized aircraft engine industry would be invited to participate. It was also considered desirable for the National Advisory Committee for Aeronautics to participate on par with the aircraft engine industry in this thing.
So a meeting then was held at the Pentagon, bear in mind this is now early fall of 1946, in which all interested agencies and companies were represented. Lieutenant General Curtis LeMay and Major General Crawford outlined the various considerations and the Air Force views on the subject.
And discussion brought forth unanimous concurrence of all participants that this thing ought to be looked into. Whereupon, now the Air Force did a very strange thing. Whereupon the Air Force asked the industry members to select one of their companies as the single manager of the group effort and as the recipient of an Air Force prime contract. After the military representatives had excused themselves temporarily from the conference,
The meeting was continued under the chairmanship of Burdette Wright, then president of the Curtis Wright Corporation. And one hour later, it was announced that the Fairchild company had been unanimously selected as the leader and single manager of the aircraft engine industry group effort. Now imagine that. The military walking out and saying, you guys decide.
These guys talk for an hour, and they decide that Fairchild should do it, and the reason they did was because Carlton Ward and Simmons had taken the lead in the thing, and they had done more thinking. These other guys, they were going to be involved, they were part of the act, but somebody else could get things going. So, you know, that is an amazing thing. Thus, the NEPA project began officially on May 28, 1946.
with the signing by General Spatz of an Air Force letter of intent, co-signed by Major General Groves, associated with Fairchild in the capacity of member companies were Allison Division General Motors, United Aircraft Corporation, Wright Aeronautical Corporation, General Electric Company, Westinghouse Company, Continental Aviation and Engineering, Lycoming Division of the Avco Manufacturing Company, Frederick Flader Incorporated,
Northrop Aircraft and Monasco Manufacturing Company. So that is how the thing was kicked off. Fairchild started these studies. They gathered together people that they could, and of course there were a lot of very capable people that had worked on the engines of
of various types. The war was over, we used engines during the war, including both reciprocating engines and early turbojet engines. It turned out the gas turbine engines, and looking at patent work later on, had a much more ancient history than I had thought.
We had a guy come in with some ideas and we were checking back on the patents. We found out a lot of the patents, issues of basic patents on gas turbines were issued to this, who was it? Was it Roteau? I think it was Roteau back in the 1890s. It was very hard to find any basic thing that hadn't already been patented in the 1890s.
Early conclusions, and they got some very, very capable people in there, generally advanced thinkers, you might say, but people who had actually worked on real engines and real airplanes during and before the war. So early conclusions there in their first report.
the first NEPA report in 1946, they said, in general, it appears that nuclear power can be used as a source of thermal energy for all basic types of power plants, including closed cycle condensing turbine, that means steam turbines. And steam turbines were actually looked at very seriously. Steam turbine has a huge advantage in that the so-called Rankine condensing
cycle and you can move from steam to condense it into water which is easy to pipe around and then you reheat it. Rankine cycle is a great cycle and that's what we use in Central Station power plants. Closed cycle gas turbine, open cycle turbine, turbojet, ramjet, rocket. The range of practical applications and the usefulness in relation to chemical energy
engines will depend primarily on the temperature at which the nuclear reactor can be operated. And that again, of course, then put the problem right into the lap of the materials people. Now, the thing is that in a gas turbine engine or in an automobile engine, the working fluid, the products of combustion,
run at a much higher temperature than the structural temperature. The structure is cooled in those things, but the working fluid runs at high temperature, and the walls, which might melt or corrode, only see the radiation from those things. By being very clever, we protect the walls.
from damage. Take the ordinary automobile engine. That combustion takes place at, what is it, 4,000, maybe even 5,000 degrees. There's a little explosion. Well, the material can't stand that temperature. But, you know, it's reciprocating for a bunch of reasons, and we're cooling the walls with water and so forth. But in the nuclear reactor,
You have to have the uranium inside a fuel element of some kind, and the heat then that is being produced in there has to go through a material. So in order to use it to heat some working fluid, you have to have that material withstand the temperature. And so consequently, you're limited then.
by the so-called heat transfer nuclear reactor. You are limited by the temperature of the metallic materials or ceramic materials to an extent much greater than is the case in other engines. We can't get into that in great detail right now.
Let us pause just for a minute and say, okay, where does this energy come from? Well, we know where it comes from in an ordinary turbojet engine or in our automobile engine. We burn some fuel in the presence of oxygen and we heat the air, which is actually primarily nitrogen.
So really, we're heating air and we're pushing a cylinder or we're pushing a turbine with hot air. We've got some energy still in the thing and we put out an exhaust that has a velocity which pushes the airplane along. Or we drive a fan from our little engine here. And the fan is kind of like a big huge ducted propeller. And so that the turbine or a turbojet, the jet,
is really hot air. That's what's pushing us, and really it's hot nitrogen, mostly, because air is mostly nitrogen. So really all we have to do in the nuclear engine is heat the air somehow or other, which we're going to do, as I said, by getting hot fuel elements. Well, there are people that didn't believe that.
members of the academic community. They thought somehow you had to have impinge on the blades, high velocity products, fission products, and they believed that in ordinary jet engines you were impinging combustion products on the blade, and that was what was turning the engine, the turbine. They didn't believe that it was hot gases. Now, I've told that to people and they can't believe that what I say is true, but believe me.
there were people, and I can mention some very prominent physicists, who just were not convinced that you could run an engine off of hot air. It's hard to believe, and it's hard for people to accept that when I tell them. And so we even had to go to the point, and this was happening at Fairchild just when I got there in 1948, they had to take a small turbine engine,
and take electrical heating elements and put it in place, the electric heating elements, in place of where the combustor was on the turbine. And the air came out of the compressor and it was heated by going over these electrical heating elements and it drove the turbine. And lo and behold, it worked.
The turbine rotated the compressor, there was extra energy that came out the back in the form of a jet, so that disposed of that problem. But, you know, it is hard in retrospect to think that that is one of the hurdles that had to be overcome. These studies went along, and by 1948, when the NEPA project was two years old, the AEC started to get interested in the thing. It had been sponsored up until that time by the Air Force.
And they established the so-called Lexington Project study, which was a study with MIT. And they got them some very well-known and capable academic people with representatives from industry also to study this. And so they did. And they said that they feel that it is feasible.
In its report, dated September 30, 1948, the Lexington Project concluded that although success cannot be guaranteed, there is a strong probability that some version
A nuclear powered flight can be achieved if adequate resources and competent manpower are put into the development. The aircraft would have a speed approaching the velocity of sound, probably 600 miles per hour, and an operating altitude of at least 30,000 feet, but probably not much over 50,000 feet. A manned nuclear power plane
powered plane, would be very large, perhaps larger than anything yet constructed, and that was true. When the program ended we were on a 600,000 pound airplane, but you bear in mind that the big transport now, like the commercial transports, are way up to 800,000.
So as the planes got bigger, it got more feasible to have these things. They said a tug-toe combination currently appears to be the most promising project. Well, that tug-toe was dropped out because somebody had a time machine and they'd been reading about this satellite that they tried to put out on this reel.
in 1996, and lo and behold, the tow rope broke and they lost the satellite. So somebody had, probably H.G. Wells had that in his book, so you know, somebody ruled out the tug-toe. So then the key thing in the lexicon pre-report. The project recommended that if it is decided that as a national policy
high cost and technical manpower, fissionable material, and money can be justified, a strong development program on nuclear-powered flight should be undertaken." And that was indeed done. This recommendation was accepted. And in February of 1949, then, it was decided that the nuclear-powered program would become a joint effort of the Air Force, the Navy, AEC, and NACA.
And so that was in 49, and it was at about that time then that the AEC felt that they better get people into the program that they were familiar with and recognized, and one such person now with a cast of characters increasing was Miles Leverett. He was a man with a lot of experience in the Manhattan Project, well respected by the Atomic Energy Commission and so forth, and they probably couldn't have picked a better representative.
to come in and represent that knowledge and so forth coming out of the AEC-sponsored programs. Well, what Fairchild did then on the NEPA project, they studied these various things and came up then with the early conclusion that the most likely ultimate successful candidate would be a reactor
a ceramic reactor consisting of a mosaic of fuel elements or a bundle of maybe beryllium oxide or things that we did indeed end up in GE working on. And Fairchild started then development on those things to heat a car to drive a turbojet and so forth.
But they also point out that there are all kinds of problems in these ceramic reactors. You've got to hold together this big bundle of things, and later speakers will talk about some of those problems and how they were encountered later on in the GE program. Well, shortly then, these were indeed big problems, the thermal expansion problems and so forth.
But then what happened is there was an invention made, and there's a patent out on it held by Walter Thompson and Austin Corbin in which they used a water-moderated reactor. The water served to cool the reactor structure and to moderate the neutrons. Moderating the neutrons is the neutrons bounce around, hitting the hydrogen in the water.
they get down to an energy where they are more easily absorbed by the uranium. And then using metallic fuel elements. The metallic fuel elements are to a considerable extent modeled after the metallic fuel elements used by the Navy in the Navy submarine program and used in these so-called swimming pool reactors. A sandwich in which you had some outer material and then inside...
Two layers of a metal, which could be in the naval reactors, could be zirconium or aluminum, because they're relatively low temperatures. And then the inner meat of the sandwich consisting of the same material mixed with uranium oxide. So here was a way of getting away from the structural problems, but still make a reactor.
in which you could gain a lot of experience with nuclear and so forth. At first, the idea was to make slabs of water, cans of water in the form of slabs and then fuel elements in between these things. Later, it was made into big concentric rings in diameter, annular rings of water.
held by aluminum, and then there were to be stainless steel or nichrome fuel elements impregnated with uranium oxide in between. That was the original concept. That was all done then, up to that point, under the Fairchild project. And whereas when the project then, it was decided then, late,
With this development and seeing a way that you could get into early development and get something started without really going to the ultimate, it was decided that there was now a time to get beyond the study phase and get an industrial contractor in there. And as Fairchild themselves had recommended, give it to one contractor. And General Electric was selected then and they took over then.
the NEPA project work, and soon after that, it was decided in their early studies that they would concentrate on the GE air-cooled direct cycle reactor. Pratt & Whitney was brought in to work on a reactor in which you heated liquid sodium in the reactor off to the side, and then you piped it into a heat exchanger and used that heat exchanger to heat the air.
to run the thing. Now later on then, GE started out and reactors were actually run using the water-moderated nichrome fuel element thing. Engines were run and so forth starting in 1956 and I might say they were exactly on schedule. They were scheduled to run in January of 1956 and on January 31st of 1956
the first modified conventional jet engine was run on the nuclear power under the General Electric Company. Something I would like to emphasize, during that period of development, there was continual, from NEPA all the way through GE, until termination in 1961.
There was a continual change in mission, which really didn't affect us as much as you might think, but the Fairchild Project said, the NEPA people said, you have to demonstrate operational feasibility. It isn't enough to be able to run an engine.
Get in an airplane and show that you can fly an airplane around, you can handle it on the airfield, and do one thing or another like that. You have to do that. So therefore, the mission would be to, okay, let's get a reactor and an engine developed, let's get a test bed, and let us fly it around for a few years, and so forth. Then a committee would come and they would propose that, gosh, we have to make it mission-oriented, so we want 60 or 100 airplanes.
Then another, if you do that for a while, then another committee would say, no, we have to have this demonstration. It didn't really make too much difference. We kind of went along anyhow. Gradually, the configuration changed from the early demonstration of metallic fuel to ceramic fuel, and that's where it ended. And the later speakers will go into that in a great deal more detail. Okay, Gunnar.
What was the motivation of the project and if it was so great, why do you suppose they terminated it? Well, okay. I think what I'll do here is I'll pick up the key page out of the Fairchild, the NEPA final report where they tried to have several
items which they state as the motivation. And then I'll jump over to the GE termination report and explain a little bit about the endings. But the NEPA, in their summary, in the summary at the beginning of the report, page five, it says that the motivation for development of nuclear airplanes
was well understood amongst the highest circles of this nation's government is attested to by the following extracts from important documents. Okay, important document number one. The President's Air Policy Commission on 1 January 1948 in its report, Survival in the Air Age, states,
Atomic propulsion. The possibility of employing atomic energy for the propulsion of aircraft and guided missiles is sufficiently important to warrant vigorous action by the Atomic Energy Commission, the Air Force, the Navy, and the NACA.
Some work of a preliminary nature has already been done in this field by the AEC, the Air Force, and the NEPA project. Immediate steps should be taken to intensify research effort in this field under a plan which would be supported by all of the above agencies and under which the project would be given
the benefit of all the background information in the atomic field actually needed by the recipients for the appropriate performance of their respective functions. Now that was not a conclusion of contractors or Fairchild or anybody else. Let me remind you that was the President's Air Policy Commission in January of 1948. So they came out at that time with strong support
of what had been going on and that the nuclear-powered airplane investigation should continue. Okay. Important argument number one was the executive branch of the government. And that was the President's Air Policy Commission. Now the congressional legislative branch of the government at about the same time. The Congressional Aviation Policy Board on 1 March 1948, two months later,
In its report, National Aviation Policy states, the nuclear energy propulsion for aircraft, or NEPA project, should be accorded the highest priority in atomic energy research and development, and every needed resource and facility should be devoted to its early accomplishment.
In the event of war or in any international situation likely to lead to war, nuclear energy for the propulsion of aircraft would be comparable in significance to the atomic bomb itself. Presently known limitations inherent in all chemical fuels make it difficult to deliver atomic bombs by air against a distant enemy. Therefore,
If the United States had nuclear energy propulsion in addition to atomic bombs, it could be the dominant factor in maintaining world peace. Until these ends are attained, the United States must depend on military weapons and techniques currently available. We must therefore devote the best efforts of the national military establishment
and the Atomic Energy Commission to the prosecution of the NEPA project to provide our armed services an effective method of accomplishing, without geographical limitations, immediate and devastating retaliation should our country be attacked. In addition to its military application, the successful solution of the problem of nuclear energy propulsion of aircraft
will include vital contributions to human welfare of enormous value to our people. Now is this the time of the Berlin blockade? I don't recall. I don't recall. The Berlin blockade was taking place and the situation with Russia was getting quite dangerous. Had the SAC programs where we wanted to keep aircraft loaded.
A constant 24 hours a day put a real pressure to develop such an aircraft. So you can see the motivation as viewed at the time. It was very clear and it was supported by both the executive and congressional branches of the government.
Now, of course, some of the things that happened here is they talk about the limitations of the present weapon system. While the project was subsequently carried on, the conventional jet engines and so forth also continued to improve. As Chris said, you had the beginning of the growth of bronze and missiles.
And the post-war development of those things. Okay, so that was the motivation that existed at the time of the NEPA project, and as summarized in the NEPA final report, and it's interesting that they make a point of that this could be the combination of nuclear energy propulsion in addition to atomic...
bombs could be the dominant factor in maintaining world peace, which is the point brought up in 1914 in the book by H.G. Wells.
2.2 R-1 Reactor
the Joint Chiefs of Staff endorsed the military necessity of a nuclear-powered aircraft. The Air Force, therefore, scheduled the flying testbed to prove the concept. In March of 51, the Air Force awarded GE a contract to develop a nuclear turbojet engine. By June, the AEC also contracted with GE to develop the nuclear reactor. After several months of study,
GE chose the direct air cycle concept using an air-cooled reactor with metallic fuel elements. Water was to be both moderator and structural coolant. The resulting GE power plant, the P-1, was to be test flown in a B-36H modified by Convair. Initial ground tests of the P-1 was scheduled for 1954 in Idaho. The test bed aircraft designated the X-6
was to be flown in early 1957. The X-6 aircraft with the P-1 power plant was to evaluate the operational practicability of this system. Areas to be tested included shielding, propulsion, radiobiology, and radiation effects on aircraft components. The P-1 power plant, shown here, with four engines,
the modified j47s with the ducting leading up to the reactor shield assembly the compressor discharge went into the reactor and out back through the ducting to the scroll of the at the turbine and out the rear and now we'll have buck jordan discuss the reactor itself but uh just another word on this uh the
Reactor essentially replaces the burner cans of the jet engine. The compressor exhausts into the reactor and back down in a high-pressure turbine through the exhaust nozzle to provide thrust for the engines. But combustor cans were located in the pipes. Yes. I'm going to speak a little bit about the reactor shield assembly.
The shield assembly approximates a cylinder 10 feet in diameter and 20 feet long. It was essentially a tank of water surrounding the reactor core and reflector, which was 68 inches in diameter. This meant that there was two and a half feet of water surrounding the entire core.
The shield assembly is made up of the center section, the forward section, and the aft section. The forward section contained the instrument wells and was penetrated by the air ducts, both fore and aft. The reflector
of this core was made up of two concentric rings, three or four stainless steel, which were water-cooled. The outer ring was a three-inch ring of stainless steel. The inner ring was a two-inch ring of stainless steel. The reflector was water-cooled. And the airflow then was from right to left. Right to left.
Correct. We might say also that we used stainless steel. Stainless steel wasn't the reflector material of choice. We would really have liked to have beryllium, but beryllium at that time could not be fabricated in the sizes and shapes that we needed, so we ended up with stainless steel. This is a schematic of the reactor assembly. The reactor assembly is essentially rings of moderator water.
concentric with and alternating with rings of fuel elements in each one of these annular spaces. The moderator water flowed through the reactor and out through these moderator water tubes. The fuel elements were in trays longitudinally.
with a reflector, stainless steel reflector, water-cooled. The outer reflector was 3 inches in thickness. The inner reflector was 2 inches in thickness. The fuel elements were in trays with 10 segments in each tray. Those were on rails, weren't they? The fuel element itself,
were 3 1⁄2 inches long, and there were 10 of them that were brazed to tracks that supported the fuel elements as they were slid into the reactor core itself. And the moderator cooled the water, cooled the structure, which was of aluminum. All the reactor structure was aluminum. These were the support struts for the moderator.
What was the nominal power of that? It was a 150 megawatt reactor. It was designed for 150 megawatts. Although 80 megawatts is probably the megawatts that would have been used. This is a picture of a typical fuel element. The sheet material was contained
78% of stainless steel and 38% of UO2. The cladding was a 4 mil, 310 stainless steel material. The thickness of the plates varied between 11 mils and 16 mils to provide power flatten across the
power distribution flattening across the core. These largest fuel elements were toward the center. This is the hydraulic diameter of the fuel element. About three and a half inches at the center of the core and nine tenths of an inch at the outer periphery of the core. The plates were eight inches long and three
and a half inches in depth. Each plate had fins brazed to the inner side and outer side of the plate. And then the corrugations were then brazed to these fins to supply support for the structure and also provide air flow area between the fuel elements.
These fuel elements were tested, airflow tested, and also burner rig tested in the burner rig facilities at Evendale. I think that's it. Even though the R1 reactor was never built,
as far as the actual hardware was concerned. The reactor core itself was simulated in a critical experiment at Oak Ridge using, at the Oak Ridge table facility, what looks like this, wherein half the core is loaded in one half of the facility and the other half of the facility, and the two halves are brought together hydraulically to make the core critical.
The matrix of the core, half the core looked like this. It was made up of blocks about three inches square of plexiglass and stainless steel impregnated fuel, UO2 fuel. Steel on the blocks look like this. The plexiglass blocks on top and bottom.
in between which was located the stainless steel and uranium fuel. These blocks were just slid in and out of the honeycomb matrix of the critical experiment. It was envisioned at one time that the actual reactor would be built up on a critical experiment, and that buildup would look something like this. This was never really done.
But it does give you an idea of what the planning of the types of testing that were going to occur in the early stages of the program. The weights and vision for the reactor, the P1 power plant, the reactor itself about 10,000 pounds. The shielding around the reactor, 60,000 pounds. The crew compartment, the shielding around it, 37,000 pounds.
The engines, 18,000 pounds. The ducting and the structure, 40,000 pounds, which made a total of 165,000 pounds that would go into the, for the total nuclear reactor system in the B-36H. Of course, fuel could be much less than the B-36. Some of the people that were involved in this were, like, I remember E.B. Delson in reactor design.
and Carl Lockhart and Bruno Benigni, Al Hintze and controls and accessories, Charlie Jones worked on the aluminum rings, which was a real problem at that time because the aluminum available was certainly not what we have today. And Jerry Moore, the World War I pilot, always helpful. Certainly John Monday in manufacturing, Buck Jordan, Bob Evans.
Any others that you can... Bill Long. Bill Long, Al Crocker. Carbon. Yeah, Austin Carbon, Wally Thompson. West Schmill. West Schmill. They could go down the list of people that were at A&P at the time because that was virtually the only thing that was being worked on. That's correct. Though a lot of the work, I guess, was done at Oak Ridge before GE moved up to Evendale. Right.
But the decisions, again, were remade by Miles Leverett and the rest of the people. Can you say something about why it was terminated? Yeah. In March of 1953, Stalin died. The Korean War was unresolved. Charlie Wilson, the Secretary of Defense, deciding on President Eisenhower's first budget, ruled that the flying testbed aircraft had no military worth. The X-6 program was therefore canceled.
along with the P-1 power plant. However, a divided shield test using a B-36 was continued by Convair using a small 2-megawatt thermal reactor. Meanwhile, GE's effort was reduced to an R&D program. However, testing of the nuclear engines, the X-39s, on a common heat source was accomplished in Cincinnati.
2.3 X-39 Engine
Hi. Today we're covering the X-39 engine that was used on the HTREs 1, 2, and 3, and also some test work down here. And we're very pleased to have with us Jack Hope. Jack, you came here with Gerhardt Newman from Lynn, Massachusetts, down to Cincinnati. Back in what, 52? Yes, well, I actually signed on with Gerhardt in the fall of 51, and we came to Cincinnati in the early...
spring, February or March maybe of 1952, and carried out the program that we're going to talk about. Good. In fact, I think there was only about four of us in the group at the first. Okay, go to it. So the first engines that was used in the nuclear experiments, in fact the only ones used in nuclear experiments, was the X-39. And we have a picture of it here, if you can switch to the picture.
The X-39 was a modified J-47, which was in production at GE at the time. And the key difference between the X-39 and the J-47 was that we were to remove air from the compressor and take it to the reactor, and then after the energy was added, heated, come back to another scroll to go out the back. This presented about a 30% pressure drop.
between the compressor and turbine as opposed to 5% in a normal engine. And as a result, we had to change the cycle. We chose to reduce the flow going through the compressor. So the X39 compressor was cast with wall thicknesses about between two and three inches thick. And the blading stayed the same so that there was about 70 pounds a second going through the compressor.
as opposed to 100 pounds per second in a J-47. And then, of course, the cycle was balanced. The greatest mechanical design job, of course, was in the scrolls. Now, in order to test this engine, we built a single-engine test cell. Cell 33, I think it was in Building 800, had a big loop in it.
So we had what we called a single-engine test cell, just a loop of pipe here with a bypass combustor in there to run the engine. And the main thing I remember about that was after working for several months, Gerhardt and I and several other people were in the cell getting ready for our test. We had a Cadillac engine that was to drive through the bullet nose to start it. But just before we were to hook that up, there sat Gerhardt doing this.
And, of course, it ended up the Cadillac was going to turn the engine backwards. So we worked all night switching the input from the bullet nose down to a gearbox to get it started. And by the next day, we had it running. We also had called one of the midgets that was at GE at the time to crawl through the pipe to make sure there was no obstruction in there. And the midget got there and looked at that, and he said, no, there's no way he was going through there. So guess who volunteered to crawl through the pipes? I was smaller then than I am now.
But we built, I think we said 11 of these, and so they were running a single-engine test cell in Evendale, and also we had a single-engine test cell in Idaho. But in addition then to running in the single-engine test cell, we built at Evendale what we called the propulsion unit test cell. And this was to simulate a complete nuclear setup, only of course we were going to run on chemical power.
have a couple sketches here that show the inside of the cell I believe at the time that was the largest engine test cell in the world the idea being another setup we had four x39 engines that the air came was collected from the scrolls went up to a simulated reactor and then back through the
bypass combustors, as we call them, to the turbine section. The reason for the bypass combustors was that in order to get the engine running, we would divert the air to just bypass the reactor, come back through this combustor, get the engine up to speed, and once it was running, then we would open the valves to run it through the, in this case, another combustor, or in Idaho, the reactor itself.
Another reason for this configuration is, as you know, when you shut a nuclear reactor down, its energy doesn't stop immediately. In fact, I think it had planned to aftercool for maybe even days. With this configuration, we could start this combustor, continue to pump air through the reactor for as long as you wanted to run the engine, days at a time for that matter. Here's another picture in the propulsion unit test cell.
that shows all four X39s there with the four legs of pipe and the simulated reactor up here. We ran a lot of tests in the PUT cell. In fact, I have another photograph here, if it shows up. In this, we just had a single loop in here. This is how we tested the bypass combustor.
We had one combustor running in front of another so that we could get simulated conditions out of the reactor to test the combustor. I remember when we finally got this fired up, and if the color will show up on the video, you can see that the first combustor is
is running very hot, supplying hot air to the second combustor, and then that's how we got the data for designing a combustor to operate under hot inlet conditions. I do also remember when we had that test running, Schultz saw it and thought, boy, he was so impressed that he sent word all over Building D, and we was running at a waiting line of people waiting to come through to see that thing running hot. We did get a lot of data from it.
Then, of course, the thing we did after this test was to go to Idaho. And a lot of us spent, I know I spent a complete summer out there helping the Indians build this up. They had a lot of Indians working in Idaho. And so our group, engine design, was responsible for the engines, the ducting, and the bypass burner, and the whole configuration.
the first unit only had two engines instead of four and i like to think that that is a unique engine in my design experience because i believe the power plant weighed 700 tons and put out about 500 pounds of thrust so the thrust weight ratio was atrocious but it also had a zero sfc so we learned a lot from that and the other
The thing I remember about that particular program was, I think it was after the first or second test, one of the fuel elements burned out in the core and scattered uranium out that big duct or smokestack. And they brought the power plant back into the hot shop. And then I hear that it's so hot, after they pulled the fuel elements out with remote handling equipment, put them in a pool.
But it was so hot underneath the core that if they put their workers from Idaho in there to reload the core, they'd get a three-month dose in four minutes. So I told Schultz, well, why do you want to burn up those guys? Let me reload it. So we chartered a DC-3 and loaded it up with people from engine design and the new fuel elements and took off for Idaho.
So we got out there, and all of us guys in engine design, we practiced a whole day in a cold shop, getting the fuel elements up, rotating them in the instrumentation plugs, pulling them back down. So we went in the next day and reloaded that core. We each had a four-minute time to do it. And I remember when we got done, Bill Long was so happy, he threw a big party for us at the hotel there.
The other thing that's interesting about these, we had initials for all these. We had a single-engine test cell, and it was called SET. Then we had a propulsion unit test cell, and it was called the PUT. And I remember the initial nuclear test we called the IET.
And the funny thing is, for a few hours one time, we had the next test where we were going to fly. We called it the flight aircraft reactor test. But that name got changed very fast because they didn't want to put the initials to it. I think that the only other aspect of the X-39 which didn't show in these photographs was the structural analysis of trying to,
transmit loads through a scroll was one of our biggest problems. Unsymmetrical pressure vessels are not easy to design. If you blow up a balloon, it goes round. So we spent a lot of money redesigning the scrolls on the X-39, the turbine scrolls, and that's what Don Riley did a lot of work on before the program was canceled. Without computers.
With slide rules. That's right. Everything was slide rule. I still have my 20-inch slide rule with a magnifying glass on it. Another item of levity, if you will, is we were all crowded for space in A&P. And I had a small office in Building D. And as we got into the program, maybe I'll use this one here, Otto.
We got squeezed, and they moved three of the desks from the unit into my office. And so I said, well, we need more room, so I called in some of the engineers. You can see them here. It's Frank Dorsler, Russ Motzinger, Don Riley, Bob McDonald, Pete Dwyer. So we pretended we were working on a two-level desk arrangement in my office and called the photographer. He took the picture. I gave it to Schultz, and all he did was laugh, frame it.
put it on the wall in his office, and I got squeezed a little more. So that didn't work. Another item of levity is we, for the first several years, we only had design groups. We had reactor design, shield design, engine design, and so forth. And we did everything. But then they changed the organization eventually where there was project groups. There was a project for HTRE one, HTRE two, and then, of course, the design groups.
Well, I ended up with a design group, and this, I'm not sure if it will, if you can see it on there, the word, but essentially I found this on my desk one morning where all the guys in my group didn't think they were being treated right. So here is a model of a two-story outhouse with engine design being on the bottom and the projects being on top. So I thought that was funny, and I've kept it all these years.
The other thing we worked on a lot in engine design was configuration for the overall engine system, both the one that ended up with the X-211 arrangement and also a nuclear turboprop study.
Water Moderated Reactor Design
At that time, a number of us became more interested in the actual design of reactors. Wally Thompson worked very closely with me and was also a very good friend. He and I came up with the concept of using a water moderated metallic reactor wherein the water kept the
metal cool below melting point and submitted a patent on that so this patent was issued by in the name of Wally Thompson and myself for the reactor which actually was the reactor that powered the first turbojet my next assignment after the efforts and reactor design after I got to Cincinnati we moved to Cincinnati in 1951 and
deep-building without air conditioning, incidentally, in the summer, my assignment was to become a part of a westward-hove task force, a task force of about a dozen people, headed up by Jack Parker, who was working for Roy Schultz at the time, to establish the criteria, the fundamental criteria for the Idaho test station, the new Idaho test station, to be used for testing power plants.
Anyhow, from those preliminary criteria, the architect engineer went on to design the Idaho Test Station.
3.1 HTRE-1
I want to first discuss the heat transfer reactor experiment number one. The original objective of the ANP program after we left Oak Ridge and moved to Cincinnati was the design of a power plant to power the Convair X6 experimental airframe. The early availability
rather than high performance was the primary requirement for this particular program. This work was authorized in 1952, but the whole direction was changed in 1953. The new objective was directed toward a broad spectrum of potentially useful nuclear propulsion systems.
And the approach chosen by GE was to perform nuclear reactor experiments using reactor types with potential application to aircraft propulsion systems. These operations were designated as the heat transfer reactor experiments, and HTRE number one was the first of the series.
HTRE number one and its successors were to be operated in a mobile facility known as a core test facility. The HTRE one is shown in this sketch, this artist's conception, as it was to be mounted in the core test facility.
The test vehicle was mounted on a four-rail flat car and consisted of a shield tank, ducting, chemical combustors, and two J-30, modified J-39 engines. The provisions were made for transmitting control information
data through an instrumentation and control plug mounted on the front of the CTF. The plug receptacle at the test facility accepted this plug and made the connections automatically. Next, there were a number of specialized jigs and fixtures.
3.2 Idaho
That's the next chart, please. Designed to simplify the handling, loading, and unloading of the heat transfer reactors into the CTF. Next one. The CTF was assembled in the Idaho Test Station Maintenance and Assembly Facility. And this is a picture.
of that assembly in progress. The heat transfer reactor number one as the first of a series was to be a highly simplified reactor primarily aimed at testing some of the key desirable features of future designs. Most important of these was power flattening so that each fuel element would produce the same amount of power
and thus the entire reactor could be essentially isothermal. The design for the HTRE 1 was based on the use of many of the materials that had been selected for the R1 reactor. This is an artist's concept of the R1. Specifically, water was to be used as the moderator.
and aluminum is the principal core structural material. The configuration differed sharply in order to provide better structural characteristics and to simplify fabrication. The R1 had been designed as a series of concentric annuli with the fuel elements contained in the air spaces between the annuli.
fuel elements themselves were to be a metallic material, a sheet metal, and in the particular case of the HTRE 1, nichrome was selected as that material. The basic design of the reactor core
was very similar to a fire tube boiler. The core was pierced by 37 tubes designed to hold the fuel elements. It was 30 inches across flats and the hexagonal
bundle itself was approximately, the fuel elements occupied approximately 28 inches of that bundle. The dynamic and shim control rods were housed in smaller aluminum tubes, which also served as inlet for the moderator water, which filled the entire vessel
except for what was occupied by the fuel elements and the beryllium reflector. Water pressure was only that necessary for pressure drop due to pumping. The water temperature was held around 160 degrees Fahrenheit. The individual fuel elements were made of rings of the clay and metallic material.
This is a sample element. These rings were joined into elements with a varying number of rings depending on their location from front to back of the core. The 18 of these concentric rings
formed a cartridge and the cartridge itself had a nose section which was used to fasten it into the core and a tail section which had access for a special tool that would go up and release the catch at the front.
to remove it the cartridges were then placed in an insulating or an insulated liner which was then inserted into the core the here's a cartridge in the process of being inserted and here is a liner
Now these liners had a steel inner jacket. They were wrapped with insulation and covered with stainless steel foil. And the elements themselves were designed to operate at approximately 1700 degrees.
produce an air temperature of about 1,350 degrees Fahrenheit. And then to achieve the power flattening, the spacing of the tubes in the hexagon were varied because the flux would normally fall toward the outside so that the tubes were stretched further apart as you moved out.
except at the beryllium reflector you had a reflux back in. So it took pretty careful adjustment to get the power right. Longitudinally down the cartridge, the number of rings were varied to hold the heat transfer roughly constant.
It was very gratifying that the predicted fuel element temperatures and those measured in operation were in very good agreement, and this method of power flattening was used in all subsequent metal-fueled reactors. The HTRE program, HTRE one program, was initiated immediately following the issuance of the
1953 program recommendations of the chart in September of 53 the button was pushed to start the program by the following August a mock-up critical experiment mock-up had been completed
and the results achieved. And in September, we started releasing drawings for the manufacturing of the actual core itself. By August of 55, the manufacturing was completed, and we hired some carriers and got a caboose and a flat car and hauled it out to Idaho.
In November of that year, we took that real reactor critical by itself. And then by January, it had been installed in the CTF and had gone to full power. By January of 57, the test series were complete. And it's very interesting.
that during this test series i need the other chart yeah but during the test series all the principal goals were achieved we ran in all a total of some 5004 megawatt hours at power levels up to
20.2 megawatts which is pretty darn good for a little thing 30 inches in diameter and 28 29 inches long 485.6 hours were above 200 kilowatts and 150.8 hours were at full nuclear power now the original goals had been to run 100 hours
at full power. Unfortunately, during the first six hours of operation, there was damage to the fuel elements. Now this damage was basically caused by the collapse of the insulation liners blocking the air passage
and allowing the material to reach melting temperatures. One thing this put to rest almost immediately were the arguments that some people had raised against using aluminum backed up by water. This view shows the worst damage cartridge from the side. Well, this was taken to the hot shop.
We unloaded it. We redesigned the insulation liners, reloaded it, took it back to the initial engine test facility, and performed an endurance test of 100 hours with the discharge temperature at 1280 degrees Fahrenheit.
which was ample to run the modified J-39 engines, and then ran it 44 hours beyond that at 100 degrees higher temperature at 1380. The objective of the test was 100 hours at full power. So all of the objectives of the program...
in terms of the design for power flattening, in terms of the design for operation, were either met or exceeded. And the feasibility of operating a nuclear turbojet engine with a direct air cycle reactor was demonstrated. And this was the first known operation of a high temperature gas turbine engine.
on nuclear power. Certainly looks its age in that picture. And that ran regularly between Wilmington and Idaho Falls. This is the site flight again with a group preparing to leave Wilmington to head for Idaho Falls. And that is the story of the HTRE One.
We're going to spend a few minutes giving definitions of what various terms mean you're going to hear throughout the discussions of the A&P program. We'll start out here with power plant.
Next is the concept of a thermal reactor. Next is the definition of what a critical reactor is.
I'm not going to really describe these terms because it's much easier to just read what it says. And of course you'll have this for reference as time goes on. What is fishing?
What is a moderator What is a reflector
What is a control rod What is a safety rod
What is a neutron? What is neutron flux?
What is a barn? What is a cross section?
In relation to cross-section, here are some cross-sections of some typical materials used in the ANP program.
What is multiplication factor?
And finally, some facts about uranium-235.
I made my first trip to what eventually became the Idaho Test Station in early 1951. I mean very early when the temperature was about 30 below zero. The selection of that as the site for GE's test station was made in the 51-52 period.
construction of facilities proceeded very rapidly and was well underway by the end of 1953 the most challenging of the facilities and it was started early was assembly and maintenance area better known as the hot shop and we're ready for that chart the hot shop
consisted of two major buildings, an assembly and maintenance area, which was known as the coal shop, and the hot shop itself, which was a high bay area with shielded walls, shielded windows, and the shielding was sufficient for any of the projected reactors.
including the aircraft reactor, to be bare within those walls. It was equipped with turntables, it was equipped with standard manipulators, and it was equipped with one special manipulator. The manipulators were operated from behind
shielded windows. These windows were approximately eight feet thick and they were a combination of leaded glass and a special liquid. The particular manipulators in the area
You had the small ones that you saw the man operating from behind the window, but there's one very special one known as the O-Man. The O-Man was designed and built in Schenectady by GE. It was one of the first of the servo manipulators in that there was actual feel.
back to the hands of the operator who wore special handling gloves and he could adjust his pressure so that with this particular manipulator you could either pick up an egg or lift several hundred pounds. Also under development at that particular time, the aid in viewing was a
3D stereo television camera, which was one of the first areas where such a device was used. Behind the windows here and at one of the stations, you see most of the upper brass of GE ANPD. This is...
Frank McCune, the atomic division general manager. Next to him is Jim Lapeer. Next to Jim is Wayne Lapeer. I mean, excuse me, is Wayne Perry.
who was the manager of the idaho test station and next to him roy schultz the general manager of the department and in addition to the general purpose manipulators and the old man the hot shop and the assembly and maintenance area were both served by the four rail track system which
went to a turntable from the turntable to the initial engine test facility and ultimately was designed to go to the flight engine test facility. Along with the permanent facilities,
There were mobile facilities like the shielded locomotive that was used to move the various test vehicles from their facilities. It needed to be shielded because the shielding on all projected reactors was to be fairly light since the crew was
to be shielded and a special crew shield in the aircraft. And there were some other early buildings built here. The single engine test facility was a facility built to just test single engines prior to assembly onto the CTF.
We also built a low-power test facility which was a critical experiment facility and next to it was a shield test facility which provided for tests of sections of shields both in a water tank and from towers where it could be suspended. The
Extremely important to the overall program was the initial engine test facility. It had an underground control room. It had a test cell and an exhaust stack, and you can see the connection on the four-rail trackage
to the turntable. This facility, the underground control room, could be accessed through a 450-foot tunnel since the radiation levels when running at full power were too high to approach it directly. The...
the test cell itself was covered by a movable shed which rolled on tracks outside of the four rail track system and it was connected to a stack through a series of electrostatic precipitators the engines on the ctf
would plug directly into the receptacle ducts that led to the precipitator and the stack, and the plug contained all the necessary instrumentation leads and control leads from the core test facility.
now this shows the core test facility as plugged in with the shed removed inside yep of the control room was what at that time was an up-to-date state-of-the-art
data processing system which could handle 500 pieces of data and could reduce them in a minute to a half dozen important parameters. Beyond that, there was a control panel from which you could control the reactor and the engines and the other functions on the CTF.
The IET was the test facility for the entire series of heat transfer reactor experiments. The final facility that was built at Idaho was the flight engine test facility, which is one of the largest single span poured concrete buildings ever built.
it allowed for a clear span of some 300 feet for the wings of the projected aircraft and a tail height of approximately 90 feet it had an underground control room it could be approached through roughly a mile of underground tunnel and it would have been
connected to the hot shop with the four rail system and ultimately it was planned to build a ramp in the front of a taxiway out to a 15,000 foot airstrip which for which incidentally all the test borings had been made and the strip was designed
Also, along with that, there were a number of studies done on flight patterns from Idaho to a clear ocean test area. As an adjunct to the FET, a shielded vehicle known as the Beetle was designed and built.
the purpose of the beetle primarily was for the installation and removal of the test engines either from an aircraft or an aircraft mock-up in the fet actually the beetle was completed even after the
closure of ANP and ultimately it was used in the cleanup of the SL-1 disaster in Idaho Falls and then shipped to Jackass Flats where it was used to pick up the spillings out of the Nerva rocket reactor. This is another shot of the beetle fully
extended and finally here's a shot of the beetle on the rollout day with colonel stander for the air force doing the honors of swinging the young lady in 1988 december of that year the ctf and the HTRE 3
were moved from the ANP site down to the old experimental breeder reactor site, which was 31 miles south. This was to make them available for public display because the ANP site, which is now known as the test area north, is still very active. This picture
Shows the transport as it went down that 31 miles of road. It was on a trailer with 144 tires, spread over 12 axles, and the darn thing weighed 600,000 pounds. So if you want to see it, you can go to Idaho Falls and visit it right now.
3.3 HTRE-2
HTRE-2 reactor design was essentially an offshoot of the HTRE 1 design. And the HTRE 1, of course, was a, as Bill Long explained earlier, was a 30-tube, water-moderated, beryllium-reflected reactor. This gives you some idea of what the overall configuration looked like. We have the 30 tubes.
in an aluminum can surrounded by a brilliant reflector and containing water which was a coolant pumped through these pipes. And the 30 fuel elements were inserted from the bottom and they occupied about 30 inches of the center section of the core right in the middle of the reflector. In the process of building up the core, the system was fitted into a series of jigs
We can see the top tube sheet and the bottom tube sheet with the control rod tubes sticking, aluminum tubes sticking up through the tube sheets. This thing was assembled gradually as the parts were made available and of course it was assembled from the inside out. This picture
It gives you some idea of what a side view of the system looks like with part of the reflector installed. This was a four-inch thick reflector loaded up. Here's basically the bottom of the core, and there was another section stacked on top of that.
Here's another view as the assembly progressed, looking down again on the beryllium reflector segment of the core. Remembering this, this is important because when the HTRE two was designed, it was concluded that in order to get the power densities up in the center section of the core, it would be necessary to increase the thickness of that reflector from its HTRE one.
thickness of 4 inches to HTRE 2 thickness of 8 inches. And we can see that difference here with a similar build up of the HTRE 2 reactor showing the center cut out of 11 and a quarter inches.
and the 8-inch thick reflector. The reflector is fully installed here. Incidentally, the manufacturing of the cover or outer section of the tank was really a unique manufacturing job. They took a sheet of either 3-quarter-inch or 1-inch aluminum and put it on a brake press and just kept pressing it until the flat sheet was circular. It was kind of ingenious.
And, of course, the HTRE one fuel element looked like this. It was a concentric ring fuel element with 18 stages of approximately one and a half inch thick concentric sections stacked onto an assembly that would allow remote insertion and removal from the reactor core.
The HTRE two, the HTRE one, of course, the design, having been completed at the time of the design of the HTRE two, pretty much of the HTRE two design characteristics were already established. The fuel elements were basically the same. We're going to be talking for the next few minutes about the HTRE-2 reactor as it was operated and built and operated during the ANP program.
that was controlled and operated by the General Electric Company. The HTRE 2 reactor was started and designed in 1956 and tested, first tested in 1957 and ran throughout the length of the program until it was terminated in 1961. A brief history of the reactors in the ANP program.
The R1 reactor was the initial reactor studied. It was a design that was transferred up from Oak Ridge, the old NEPA project, in 1951. It was studied for about two years, at which time the program was redirected. And the HTRE 1 reactor was started. The design started in 1953. The objective of the HTRE 1 reactor was to try and demonstrate that we could actually operate a turbojet engine on heat produced by a nuclear reactor.
That reactor was tested in 1956, at which time two other reactors, the design was started on. The decision was made to move away from a water-moderated reactor because of the inherent problems in flying water-moderated systems. And secondly, HTRE two design was started.
in order to test larger segments of reactors compared to what could be tested in the various materials test reactors and the engineering test reactors at Idaho. The HTRE 3 reactor was the design envisioned to eliminate the water, which used a hydride zirconium moderator instead of water, and that design was also started in 1956.
The HTRE 2 was operated, as I pointed out, from 1957 through 1961, and it provided a whole series of materials tests, the results of which were used to design not only the HTRE 3 reactor, but also the ultimate reactor in the GE program, the ceramic reactor 104E. The design of the 104E was started in 1960 and was underway.
just prior to termination of the program, and of course everything came to a halt at that point. It'll take a few minutes to look at some of the design features of the HTRE 2 reactor. Most of what we're going to be talking about the rest of the day are what the HTRE 2 design looked like and what the tests accomplished. It was one of the most successful programs in the ANP program as far as total reactor operation is concerned and as far as the results generated is concerned.
The HTRE two reactor is basically a redesigned HTRE one reactor with a center plug cut out of the HTRE one plug and a center fuel and moderator section cut out of the center of the HTRE one core. The idea being, of course, to be able to build an insert assembly that could be removed remotely and have the insert materials changed and reinstalled in the reactor and operated again.
to get test results. The actual plug, the photograph of the actual plug looks like this, where we have the plug assembly made out of steel with cooling water and an insert assembly, which could be removed and changed at the designer's will, and instrumentation section up here to record results. This device, this insert, of course, being removed and installed
remotely because the materials after they're exposed in the reactor are pretty hot nuclearly, making a lot of gamma radiation. The top tube sheet of the HTRE 2 looked like this. There were 30 tubes of the HTRE 1 design, exact same design configuration. The center section was a hex segment was cut out to allow the insert to be installed through the top of the core.
this was about 12 inches across the whole core is a little over 30 inches across the fuel elements in the outer 30 tubes are identical to HTRE 1 fuel elements modified slightly to take account of HTRE 1 test results nichrome cladding over uranium oxide fuel to get some idea of the relationship between the so-called inserts and the parent core we've got a chart over here
that shows basically what the HTRE 2 top tube sheet looks like, realizing that this originally was a HTRE 1 top tube sheet with the center section cut out to allow the insertion of the insert sections. This cut out was about 11 1⁄2 inches. There were three basic types of inserts tested, the Insert 1B, Insert 2B, and the so-called cartridge-type insert.
The insert 1B, of course, being the seven tubes of zirconium hydride and metal fuel elements. The insert 2B being an insert made up of round beryllium oxide-fueled and unfueled tubes with beryllium oxide slabs. And because of the fact that we're getting relatively low power densities in the insert region for these first two inserts,
We designed an insert that would give us a boosted power density in a smaller insert section by building a metal or aluminum can with beryllium slabs that would boost the neutron flux in the central region. And you can see by comparison, we're talking roughly a four and a half foot section compared to an 11 and a half inch. This is a four and a half inch section.
compared to an 11 inch section up here and the types of testing that were done here were the so-called cartridge tests of which there were approximately nine and we'll be discussing those in detail later the first insert tested was as i said was a zirconium hydride insert we had seven zirconium hydride tubes uh seven uh fuel elements
They're essentially basically the same design as the HTRE 1 concentric ring, nichrome-type fuel elements that look like this. The second insert tested was a ceramic insert test which involved testing of beryllium oxide
round tubes, some of which were impregnated with uranium oxide fuel. These tubes in a bundle looked basically like this, stacked up on top like this, and there were about 10 layers, seven of which, no, there were about 12 layers, seven of which were, no, 10 layers. There were seven of which were fueled, and the top and bottom layers were unfueled, acting as reflectors.
and there was beryllium oxide slabs containing the whole structure. This was envisioned as the basic design segment of the 140E reactor, the full-sized ceramic reactor. In looking at the slab and tube design, this particular design did come out of the blue. It was a section, the insert 2B.
was a section of a full-size reactor that was basically designed to look like this. You can see that this center piece here pretty closely represents the way the insert 2B looked as it was inserted in the reactor. And, of course, the whole reactor itself was built up of just layers and layers or segments and segments. This is a mock-up for mechanical testing purposes.
At a point in time, we came to the conclusion, after testing those first two full-sized inserts, we call them full-sized inserts, we came to the conclusion that we weren't getting high enough power densities in the insert region. We would run the reactor up to power, and before we could get the test conditions that we desired in the insert, we'd reach a limiting fuel element temperature on one of the outer fuel tubes in the parent core.
We came up with a solution to that problem, and that was to build an insert that had the basic same size to fit the hexagonal hole, but included a can that had a bunch of beryllium slabs stacked in the can and water-cooled, and with a central hole that would contain the test cartridges. And of course, these test cartridges were considerably smaller than the full-size insert, but we felt that the process was needed in order to get the
The power level's up. The relative size of that test cartridge looked like this. This is a bottom tube sheet off of one of the ceramic test cartridges we'll talk about a little bit later. The tubes being stacked up on a cartridge that looked like that. This is still a lot bigger than what we could test under the kind of conditions that we wanted to test.
in the engineering test reactor or the materials test reactors. Of course, the testing was conducted in the same facility that was used to test the HTRE 1. This was the so-called core test facility, which was a big platform. I'll show you a picture of it in a minute. That contained a vessel in which was housed the basic HTRE 1 type core. This is a picture of the HTRE 1 setup.
with the turbojet engines supplying air through ducts into the top of the core, the air going through the core picking up heat and being ducted back out to the turbine of the turbojet. The turbojet, of course, had to be started on chemical fuel, and as the reactor power was brought up, the chemical fuel process was turned back, just like turning down a burner. To the extent that
At full, at the nominal powers that we're running the HTRE to at, we were essentially providing all the heat to a single turbojet engine. And photographically, this is what the nice pretty color picture of what the core test facility looked like. This is the jet engine we were talking about. This is the duct that took the compressor discharge air into the top of the
reactor core, and this is the plenum that brought the hot air back into the turbine of the engine. This whole device was moved on a set of railroad tracks by a locomotive here that was especially designed for this program. It's heavily shielded to protect the operator as he moved the core test facility up and down the track. Of course, the reason for the track was the
The actual test building was about a mile and a half or two miles away from the so-called hot shop where all the examination and disassembly of the cores was done. The Idaho Test Facility that we talked about is the place where all this testing was done. This picture is not a very good picture, but it shows you. You get some idea. It's out in the desert.
There was an underground bunker that was covered with dirt where people entered, went down to a control room. The control room had access through tunnels to the initial engine test facility, which was a sheet metal shed that could be slid back and forth. And the discharge of the jet engines went out through a long pipe to a smokestack.
During the operation of the reactor, it was very important to pay close attention to which way the winds were blowing. There was always a potential of radioactive release from the operation of the reactors, and we wanted to make sure that if whatever release got out would go be deposited on those parts of the desert that there was virtually no populations. The types of testing there were
Two types of testing, as I alluded to earlier. There were full-size inserts, and then there were so-called cartridges, which are those smaller segments that went into that redesigned insert with the beryllium slabs. We tested four inserts for a total of 256 hours, one of which was ceramic and one of which was a zirconium hydride moderated insert. We tested nine cartridges.
for a total of 1,087 hours, with a total of megawatt hours of 12,332. The nominal reactor power during all this testing was somewhere between 12 and 15 megawatts. As I mentioned, the testing of this reactor provided essentially
all the design information from the standpoint of materials tests that were used in all the future reactor designs for the aircraft propulsion at GE and Evendale at the ANP program. Let's spend a few minutes talking about what the results of some of these tests were and how they impacted the program. The insert 1B was the seventh section hydrated zirconium insert with
seven stainless steel clad zirconium hydride bars loaded with the concentric green fuel elements. As this reactor was started and testing proceeded, it was very apparent that we couldn't get the test conditions, especially in the moderator material that we desired. And we were looking for temperatures in the neighborhood of 1650 degree F, moderator temperatures.
And the general conclusion was that there wasn't much reason to continue this test, and it was concluded early and sent back to the shop for examination. The second insert was a first in a series of many tests on beryllium oxide tubes. As the examination or design analysis went on,
and program philosophy evolved, it became apparent that probably the only way we were ever going to get something that really approximated a true flight-type reactor was go to a ceramic-type fuel element. There was therefore a lot of emphasis placed on the development of ceramic-type materials. The first testing was done on these round-type beryllium oxide tubes.
impregnated with uranium oxide that testing was commenced and at the outset we generally we came to the conclusion that was a gross misdistribution of air between the parent core and the insert we just couldn't get the temperatures up in the central part of the insert assembly the reactor was returned to the hot shop several times to provide plugs to the
to the tube sheet on top of the core. This is a tube sheet made out of Inconel with little holes punched in the top to allow the air to get through to get into the BEO part for cooling. Eventually, we were able to get temperatures up in the core in the neighborhood of 2750 degrees F, which was our design target.
And we commenced to operate the reactor. As we monitored the test results, two things were happening. The temperatures on the thermocouples that we had available started to increase, and the thermocouples started to fail. And also we started picking up fission product, indication of fission product release on our stack monitors. This became a point of concern. There was always a...
a concern that with a an uncoated fuel tube that we probably would get fission product released it was decided however to continue the reactor operation in order to accumulate time on that particular type of material and as time went on we the temperatures continued to increase and at some point we finally lost all the fuel all the thermocouples in the core we did however complete 100 hours of testing
25 of which was at temperatures in the order of 2550 degrees F and 100 hours at temperatures over 2800 degrees F. The core was returned to the hot shop for examination. As the reactor was unstacked layer by layer, we started to discover the first thing we observed was there was
almost a total loss of fuel in the lower fuel tubes in the insert section. The tubes were essentially bleached white, which indicated to designers that that part of the core was probably operating over 3,200 degrees, which was considered a temperature required to boil that fuel out. As we continued to unstack the core, we started to notice the presence of a lot of crystals.
In fact, crystals in a quantity sufficient to essentially block the air to the fuel tubes. And it became apparent that this was one of the reasons that their temperatures were increasing as time went on, that the tubes were being blocked and the cooling air was prevented from getting to the fuel elements. As a result of that, of course, the big question is why was that happening? And boy, that's not a good thing to happen.
if you're going to design a ceramic reactor. And the eventual conclusion was, after the materials guys had done a pretty in-depth analysis, was that we were basically hydrolyzing the beryllium oxide at the top of the core and then redepositing it in cool regions in the hot section of the core. And these crystals especially were noted at the joints of the tubes.
If you take a look at what I mean by a joint, where the tubes come together, the stacks of tubes come together, they form a joint. And it was apparent after some analysis that the air was getting down the interstices of the tubes. What I mean by that, that if you hold a bundle of tubes up, you'll see that there's a little gap right where the tubes come together.
And the air was getting down in between the tubes, and it was coming out at the joints. And where it was coming out at these joints, it was cooling the joint. And therefore, we were getting these crystal buildups. Well, of course, this is a revelation. Obviously, we're not going to build a reactor that's going to operate at the temperatures we'd like it to operate at and kick out uranium fuel and fission products.
So the emphasis was placed on trying to find some way of coating the tubes on the ID to prevent this efficient product release from taking place. The next insert we tested was another zirconium hydride insert, but in this insert the stainless steel cladding was removed and the tubes, the zirconium
Hex sections were slotted to improve the heat transfer. This reactor was operated. We got 1200 degree F on the moderator. However, we almost immediately ran into a problem. We started getting indications of fission product release. And the fear was that we had ruptured some fuel elements in the concentric ring fuel.
So the reactor was taken back to the hot shop and it was examined. And it was discovered that the fission product release was basically due to the deposit of uranium oxide in the lower plenum of the core test facility due to the previous test, the ceramic test where we had boiled off all the fuel from some of the lower fuel tubes.
And this fuel had deposited, UO2, had deposited in the plenum and was actually fissioning in the lower plenum. The inability to achieve the desired moderator temperature was again determined to be partially due to the fact that we couldn't get the power distributions, the power densities that we'd like to get in these full-sized inserts.
The last of the full-sized inserts was essentially a copy of the zirconium hydride insert with the unclad moderator, except this time a beryllium oxide or a beryllium bar was inserted in place of the central fuel tube. And the desire here was to increase the core flux.
And also, the primary reason this particular insert was run was to evaluate a hazards condition that would result from the total blockage of air to one of the fuel tubes. This test was run. The air to the tube was blocked remotely by an automatic valve over one of the tubes. And the reactor was taken back to the hot shop for examination.
The examination showed that in the process of blocking this one tube, it had essentially overheated to the point that a lower section of the tube dropped out, like the support structure had failed and the tube had dropped out and was caught in the lower support structure of the fuel element. And in addition, the fuel elements had melted, some of the melt had gone out into the reactor.
And in general, the testing essentially revealed what the characteristics of what a melt would be. Of course, this went into the history file for future evaluation at the time the HTRE 3 Reactor Act, HTRE 3 Reactor, went to test or would go to test. The conclusion.
of the reality that we were having severe problems in getting our power densities up in the full-sized inserts, of course, led us to go to the design of the cartridge-type insert. The first of the cartridge tests was going back to the ceramic reactor concept again, using round beryllium oxide tubes stacked in a hexagonal array around the
the aft tube sheet and this this again i'll just show you this is a the concept of what the stacking process looked like we had a the tube sheet and these these tubes were stacked over these holes one on top of the other until we we had our full insert assembly the reactor was operated uh for 100 hours at temperatures above 2500 degrees f
And of course, the primary purpose was to quantify what was going on in the previous insert from the standpoint of the crystal growth, the air blockage, and the temperatures that were involved. And again, we were trying to duplicate the beryllium oxide transportation, or the hydrolysis of the beryllium oxide and the transportation down and redeposit as crystals farther down the tube.
And also there's a question of what's a realistic operating temperature for an uncoated beryllium oxide tube. The test was completed, and the insert was taken back to the hot shop and the cartridge removed. And we observed the crystal growth in the same places that we observed it in the previous test on the full-sized insert.
Again, the trailing edge indicating the cross flow in the round tubes. And again, we measured efficient product release as time went on. As testing increased, efficient product release increased. Well, this is kind of significant from the fact that you conclude there's migration across the fuel tube, and as the temperature
time history of the core increases, you get efficient product release increase. This essentially confirmed the notion that we probably weren't going to operate any ceramic reactors with uncoated tubes. And the desire then was to try to find out something you could coat these tubes with so that you could prevent not only the hydrolysis of beryllium oxide, but also the efficient product, prevent the efficient product release.
The second test of the cartridge type was with a zirconium hydride moderated fuel tube assembly again. But this was the so-called XMA1, which was the first zirconium hydride type flight type reactor design. The specs on this particular cartridge were the specs.
that would be given for the flight type reactor. This particular design had a round concentric ring fuel element, but the materials were a little bit different. They were chromium UO2 titanium core material, which previously was essentially UO2 and nichrome, with an iron chromium yttrium cladding.
This reactor was operated for a period of time at peak temperatures up to around 2,100 degrees F. We did detect fission product release again, and this again resulted in the termination of a metal ring test where we were again afraid of blistering or rupture of the metal rings. The reactor was removed back to the hot shop again for examination.
And this examination did show that many of the fuel stages, the concentric green fuel elements, had blistered. And in fact, some had ruptured, which, of course, would have explained why we were getting efficient product release throughout the period of the testing. This information was fed back to the materials guys who used it to evaluate
design characteristics and the effectiveness of this new type material, fuel oil material. The next cartridge went back to round ceramic tubes again, but this time the tubes were coated on the ID with aluminum oxide. The coating, I think it was approximately three mils thick. They were both unfueled and fueled tubes, the unfueled tubes acting as top and bottom reflectors.
The tubes were contained in an insulation liner in order to minimize the heat loss to the parent core or to the parent insert in this case. And this cartridge was operated at a maximum temperature of about 2,500 degrees, producing a maximum power in the insert of about 300,000 watts. This in itself is a significant thing because you've got an insert that's
a fuel section of a cartridge that's essentially 30 inches long and about four and a half inches across its flats. If you can imagine that size of a piece of material, UO2 fuel producing 300,000 watts, you come to the realization of just what the power producing capability of uranium fuel is, enriched uranium fuel. This particular cartridge had
also had a radial variation in the fuel loading. And the objective here was to try to obtain a flat temperature profile across the cartridge. Of course, results of earlier testing on cartridges indicated that we were getting a slight distortion in the radial profile. And the desire was to try to flatten that out. The test was operated for 106 hours at a temperature
above a 2,500 degrees F in order to evaluate what the effect of that coating was. Reactor was returned to the hot shop for examination. No beryllium oxide migration was detected, and no crystal growth was found, which is all right. We got a nice indication that the coating is working pretty good.
However, we did find streaks, tan streaks, outside of the tubes, again suggesting that we're still getting this interstitial flow through these round tubes. We did get some fission product release, though it was much, much less than previous testing. I've got a little picture here to give you some idea of what some of these tubes looked like.
previous test where we're talking about the crystal growth. And you can see at the trailing edge of these tubes this white material that's building up in these areas. This is the kind of blockage that we experienced during this type of testing. Well, the general result of this particular test and the previous two tests was that
Interstitial flow is a problem. Round tubes are a problem. So the obvious choice was to try to come up with a tube that would essentially eliminate the interstitial flow and this would be a hexagonal tube which would essentially fill up all the gaps. This chart gives you a picture gives you an idea of how the round BEO or the hexagonal BEO tubes were stacked up to
These are incidentally fuel tubes. You can see the dark color that the tube turns when the uranium oxide is added to it. These tubes are also coated on the ID with, I think it's alumina, aluminum oxide. And this is what the hexagonal type tube looks like. And you can see that when you stack them all together, there's no interstitial little holes for the air to move down.
So the process of going to a hexagonal type tube was the next point on the development curve of the ceramic tubes, ceramic reactors. And the cartridge was built with hexagonal beryllium oxide tubes, which were coated on the ID with zirconia this time, ZRO2. We're trying different types of coatings to see which is more effective, aluminum, zirconia.
Again, about 2 or 3 mils thick. The reactor was, of course, the purpose of the test was to evaluate a zirconium oxide coating as compared to aluminum oxide, relative efficient product release, and hydrolysis generation. This reactor was operated for a total of 102 hours above 2,500 degrees and 102 hours above.
2,600 degrees. However, there was some flow restriction required in order to achieve these test conditions. The little pins had to be put in to get the temperatures correct. This reactor was taken back to the hot shop and examined, disassembled and examined. The core was, or the insert was. We found no blockage, no crystal growth, no cracked or broken tubes.
We did have some reddish color on the ends of the tubes, but we couldn't identify what was causing that. There was some blistering observed between the fuel and the zirconium coating. This information was fed back, of course, to the materials guys to figure out what was going on and see if they could improve the design. The next cartridge tested.
was basically the same as the previous one, except instead of a ZR, a zirconium oxide coating, it had an aluminum oxide coating. So now we have a hex tube with an aluminum oxide coating. We also went to a slightly smaller tube, and we did another thing for the first time, and that was to stagger the rows. And by staggering, I mean that whereas if you take a bunch of tubes and stack them,
you'll have two rows essentially budding, two stages essentially budding each other like this so that you have a gap. You could have a gap between the two stages. By staggering the tubes, every adjacent row was set at a little different height so that no two tubes ever lined up at the joint.
And of course, this was done to try to minimize the crossflow process. The loading, again, was varied radially to try to flatten the temperature profile. Again, we were trying to evaluate what the effectiveness of aluminum oxide coating was. The reactor was operated for 46 hours at 2,500 degrees F and 99 hours at 2,600 degrees F.
When we talk these reactor operations, we're talking 200,000 to 300,000 watts generated in those insert sections. We did achieve a flat temperature profile for the first time. And of course, this is much desired because you can get a better evaluation of what the tubes are doing if they're all running at the same temperature side by side.
Post-test examination showed no crystal growth, no cracks were observed. There were some cracks observed in some of the tubes. And after the test, it was generally concluded that this was probably due to an inferior grade of beryllium oxide. We did have an increase in fission products during the test. This was probably due to the cracks, the cracked tubes, fission products leaking out through those cracks.
The next cartridge in the line of beryllium oxide testing, practically everything from here on throughout the remainder of the HTRE tube program was beryllium oxide testing. We're getting to the point now we want to finalize the basic design and materials content of the 140E reactor. Again, this was hexagonal fuel tubes and unfuel tubes, unfuel tubes acting as the...
top and bottom reflectors, with a zirconium coating on some and an aluminum coating on others. This particular test was operated in order to evaluate a hazards condition if some of the tubes had been plugged or were plugged due to inadvertent damage or plugging at the air source for some other reason. The reactor was, the testing was terminated.
After about 850 seconds, when we had got an inadvertent scram, the sensors, we believe, detected some kind of an excursion due to efficient product release coming out the back end of the reactor and interpreted that as an unacceptable period on the increase of the reactor power and shut the reactor down. This test was repeated.
except the desire was to essentially change the cooling characteristics a little bit. The central tubes were still plugged, but the outer two rings, some air was allowed through those outer rings. No, the outer rings were unplugged, and some air was allowed through the central plug region. There was about 10% of the air was allowed to go through it.
The hex tubes had no ID coating. The object here was to see what the differences between an uncoated tube and a coated tube would be under these meltdown conditions. The objective here was the same as the previous test. We wanted to find out what the effect of a blockage condition would be on the ceramic tubes.
Maximum temperature in that particular region of the reactor probably was in excess of 3,900 degrees. Yeah. Though for this particular test, we measure reactivity on these limited mill experiments to find out what the effect of losing fuel is on the reactor core also. We want to know if we lose fuel, how much is that going to cost us in excess reactivity?
And this particular test indicated that we didn't lose any reactivity. So the indication was we didn't lose any fuel. Hotshop examination showed that practically all the tubes showed no, the tubes outside the unplugged tube showed no evidence of milling.
The block tubes did have heavy deposits of white crystals. Now, of course, remember, these were unfueled tubes, or uncoated tubes on the idea, so we were pretty confident we were going to get the hydrolysis problem reappearing. The white crystals turned out to be, of course, beryllium oxide, in some cases yttrium oxide, which was...
installed in some of these tubes as a binder or as a mixture with the uranium oxide. The rear grid plate for this test was intact, indicating that we had not milled a lot of fuel out of the center region.
Some fuel tubes were cracked and some were shattered. But the basic result of this testing was that we gained a high degree of confidence that if we did block the air to a small section of the reactor for whatever reason, we could feel reasonably confident that that propagation wouldn't spread throughout the reactor. It would be relatively well confined.
The final test in the cartridge test series basically brought us to the end of the ANP program was kind of a proof test on hexagonal fuel tubes with an ID coating of zirconium oxide. All the fuel distribution is varied across the flat to give us a flat temperature profile. None of the tubes were blocked.
The reactor was operated at temperatures in the neighborhood of 2,650 degrees F for a period over 300 hours. The desire here was, well, we thought we were pretty close to having what we think is a final design on a ceramic fuel tube. We wanted to accumulate a lot of time, measure efficient product release as a function of time,
measure, and of course after a shutdown, determine what if anything happened to the tubes operating at this temperature. We had one small problem during the operation. As time went on, and we had a couple of shutdowns during the testing, and this is 200 type hours, we were getting a xenon buildup in the parent reactor core that became high enough to prevent us from making the reactor critical.
We couldn't override the xenon with the control system we had. But fortunately, as we had talked earlier in previous discussions, we were able to raise the moderator temperature to about 200 degrees F. And since this reactor has a positive temperature coefficient for reactivity, we were able to achieve enough reactivity in the system in order to get it up on its period, or get it up to power.
Now, of course, once you get to power on a reactor, you tend to burn that xenon off, and you gain your excess reactivity back. We desired to measure fission product release as a function of time and temperature. We found, as a result of this, that we had increased the fission product release by a factor of 2 between 2,600 and 2,700 degrees
but we increased it by a factor of 10 and going from 27 to 2,800 degrees F. Of course, this is a significant, we're still releasing fission products. This is going to be a continued design problem. When you put a type of reactor like this in the air, chances are pretty good you're probably going to release, continue to release fission products, unless a new coating could be discovered that would totally prevent fission product release.
The general appearance of the tubes was good. There were no broken tubes. There's no broken unfueled tubes. There were some broken fuel tubes. There was some blistering in the cladding. Again, all these things we had observed. In this particular test, there was an unfortunate occurrence. After the examination and post-test analysis was started, we discovered there was a small piece of fiberglass or glass insulation laying on top of the cartridge assembly.
which could have affected the airflow to some of the tubes. And although the general result was very good, there was, I think, enough confidence as a result of this test to continue to proceed with the design of the ceramic reactor, which, of course, was the intent with the 140E reactor design, which was started in 1960. Program was terminated in 1961.
and all testing on the HTRE 2 reactor and any other reactor was also terminated. We have here a picture of some of the people, especially in the manufacturing end, that worked on the HTRE 2 reactor. A lot of them you probably recognize. Some of them I don't recognize, but, for example, here is John Tenenfeld. Here is Jake Keeney. Here is...
Hal Mendenhall. Here's Jerry Moore. Here is Buck Jordan. Here's John Mundey. And Harry Teeting. No, that's not Harry Teeting. That's Hubie Keck. That's Harry Teeting. And Joe Messenslager. And in back, you can see the HTRE 2 core and its erection dolly. The skin is on.
And up here, of course, the control system.
3.4 HTRE-3
We're fortunate today to have with us Charlie Schiltz, who worked for Franklin during the HTRE three days, the heat transfer reactor experiment number three. And Charlie will go through that particular program, which is our final metallic reactor test that was done in Idaho. Charlie, welcome. Thank you, Otto.
Heat transfer reactor experiment number three was the third in a series of four ground-based nuclear reactor tests leading ultimately to a flying model which was going to look something like this. This shows the reactor in the center flanked by two jet engines below and all the ductwork connecting the two of them.
It was by far the most advanced test to that date, incorporating the latest developments and features consistent with an actual flying model. Its primary purpose was to evaluate and confirm performance characteristics of the reactor, which had been predicted largely by the analysis and non-nuclear testing.
It was affectionately known in those days as HTRE 3, which is a play on the letters HTRE-3. This power plant had the look and feel of a real flying system in terms of shape, size, weight. With the reactor positioned horizontally for the first time, and two jet engines,
slung below, it almost looked like it could take off. Even into external structures, nuclear shielding and air ducting were representative of the real thing. However, a non-requirement of the experiment was the power and flow levels required to propel a large aircraft.
Instead, these two parameters were set at relatively low fractions, such that maximum fuel element temperature and core discharge temperature would be consistent with those produced by much larger engines. This strategy allowed for proper testing of the two main threats without having to rely on yet unproven engine technology.
And furthermore, by being the correct geometry, it would provide a base for extrapolation to higher power and flow levels later. The reactor itself incorporated innovative new design features which had never been tested before in both a nuclear environment and a horizontal latitude. Among these were the new solid uncladded
zirconium hydride moderators shown here. This is the leading edge, tailing edge, hexagonal shape, no cladding. It's just bare zirconium hydride. All previous tests had used circulating water as a moderating media, the mechanics of which were a spatial nightmare.
always in the way of everything else. One only has to imagine pipes, manifolds, and headers competing for space where carefully balanced air flows were required, not to mention the risk of a pinhole leak, which would have destroyed the test. A very significant breakthrough was achieved in learning how to diffuse hydrogen
into the metal to provide not only better moderation, but also a much improved capability for power shaping and flattening over monolithic water. Hexagonal and cross-section with a circular bore, as shown here, the moderators were equipped with longitudinal grooves around the perimeter.
strategically placed for cooling by forced convection. Between the inner and the outer surface, between the inner surface and the fuel cartridge, an insulating sleeve was provided to limit both radiation and mixing from the fuel element channel. In the reactor experiments, several important design refinements were also
incorporated in order to maximize exit air temperature while not exceeding allowable fuel element temperatures. Some of these include the following. A highly engineered inlet plenum designed to provide uniform air distribution over the face of the core. In this way, air flow could be more properly metered
to each component by its own design. Second, the hydrogen content of the zirconium hydride moderators was ramped upward toward the outside for power flattening, towards the outside of the core. Power flattening in fuel elements was done by increasing UO2 content toward the outside rings.
Fourth, heat flux at the tail end of each cartridge was increased by additional rings to compensate for a lower temperature differential between metal and air at that location. The moderators were extended well forward of the leading edge of the fuel cartridges in order to shift axial power peak forward, which was always the intent. The more you could get the power forward,
the better the thermal performance would be. Sixth item, circumferential power scalloping around the reactor perimeter was trimmed with boron steel shims at the peak locations. Number seven, all of the variables associated with control rod worth in a brand new environment and with...
manufacturing or control system tolerances had to be systematically dealt with during the development of this experiment. And finally, while secondary nuclear heating in non-fuel components was not significant to overall system performance, it was clearly vital to their structural integrity. Calculation methods for internal heat generation
including Monte Carlo, were developed and used extensively for this purpose, and they are still being used today. Now for the reactor. The reactor had a nominal diameter of 51 inches with an active core length of 30.7 inches and an overall length of 43.5 inches. This is a...
View from the tail end showing the cross-section with some of the zirconium hydride moderators already hung in place. No fuel there yet. The core consisted of a hexagonal array of 151 moderator cells, each 4 inches across flats and with a 3-inch bore through the center.
Surrounding the array was a beryllium reflector of varying thickness, filling the margin between the outermost moderator cells and the cylindrical pressure vessel, seen here. All components were supported from the front tube sheet, obtaining lateral only support from the rear tube sheet.
All axial thermal expansion was thus accommodated to the rear. Fuel cartridges, shown here, consisted of 19 stages of fuel elements. You see, and they looked something like this. This is a
reduced size model, which was used in a HTRE two experiment. And then this was the same general geometry, only bigger in diameter, as you can see here. Each fuel element was composed of 12 or more concentric rings, each one and a half inches long.
The rings shown here were made from a blend of 80% nickel and 20% chromium, also known as nichrome, and fully enriched uranium, all cold-pressed together, centered, and the fuel matrix was then sandwiched between two layers of pure 80% nickel, 20% chrome cladding,
and they were edge-sealed by brazing. The composite was then rolled, hot-pressed, roll-formed into a ring, which is in turn brazed to the supporting hardware shown between the rings. While the design manufacturing process had been largely completed, prior to HTRE 3,
And an enormous amount of work was done during this phase to minimize process variability and maximize performance reliability. Shielding consisted of lead steel and borale. In this application, water cooled for both simplicity and convenience.
Having been designed to be airworthy and capable of supporting flight loads, the resources were just not spent on the design of an air cooling which would be needed later for a flight prototype. Controls for the two engines consisted of two independent speed controls governing speed by throttling fuel flow. Once reactor air temperature increased sufficiently,
Fuel flow was gradually cut back, and the engines were run at constant speed only. Later in the test series, a coal startup using the reactor motors, starter motors, and reactor heat only was successfully demonstrated, and we'll deal with that later. Reactor controls consisted of two primary independent but closely coupled systems.
Reactor power was automatically controlled by the nuclear flux system, which in turn moved control rods in and out to maintain a set power level with an accuracy of plus or minus 1% over the full range. The temperature control system maintained a prescribed core discharge temperature by modulating air flow
through a bypass valve in the external ducting. Backing up these two control systems were the safety systems monitoring temperature and flux levels at many strategic locations. The safety and control system loops were designed to be coincident type logic where two out of three circuits governed any safety action. However,
the failure of any one of the three circuits would activate one of these safety modes. One, the interlock mode, which prevented any increase in power demand, and two, the override mode, which reduced power level, and three, the scram mode, which caused immediate power shutdown. In addition,
The scram event was designed to be completely independent of the control system requiring only a signal releasing the spring-loaded rods to drive themselves into the core. Overall, the control system was fairly sophisticated for its day, but archaic by today's standards. How many times better it could have been with a fully integrated computer control system?
and better instrumentation. Now for the power plant. The turbo machinery consisted of two modified J-47 engines, which were in current production at that time, with their combustion sections replaced by compressor discharge scrolls and turbine inlets fitted with special inlet annulases.
Each was equipped with a separate combustion chamber for full fuel operation. The reduced airflow for this application required the design of a brand new engine control system. Now while the experimental hardware was being designed and built in Evendale, activities at the Idaho test site were at a feverish pitch.
All kinds of remote handling equipment and skills were being developed for disassembly and possible reassembly of very hot components, nuclear hot. There had to exist a proven method for removing and replacing every last nut and bolt by non-human hands, a huge task. A platform had to be built.
shown here, which could transport the test hardware from the assembly point to the test site, make all the necessary connections, manage the test, and finally return the system to the hot cell for disassembly and evaluation. A major accomplishment was the construction of the test dolly whose weight and span were supported by four railroad tracks,
A rough idea of the amount of hardware involved in this test rig was given by the total loaded weight of almost 270 tons. I remember one of our competitors took that picture and put a caption under it, this is supposed to fly, or something like that. You remember that? Okay, now on to the test.
After assembly, the reactor was tested at very low power levels between April and June of 1958 in order to confirm nuclear design parameters and to provide basic data for the as-built system. During the series, the reactor poison profile was altered four times to achieve the desired flatness of the radial profile.
On September 8, 1958, the assembled experiment was finally wheeled out to the initial engine test facility, known as the IET, for a shakedown run of the overall power plant. During this period, a number of adjustments were made to operating profiles again due to differences in performances at elevated temperature. Unexpectedly,
A power excursion occurred during these tests, which resulted in the over temperature and collapsing of fuel element rings. The positively identified cause of this problem was an erroneous reactor power indication during the withdrawal of control rods. It was a false signal.
The false signal was associated with a high-series resistant noise filter in the output of each ion chamber's power supply. It had nothing to do with the instrument itself, but it had to do with its power supply. The over-temperature condition appeared to have happened so fast that some melting of fuel elements occurred before the automatic scram function,
took place. Immediately following this failure, the system was returned to the hot shop for disassembly. Except for the fuel elements, no significant damage was found in either the reactor or the rest of the power plant. However, a six month delay was inevitable before testing could be resumed.
On June 8, 1959, the refueled core was returned for low-power testing again and was found to duplicate the system parameters of the initial core within specified parameters. Following a minor alteration of the poison configuration, the reactor was returned to the IET
on July 1928, July 28, 1959, and went critical on October 9th of that same year. The reactor was brought to a power level of 10 megawatts and controlled stepwise increments with thermal and nuclear measurements made at every step. After completion of this series, the plant was examined for defects in the hot cell
None were found. The analysis of test data verified that performance had met or exceeded all expectations under these test conditions, and the system was now ready for full-powered testing. On December 14, 1959, the plant was wheeled out to the IET for endurance testing and was successfully operated for
166.5 hours at very low power levels. Then the reactor was brought to full power in six incremental steps, ultimately reaching 31.8 megawatts. During this series, data was gained on the relative contributions of chemical versus nuclear energy to the engine performance.
With this data now available, six transfers to full nuclear power made. Further testing was done over a wide range of engine speeds and nuclear power levels. Start-up and cool-down transients were also studied. Probably the least trivial accomplishment of all during this series
was the accumulation of 126 hours of endurance at full power, including one run of almost 65 hours. The icing on the cake was on disassembly, finding all components in excellent condition and ready to run again. However, the most valuable product of these tests were
is found in the mountain of information and tons of operating data, all of it worth its weight in gold to the engineers and scientists who only a short time before had only theories and calculations to go by. While HTRE 3 lay dormant for about 10 months, when it was again wheeled out for a special test to verify the power plant,
could be started on nuclear power alone instead of having to start on combustion and then switching to nuclear. This was an extraordinary tricky assignment, requiring the simultaneous management of a large number of variables, all in the transient mode. While an analog simulation model was constructed,
Only a human being in those days could be trusted to watch all the gauges and move all the dials exactly the right time and the right amount. A very proficient operator was found, and he did it not just once, but three times in December of 1960. A post-test analysis revealed that manually controlled transients
closely followed the simulation model. And this is, this feat is approximately the equivalent of manually piloting a modern aircraft that can only be flown by wire. It's that difficult. Between the second and the third nuclear startups of that previous test series,
The reactor was operated at 29 megawatt for one hour. Then it was manually scrammed to simulate an uncontrolled shutdown. In the absence of turbojet airflow, this condition might have resulted in serious consequences, including over-temperature of some components. But with one blower furnishing a modest amount of after-cooling air,
all measured temperatures declined steadily and relatively uniformly with no damage resulting. Finally, an elevated temperature test was performed to determine system integrity under a temporary worst-case scenario where maximum fuel temperature was forced to run at or near
2,030 degrees Fahrenheit. The test lasted 20.3 hours with all other parameters remaining stable and a clear indication that normal operation could have been resumed. All testing was thereby concluded, leaving a lasting impression that nuclear
powered flight was indeed possible the htre3 program had clearly met or exceeded all of its technical objectives but much more significant was a confirmation that dedicated hard-working intelligent people could transpose an idea from the depths of unknown knowledge to a real
working apparatus some of the key players in the design of the reactor including the following franklin who was the program manager bruno benigni who i believe was materials and procurement manager jack simpson supervisor of thermal nuclear analysis aubrey smith
was principal engineer of mechanical design. Henry Bermanis, also known as Hank Bermanis, thermonuclear analysis. Mike Goldstein, nuclear analysis. Fred Robert Shaw, our materials expert. Ray Cook, who did all the design work on the fuel and the moderator.
John Mundy, our manufacturing engineering specialist. Charles Jones, who did reactor component design. Roy Eckert did reflector design. Aaron Spivak handled some of our stress analysis. Jack Cole, also stress analysis. Bill Chenault, who was responsible for...
some of the core hardware items. These are only some of the key players associated with the core design. Many, many more people made major contributions to the power plant and its testing, and here are a few of them. Many of those faces I don't recognize. Some I do.
In the fullness of time, of course, nuclear power flight was not destined to become a reality. But that does little to diminish the success of HTRE 3 experiment. And for the rest of their lives, the men and women who made it happen believed that it would have succeeded. And along with that...
goes a legacy of pride and satisfaction and confidence not commonly shared by everyone. If anybody were to ask you what the most important aspect of the A&P program was, the answer would come back, materials, materials, materials, materials. And one of the most important feature of materials was what materials do you use?
materials are acceptable in nuclear reactors. Long before A&P got started in the days of the atomic bomb development, a intense program was conducted at the Brookhaven National Laboratory to measure what is called cross sections of nuclear materials. These cross sections essentially
involved the impingement of a beam of neutrons of various energies on the material itself. And the objective was to measure what the absorption was of the neutrons, what the scattering characteristic of the neutron, in the case of uranium and other fissionable materials, what the fission cross-section was. And from that information, the materials scientists
were able to determine the kinds of materials they wanted to use in the reactor. In some cases, for a control rod, you want a material that absorbs neutrons. In the case of a reflector, you want a material that's going to not absorb neutrons, but bounce them back into the core. And of course, in the case of materials you're using for structures, like aluminum or stainless steel, you want the material not to absorb.
thermal neutrons so that they are available ultimately for the fission process. These measurements were compiled in a book that looks like this. It was available at A&P and in essence it's a compilation of the cross sections of all the materials and their various isotopes that were available at the time. In the case of like a control rod,
There are several materials that were looked at. You look through the book, for example, and you can see what a cross-section for boron is. Boron being looked at because it has a very high absorption cross-section. You can see it starts out over here at the higher energies at about, oh, 15 or 16, no, at about six or seven barns.
barn being the measure of the cross-sectional characteristic of the material. And as the energy of the neutron slows down, you move over here, and you get up into the neighborhood of 3,000 or 4,000 barns, which, by comparison to other materials, indicates that boron is a very good absorber. Oh, there are even better absorbers for neutrons. For instance, europium, a rare earth, looks something like this.
You get down into the thermal neutron region over in here, and you're up in the neighborhood of Europium. It happens to be on this page. You're up in the region of down in here in the thermal neutron region. You're up in the neighborhood of 30,000 barns. Compare that to 2,000 or 3,000 for boron.
It makes geropium look like an excellent control rod material. The question is availability and cost. In the ANP reactors, we more or less defer to the cheaper material, boron. So from the standpoint of what went into a reactor, this particular set of information became more or less a bible.
You consulted this every time you wanted to incorporate a new material or a different type of material in the reactor core to see what its effect on thermal neutrons was going to be.
4.1 Mission Studies
While the heat transfer reactor experiments were being designed and fabricated and developed for further work and then tested in Idaho, Advanced Engineering and the A&P Laboratories were diligently working to identify the power plant that should fly. The mission, however, proved to be a moving target. We have Tom Hunter with us today, who was a member of this Advanced Engineering team. Tom, welcome aboard. Thanks, Otto.
Well, you're right. It was a moving target. I'll try to put it in context for you here if I can today. The General Electric Company, based on its experience in the development of turbojet engines, took on the task of developing a nuclear power plant for a manned aircraft. The system that would be utilized by GE was referred to as the direct cycle. By that, the direct cycle means that the air from the compressor
of the turbo machinery was passed directly through the reactor where it was heated and then into the turbine and into the nozzle and thereby providing the necessary thrust to meet the mission requirement. Pratt and Whitney then was asked to look into an indirect cycle. Now the indirect cycle essentially took the heat from the reactor, transferred it into a liquid metal circulating loop.
This loop, then, went through an intermediate heat exchanger where a secondary loop of liquid metal was discharged in the radiator into the center of the turbo machinery. And in the same manner, then, the compressor discharge air was heated by the radiator, went to the turbine, out the nozzle, and provided the necessary thrust. These early years, the 1951 and 1955,
were primarily research and development, R&D. The unknowns were many disciplines that had never been explored before. Nuclear radiation problems had to be combined with thermodynamics, structural problems, shielding problems, reflectors. All these were new materials to the ANP program. The shielding had to be such that it could protect the crew and the aircraft parts from radiation damage. Design studies
of various configurations were prepared in order to develop the concept. As a mission, the general concept was to provide extended range and endurance on nuclear power alone at medium altitude, high subsonic speed, with chemical augmentation to give a dash of Mach 2.5 at 65,000 feet. This general relation provided an early guideline in the R&D program.
The early work, however, gave an indication of the difficulty in obtaining the necessary temperatures to provide sufficient thrust. Also, shields had become quite heavy to provide the necessary protection for the crew. Studies on how to maintain the system from a remote handling standpoint also provided to be more difficult than we had imagined. The general arrangement to optimize these problems led to the utilization of a single reactor
with two turbojet engines, one on each side of the reactor. This became known as the XMA-1A series. Two of these systems were necessary to provide the proper propulsion power for one aircraft. This meant that two of these would result in a flight weight of approximately 500,000 to 600,000 pounds for the aircraft. Now this seemed to be pretty large and pretty heavy.
in those time period. But today, it's not unusual to find a C-5A or an extended range 747 approaching a million pounds in gross takeoff weight. By 1955, the Air Force, the DOD, the AEC, all felt that it was necessary to focus the effort towards nuclear flight. Accordingly,
The 125A weapon system requirement was established. This operational requirement was referred to as number 81, and it called for a cruise-dash-cruise mission. That was specifically to provide cruise on subsonic on nuclear alone with the dash at high altitude, supersonic, utilizing the nuclear plus chemical fuel.
hard objective put on the program. Ballistic missiles were under development. In-flight refueling of bombers was being developed. These events were making it necessary to establish some definite goals for the nuclear-powered aircraft. However, there were still many problems to be solved. There were limitations on available materials to meet the high temperatures. This resulted in a reduction in the predicted performance.
The dash radius was less than required. Shield weights had increased. However, there were many accomplishments in this same time period. The HTRE experiments, the heat transfer reactor experiments, provided many accomplishments as their testing was accomplished at the Idaho testing station, where they were evaluating new materials. These included metallic and ceramic fuel elements, coatings, moderators, reflectors.
all these leading to obtaining higher temperatures. The HTRE tests obtained the necessary airflow for testing by operating the attached turbojets on nuclear power and could also operate on chemical power. During the entire HTRE series, many hundreds of hours were run on nuclear power alone. The large X-211 turbojet engine designed to match the reactor was being developed at Aircraft Engine Group.
Convair was under contract to modify a B-36 bomber. It would now be designated the X-6. This was to be used for flight tests to evaluate the divided shield concept. A reactor of low power was installed in the aft bomb bay, while a separate crew shield was provided in the forward part around the crew compartment. The reactor of low power was operated in flight.
About 40 flights were made with 24 of those being with the reactor operating to measure the effect on the crew compartment and to evaluate the activity on the various aircraft components. Also, valuable experience was gained by studying the remote handling problems of the radioactive reactor system. However, the problems of meeting the 125A mission resulted in redirection of the program.
back to an R&D effort. There was still a strong interest in the extended range aircraft, but the ballistic missiles and the in-flight refueling systems were moving ahead. And accordingly, it was necessary for increased the performance of the nuclear power plant to be competitive with these other systems. Accordingly, work was directed toward materials which would provide higher temperatures. At the same time, there was conflicting directions as to the next goal.
Some wanted an early flight with a prototype. Others wanted to wait until a definite mission objective could be established. The Strategic Air Command, SAC, proposed an operational requirement for a CAMEL mission. CAMEL stood for Continuously Airborne Missile Launcher and Low Level Weapons System. This CAMEL mission had these specifics. It was to cruise at 30,000 feet
at Mach .9. It was also to have capability to fly on the deck at Mach .9, which is more difficult than flying 9 at altitude. Convair won a contract against Lockheed to work with General Electric on the initial design of a nuclear-powered manned bomber to meet the Camel mission. There was no airplane development program.
authorized at that time. The prospects in 1959. General Electric was satisfied with the requirements of the Camel mission and worked with Convair on design studies based on utilizing the XMA-1A metallic fuel element early with later use of the XMA-1C with the ceramic fuel element. GE proposed that the aircraft check out beyond chemical fuel only.
with the first flight being made on the XMA-1A and next flight being made on the XMA-1C with the ceramic elements. R&D efforts should be pressed ahead to develop higher temperature fuel elements in order to get higher performance. The XMA-1A resulted in turbine inlet temperatures of about 1500 degrees, while the XMA-1C with the ceramic elements were able to provide T4
turbine inlet temperatures of 17 to 1,800 degrees F. By August 1959, the XMA1C, XMA1A, had been canceled as not being able to meet the requirement, which was now asking for longer life and increased performance. In February 1960, Convair and GE completed a study of propulsion systems evaluating the two-engine, one-reactor, the XMA1A series,
as opposed to the single engine, one reactor system. GE recommended the single engine and now designated it as the P140. Several configurations were studied. Asymmetrical design with the compressor shaft going right straight through the center of the reactor core to the turbine. This symmetry made for a simplicity of design relative to structural requirements.
A second version was looked at where the reactor was pulled away where it could be shielded more effectively, but this then required the compressor air to be ducted into the reactor and then back out of the reactor and into the turbine and out through the nozzle. A third configuration took another look again at what we call the twin engine, the two engines on one reactor. This is similar to the XMA-1A series.
Took this into an aircraft installation, we had to have two of these, which meant we had a very broad power plant installation problem. The chosen design was the shaft to the center of the core. This arrangement, similar here in this mock-up, shows the inlet with the compressor discharging through the reactor, through a combustor, and into the turbine, into the nozzle.
The power plant could be run on chemical alone, using the combustor, the chemical combustor, or nuclear alone, or a combination of both. This symmetrical design then was the chosen design. So what could we do with the aircraft? GE had chosen the P-140. I want the model. Convair came up with an aircraft which they called the NX-2.
The NX-2 is a canard configuration. By canard, it means that there's no tail feathers, no stabilizer and rudder. There is a small elevator in the front. This concept was quite effective for a divided shield concept in that the power plants then could be mounted as far aft as possible with the crew as far forward as possible, giving the greatest separation between the power plant, the radiator, and the crew.
the advantage of being able to put the three engines together. Next one. As you can see on this configuration drawing, the three engines are mounted side by side. We refer to this as being Siamese. It has a definite advantage in addition to being far separated from the crew. Next one. It has the advantage of doing shield sharing. Now by shield sharing,
This simply means that if you look at the NVO, the power plant, we could lop off the side shield on the center power plant and the inboard side of the external power plant so that this shield was sharing here and this one here. This resulted in a net savings in the shield weight. In order to get the true evaluation then, the combined studies were based on these kind of requirements for the NX flight.
a Mach 0.8 at 35,000 feet with a minimum thrust of 8,120 pounds, with an engine life of 1,000 hours, with a turbine inlet temperature of 1,800 degrees F. This set of requirements then was transposed into a typical mission for the NX-2 weapons system. This flight profile, this being the runway, shows that
The aircraft would take off on chemical fuel alone. On the climb, it would switch over to nuclear fuel and continuous climb on just nuclear power until it reached its own station here at Mach 0.8, 35,000 feet, whereby if it were required to make a penetration, it could make a penetration at altitude or on the deck. If not, it would continue to orbit on station for approximately 120 hours. At the end of that time,
It would then return to base, where it would transition back from nuclear power to chemical power, and would make the landing under chemical power. Thereby, both the takeoff and the landing were under chemical power, thereby minimizing the radiation effect on the flight station. The last chart, pretty complicated, but it shows the degree to which these studies were made, where we looked at the different flight profiles.
The ground checkout, the chemical takeoff, the climb to station, the cruise on station, the mission on maneuver on station, the two-engine operation, and so forth. These parameters were all studied to evaluate the concept of the NX-2 with the P-140 engine. Typically, at the on-station flight profile here of 0.8 at 35,000 feet, the thrust developed by each power plant was 8,120 pounds. At the time, it was using a turbine inlet temperature of 1,740.
degrees F. Take it down. These design studies, for the first time, indicated that a P-140 power plant in the NX-2 aircraft could meet the requirements of the CAMO mission. Development of the ceramic fuel elements was progressing very satisfactorily. The NX-2 aircraft was ideally suited
for a prototype test bed. However, the program was canceled in 1961 after finally achieving what was considered to be a prototype configuration that had reasonable possibility of being successful as a manned nuclear-powered aircraft. Hi Garth, good to see you again. Nice to see you Otto. In 1955
X-211 Engine
the decision, the major decision was made to build the X-211 engine and also to have the twin engines for one reactor. Could you give us some background on that? As I remember, Otto, the single engine design was proposed in early 55, possibly late 54.
And not too much work had been done on it compared to the twin-engine version, which had been studied for a long while. There was a study put together to compare the performance of the single-engine configuration and the twin-engine configuration. And I might mention as a preamble that it was going to be a turbojet. We wanted potentially supersonic flight.
and there were no fan jets in existence at the time, and, of course, propellers wouldn't give us that speed. The comparison study lasted for several weeks, as I remember, and Wally Thompson suggested we use a 10-point must system as the fighters use. In other words, whatever parameter you were comparing, you give it 10 points, and if the, for example, the twin engine
had a lower weight than the single engine, then the twin engine would get 10 points, and the single engine, 9 or 8 or 7, whatever you thought. So there was a fair amount of judgment in this comparison. On the other hand, we had not enough engineering data to do a detailed comparison in terms of all the parameters you're interested in. The primary one, of course, was the thrust-to-weight of the engine.
The reactors themselves were not that different. The major difference, of course, was the single engine had the shaft through the core. The twin engine system had very long shafts outside the core, as well as a complicated ducting system. Many of the arguments relating to the comparison involved parameters that were difficult to quantify or even measure in any way,
which one is going to be easier to build, easier to repair, easier to maintain, that question. The end result of the study was, as I remember it, the twin engine configuration was selected as the major configuration to proceed with. Now there was another besides the shaft through the core for the single engine, there was another one where the reactor was ahead of the engine.
It was a twin, but there were two engines and the reactor was up in front. And Jack Hope had... I don't remember that particular configuration, so I... It may have been discarded prior to this decision. Yeah, the one we were actually studying in some detail was simply the one that became the XMA one, the twin engine with a reactor and the two engines on each side. Right, right.
And this is before we tested even the first reactor, the metallic HTRE one. Before that test, yes. Because that was in January of 1956. Right. Well, thanks, Garth. Okay, I can't think of anything else to say about it at this point. That covers it.
4.2 X-211
41 years ago, the Air Force proposed the 125A weapons system with the following objectives for the aircraft. Continuous cruise for 40 hours on nuclear power at Mach 9 tenths above 20,000 feet and sprint near the target area for 2,000 miles at Mach 2 1⁄2 and 55,000 foot altitude, including a low-level penetration.
at Mach 9 tenths on the deck that is about 500 feet. The turbo machinery designated for this was called the X-211 and was placed under development in the large engine department of General Electric in Cincinnati. We're pleased to have with us Bob Miller who was the mechanical systems engineer on this X-211 project. Welcome Bob. Thank you Otto. I'd like to describe the
total machinery program to you today. I'll use two charts and then several photographs that will actually show some of the hardware that was used in the program. Now on the first chart, you can see there were two configurations for the X211. The first was called the twin because there were two sets of total machinery
operating on a single core reactor. This was done because it was not feasible at that time to run a shaft directly through the core. Later on, with a new reactor, it was determined that they could run a shaft directly through the core, so then we came up with the inline version. The program was initiated
with a task force in June of 1955, and then followed the design and manufacture of the hardware with the first engine to test for the twin in January of 1958. The first engine, however, was a single active engine on one side with a dummy on the other side.
for the first operation. The first real twin with two active engines was in January of 1960. The first engine to test was in January of 1958. That was the twin configuration, but only one side of the engine was active. The other side was a dummy simply to represent the weight and inertia of a set of turbomachinery.
The first twin operation was in January of 1960. The first of the inline engines ran in December of 1959. And we'll see some of this turbo machinery later on. The program was terminated, unfortunately, in March of 1961. This next chart shows some of the design parameters for the engine.
in that the airflow was 425 pounds per second per set of turbo machinery, which means that the twin engine core was operating at 850 pounds per second. The turbine inlet temperature at military power was 1700 degrees Fahrenheit, and that was to be increased slightly to about 1800 degrees for the inline version.
The max thrust was achieved at 27,370 pounds, and at that time, that was the highest thrust achieved on a turbojet engine. Now, as you will see in more detail later, these two configurations required quite different combustion systems. The twin engine required parallel combustors running around the outside of the reactor.
However, in the inline version, combustors were built in behind the reactor core. We tested six of the twin engine configurations and four of the inline configurations for a total of 758 total test hours. I don't have the detail to split that down. I'd like to remind you now.
of some of the management people involved in the program. Bruno Bruckmann was the manager of the engine program, and he is here shown discussing the 1 tenth scale model of the twin configuration with Roy Schultz, Herman Miller, and Don Berkey. It was hard from that
picture to get a real sense of the size of the x211 turbo machinery so this picture shows a comparison of the x211 1 10 scale model and the j79 1 10 scale model the j79 was 207 inches in length with a 34 inch diameter compressor the x211 has a 54 inch diameter compressor
and a 510-inch length. And 510 inches is 42 1⁄2 feet, which gives you some idea of the size of the engine. Of course, while the engine design and manufacturing was in process, it was necessary to build a new test cell. This was built at the Evendale plant in the back of Building 500.
And here you see the test cell with the inlet air duct and the air discharge duct. Now I'd like to talk about the actual turbo machinery a little bit. This picture shows the simulated core used for the turbo machinery development.
assembly has in the front of it a valve which is used to control the pressure drop and also to shut off the core flow when it's necessary to operate on the parallel chemical combustors. This next picture shows the core assembly from the back. So you can see the
the chemical burners that were used to simulate the reactor heat addition. Also, this picture shows the bearing supports for the shafts which ran on each side of the core assembly. This picture shows the engine in the assembly area.
which gives a much better visualization of the overall size of the twin configuration. You see this one engine here with a dummy, in this case, on the left side. And we see here the 16-stage compressor with its variable stator mechanism. And then there's the collector and the reactor simulator, or core, the turbine collector.
turbine inlet collector, and the three-stage turbine, the exhaust pipe, and jet nozzle. The chemical parallel combustors have not yet been installed at this particular point. This is a similar shot, except this time from the aft looking forward, so that you can see the variable exhaust nozzle, the exhaust pipe, and again, the collectors and the
combusters and the compressor in the distance. This picture shows the twin configuration being rolled into the test cell, being pushed by the dolly that you see in the front. Off to the side, we have a compressor inlet screen, which is rolled in front of the engine to prevent foreign objects from going through the engine.
anything that might be loose or break loose from the cell during the actual operation this picture shows the first of the inline engines that's assembled and so you see again the compressor and simply a frame going from the compressor to the core and the same cores we had before
perhaps slightly modified, and a frame going to the turbine, the three-stage turbine, the exhaust duct, and the variable jet nozzle. And you can see that it's greatly simplified because we have been able to eliminate the two collectors, the compressor collector and the combustor discharge collector, or turbine collector. And these were the two most complex pieces of equipment.
that were required in the twin configuration. And of course we were happy to be rid of those. This last shot of the turbo machinery shows an inline engine in the test cell with the inlet screen in position and with the test equipment starter connected to drive and start the engine. In the final version,
the starter would be an air turbine starter, which would be built into the compressor bullet nose. This last picture shows a group of the management people on the program standing in the exhaust collector of the test cell. You may not be able to recognize these people, but I'd like to point out just a few. There's Gerhard Neumann.
and Bruno Bruckman, and Stu Beekman, and Jim Harrison, who was the engine test manager, and several others that I won't introduce at this point. Now, this is the one-tenth scale model of the X-211 twin configuration.
doesn't look very large from here, but recognize this man and the fact that this compressor rotor tip diameter is 54 and a quarter inches and the overall length of the engine is 510 inches or 42 and a half feet. Let's go quickly through the twin configuration. There are two sets of turbo machinery, obviously.
working through a single reactor. Each set of turbo machinery consists of a 16-stage axial flow compressor with variable vanes. In the final version, there were six sets of variable vanes, but in the initial test configuration, all stages were variable in order to
establish the aerodynamic performance of the compressor. The compressors then feed into a compressor collector and the two compressor discharges are gathered together and fed through the single reactor core. Now we'll come back later and talk about the chemical burners which are on the outside and you can see there are three ducts.
taken off of the compressor collector. Those are split each into two, so we have six chemical burners which bypass the reactor core. These chemical burners are necessary for starting and for operation on chemical for takeoff and so on. These, the core and the chemical burners discharge
into the turbine collector. Now, the fact that we have a twin configuration introduces some complications into the turbomachinery, namely these collectors, which are a complex pressure shell, as you can imagine, and particularly this turbine collector required internal insulation to provide a shell that was cool enough to have adequate strength to support the turbines.
Also, the size, the spacing of the compressors and the turbo machinery is determined by the size of the reactor because the drive shaft for the compressor from the turbine runs right along as close as we could to the reactor. The shaft is definitely on the outside of the core.
Well, I don't know if it's outside of all of the shielding, but it is outside of some of it anyway. Now, for factory test purposes, we built a chemical burner inside of a simulated reactor so that we could operate on the reactor simulated by the chemical burners, or we could operate on the parallel chemical system.
Now, in order to do that, you also had to be able to operate either engine independently, this one or this one. So that meant, say, if you wanted to operate that engine, you had to shut off this compressor so the airflow couldn't backflow out through the second compressor. And similarly, you had to prevent flow from going out through the turbines so that it was necessary to establish several different valves throughout the engine in order to make operation possible.
Also, there were valves in the combustor air inlet ducts in order to shut those off, so then you could operate through the core system. Also, you can see the bullet nose where we had the bevel gear drive that drove the accessories, which were located down on the bottom, and the starter motors on the outside, as you can see here.
The aerodynamic and mechanical design of the basic turbo machinery was based on the J79, which was a contemporary in the design phase of this X211 turbo machinery. But, of course, the J79 compressor inlet diameter was like 29 inches, something of that order. There's a picture here that we perhaps should have mentioned earlier, but it gives a quick overview of...
a typical turbojet, really. It has a simple axial flow compressor with a heat source, which may be a heat exchanger in this case, and a core reactor, or it could be simple chemical burners as it is in a normal jet engine, then the turbine and exhaust. So that's the basic system. You can see the variable stators in through here, and this was the actuation system.
and primarily you can see better the combustor arrangement for the bypass or parallel flow system. Fuel nozzles were located here for each burner, and valves here so each of these systems could be shut off when you were operating on the reactor. There are six chemical burners, internal liners here to provide for the combustion flowing into the turbine collector.
through the three-stage turbine and into the exhaust, and perhaps you can see the variable jet nozzle here. Initially, it was simply a conical nozzle, but in the later versions, it was going to be a conical convergent-divergent nozzle. But we never actually got that far. The engine was mounted from the reactor. Of course, the reactor was...
The reactor was bounded from three points, apparently here and two back here. But all of the turbomachinery components were suspended from the reactor.
The X-211 Engine (minifilm)
The basic objective of the X-211 project at General Electric Evendale is the design, development, manufacture, evaluation, and supply of turbo machinery for aircraft nuclear power plants. The immediate objective is the development of turbo machinery for an advanced reactor core test program. A further objective is the continuation of a program of engineering study, design, and component testing for flight and use with reactors having substantially improved performance.
The X-211 engine being developed for the advanced core test is an inline shaft through reactor direct cycle engine with an overall length of approximately 37 feet and inlet diameter of 54 and 1 quarter inches. The military thrust is 23,900 pounds without afterburner. The engine, which you will see in the following sequences,
is a Z configuration. The Z engine is built up of parts developed for a former twin engine configuration modified as necessary to adapt to the single inline configuration. The Z engine is a test stepping stone between the twin and the engine being developed for advanced core testing at Idaho. In this diagram, the center section containing the reactor has been omitted because of security restrictions.
The compressor has a 12.6 to 1 compression ratio and airflow of 425 pounds per second. The compressor stator vanes are being machined while assembled in the casing in order to gain optimum concentricity of the vane tips with respect to the rotor.
The engine parts are thoroughly inspected prior to assembly. The compressor rotor is turned over slowly while the inspector receives an accurate reading of the tip radii in order to ascertain being within print tolerance. Eccentricity of rotor parts is also checked in this fixture. Blades in all 16 stages of the rotor have been instrumented
with strain gauges in order to assure optimum use of engine hardware and test time in the cell. Let us look closer at an instrumented disk. In this case, a disk to be used in a component spin test. Lead wires are
are carefully nichromed to the surface of the disk. Wires in engine testing are let out the front end of the engine where the strain gauge signal is taken out through the slip ring.
One gauge signal is read on an oscilloscope. The compressor front frame contains the inlet guide vanes which can be controlled as to angle. All 16 stages of compressor stator vanes are also variable.
proper design to withstand the forces encountered in later service, the inlet guide vanes are subjected to hailstone tests. An ice ball of standard size is enclosed in plastic and inserted into the air gun. The vane is placed into the fixture and the gun fires the ice ball at a velocity of approximately 490 miles per hour while a high-speed camera photographs the effects of the ice ball on the vane.
Results are analyzed and tests such as these lead to design of optimum strength, weight, and reliability throughout the engine. In the sub-assembly area, where the compressor front and rear frames and stator casings are mated to the compressor section, some of the details of compressor construction can be seen, such as the 16 stages.
Notice the large movement of vein possible by this unique design. Notice also the shrouds on the early stages of veins. On the engine, the vein movement is checked out prior to using the engine in the test cell.
The X-211 engine can be capable of both nuclear and chemical operation. The use of a test rig is being employed to develop an optimum burner for chemical operation of the engine.
continues into the night. The turbine shaft passes through the center of the reactor
and as such is exposed to high temperatures. The shaft is air-cooled. The X-211 turbine shaft of approximately 17 feet is extruded by a subcontractor.
are being made to balance the turbine shaft prior to installation on the engine. The three stage turbine rotor is balanced by
by optimum distribution of the blading at assembly and then final balance by balance washers at the last stage disc.
onto a fixture prior to the casing assembly. Stator rubbing strips are assembled into place.
be a necessity in assembling some of the components in the later phases of testing at the Idaho test site. The turbine casing remote handling development program is shown here in its initial phase.
operation of controls and accessory components under severe conditions, continual development testing is run on this test rig.
and assembly, the engine is ready to go to test in cell X1 in the large jet engine department. The present engine testing is accomplished with a chemical test burner simulating a reactor to permit independent development of the turbo machinery. Air resistance simulating the reactor is accomplished by varying the inlet area through the porous plates shown.
Combustion is accomplished in the 19 burner cans. The assembled engine is rolled on the dolly into the cell where final instrumentation will be accomplished. Notice the length of the Z engine, approximately 42 feet. Incidentally, the Z engine has already developed on test stand at Evendale.
the highest net dry thrust of any jet engine in general electric history and has exceeded thrust requirements for the advanced core test engine.
of instrumentation wires leading from the engine are to ensure maximum utilization of test time and to protect the engine in the development phase of testing. The engine starter used here is a 500 horsepower steam turbine, later to be replaced by a prototype air starter. The variable area jet nozzle is checked out.
to make sure of proper functioning prior to firing. Last minute checking on the engine is accomplished. The testing is underway.
with all members of the evaluation and test teams at their stations. Data is gathered to verify pretest predictions by systems analysis.
of the X-211 engine continues, directed toward achieving nuclear flight in planes of this type being designed by Convair, the NX-2. Achievement of aircraft nuclear flight will open an entirely new era in aviation.
4.3 XMLA-1A Reactor
I want to talk today about the XMA-1A reactor, but I think I should start by giving you a brief overview of the total XMA-1 power plant, including the X-211. This first chart shows the overall power plant. The XMA power plant was a nuclear turbojet system designed to operate on both
chemical and nuclear fuel. It consisted of a single reactor shield assembly shown here in black and two turbojet engines with the center section taken out to make room for the reactor. I wanted to start out first talking about some of the significant differences between the HTRE number three
and the XMA1A power plant. Just to refresh your memory, the HTRE stood for the heat transfer reactor experiment number three, and it was generally referred to as HTRE three. HTRE three had 19 fuel elements, which were an inch and a half long, whereas the XMA1 power plant had only nine segments, but they were three inches long.
This additional length was introduced to improve the aerodynamic flow through the engine and the heat transfer from the reactor. The moderator for both of these was zirconium hydride. The difference being that in the HTRE three array,
the fuel elements were pushed down the center of a solid zirconium hydride bar, whereas in the XMA1A, the fuel element was integral in itself, and a tri-fluted arrangement of moderator was provided around the fuel element. Both the reflectors were beryllium, and the shield in the HTRE
three was lead in water, whereas the XMA-1A, since it was to be a flight type version, we had to get rid of the water and we went to borated beryllium and tungsten in the front shield, lithium hydride in the side shields, and borated beryllium oxide in the rear shield.
This shows some exploded view of the reactor itself. And these elements are pulled out of the shell. This would be the reactor itself in the part that we saw that was in black on the initial picture. The first thing that's inside there is the core. And then in order to...
contained the radiation of the reactor core it had the side shield around it a rear plug came in from the rear a front plug came in from the front and this is the bypass valve which i talked about before in addition the the side shield was not complete because it ran
in the way of the engine shafting so there was a bearing support beam on both sides which completed the shield and provided support for the turbo machinery shafting. The bypass valve
as shown in this chart, was located at the front end of the shield and its function was to cut the air off through the reactor when the engine was running on chemical fuel. In other words, it was going through those side pipes. The valve consisted of a nose section in which there were cylinders put down through it
And in each of the areas where the air flowed, this cylinder had a hole through it so that when it was shut off, it was twisted in this direction and the air ran into a blank wall, so to speak. When they wanted to open it up, that was turned 90 degrees and it presented holes where the face was and that allowed full flow through the reactor.
This type of design allowed for a very low pressure drop across the valve when it was in its open position. The valve was supported at the front shield through connections in the nose caps and the wavy walls and four splines on the valve flange ring that engaged matching fittings in the front flange.
The bypass valve was actuated by a push-pull linkage system. Two sets of levers, one at the top and one at the bottom, provided the required motion for the blocking cylinders. And they were connected from an actuator cylinder through a linkage mechanism, one of them being located on the bottom and one of them being located on the top.
The XMA reactor core was a cylindrical assembly 62 inches in diameter, 38 inches long, and these dimensions encompass both the side reflector and the forward and aft reflectors. The core consisted of 151
holes for fuel elements, and those were within cartridges in a matrix of unclad zirconium hydride moderator bars. In addition, to the 151 fuel holes, there were 129 other holes that went clear through the reactor, and these provided the paths
for the control rods, there being 129 of the control rods. And it also provided a structural connection between the front and the rear of the reactor core. This design permitted the air leaving the compressor to pass through the full length of the core
in a single pass to cool the core components and pick up the reactor heat. This chart here shows a summary of the reactor core. 38 inches in length, 61.9 inches long. It weighed 11,900 pounds.
It had 151 fuel elements in it, which were spaced at a 4.387 center-to-center distance. It also had 129 control rods. Now, of the 129 control rods, 122 of those were what we called shim rods and seven were called dynamic rods.
In the typical operation of the reactor, the shim rods would be pulled out until the reactor went critical or achieved the required level of power, say it was 25% power or 50% power. And these rods would be pulled out until that power level was reached. At the same time, the seven dynamic rods
were more or less positioned in their central position so that the shim rods established the power level and the dynamic rods took care of minor variations around that power level. Now the shim rods would have to be pulled out various distance depending on the age of the reactor, the time since shut down, the temperature of the reactor, so that it had a
could travel the entire 20 inches. And it may start out early in its life, maybe four or five inches retracted. And towards the end of the reactor's life, they would be almost fully retracted. This is a chart of the XMA-1 fuel cartridge.
And that consisted of nine 3-inch segments of the fuel elements, which were somewhat similar to this fuel element here. This happens to be a 1-1⁄2-inch fuel element. But they were similar to this with the exception that in the XMA-1, there was a large hole down the center.
of the fuel element for a moderator bar to be put inside. That was one of the things that was learned in some of the earlier experiments that we needed moderator inside the fuel element bar. The center moderator bar was mounted in the center of the assembly. The outside surfaces of the fuel element was covered with a
fairly thin insulating material to minimize the heat from the fuel elements into the moderator and to keep the moderator at a reasonable temperature. The fuel cartridges were designed so they could be remotely inserted and removed from the rear of the core. A fuel latch in the nose piece
attached the fuel element to the structure. And during a disassembly or overhaul, a rod was pushed up through the middle, which activated this latch, disconnected it, and the fuel elements could be removed towards the rear. The total fuel content of the core was 500 pounds of uranium-235 in the form of fully
enriched uranium dioxide contained in an 80 nickel 20 carom alloy. The weight fraction of UO2 in this mixture varied from 0.42 to 0.27. The first three stages of the core had a loading of fuel 1.5 times the average.
The fourth and fifth stages had a loading of 0.9 times the average. Both of these variations were to allow the core to operate more isothermally. That is, in order to maintain somewhat of a constant temperature down the entire length of the core. Provisions were also made
for boron steel strips to be placed outside the fuel support tube for the purposes of fine power shimming. In the reactor assembly, the diameter of that center moderator bar
incorporated in the fuel cartridge was varied at seven different radial positions within the reactor. And again, this was done in order to get the radial power flattening. The idea in the power flattening was to have the fuel element operating at its maximum temperature evenly
longitudinally and evenly radial because then you can get the maximum power density out of it. The individual fuel elements therefore in each of the zones had a different number of rings and the number of rings was varied along its longitudinal length in order to achieve this power flattening.
I want to talk now about the trifluted moderator. The moderator sections of the XMA1 core were designed as unclad zirconium hydride triflutes, having the shape pretty much shown in that section of the picture there. This picture shows a complete
hydrated moderator assembly. In the core assembly, the moderator cooling air passes through a series of four holes in the moderator section through the cooling holes in the tri-flute and exhausted to the plenum area at the rear tube sheet and at the end of the tri-flutes. The moderator assemblies
like this, consisted of six full triflutes at the center of the reactor, or there were segments made up of two full triflutes, two partial triflutes, and two arched segments clustered around a single support tube. And there were 55 assemblies of this type in the reactor. We have a photo here.
that shows, this is a full tri-flute here with its cooling holes in it, and this is a partial tri-flute that fit in various other sections, also showing its cooling holes. Side, no, we're done. No, side reflect. The XMA-1
side reflector was designed to be fabricated from unclad reactor-grade beryllium. As shown in this chart, the reflector was constructed of 84 separate segments or blocks for structural stability. Each block was supported by two bolts, one of these bolts being a stationary bolt and the
other one being what they referred to as a swing boat. And that boat was designed to allow for the thermal expansion and contraction of the moderator shell. Like the rest of the moderator, each of the reflector assembly was internally cooled by air that passed through its longitudinal hole, this air being the same air that went through the reactor and through the engine.
The reflector, a whole shell, was a one-piece right circular cylinder with support flanges at each end. The shell was designed to cantilever from the forward tube sheet and this support the aft sheet and the side reflector. It also provided the air seal for the shell coolant and the core void.
The shell was fabricated of aged Inconel X for operation at temperatures of 1,100 degrees Fahrenheit. From our experience with HTRE one and HTRE two, it was found that we experienced quite a bit of thermal distortion of the control rods. The control rods
were in fairly narrow clearance holes, and there was always the concern that uneven heating would cause them to bend and jam them in their own holes. In consideration of the problems of non-uniform heating and cooling in the reactor control rod, in a control rod guide tube deflection and distortion,
A rod designed was conceived that would move freely in a distorted tube, even though it itself was distorted. In this design, the control rod was made up of individual sections. There were four sections, excuse me, five sections, four inches long a piece.
that made up a total poison section of 20 inches. Let's look at another chart here. The individual segments were 4 inches long. This chart here shows 4.8 because it's from a difference from the X140. But what I'm trying to show here is that these sections were joined by straps
diametrically opposite from each other. These straps had enough structural integrity so that you could push on them and on the whole rod without having it collapse. But at the same time, they would not bend in their long direction, but they would bend in their narrow direction. And this gave the rod the ability to maneuver itself around
any thermal distortions in the reactor itself.
4.4 Ceramic Reactor
My name is John Tenenfeld. Today we are going to cover the design and development of the ceramic reactor at GE Aircraft Nuclear Propulsion in Cincinnati, Ohio during the 1950s. Our narrator today is Otto Wojcicki, who was an engineer in advanced design at that time. Otto? John? At GE A&P, a number of ceramic materials
were recognized as having potential value as fuel-carrying materials due to their inherent refractory properties and good strength above 2,500 degrees Fahrenheit. Thus, feasibility studies toward a ceramic fuel element began in 1954 with a nuclear ramjet missile as an initial application under Austin Corbin. In this chart, we have shown the silicon carbide, molybdenum disilicide, beryllia or beryllium oxide, magnesia,
the zirconia, and alumina. These were the primary refractories that were studied during the 53-54 time frame. By 1955, the R&D effort on fueled ceramic materials was accelerated in the A&P research labs. Focus was now on Borrelia as the moderator, fuel carrier, and structure.
Preliminary designs followed of a slab and tube ceramic reactor proposed by Lou Feathers. This study continued through 1958 as a D-101 test reactor. It was vertically mounted and was to use the core test facility from HTRE one and power two J-47s modified turbojets.
With a compressor discharge temperature of 400 degrees, the core was capable of 1,660 degrees Fahrenheit, when the average fuel max surface temperature was 2,400 Fahrenheit. The power level is 44 megawatt. A typical cell for this reactor, there are slabs of BEO.
in triangular form around here, and there's another triangular all around here, and then at the interstices, there's a control rod and a tube, an Inconel X tube, which is a support tube, which we'll show you later in a sketch. There are gussets which hold the slabs together during the assembly, and the fuel elements
are loose in here. And they have to be loose because when this goes to power, the fuel gets up to temperature quickly. And the slabs with much more material here, they take longer to get to temperature. So this is one cell of a typical cell of the slab and tube reactor. And like I mentioned, the control rod is about 3 tenths of an inch, just a shade larger than the one in here. Thank you, John.
The cross section of the 101 reactor, you can see the typical cells with the few elements in them. And the slabs, there are 90 cells. And this is a quarter section. There are 90 cells, equivalent cells, in the reactor. There are some half cells out every 60 degrees. There's a 59-inch diameter with beryllium blocks around the outside.
beryllium oxide slabs in the reflector, and the core diameter is about 44.7 inches. There were almost 7,000 fuel channels in the cross section, and the number of fuel tubes in the entire reactor was almost 63,000. The number of unfuel tubes was about 28,000. The unfuel tubes were
above and below in the reflectors. An axial isometric shows this particular reactor with in red over here is the active core with the fuel elements and then above it are three rows of fuel elements and moderator slabs and down below at the aft reflector
This is the single row. The diameter of the total reactor is 59 inches. The core diameter in here, where the fuel tubes were, is 44.7 inches. And the length of the reactor is 56 inches, but the length of the core was 38 inches. There were 37 control rods in the active core.
There were nine tiers of slabs and 90 cells per tier. A three-tier mock-up, a full-scale mock-up, was built of this reactor. Again, you can see the 90 cells with the
control rod locations. Now instead of control rods, these are the tubes that the control rods would ride in. They're Inconelx tubes. Now this slab mock-up is actually plastic, where the plastic simulates the actual weight of the BEO. There is a weight adder put into the plastic when it was cast. Again, you can see the beryllium oxide section around through here and the beryllium mocked up around the periphery.
So this is the mock-up which was used for assembly purposes and it was used for some of the testing that was done on this. A two-tier mock-up was built with 24 cells using actual BEO slabs. Circumferential bands provided radial loads onto the BEO slabs.
Gussets and pins were used for locking these together and were left in for the testing. The BEO tubes filled one of the cells. It's pretty hard to make out, but it's in this particular cell here. During vibratory tests at 1.5 Gs, relative motion occurred between the slabs and between the tubes from 20 to 200 cycles per second. Test results showed chipping of the BEO slabs and
breaking of the gussets and pins. The gussets and pins were removed in a second test run. Shifts in slab positions of about an eighth of an inch were seen, as well as more chipping of the slabs. Next, a single tier was assembled with a better structure around the outside. It's a cylinder, but it was a single tier deep. And again, you can probably see the tubes within this particular cell.
Now the gussets and pins were left out of this particular assembly. At two and a half G's core rotation occurred and some slab intersections shifted up to a sixteenth of an inch, especially at the half cells. A high degree of vibration was apparent among the fuel tubes.
With additional development, the D-101 might have been satisfactory for use in the HTRE 1 facility. However, for close coupling with the X-211 engine with vibratory excitation as well as an increase in power by about two and a half times, the D-101 was designed, as designed, appeared inadequate. When did you actually start on the ceramic reactor, Otto? Well, I transferred to advanced engineering in 1957.
My assignment was to come up with a competitive ceramic reactor design. The A&P research labs, under the capable leadership of Clay Brasfield, had made significant strides in ceramic technology. Probably the two most significant were the use of yttrium oxide to stabilize the fuel and the use of zirconium oxide coatings on the inside of the fuel element to suppress the water vapor corrosion. Both these material developments will be discussed by Clay Brasfield.
Fabrication of the zirconia clad fuel tubes was by co-extrusion. To this process, the powdered oxides of the clad were mixed with suitable plasticizers. The fuel element mix and the cladding mix were simultaneously extruded to produce a tube, either round or hex, with a circular bore coated with the clad. This next chart shows an extruded length being collected on a half shell.
half shells used for controlled humidity drying of the ceramic material. The tube is then cut to length. The next chart shows the tube size reduction. In the centering process, there's a 21% approximately reduction in size in both length. This is after centering, and this is before centering, both length as well as the cross flats or diameter dimension.
If tighter tolerances are required, then what is achieved here after the centering operation, an automatic grinding operation, a straddle grinder, can be used to bring the dimensions down to within a thousandth of an inch. Now here's another view of a piece before and after centering. Although much testing remained to prove the fuel tube,
the labs that lead people like Ed Aiken, Al Clavel, and Pick Turner, and Fred White, were on the threshold of a 2,500 degree Borrelia fuel element that would provide 1,800 degree Fahrenheit mixed air to the X-211 engine. A dependable reactor design was now needed. Now here is Clay Brassfield looking over one of their products with Fred White and John Collins.
Now, this new design, my initial hope was for a hex fuel element with about 19 cooling channels and one and a half inches across flats. I would even settle for one inch across flats and seven holes. However, projected rejects during the extrusions, during the fabrication, made these what-ifs positively prohibitive. Too costly. Otto, from the changes in the aircraft mission in the late 1950s, it soon became apparent that a
1800 degree Fahrenheit turbine inlet temperature was needed. The X-211 engine itself was capable of meeting this requirement. However, the development of the metallic reactor was falling far short of this requirement. Is this where the XMA-1C came in? Yes, John. That's the original work that we were aiming for, for the 1C, for the twin engine, for two engines on the...
one reactor. After several false starts, the design and development efforts led to the two-bundle reactor. The initial application was the XMA-1C to power the twin X-211s. Because of a steep rise in project activity at the time, I was limited to one designer, but I was assigned the best, Warren Berard. He personally provided all the initial drawings. His zeal and optimism were contagious.
The other key member of the design team was Warren Pollington, a contract engineer. He provided the complex analyses. Both were sound idea men. The initial design, and especially the three-tier development work, was based on the XMA-1C. However, to simplify this narration, I'll concentrate instead on the follow-on design, the D-141. It was designed as a static ground test of a ceramic reactor.
sized to be compatible with a single set of X-211 machinery, and to be operated at the Idaho test station. In order to minimize costs, the reactor was to be operated in the HTRE 3 facility. First, I'll describe the D-141 reactor, then cover significant analyses, and follow with the development testing. On the chart, we have the reactor shield assembly of the D-141, and this is sitting
If you remember, it's horizontal, and it's sitting on like a second floor of the facility. And down below it is the X-211 turbojet, where today, or at that time, the J-47s were on the ground floor. The air coming from the compressor then goes through to the front end of the reactor, through the reactor, out past the plug over here, and goes out back down again to the turbines.
This is the shield plug in the front end. It's lead and water. And the aft shield plug, likewise, lead and water. And the side shield's lead and water. Again, these were not the purpose of the design. The design was to get the reactor going. The next chart shows a cross-section of this two-bundle reactor. It was built up.
of thousands of these quarter-inch tubes. There were about 141,000 tubes in the active core region. And these are some of the tubes that were used. Spaces were provided for 37 control rods within the active core.
And there were 55 longitudinal support tubes throughout the reflector and the core. And the control rod ran within the guide tube or the support tube. The Inconelx guide tubes then were located, there was insulation around the guide tube, and then they were located inside of a ceramic BEO arch, or it was really just a large.
hex tube. It was about an inch and 3 eighths in ID. The support tubes were externally insulated, internally air cooled. The outer reflector was 4 and 1 half inches thick. It was built of hex bars. About 90% of the tubes in the reflector were solid, and about 10% actually had cooling air going through them to cool the entire reflector.
I guess I could walk around here a little bit, 41-inch active core diameter, 50-and-a-half-inch diameter around the reflector bars, and around the periphery there are pads, or at that time Inconel X, later stainless pads, which distributed the load from springs.
which were leaf springs around the periphery over here. And the leaf springs reacted against the shell, which is about 53 inches in diameter. There were spring adjusters at the outside in order to adjust the spring load during assembly. Also, you can go through the adjuster to pick up the inside part of the spring and pull it out for a core assembly.
The next chart shows one of the hexagonal beryllium oxide tubes, or what we call the radial arch. And this was a short in distance and about 1 and 3 eighths in ID. The next chart shows one of the control rods.
Now, John, I believe you worked on some of these control rods. Yes, I did. More on the control rod actuators rather than the control rods themselves. But this is what you call the articulated control rod, where there were sections like every four inches, and then there was a break. So it was completely encapsulated. It was uropium oxide for the poison material, poison to the neutrons, that is. And then there were straps, a strap across here and here.
which would not bend in this direction, but would bend in the other direction. And likewise over here, then, at 90 degrees, the straps, which you can't really see, they sit right over this way. So every other one of these, they have the straps 90 degrees apart. So it does become a flexible control rod to follow the guide tube.
Some of the data of the D-141, the pressure shell diameter is 53 inches. The reactor reflector was 50.5 inches. The length of the reactor was 39.6. The core diameter is 41 inches. The core length, 31 inches. That's the fuel area. Then the forward reflector was 5.5 inches of beryllium, and the rear was 1.5 inches of BEO.
And the side was also Borrelia of 4 and 1 half inches. And like I mentioned before, it's rhodopium oxide for the matrix material in the rod and clad on the outside for oxidation resistance. And then 37 of these in the core. The next chart, sort of busy, but it does show the overall reactor.
with springs around the outside, a shell is supported by keyways, just like in the HTRE 3, to the shield structure, the shield and also pressure vessel, and it's also supported at the aft end by keys. The springs are about every, they're two inches wide almost, and then
They react against the shell, and they react against the pads that I mentioned earlier. And this area, again, the red shows the active core region. This is filled with beryllium tubes. And you can see the area where the control rod would come in from here to here. And there were forward beryllium blocks. And aft, there was a
grid plate essentially it's a tube sheet small tube sheet and hexagonal in shape and the forward shield see it over here so this is the plenum chamber and then the aft plenum chamber over here the leaf springs
I think you can probably show me against my shirt here. The leaf springs, these were actually for the mock-up, which was slightly different from the 141, but not much. And later, I'll show you some of the spring rates and things. But it's rather stiff. It was picked because of the dampening effects of the leaves, for one thing. However, as far as the space allowed and what kind of spring design would
would require the least space within that circular annulus. And this is really one of the most efficient springs. Now, how in the world do you assemble these loose tubes when you want to build up and knowing that there are different regions of this reactor where
There are different hydraulic diameters, different IDs, and also different fuel loadings in some of these. Well, if you visualize a cartridge or a small section of that reactor, it wouldn't necessarily have to be hexagonal in size because this gets heavy. However, it's a good way to look at it. This is an area where there's a support tube coming through, the tube sheet at the aft end, the beryllium block on the front end. And these tubes are staggered. There's a half tube.
essentially, and then a hole tube. And then they're all hole tubes, so you get the other end, and there's a half tube at that end. So it's just like concrete block when you're laying block at your garage or somewhere. But it's done for shear strength, just like it is for concrete block walls or brick walls. Now, as far as locating the particular fuel element in the right spot,
A color coding, first of all, is a system that's used to, it's like electrical resistors, that's used to identify the particular fuel element. And I believe there were 15 different variations of the fuel element. So with the color coding system, and then you apply glue at the end of each end, just a little tab, no more than a tenth of an inch wide, at those ends to hold the ceramics together until the final assembly of the reactor.
So you have it in a jig or a fixture, and then you simply lay these tubes. All the tubes within one length are the same kind of tubes, except on the ends where some are shorter. So there is a very easy way of putting them together. And it's actually the glue only serves a purpose until the final assembly.
put into these tubes, into the cartridge that I showed earlier, and they poison out the reactivity at final assembly. At 600 degrees Fahrenheit, the cement, which was an isobutyl methacrylate, it vaporizes, and the cartridges lose their identity. So then you have the complete bundle, and this includes through the reflector. The aerothermal and nuclear analyses of the ceramic reactors went smoothly under the capable leadership of
Frank Dawson. Their parametric sizing graphs in particular were useful during the advanced design. Got some selected data here, 141 reactor data. The total power to the air was 107 megawatt thermal. The core pressure ratio is 0.846. Average film at max surface temperature is 2,300 degrees. I'll show you later a curve of that.
The Mach number of the fuel exit was about 0.253. The airflow was 350 pounds per second. The pressure, 156 psia, and with the exit, 132 psi, the exit of the core of the channels. The air temperature at the inlet was 700 degrees Fahrenheit from the compressor. And the exit air to the turbines, which is a mixed air, in other words, the cooling airs that bypass and go through the core,
And the 1870 degree air coming out of the fuel channels are mixed to an 1800 degree at the turbine inlet. I want to next get into what the sizes and some of the loadings and things like that. The flats is 0.256 inches. The ID was 0.18 inches, just a shade under 3 16ths. There were 12 variations of the ID.
and there were 15 different regions within the core going radially. The ID varied from minus 8,000ths less on the diameter to 18,000ths higher on the diameter, in other words, a thinner wall. The clad thickness, the zirconia clad thickness was 3,000ths of an inch, or 3 mils, and the length of the tube was 4.428 inches. The max loading, the UO2, was 6 weight percent.
Later, you'll notice when you get into the 140 reactor that by that time, Clay had developed enough work on the fuel element to know that he could get to 10.8% and still have a good, strong fuel element. At this time, we were limited to the 6%. The number of fuel passages was 20,200, essentially, and the nuclear life was 55,000 megawatt hours,
1,000 hours at Mach 9 tenths, 30,000 feet. So that was our design goal. The next chart shows the, this is the, on the axis, the length of the reactor from the zero point to the one is really the front of the core of the fuel to the aft end of the fuel. And showing coming in the air temperature at 700 degrees and going on up to 1870.
from the fuel, remember I said it's mixed 1,800 degrees at the turbine. And then this is the surface temperature of the average fuel element. And this average fuel element reaches between 2,000, 2,400, 2,300 degrees Fahrenheit. And then slows down, comes down to about a little under 2,200 degrees as it leaves the fuel.
The next chart. Because of the importance of the fuel tube stresses, I'll next discuss the allowable tube bundle deflection, the minimum radial pressure required on the bundle for stability,
Max pressure, as well as the three methods of radial support that we finally evolved for this tube bundle. The control rods, since the tube bundle is sitting on springs, it will be essentially bouncing for G-loads. If you get G-loads, we're designing for a maximum of four Gs down, say, and one G to the side. There will be occasional bouncing. As you can imagine, it's flying in an airplane. You'll be bouncing. So would the core.
Now, there's got to be a limit on the amount of bouncing of, say, the center line, because you've got control rods feeding from the stationary, if you will, forward plug to the bundle of tubes. I set an eighth of an inch as being the maximum deflection that we could stand. And we designed the control rod guide tubes and the shield plug and so forth for that.
The allowable displacement of the tube bundle within that shell was established by control route alignment. It was set at eighth of an inch. By dividing this bundle weight by this displacement, a system spring rate is found, which remains invariant in the design. If chosen too small, it can cause large increases in bundle pressure and fuel unit stresses. But there's a minimum radial pressure, which occurs when the radial spring gap is the largest.
and the spring force is the least. It occurs at the highest temperature that the support system and the tube bundle have in common. In other words, when the temperature of the support shell, the Inconelx shell on the outside, your main structure, and the tube bundle itself are at the same temperature, the ceramics will expand much less than the metal.
and so the gap will increase. So that is the critical point where we have to maintain enough spring pressure to keep the bundle acting as a unit. The maximum radial pressure is the algebraic sum of the minimum pressure, that is the allowance for the inertia, plus the assembly tolerances, plus the differential expansions of the shell and the tube bundle, plus the autoclaving from static pressure in the spring cavity.
plus the fuel tube creep and spring load relaxation. The latter two are functions of time, temperature, stress, and NVT, or flux or fluence. All of these variables, which affected the spring gap, were investigated and were controllable. As a tube bundle deflects due to the G-loads, friction between the springs and the pressure pads would most likely prevent the bundle from recentering itself.
returning back to where it was in the center. So another design approach was to support the tube bundle at its periphery by a shear distribution to that structural shell. If keys were used, like we key tube sheets and things, lateral bearing pressures on them would lead to galling and seizing. However, the elliptical spring could act as a flexible key.
Otto, could you discuss the spring rates and the hardware, the materials, and the spans and things like that? Sure. First, I'll mention if keys were used like this from the pads to the shell, there would be galling and seizing, and certainly it just wouldn't work. Again, corrugated shell might be used, and again with keys around through here. We looked at both garter springs as well as a shell, which is...
corrugated, and it was much heavier and did not help us at all. What we did do is take the spring and use the spring itself as a flexible key between the pad and the shell. So John, if you'll notice, we have here a key that's built right into the spring.
It's welded right on. And it's on one end only and on the bottom over here. It's not on the other end. So when this gets locked to the shell and this gets locked to the pad, and again, the shell is this way, the other end is free to expand like this. So we really have two different spring rates. We have one which is keeping the bundle together as a unit and the other which is keeping the bundle at the periphery, at the outer diameters.
from slipping, from moving. On this chart, you can see the spring rate here is like, this is for the three-tier mock-up which we built. Spring rate is like 100 psi, and it's a function of the distance, really, on the gap, the spring gap. And this is the tangential, or the pressure this way. And you can see we're up 900.
to over 1,200 pounds per inch on the spring. In pure shear suspension with the bundle displaced bodily, tangential loads are developed at the horizontal diameter. Internal shear strength is required to hold the shape of the bundle. Now a lower radial spring rate can be used.
When this is supported here by these keys, if you will, or that spring with the keys on it, it's like opening up, trying to open up the core and opening here and trying to break it apart where the maximum pressure is on the top of the tube bundle. Now, in the hydrostatic system before, the highest pressure was here at the bottom.
where we allowed this point to almost reach zero. So we've relocated the maximum stress point from instead of being down here, we've put it over here. Now the obvious next step was the design for both the hydrostatic and the shear support of the tube bundle.
in an integrated or combined support. Again, you can show here the tangential support here to keep that deflection down. And again, the radial support or hydrostatic support over here. So we now have combined, if you look over here, for the tangential, this is the support we get around the tube bundle to keep it from
deflecting too far, and likewise over here, and from getting too much load is really what we're after. And then for the hydrostatic, here and here. So that is the integrated support. Using that hydrostatic rated support, the internal pressure to the G loads is highest on the lower tubes, like I mentioned. For the shear support, the internal pressure acts tangentially in our maximum on the upper tubes.
For that integrated system, a reduction in both the minimum and the maximum pressure occurs. The integrated system, or the balanced support, was the design goal. What about the axial alone and longitudinal support of the tube bundles? Well, the axial support of the tube bundle was a modification of Lou Feather's design for the slab and tube reactor. The retainer plates, support tubes, and forward reflector block
made up this system. The primary objective was to resist the axial deformations of the tube bundle through all operating conditions. At reactor startup, the axial pressure forces are resisted by frictional forces from the radial support system. In other words, you've got a pressure on the forward end, which is higher than the pressure at the aft end, and so the tube bundle wants to slip aft. Also, there is a 1g
pushing in that direction, which is small. So we're talking roughly 28, 30 psi, 28 here, and then maybe another 3 psi, which is an equivalent for the 1g. And you've got around the periphery over here, you've got the friction, which is preventing, trying to prevent that bundle from slipping aft. That total aft pressure force is proportional to the radius squared
while frictional resistance of the tube bundle movement is proportional to the first power of the radius. Thus, the tube bundle will slip afterward, not down through here, but at that radius, at the outer radius, until the retainer plates and the support tubes take up the load. A redundant axial structure exists because you've got 55 tubes, and they're not all the same length. They're not all exactly the same temperature or same tolerances and stuff like that.
So the 55 support tubes pick up loads at different times, depending on tolerances, temperature differences, and so forth. A load relief device, a spring on the forward end of it here, was placed at the forward end to prevent axial distortion of the tube bundle and to assure almost equal loads on all these retainer plates. The total preload on the tube bundle must exceed the side restraint imposed by the peripheral friction that I mentioned.
at the pad. And it's important that the tube bundle returns to its original position at reactor shutdown. Otto, there was some concern about unequal load sharing in the load on the tubes or pressure magnification. Could you amplify a little bit on that? OK, I'll try. That's one area that was really a puzzler at first.
The question of bridging and going around a number of tubes, well, what about the tubes that are doing the bridging? What kind of loads are they experiencing? This pressure magnification within the tube bundle was a major concern. The ceramic elements within the tube bundle vary within the manufacturing tolerances and so forth. This permits some elements to accept more load than others in the form of local pressure concentrations.
Both analytical and experimental means were used in attempts to assess the magnitude of these pressure concentrations. Now this is within the tube bundle itself. One analytical approach assumed that undersized tubes may be assembled adjacent to each other to produce cavities within the tube bundle. Pressure concentrations are expected to occur around the cavity due to the bridging of the load. By applying the analogy of a disk,
with a hole in it. This would be then the load depending on the size of the hole. This would be the loads that would apply to these tubes around that hole. The stress at the edge of the hole can be determined as a function of the hole size relative to the size of the disk. When a large cavity is introduced with a radius up to a half of that of the tube bundle radius, which is rather large,
The tube bundle pressure might approach three times the radial pressure. The inward radial pressure, of course, is zero. In another analytical approach, an adjacent pair of undersized tubes might introduce a pressure concentration factor of three times that average pressure. The largest load multiplier found during analysis for that D141 was a factor of three. Several experimental tests were run, but were largely unsuccessful.
Strain gauges attached to various small and large tube bundle mock-ups failed to measure dimensional changes as the radial pressure, shock, and vibratory forces were varied. Another test introduced successively larger cavities into the pressurized tube bundle and was considered a count the cracked tubes approach. The stability of the tube bundle in redistributing the loads and bridging around the cavities was impressively demonstrated
but the tubes did not fail. The tests gave confidence that excessively large concentration factors did not exist. A value of three was therefore chosen for the design. Otto, what about the stresses in the fuel tubes themselves? Thermal stresses are a function of the overall geometry of the fuel element, and the exact analysis is rigorous. However, it was found that the D141 elements could be treated as an equivalent round tube.
with the same materials cross-section. The max calculated thermal stress using elastic theory turned out to be about 9,700 psi, and it occurred in the region of the tubes with a minimum hydraulic diameter, that is, the greatest wall thickness. The mechanical stresses in the fuel elements resulted from radial and axial loads imposed on the reactor. The radial loads were a primary concern. Beam bending with loading at the center
was the most significant mechanical stress. It resulted from the radial pressure between two or more elements with appreciable camber being squeezed together and restrained. Ring bending, in other words, squeezing the tube this way, depended on the type of loading, whether two or three or four, six sides were loaded. But the photoelastic analysis showed the worst stress when opposite sides of the fuel element were loaded.
Plaudits are deserved by Jim Mcconnelly and Art Ross for reducing the complex stress analysis of the fuel tubes to practical use in this coming section. The total stresses at any point in the fuel element were obtained by combining the individual stress contributors algebraically and using the principle of superposition. The time of occurrence, the location, and the direction was considered in the combination.
In other words, you have to add tensils to tensils and whatever. Maximum tensile stress for ceramics was of more concern than for the compressive. Total combined stresses were found first, the initial elastic tensile stresses occurring during operation, and then the residual tensile stresses at room temperature. You have that tube there? The residual stress, if you would imagine, when I
When we first put the bundle together at cold, we just have a straight fuel tube, say. And then as it heats up, the core expands. And you can imagine now you've got different kinds of thermal stresses as well as mechanical stress. But I'm just going to show a mechanical stress. At temperature, the tube might be like this, where there was no stress in either the outer or the lower fiber.
When I bend it like this you've got a tensile stress on the upper and a compressive stress in the lower fiber Now at temperature if there's enough creep What would happen then if you come back down to room temperature? The spring will press down and actually force the thing to reverse it in so that the lower now would become tensile and the upper one would be in compression So both of those have to be looked at
The combination of residual temperature stresses assumes that complete relaxation had occurred and consequently gives the max value. The design criteria for the D141 fuel elements required that the ultimate stress be compared to the strength available, that is, the average modulus of rupture minus three sigma, or the standard deviations. The validity of the fuel tube and of these calculations was demonstrated
by extensive development testing, which will be discussed in the ceramic material section later by Clay Brasfield. The ceramic reactor development tests, there was significant development testing in support of the 141 design program and is mentioned to illustrate the degree of confidence with which the GEANP department recommended the ceramic reactor in 1959. Nuclear tests of a preliminary critical experiment
were performed by Jack Simpson's engineers to obtain experimental data on shield worth, control rod worth, worth of structural materials, power distributions, and fuel loading. The tests included three successive assemblies. The nuclear analysis was led by Frank Dawson and included Gene White and others. Typical of the aerothermal development testing in support of the 141 were the airflow tests required to experimentally determine air channel friction factors.
These tests needed by Mel Lapides and Bob Spira in design covered the range of Reynolds numbers from 7,000 to 35,000 and were made on a series of 12 stages of tubes. The three-tier structural tests will be next. The three-tier ceramic development mock-up was 56 inches in diameter for the bundle and 14.5 inches long, half of the core length. And again, that was the dimensions for the XMA1C.
This tube bundle weighed as much as my Saturn station wagon. The structural tests explored first the assembly procedures and the tolerance stack-ups with .372 inch tubes, containment of the bundle within an eighth of an inch, pressure load distribution within the bundle, especially load multipliers acting on the fuel tubes,
behavior of the assembly in separate tests of shock loads, vibrations, sustained G loads, axial, aero, and friction loads, and temperature cycling to 900 degrees Fahrenheit. This is a shock test setup. It's probably difficult to see. However, this is the shell and shows the adjusters on the outside to get the proper spring pressures and change the pressure depending on the
the requirement and the shock comes from here and there's reinforcement even on the shell to make sure the shell can be understood as far as the G loads on it. It's a high G system. The leaf springs covered 10 degree arc. So the leaf spring covers a 10 degree arc.
here the pads were five degrees and the pads on the inside towards the bundle locked in with the tubes so that there was always a connection a shear connection with the tubes rather than putting just half tubes or something like that on the outside we wanted to actually get a lock between the pressure pads and the the hex bars on the outside the
Reflector hex bars measured statistically 0.365 inches on the flats while the core tubes were 0.372. Slight radial cracks uniformly around the bundle appeared. They were conservatively set approximately twice the expected differential expansion.
And again, the differential expansion was the cool core and a hot reflector at shutdown. Longitudinally, in the buildup of the three-tier mockup, there were three tubes axially. However, then in order to get that shear connection, we put in a smaller tube on each end.
for every third, I believe it was, tube, although we actually assembled 19 tubes at a time. The tubes were assembled into cartridges, like I showed you before, and using an adhesive, which was that same isobutyl metacrylate, and was removed by heating to 600 degrees Fahrenheit, and the tubes were free.
The ceramic tubes that were used were ALSAMAG number 196 because of its low cost and low toxicity factor. Its density and low temperature mechanical properties were similar to BEO. Accordingly, room temperature stresses in the mock-up corresponding closely to room temperature stresses in the reactor tubes, and that was pretty important. Otto, did the two bundles of the three-tier test mock-up
stay within the 1-8 inch design allowance for the control rod tubes? Well, a good time for a question like that. The shock tests we'll mention next. Measurements were made of the three-tier mock-up acceleration and radial displacement. In some cases, a load cell replaced a radial arch to measure the radial pressure on a tube within the bundle.
The load cell data showed a load multiplier acting on the tubes of approximately double the average radial pressure at the core periphery. The first series of shock tests used a hydrostatic support. We did not use the keys on the springs. Loads up to 4 G's cause a peak deflection 82 mils, or 0.082 inch.
Residual deflection or hanging up from friction forces were within 20,000, but appeared to be increasing. In other words, it was not completely back to center. It was within 20,000, but it looked like it could go beyond. And when you're dealing with friction, you could certainly see where that would continue like that, although it would bounce around. The core was disassembled and modified to include the shear ties of the springs.
After a series of about 25 shock tests, again up to 4Gs, the peak deflection remained below 70 thousandths, while residual hang-up or re-centering of the bundle relative to the shell remained below 10 thousandths of an inch. Now that's pretty darn good when you're thinking of a five-foot bundle weighing as much as my car. These tests, these shock tests were verified.
The deflection of the bundle varies linearly with the acceleration, or the G-loads. The shear ties contribute to retention and centering of the bundle. The undersized reflector bars resulted in gaps in the region. These gaps permitted some movement within the bundle. However, high-speed movies failed to produce any detailed knowledge because of the very small motion involved.
It should be noted I had set the reflector bar undersizing at twice the value expected. This was done to exaggerate the effects of the radial gaps during tests to better understand whatever the effects were going to be. The axial loads were applied via pressure through a neoprene diaphragm. The simulated air drag load was carried from the aft retainer plates, axial support tube, and to the forward preloaded springs.
Five series of tests showed that the tube bundle acted as a unit assembly as predicted. A series of hot gas flow tests were run using the three-tier mock-up. The forward end of selected air channels were plugged to increase the overall pressure drop across the core. The delta P's of 20 PSIG and temperature of 900 degrees Fahrenheit
were simultaneously applied to the reactor. Air flow was cycled incrementally under steady state temperature. Radial displacements and loads in 12 of the axial support tubes were measured. The results were the core did not settle. It remained centered. The springs worked. The axial support system limited the peak load in any tube.
It would also absorb the differential thermal growth between metal and ceramic parts. The bundle returned axially to its cold position. Exploratory vibratory tests were conducted on the three-tier mock-up at GEMSD, the Missile Space Division in Philadelphia. The mock-up was excited at various frequencies and amplitudes to a maximum of 3G.
Sweeps were made over the full range from 20 to 120 cycles per second. The engine ran at 5,000 RPM. The sweeps were then followed with three one-hour steady-state endurance runs. Tests were performed for two radial loads, 17.5 and 10 PSI. The results?
no deterioration or gross dimensional changes were detected. A final vibe test consisted of removing tubes to form two cavities through all three tiers. The result, the structure remained sound and showed no evidence of progressive disruption. We did the component tests on the aft reflector, aft retainer plates, the support tube insulation, springs, the coefficient of friction among various materials.
The materials we're interested in here, as well as the final reactor, including the alzheimer tubes. And I sure have to thank Ernie Ratterman and Spud Eckert and all these other test engineers. You probably mentioned a bunch of others, John Blake, others. They did a commendable and thorough job. Well, then, how did the single engine shaft through the center of the core design come about?
In the summer of 1959, Garth Leith asked me if a pipe could be placed in the center of the tube bundle to allow space for the X-211 midshaft. Garth had already determined that the center island was feasible from a nuclear and thermal standpoint. I said a thin liner should be ample since it's constrained by the tube bundle. In other words, I demonstrated with a round wastebasket. You can squeeze it this way, but if it's...
kept from moving this way, it'll do the trick unless there's a third kind of cycle that comes in for buckling. The chart here shows a buckling test of the duodecagon on a periphery liner. And you can notice it's quite thin. And it failed at 285 psi, which certainly is quite a bit above what we were talking about, what we were expecting. The next chart
shows the initial advanced design of the XNJ140 core with the liner here and then the shaft through here. And these now are just tie rods. They are not control rods in the active core region. The control rods are out in the reflector region. And again, we got the shell and the springs and the pads. These springs covered rather than 10 degrees like we had in the
In the mock-up, for example, the springs in this case covered 15 degrees around the periphery. Because we were sure of the integrated system, we could have a lower spring rate. So obviously, the farther apart you make these, the longer you make the spring, the smaller spring rate you have. Various tests of a liner in the three-tier mock-up were successful. Here is, again, a Duodecagon, the liner.
And it went through all the usual tests of shock tests and everything that we mentioned before. And it was very successful. There wasn't anything unusual at all about it. The final chart shows a three-tier with a liner and with a 10-inch cavity.
after completion of a 5G shock load. So here's the liner. You can see some of the instrumentation on it. And these tubes were pulled out before the shock test with the idea of they seemed to be tight. And so he pushed out all the loose tubes they could during vibration. They pulled them out. And so we knew that this here was bridging around here.
The cavity was introduced before the test and remained unchanged. It amply demonstrated the self-healing and bridging in the ceramic tube bundle design. In the fall of 1959, the ceramic reactor design and development was transferred to final design. The new owners scrutinized the design and testing and found it sound in all major areas.
Arnold, I think to close out, it would be nice to acknowledge some of the engineers who worked on the final design. I put a few of them down on the next chart. I just hope that I haven't forgotten too many of them. Well, it's always a problem. But the players alphabetically are Ken Anstead, who worked on the radial support system, Ray Cook, Roy Eckert, Ernie Elevick, Pete Flagela.
Dave Flitner, Jim Hagan, Bill Hunter, Buck Jordan, Jim Kepler, Dick Lorke, Dick Molman, Cal Moon, Charlie Schultz, Aubrey Smith, Glenn Wessendorf, Dick Wilkie, Jack Cole, all their others. Well, with that, I guess we conclude.
4.5 XNJ 140 Reactor
Proceed with the description of the reactor. Symmetric view here, which shows some of the main features. It shows the engine shaft passing through the axis of the reactor. It shows the active core, which is about 30 inches long. It shows the control rod guide tube. I believe this is the control rod.
penetrating the inner part of the reflector, and there are 48 of those control rods around the periphery. There's a front reflector which is 4.75 inches thick, and there's a rear reflector which is one and a half inches thick. Okay. Well, we'll get pretty smooth about this one.
In this view, you see the fuel elements in the active core, which is a radial segment. I'm sorry, an annulus. The control rod guide tubes and the control rods are shown on this circle. There are 48 of them. This is the outer reflector, which is mostly solid beryllium oxide with appropriately
spaced cooling holes, cooling channels. And this reflector grades the temperature from about 2200 F in the active core down to about 1100 F at the outside. So this retaining springs out here can run at 900 F. At the inner radius there's an aluminum oxide reflector.
which is about one and a quarter inches, one and a half inches thick. There's an aluminum oxide reflector, which is about one and a half inches thick, and which grades the temperature from 2200 F in the core.
to about 1400 F at this Inconel shaft liner. Now the active core is made up of fuel elements which are beryllium oxide. The active core is made up of fuel elements which are beryllium oxide impregnated with fully enriched uranium dioxide and stabilizer materials. Beryllium
is a good moderator and is frequently used as a reflector for reactors. The beryllium oxide is a stable high temperature ceramic and so this fuel tube is used as the fuel carrier, the moderator, and the structure for the active core.
This is an unloaded fuel tube. If it were loaded with UO2, it would be black. But basically, it's a small hex tube about a quarter of an inch across flats with about a .167-inch hole, and it's about 4.38 inches long. There are a total of 25,000 channels in the core. The fuel element...
is internally clad with a three mil coating of zirconium oxide. This coating is present to counteract a hydrolysis reaction between BEO and the moisture in the air. And this reaction product is a white crystalline growth that tends to grow into the airstream. The cladding doesn't suppress fission product release. The fission products that recoil from the fuel
near the surface enter the airstream directly but there is no evidence that fission products or fuel diffuse from the interior of the tube so let's go to the next table this table shows you some of the gross characteristic of this characteristics of this power plant the
number which should calibrate everybody is that the reactor power is about 50 megawatts and this number applies at 30,000 feet with the airplane flying at Mach 9 tenths. The reactor inlet temperature at that point is 582 degrees F and this turbine inlet temperature is 1740 degrees F.
The reactor pressure ratio appears on this chart. It's 0.857, and the pressure ratio is a very important variable and one in which quite a few trade-offs are involved. The nuclear designer would like to have the core solid, so it takes less fuel, has less of an inventory, but the airflow thermal designer
wants to have the core with lots of porosity so that the pressure drop is small. In this case, the trade-off was made at this point and the frontal area is 47% air and 53% fuel.
The nuclear characteristics of the reactor are expressed kind of by the fuel inventory and the excess reactivity requirements. So let's go to the next chart. The fuel inventory for this machine is 87 kilograms of uranium oxide, fully enriched. The excess reactivity at that loading is about 5.9%.
1.5% is needed for equilibrium xenon at 50 megawatts. And burn-up and fission product poisons over the life of the core account for about 2.9%. The temperature effects are relatively small. In fact, listed as zero in this chart, there's a margin of 1.5% for error. And the control rod worth,
is 10%, which gives us a fairly comfortable margin for SCRAM. The next figure I'm going to show you is the longitudinal power distribution in the reactor. Power distribution is all important in a high performance machine of this type, and it needs a lot of discussion. I'll switch.
curve shows the longitudinal variation of the volumetric power generation watts per cubic centimeter or whatever units you want to use on a relative basis with one being the average and this solid curve pardon me but at in this figure the control rods are
inserted about 18 inches, a little more than halfway, and that would correspond to the clean startup condition of the reactor. Now near the inner radius, the curve is quite smooth and it's typical of the classical distribution you get with a fairly thick front reflector and a fairly thin rear reflector. As you go out on the radius next to the control rods, the
the power is peaked further aft. But this is of not too much consequence, sort of a second-order effect, as I'll show you in the next slide. Don't change it. This is a calculation of the longitudinal air temperature distribution for an average tube. And this...
peak here is about 2200 degrees and the exit is about 2010. The peak occurs near the aft end of the reactor because the air has to be heated all along here.
and it reaches its maximum temperature at the end. So whether the power peaks at this position or this position is kind of a second-order effect on this temperature distribution. Over here I have a schematic plot that I drew that shows what the neutron flux distribution would look, what it would look like in the radial direction.
for a reactor of this type. There is a fairly substantial peak near the reflector, and this is because the fast neutrons generated in fission stream out into the reflector. They get slowed down, and about half of them come back towards the active core, but the fuel has such a high fission cross-section for thermal neutrons that the
neutron flux is quickly attenuated. That's the physical picture. The difficulty is that if the fuel elements were sized to handle this maximum power, they would be loafing in the center of the core and therefore the performance wouldn't be very good at all.
Well, the homogeneous nature of this reactor gives us two ways to try to equalize the temperature distribution across the radius. One of these is to vary the hole size. So as you put small holes in the low power region and big holes in the high power region, the other way is to vary the fuel concentration. And variation of fuel concentration
was the mechanism that was actually selected for this power plant. And I believe this chart will show... This shows the fuel concentration in the active core versus the radius. What this means is that the entire length of the core is run at whatever
concentration you see here. Maximum is about 10% where the dip in the neutron flux would occur, and the minimum is about 4.5% out near the reflector where this large peak was shown. The radial power distribution now looks like this. And this is plotted for three different control rod positions. This curve over here
at the top is for control rods inserted 18 inches. The power is highest near the center of the reactor and falls off near the outside. The lower curve here is for control rods fully withdrawn at the end of life and it is low in the center of the reactor and peaks near the reflector as we saw before.
curves are hash-mark shaped because there's constant fuel concentration in each of these regions with a varying neutron flux distribution. Okay, now, having adjusted the power distribution by means of variable fuel loadings, we now can ask what effect this has on the fuel element temperature distribution switch. It shows the
radial variation of temperature with the variation temperature with radius it looks a lot like the power curve but this is now the maximum surface temperature along the length wherever it occurs and you can see again that for early in the life of the reactor these peaks here are about 2300 degrees
And it's very crowded over here, but the solid lines are falling off near the reflector because, as the power did, near the end of life, the center of the core is running at about 2,000 degrees, and it peaks near the outer reflector at that point at, again, about 2,300 degrees. So there's...
From the average temperature, which is 2200 to the maximum temperature, there's about 120 degrees imposed by the small variations in power within each fuel region. It needs to be emphasized that these power distributions were all obtained from a critical experiment.
And the critical experiment was partly loaded with actual fuel tubes. It mocked up the variable fuel distribution. So these are not horseback calculations. These are good measurements. Airflow tests have been run to establish friction factors and heat transfer coefficients. And therefore, the predictions of fuel element temperature that we've just discussed have a
pretty high confidence level now there are other fuel element temperature perturbations that have to be considered uh at the top where we have 120 degrees that's due to the uneven power as we showed you in the previous chart we now talk about fuel element
size tolerances. One mil on the hydraulic diameter is worth about 20 degrees. One mil on the flats dimensions about 15 degrees. A 3.7 percent allowable variation in fuel loading is worth about 40 degrees. All of these effects from tolerances add to 100.
There are other things that have been taken into account in terms of the possible uncertainties left remaining in the measurements that were made in the critical experiment and in our knowledge of the heat transfer coefficients and airflow characteristics and in the lineup end-to-end of the tubes in the channel and these numbers add to about 124 degrees.
and then for the internal temperature rise from surface to corner in the fuel element, we have about 30 degrees. These temperature perturbations are to be added to the average fuel element temperature, which is 2200F, and the 120 degrees built in
from power flattening as shown here, that gets you up to 2330. All these allowances for fabrication tolerances, measurement uncertainties, bring you up to 2530, which is the stated maximum fuel element temperature at cruise. The maximum fuel element temperature
tends to be a rather controversial number. And people tend to regard it as a locked-in thing where if you exceed it, you're done, and if you don't get to it, your performance is bad, and so forth. But as we saw in the preceding charts, for one thing, the maximum temperature walks around the reactor during its life.
And for another thing, the estimate that we have is based on the direct stack-up of a lot of tolerances, which might be better added statistically. So that I would like to convince you that this is actually a fairly conservative design. And this chart, this chart shows the
percentage of fuel elements at a percentage of volume at a given temperature. You can't read it, but along the axis is temperature. This is about 1200 degrees, and this is about 2400 and 2500 degrees out here at the at the tip of the curve. The average temperature of 2000 degrees is about here, and so the bulk of the reactor is
running at fairly low temperatures and only a small portion actually exceeds the average temperature. I think another mitigating factor I would like to mention is simply the fact that there has not been enough experience with this kind of fuel element to understand just what happens at the end of life.
whether the failure mechanism is sudden or whether it's gradual and what just does happen if you exceed this maximum temperature for any extended period of time. This isn't to trivialize the fact that one hot channel may plug up and begin to cause progressive deterioration in the core. The point is that
None of these things can be considered hard and fast until we have further operating experience. There are some auxiliary cooling functions which are necessary, and they detract somewhat from the performance of the system, and I'd just like to take you through those real quickly.
This chart shows you what the operating temperature of some of these other components is. And for reference, I've also put the inner radius, the thickness, radial thickness, and the temperature. The shaft tunnel liner, for instance, runs all together at 1400 F. Now the inner reflector, which is 1.8 inches thick,
has its maximum temperature next to the core, 2200, and the minimum temperature is 1400 next to the liner. And this temperature is dropped from this level to this level by judicious placement of cooling air tubes. Obviously, there's more cooling air near the periphery than near the core.
Here's the active core region. It's 13.9 inches thick, and as we said before, runs at a maximum temperature of 2530 and a minimum of about 1890. These are different kinds of numbers from these up here. The outer reflector is 8.5 inches thick. Its temperature is 2200 degrees near the core.
And that is graded down to about 900 degrees so that the support springs can run at 900, which is their operating temperature. I'll just go... And can you come in a little bit more on that? Okay. All of this auxiliary cooling takes bypass air.
And the way this system has been configured to date, the fuel elements only carry 80% of the air that eventually gets to the turbine. The inner reflector carries 9 tenths of a percent. The outer reflector carries 4%. And by the way, the exit air temperature is 1240 at the inner reflector and 1120 at the outer reflector.
The springs take up 1%, pressure pads 1.3, the control rods and their guide tubes 1.5%, and that air exits at 1050F, and the aft retainer assembly, which is mostly metallic, operates with air at about 1025 degrees F. And this gives you a mixed air temperature from the core of 1800F.
Now another 9% is taken away to cool the shield and some of the engine components. This is about, as I said, 9.1%. It eventually comes out at 1120F so that mixed turbine inlet temperature becomes 1740. Now you can come back to me. Now that table...
shows that the amount of bypass air required to cool the auxiliary components is significant. And on the other hand, the knowledge of heating rates in most of these components is quite a bit less accurate than the knowledge of the fission heating in the active core. The heat that gets to these components partly comes from conduction. It also comes from neutron scattering and gamma-ray absorption.
and these are both more difficult to calculate and to measure than the fission heat. So more accurate knowledge of some of these heating rates and air flows and perhaps an upgrade in the temperature capability of some of these components may lead to further performance improvements, even if they are somewhat small.
Well, the overall results of these nuclear and thermal studies indicate that the predicted performance can be adequately achieved within a thousand hours life of ceramic fuel. The nuclear results are especially good because they're backed up by the critical experiment and prediction of the maximum operating temperature of the fuel is dependent on the estimation of many perturbations
On the other hand, the end of life mechanisms for the fuel are not totally understood. But at this stage, the achievement of the operating characteristics of this power plant looks good. It looks highly possible.
XNJ-140 Power plant
In mid-1959, a decision was made by the Department of Defense to change the objective of the aircraft nuclear propulsion program from the development of a power plant for a specific weapon system to applied research and development of a system capable of propelling an aircraft at a speed of from Mach 8 to 0.9.
Mach 0.8 to 0.9 at an altitude of 35,000 feet. The reactor in this system was to use materials capable of providing a life of 1,000 operating hours at the specified performance level. Although we didn't know it at the time, this was to be the last of many changes in program direction.
for which the ANP program leadership was to be severely criticized in an evaluation report issued later by the General Accounting Office. In view of the revised objective, it was determined that the XMA-1A power plant could not achieve the necessary performance level. Studies were initiated
to evaluate various power plant configuration and potential reactor materials. These studies culminated in March of 1960 with the decision to proceed with a system consisting of a single X-211 engine with its shaft penetrating a beryllium oxide reactor. This power plant
was designated the experimental nuclear jet 140e and it appeared to have the best chance of meeting the desired performance objective and still exhibiting the potential for future growth its development was approved and a target date for a ground test in the flight engine test facility
at the Idaho test station was set for December 1962. This test was designated the advanced core test since the X-211 turbo machinery had already by this time undergone extensive chemical testing. A date was set for flight testing in the Convair NX-2 prototype aircraft
for 1965. Based on work with Convair, design requirements consisting with the Department of Defense guidance were established. These were conditions of airflow, temperature, pressure, thrust, and other parameters. The design points were then distributed
over the required 1,000-hour reactor life. These various design points are shown on our first chart, where the ground checkout time was 20 hours. Five hours was allocated for chemical takeoff.
20 hours for climb to station, and 885 hours for cruise on station. 20 hours was allowed for maneuvering while on station, and 50 hours allowed based on two-engine operation or a single engine out on a three-engine configuration.
The power plant also had to be capable of operating at any speed or altitude up to Mach 1 at 45,000 feet. As a general rule, extended operation at low speed and high altitude would have extended the life beyond the 1,000 hours, while extended high-speed operation at low altitude
would have decreased life expectancy. It was assumed that the reactor would be replaced during normal engine overhaul at thousand-hour intervals. Once the basic design requirements had been determined, a number of studies to fix the material selections and design configurations were made.
The selection of materials and the configuration of the reactor has been covered previously in the discussion of the 140E reactor. One point that should be made, however, is that the shaft penetrating the reactor received heat from both thermal causes from the...
hot active core as well as radiation heating in the material therefore significant airflow was provided through the shaft for cooling the configuration study was done in depth with a preliminary design made for each of three major configurations
The first of these was basically the XMA-1 configuration consisting of two engines but using the beryllium oxide core. The second was an integral single-engine configuration where the shaft of the engine penetrated the reactor. The third...
had the reactor in front with ducting back to the engine. As far as performance, each configuration was to provide exactly the same thrust as a function of speed, and hence they were identical in performance.
However, the twin engine configuration suffered a distinct problem in that engine failure robbed the system of the thrust from two engines. Both of the single engine systems suffered some weight disadvantage because of the fact that it is less...
It requires less shielding for one reactor, providing the same power as two. However, it was offset to some extent by placing the engine side by side, and in the integral single engine configuration, the turbo machinery itself, particularly the compressor, provided significant additional shielding.
Further, the integral single engine unit had the advantage that the elimination of any internal ducting significantly reduced the weight. Ultimately, it worked out that two single engine systems weighed approximately
approximately the same as the two engine configuration. The maintainability of the single engine inline system was enhanced since the system was compact and comparatively easy to handle. Externally mounted components and disconnects were unobstructed by ducting and vertical assembly and disassembly
was readily achievable. Consequently, the integral single configuration was judged to be the most suitable from the maintainability standpoint. The result of the advanced configuration study was a recommendation and subsequent approval of the single engine integral configuration for development to meet the department
of defense objectives. The configuration is basically as shown here and on the model. Originally, the thing was designed to have some integral interburners for chemical operation.
Studies were made to remove the interburners with the result that the shielding weight could be reduced and the shaft length could be reduced. The chemical system would then depend on afterburners. This configuration of the engine
was identified as the XNJ-140E-1 for the advanced core test version and as the XNJ-140E for the flight version. These are the major characteristics of each of the two power plants. The objective version, the XNJ-140E,
was reduced by about 4,000 pounds in weight and the turbine inlet temperatures were increased for various operating conditions from 35 to 45 degrees. The thrust of the objective engine was really way up there for this period of time.
when we looked at a 35,000-pound thrust engine. The Air Force, in November of 1960, issued their advanced development objective number 20. And the immediate objective was to choose...
to achieve nuclear flight in a military prototype aircraft at subsonic performance levels matching the Department of Defense guidance. The initial system would have been used to evaluate the practicality of subsonic nuclear missions of long endurance, such as air alert, missile launching, low-level penetration.
logistics, reconnaissance, air early warning, anti-submarine warfare, and airborne command post. The subsonic system would also serve to develop a basic equipment and operational technology leading towards supersonic nuclear aircraft capability. As a result of this...
advanced development objective, a number of application studies were made for installation of the XNJ-140 power plant in the Convair Model 54, the NX-2 version. Both three- and four-engine configurations were studied, and all versions were designed to meet the Department of Defense guidelines.
The three engine version is shown here and primarily in the model where the three engines are placed side by side with the air intakes in the leading edge of the wing and two chemical engines mounted to assist in takeoff and in nuclear engine out operation.
The differences between the three-engine and the four-engine configuration are shown here on this chart. Neither one of them quite made the Department of Defense requirement. While the three-engine configuration could cruise with payload at Mach 0.8,
at an altitude of 32,800 feet, and the four-engine operation mocked .85 at 33,000 feet. You'll note that the rate of climb with a reactor out on a hot day is pretty pathetic. Three-engine version with payload at 130 feet per minute.
and the four-engine at 340. The ground runs were well within the limits of the runway that was proposed for Idaho Falls, which was 15,000 feet in overall length. You'll notice the rate of climb with an engine out and flaps and gear down is not bad.
The aircraft gross weight is well within the limits of aircraft flying today. The 474,000 pounds for the three-engine version and 576,000 for the four-engine version.
Also, the scheduling in response to this advanced development objective called for providing the engines for operation of this aircraft by 1965. And this schedule involved a major engineering test program, including the advanced core test, a flying test bed program.
and the engineering and operational test of the NX-2. Generalized planning of the various phases of the engineering test, which would have made up the Advanced Core Test Program, were well underway in early 1961. This planning was directed toward, one, evaluating the types, the numbers, and the location of instrumentation sensors.
to be used in and around the E-1 engine. Secondly, establishing the design and procurement of test support equipment. And finally, three, identifying, specifying, and actually procuring the necessary auxiliary services and systems. Detailed planning was scheduled to begin in the summer of 1961.
in order to meet the 1962 ACT test date. Another key element in the overall XNJ-140E program was the proposed flight test of the power plant in a B-52G to be used as a flying test bed.
Working with the GE Flight Test Group from Edwards Air Force Base, a program was developed to begin in 1963. The first phase of this program was to be conducted with a chemically powered version of the power plant. The second phase would call for one single nuclear power plant mounted on one side of the aircraft with a final phase.
consisting of the aircraft carrying two XNJ-140 engines. The B-52G performance could have been capable of full nuclear flight while carrying the two power plants and the eight J-57 engines. This recommendation...
as well as the entire A&P program, was terminated following the President's annual budget message to Congress on March 28, 1961. The program termination was based on the fact that there was not considered to be a specific military requirement.
for a manned aircraft with the characteristics of the subsonic long-endurance mission that was under development. At the time of termination, the XNJ-140E program was on schedule. The basic turbo machinery had been operated on chemical heat sources for a total of 758 hours.
using several different engine build-ups. Structural modifications were in process to adapt the turbo machinery to the specific configuration of the power plant. Design of the reactor shield assembly and sub-assemblies had been completed. Manufacturing drawings of the parts had been completed or were in progress.
Prototypes of the critical components had been proof tested. Beryllium oxide fuel tube assemblies for use in the reactor had been tested for a total of over 10,000 hours in the materials test reactor, the Oak Ridge research reactor, and as inserts in the heat transfer reactor experiment test vehicles.
Reactor critical experiments had verified the fuel element loading specifications. Permission had been requested to proceed with manufacture and assembly of the reactor and propulsion machinery. Ground test schedules, as mentioned before, were scheduled for 1962, flying test beds for 1963.
and NX-2 operations for 1965. Overall, the development effort expended on the XNJ-140E in the short period from mid-1959 to the termination in March 1961 produced remarkable results. Developments in the design of high temperature nuclear fuel elements,
in lightweight shielding, in nuclear reactor engineering and design, and in the accommodation of the crew in a manned nuclear aircraft for long periods of time, and further in the design of facilities. These push technologies to new limits, and I think that all who participated in this program
could and should be very proud of the many accomplishments of the effort.
The Soviet program
I'm holding here a model of the Russian or Soviet nuclear aircraft at some artist's conception. Now, this was probably made about 1964, somewhere around that. Probably earlier than that. I think it was probably around 1957 or 58, that early. And it was presumed that the Russians were building such an airplane. Turned out later that they were not. Not true, right. There was a textbook written in...
Russian, which a number of us got copies of there at GEA&P. This was probably in Aviation Week at the time. Aviation Week. And then a local model company made models and sold them at the 5 and 10 cent store. Aurora, they made a lot of models. Some comments follow regarding the Soviet aircraft nuclear propulsion program.
About 1956, a Russian research paper was forwarded from Washington, D.C. It mentioned both zirconium hydride and yttrium hydride. This came as a surprise. Both moderators were handled at A&P as secret restricted data. A breach in security, possibly at a vendor, was suspected. FBI and others looked into it. We never found out anymore. Later, a Russian paper mentioned their success with barylides.
A&P Laboratories thoroughly investigated but found nothing useful. The paper was probably planted by the Russians to mislead us. In an effort to understand the Russian literature in the spring of 1958, after hours and after hours course in scientific Russian was offered at A&P. Nathan Gilbert of Russian descent and a professor of engineering at the University of Cincinnati kept the course hopping.
Howard Eagle remembers that Fritz Metzger, with his physics background, was the star in a class. Others were John Motiff, Fred White, Miles Leverett, and probably Bob Van Houten and Austin Corbin. Later in 1958, Aviation Week, the respected technical magazine, broke the news that the Soviets had flight tested a nuclear-powered bomber.
In his editorial, Lou Holtz said, The Soviets have needlessly beaten us to a significant technical punch. U.S. Defense Secretary McElroy was highly skeptical. President Eisenhower said, There was no reliable evidence of any kind, and he was not changing the A&P program to achieve early flight. After the Cold War in the early 90s, U.S. intelligence agents
visited russian facilities and spoke to knowledgeable russian engineers about the amp program their conclusion was the russians did not build an atomic powered plane in 1953 or thereabouts there was a big study of advanced different types of reactors with potential applications to the nuclear powered aircraft and the advanced
design group headed up by Ed Schmidt at the time, working for Miles Lever. The various people studied different concepts. I started working on the concept for a ceramic reactor and studied the possibility of such things as sodium oxides or sodium oxide and silicon dioxide and concluded that a reactor possibly out of
beryllium oxide was feasible. This is working in close conjunction with Vince Calkins and Clay Brasfield. Clay then developed a technique for extruding beryllium oxide fuel elements and mixing the uranium dioxide into beryllium oxide, as he told me, as a lead solution, and perfected the technique of producing such a fuel element.
In the meantime, application studies were continuing, and it was concluded by me and some other people, at least, that a nuclear-powered ramjet was a logical application for nuclear power, although the A&P department was, at that time, dedicated to the turbojet application and the manned aircraft. The Air Force then initiated a series of studies, contracted with Chancellor of Aircraft,
North American Aviation and Convair San Diego to study the feasibility of a nuclear ramjet. Chance Walt had the winning solution to the ramjet, and this is a model of the SLAM missile that was developed at the time and a road carrier to play the shell game on the western part of the United States with a nuclear-powered missile.
with the ramjet inlet, the boosters, and the ceramic reactor for the propulsion. The avionics was in the front end of the missile. This missile was about 90 feet long and some 5 feet in diameter and was designed to fly at Mach 3 at sea level.
or essentially on the deck. The program lasted longer than the nuclear-powered aircraft program because it was technically considered to be feasible, but very provocative. The provocative aspect, by 1964, caused the program to be terminated. And Chance Royal Aircraft, at the time, had been taken over by Link Technical Vault.
produced a film called Muscle and Mothball, or actually a history of the SLAM program, which lasted for about five years. You can now see this video.
Muscle and Mothball (SLAM film)
Nuclear-powered missile, which would be boosted with solid rockets to high altitude, then commence nuclear-powered flight at Mach 3, let down to treetop level to penetrate enemy defenses, virtually impossible to detect and intercept, and proceed to deliver successive warheads to selected hard targets with precise accuracy. The development of a suitable reactor, the last and major problem of the identified high-risk areas,
continued at a rapid pace at the AEC's Lawrence Radiation Laboratory. The Torrey 2C reactor was a full-scale, 54-inch diameter, flight-type reactor with a beryllium reflector. Unlike the earlier Torrey 2A1, Torrey 2C was controlled by rods inserted into the reactor core and driven by actuators mounted in the main air duct feeding the reactor.
A new fuel element fabrication technique was used for 2C. A new fabrication plant was constructed at the Coors Porcelain Company in Colorado. The powders of beryllium oxide and uranium dioxide were first thoroughly mixed with a binder material. This plastic mass was then extruded from large extrusion presses. Drying and sintering causes a 30% shrinkage.
The fuel elements are finally ground to the external hexagonal shape. Assembly of the 600,000 fuel elements, the high temperature coated columbium base plates, the control rods and actuators into the Torrey 2C reactor took place in the shops at the Lawrence Radiation Laboratory. It was then moved to the test site in Nevada.
The air supply at Jackass Flats had been enlarged to supply over one million pounds of air and facilities installed to preheat this air to 1,000 degrees Fahrenheit, enough capacity to simulate a five-minute flight. In May of 1964,
Full power operation of this full-scale flight type reactor took place after an initial intermediate power run. All test objectives were completely met and conclusive feasibility was established. Exhaust air from the reactor contained less than one-tenth of one percent of the generated fission products.
The thermal energy equivalent of 640,000 horsepower was unleashed on the Nevada desert. This rate of energy release could have been sustained for hours, equivalent to a missile fight around the earth or more, had sufficient cooling air been available. Adequate thrust for the nuclear missile had been demonstrated.
5.2 Ceramic Reactors
Hello, I'm Bert Chandler and I have the privilege of introducing Clay Brasfield who's going to give us a discussion
on the ceramic materials development for the aircraft nuclear propulsion program. Clay? Thank you, Bert. Fifty years have elapsed since the start of the effort to develop and build a nuclear-powered airplane. The awesome amount of power which could be produced from fissioning a small amount of uranium-235 suggested the potential of virtually unlimited range.
thus providing the United States Air Force with tremendous flexibility in the use of such an aircraft. Before presenting the specific ceramic materials R&D, it seems appropriate to briefly indicate the status of materials science and nuclear technology as they existed 50 years ago. Beyond a doubt, both were in their respective infancy.
Nuclear technology existed mostly in secret reports of the Manhattan Project and with the relatively few engineers and scientists associated with that project. Nuclear technology was alien to most ceramists, metallurgists, chemists, etc. Terms such as thermal neutrons, fission cross-sections, criticality, critical mass, moderator reflector materials were new.
As a consequence, there was a reliance on the more knowledgeable nuclear physicists for selection of potential acceptable materials. Material science, as a separate discipline, was just starting to be introduced into the curriculum of engineering schools and universities. Almost no knowledge existed on the behavior of materials subjected to the nuclear thermal
mechanical thermal stress environment that would be expected in a nuclear reactor suitable for aircraft propulsion. It was the day of the piston engine, jet engine technology accelerating and requiring materials capable of providing higher and sustained temperatures if its potential was to be realized. Use of brittle ceramics in a dynamic environment
mechanical thermal stress, thermal cycling, vibration, et cetera, was literally nonexistent and is still very limited 50 years later. The computer age was a few years in the future. It was the day of the seven to 12 place Monroe and Frieden calculators, vacuum tubes, personal computers barely imagined. It was in this overall technological atmosphere that the NEPA nuclear energy
for propulsion of aircraft with Boeing. Obviously, materials capable of operating at high temperatures would be required. In addition, the nuclear physics desired materials having moderating capability and low neutron absorption. Materials with potential to fulfill these criteria were mostly in the beryllium family.
beryllium carbide, the beryllium carbide and graphite, graphite itself, beryllium oxide, and silicon carbide. Silicon carbide is not a moderator as the others are, but it has excellent oxidation resistance and was considered a potential fuel element.
It was hoped that these materials would be operable in the 2500 Fahrenheit temperature range, and the beryllium-based materials were capable of being used as a moderator and reflector material, as well as the graphite. The relative moderating values for some of the beryllium-based materials are shown in this chart.
What we have here is 88% beryllium carbide with 12 carbon or graphite. It took 46 pounds for this particular reactor. And the 70% beryllium carbide, 105. The BEO, quite a jump to 140. And zirconium beryllide, 115. And beryllium metal by itself, 110.
calculation was based on a square cylinder sodium cool. But the relative value for the various materials would have been the same for an air breathing reactor. Beryllium carbide was, at best, a laboratory curiosity.
Subcontracts were placed with Battelle Fan Steel Metallurgical Company and Brush Brillium Company to synthesize high purity beryllium carbide in modest quantities. The two routes for production of it at Battelle were the react beryllium metal with carbon and at Fan Steel to react the beryllium oxide with
two carbons giving the uh bryllium carbide plus co gas the early production material was only about 98 96 purity the main impurities being free carbon nitrogen and oxide the nitrogen was particularly undesirable because nitrogen has a large absorption cross-section for thermal neutrons
Nitrogen also has a large absorption. Nitrogen also inhibited fabrication. Much of the early effort was conducted using this material. By late 1950, Brush was producing material using the metal carbon reaction of greater than 98% purity with relatively low nitrogen.
The way that you could fabricate these materials, NEPA was hot pressing billets and diamond sawing and grinding to get shapes for mechanical tests and plates for coding studies and what have you. The tail was isostatically pressing and centering and large billets. And then they would
diamond saw and grind also. Centering of the beryllium carbide was really never successful. Densities around 90-95% at best were obtained and later on when we started adding graphite to it, they were even lower.
On the beryllium oxide in those days, there was no really highly centerable beryllium oxide available. The material was really good purity, but it wasn't very centerable. Silicon carbide was only fabricable to any kind of a density by hot pressing.
One of the big concerns for the ceramic as a fuel element was the ability of the ceramic to withstand the thermal stresses generated under the high power throughputs that were going to exist. And the calculations at that time showed that the ceramics were going to be subjected to around 35,000 to 40,000 BTUs per hour per foot.
and that was quite concerning. The test people developed a test for evaluating thermal stress, which consisted of a thick-walled cylinder of the material with a water-cooled tube going through the bore and the outside of the cylinder heated by radiation.
Then the mass flow and the temperature rise gave the power throughput. The beryllium carbide, 100% beryllium carbide, failed at a throughput of about 3,500 to 5,000 BTU per hour per foot, which was about an eighth of what it would need in the reactor. There was quite a study on...
to bryllium carbide, but the only one that was really successful was a 30% addition of graphite, and that composition withstood 35,000 BTU per hour, so it was close to what it needed to be, and that became the composition that practically all the studies from then on were on.
drawback to beryllium carbide and particularly beryllium carbide graphite was the requirement for a coating because they were not oxidation resistant, literally burn up whenever they were heated to the 2500 F temperature. There was a pretty intense effort conducted on these materials.
and about every coating system that you could think of, ranging from platinum metal cermets, metal mixtures, vapor-deposited silicon and silicon carbide, and glass-based enamel glazes, to name a few. Almost any combination deemed to have the potential oxidation resistance at 2500F and above.
While some coating systems provided protection for some test specimens for as long as several hundred hours at 2,500 Fahrenheit, under more or less static conditions, none of the coatings or coating systems were reliable. Failure times ranged from almost instantaneous to a few hundred hours. The most successful for the beryllium carbide were the ceramic glazes and for graphite,
and silicon carbide and siliconized silicon carbide. This is typical of a glaze composition that was used on it. It has about 50% silica, 20% to 30%, 20% to 10% calcium oxide, and then the various other oxides. Those are a little different from normal ceramic enamels, and they are much more refractory.
They were compounded by mixing and blending and heating up in the furnace until they were molten and then quenched in water and then ground down to fine particle size and put in a slurry and then either the beryllium carbide could be coated by dipping or by spraying the coating on. Then they were fired at around 2500 to 2700.
in either an inert atmosphere, in some cases in air, in order to mature the coating. Even those were not particularly successful in that they tended to, with thermal cycling, crack and sprawl, and once you got a pinhole, blistered intensively. Brillium oxide was also
investigated, but not as intensively as the beryllium carbide because it didn't have the moderating capability. And in those days, a little bit of an aside, in those days one of the real major concerns was the amount of U-235 that it would take to build one of these reactors. So anything that was
would reduce the quantity required was given priority over the other materials. But in the beryllium oxide, it was quite a bit better thermal throughput wise than was the beryllium carbide in that it didn't fail to around 15 to 20,000 BTU per foot per hour, a factor of three to four times better than the
beryllium carbide. But the beryllium oxide powder available was not very centerable to high densities. Some small additions were tried, and one of the additions that was successful was a feldspar, and about the 1% level produced material of over 98% density.
The other thing that was known at that time about the beryllium oxide was that at high temperatures it tended to be corroded by water vapor. And there was some coating work done on beryllium oxide under NEPA, but not very intensive. The coatings that were investigated were silicon metal alumina, silicon zirconium oxide, chrome alumina, zircon,
and plain old aluminum oxide. And test results were very erratic, probably because of the low centerability of the BEO. The most successful systems were the silicon zirconia and the zircon. Significant improvements resulted from re-coating with the same system. The interest in the silicon addition
was because of its excellent oxidation resistance up to its melting point, which is about 2,600 Fahrenheit. The development of adequate coatings for BEO awaited the very intensive effort of GA and P and the availability of a more center beryllium oxide. NEPA performed design studies for both air and liquid metal cooled reactors.
graphite, beryllium carbide, beryllium carbide, graphite, and BEO plus other ceramics were evaluated for compatibility with sodium and lithium. The graphite completely disintegrated when exposed to lithium at 1350 Fahrenheit for 25 hours and was severely attacked in sodium at 1500 Fahrenheit.
The high purity beryllium carbide appeared to be stable in sodium for 100 hours at 1850 F. The 70-30 mix result were similar to the graphite in both sodium and lithium. Beryllium oxide was stable in sodium up to about 2000 F for 100 hours, but lost up to about 5% of its weight in lithium.
at 1350 Fahrenheit in 17 hours. The NEPA project was terminated in May of 1951 and GEANP was formed to continue the program. When we were talking earlier about ceramic development, you mentioned that the work at NEPA continued at GEANP through 1954 and maybe into early 55 on the fuel ceramic systems of beryllium carbide.
molybdenum disilicide and silicon carbide as well as beryllium oxide. Then when the Air Force changed the ANP mission in March of 55, which required higher fuel element temperatures, the beryllium oxide system was best suited because of better oxidation resistance while still having acceptable moderating characteristics. Clay, could you just continue your discussion of ceramic development?
and focus on the years 1955 through early 1961 when the project was canceled. Except for the event or happenings, the success of the metal fuel element and the promise of hydride moderators may have caused ceramics to fade from the scene. The first happening was that the Air Force changed the early flight demonstration to a program oriented more to
development of a high-performance system. Secondly, the advanced design team, Austin Corbin, Tom Zeckley, Lou Feathers, Mel Lapidus, to name a few, had been designing reactors to meet the longer-range requirements. Austin, for one, was adamant that fuel element temperatures in the range of 2600 to 2750 would be required to provide a bulk air temperature to meet the long-range requirements.
The advanced design team was also studying unmanned nuclear-powered ram jets with the desired 3,000 F fuel element and a 10-hour life. When the Air Force changed the A&P mission in March of 1955, the ceramic effort immediately had some priority in visibility.
Brillium oxide was the one material which was intrinsically oxidation resistance and had acceptable neutron, had acceptable moderation, and was a material being considered in the advanced design studies. The fact that BEO was subject to water vapor corrosion at temperatures of interest was of concern, but was not thought to be a major problem.
While beryllium oxide was oxidation resistance, none of the uranium oxides were stable throughout the operating temperature range. UO2 with a melting point above that of BEO, 4500F, reacted with air and oxygen over the whole temperature range. A second very important factor was the non-availability of a high purity readily centerable BEO.
The revised ceramic effort concentrated on two areas. Development of an oxidation-resistant fuel and the development of a high-purity, centerable VEO. And both of those really go together. If you don't have one or the other, or you don't have anything, you have to have them both.
The Centerville BEO program was initiated by placing orders with Brush Brillium Company and Brillium Corporation of America for 1,000 pounds from each company for BEO of at least 99% purity and Centerville to over 98% theoretical density at temperatures 2,800 to 3,000 Fahrenheit.
Three general approaches were produced to develop a stable uranium fuel additive. Let me backtrack a moment here and show you what the problem with UO2 is. UO2 has a density of about 10.95 grams per cc, but when you start heating it air, it picks up oxygen, and the density actually increases.
It goes up to well over 11, something like 11.1. You continue heating it, it picks up more oxygen. The density goes way down, and there's roughly a 20% volume increase in going to its highest oxide, U3O8. You continue to pick up oxygen to a higher oxide, UO3, and the UO3 is actually volatile.
So if you try to operate in air with UO2, you have a big volume change which disrupts the fuel element. Secondly, you lose a lot of your fuel. The three approaches to the stable fuel, one was to look at uranium compounds, particularly the alkaline urinates.
And the best ones that we found in those were calcium urinate and barium urinate. And neither one actually had the fuel retention. They did give good volume stability when incorporated in BEO. The second was to try to see if we could incorporate the UO2 into oxygen-deficient systems and then develop oxidation resistance around the UO2.
That didn't work either. The third was investigation of solid solutions with just about any oxide that would form a solid solution with UO2 was investigated. Ternaries modified some of these solid solutions later on to get rid of some of the problems with the binary.
In total, in the course of the investigation, over 200 compositions were compounded from about mid-1955 to the termination of the project in 61. The only one really system that worked was the Yitria UO2 system.
And it had two aspects of it. One, you could get a fairly high fuel content with the YITRI. And secondly, the loss in the volume stability was excellent. It would pick up oxygen if reacted or when the tube was fired in hydrogen.
But that pickup of oxygen actually led to an increased density, and you had good volume stability. Now you're going to be, the rest of the talk, hearing about various fuel compositions, and I want to identify the codes before we go on. F48 is a mixture of 40% UO2 plus oxygen.
and 60% Y2O3. And the F116 has another 5% UO2 and 55 yttria. Then once the fuel additive was added to the beryllium oxide, it got a BF number with a numeral, and the numeral signified the amount of UO2 in the BEO added as whatever the fuel
number was 48 or 16. If you look at the actual compositions, what was important was trying to keep the maximum amount of beryllium oxide in the system that you could. And if you take a look at 10 weight percent 116 fuel additive, you've got 89 plus volume percent of BEO
in the system. If you're down lower than that, of course, your volume percentage of BEO will increase. And if you use the F48 composition, then it decreases. So for moderation and the lowest uranium investment, you wanted the highest fuel content of your system that you could get that was stable.
The 48th fuel additive in this system was the F48, and that additive at 2500 F for 10 hours showed virtually no weight loss, and the same result when incorporated with 6% BEO. The volume stability was excellent. The test temperature
was then raised to 2750F, and a 10-hour test had a weight loss of about 0.3%. As a result of this development, the BEO fuel element effort accelerated. While other solid solutions and modifications of solid solutions with other oxide continued to be evaluated, a major effort was implemented to gain an understanding of the U02Y203 system.
Phase diagrams studies at Denver Research and GE General Engineering Laboratory were initiated and uranium vapor pressure measurements for various UO2Y203 compositions were made concurrent with the progress in developing stable fuel additive. Brush and Brilco were delivering trial lots of BEO powder. The brush material contained up to
half a weight percent sulfur in the Bril Co up to about 0.4 weight percent sodium. Both materials, though, were centerable to over 98% TD in the temperature range of 2750 to 3000 F. Both the sulfur and the sodium were greatly reduced during centering, but as will be discussed later, led to fuel and centering problems, primarily dimensional stability. Got to have a drink.
In the ceramic reactor design area, design using BEO in the form of round, thin-walled tubes and BEO moderator slabs is well along. In this time period, about mid-third quarter of 1955, the effort on the ceramic core reactor was accelerating rapidly in both the design and materials activity. The decision was made to run a large test in the HTRE.
facility in the large central cavity for testing the various advanced fuel element design. The 2B insert was placed on test in February 1956. The testing has been described by Bob Evans. The mechanical integrity of the fuel tubes is excellent. However, significant BEO crystal deposit
were found in some of the tube bores starting at about the end of the fourth stage, increasing through the fifth stage and part of the sixth stage, and significant decrease in the seventh last stage. As I've been previously indicated, water vapor corrosion of BEO is well known, but the extent was not thought to be a significant problem. Water vapor corrosion experienced in NSERC 2B demonstrated this optimism to be a considerable misjudgment.
Within a month's time of the discovery of the extent of corrosion in insert 2B, a chaotic effort, and I do mean chaotic effort, was underway to develop a coating to prevent the corrosion. Thermodynamic calculations to define the relationships of temperature, water vapor content, mass flow rates, and to verify controlling mechanisms were also in progress.
This later effort was by Mel Lapides, Bob Spare, and Julie Malkin. Coating task forces were formed. Materials personnel who had been working on metal hydride moderators, Bob Van Houten for one, and oxidation-resistant metal alloy development were put on the coating problem. No coating system was too...
too bizarre to be studied. Whether or not the coating had high potential to degrade the temperature capability of the BEOU02Y203 system was a minor consideration. It's not my intent to toot my own horn, but my input from the beginning was that only two coatings should be considered. High purity aluminum oxide with at least minimum additive possible.
And a half neofreeze zirconium oxide stabilized with yttrium, not calcium oxide. Calcium oxide is a common material used to stabilize zirconium oxide from crystal transformations, but it has a very low eutectic with the beryllium oxide. I believe it is fair to state that only these coatings were being considered after a few months.
The major effort on zirconia coating was core extrusion because of its inherent advantages. Fred White and Al Covell were instrumental in its rapid development. The core extrusion process was developed to the extent that uniform 3000 to 5000 thick coating could be applied to the ID of the fuel tube and this coating provided
corrosion protection to at least 2,658 Fahrenheit. Starting with the revitalization of the ceramic effort in mid-1955 to the third quarter of 1958, tremendous progress had been made in both the ceramic materials effort and in the reactor core design using the ceramics. A volume-stable fuel additive with acceptable fuel retention when incorporated in BEO was available.
High-purity, centerable BEO was also obtainable in production quantities from both brush and brill coat. The water vapor corrosion problem appeared solved. Results from numerous reactor tests of fuel tubes were positive. Tests in the MTR and ORR showed the fuel element could operate at design power densities. While there were still many unknowns, the ceramic system BEO, BEOUO2, Y203,
And ZR02Y203 showed high potential to provide a high-performance power plant. The third quarter of 1958, the Air Force changed the nuclear-powered aircraft mission again. A part of this change was that propulsion system capable of Mach 0.8 to 0.9 with 1,000-hour life be developed.
The system was also to have potential for development to higher performance. The development status and potential of both the metallic and ceramic reactor materials were extensively reviewed. Because of the potential of the ceramic system to meet the longer range performance, the BEO system was chosen for development.
Results of vibration and g-testing of the D1OE tube and slab design had shown unacceptable displacements of core components, and a new ceramic reactor design was in hand. Clay, that was an excellent review of the development of the ITRIA-stabilized zirconia ID-coated BEO UO2 fuel element.
You mentioned that you wanted to summarize the status of the D-140E fuel element at the termination of A&P. You mentioned these seven areas of interest to be summarized. Fuel stability, water vapor, corrosion, fabrication and mechanical properties, radiation effects, self-bonding,
reactor testing, and fission product release. Because of time restraints, would you concentrate on just the fuel stability, water vapor, corrosion, and radiation effects? Then if time permits, we can quickly get the status of the other four areas of interest. Clay?
This is a plot of the fuel additive system, and this is the fuel additive composition chosen for the final fuel element. It has a 45 weight percent UO2. It has the lowest fuel loss of any except at the very high yidria content.
But it is a composition that maintains the most volume percent BEO that we could get into the fuel element. When you look at the vapor pressure in air of U02, which goes to U308,
It's up here. If you look at these additives, this is the F48 additive, which was the first one that really showed promise. This is the F116, which is the composition of interest. You'll notice there's about three plus orders of magnitude difference in vapor pressure. When you incorporate that into beryllium oxide,
you get roughly another order of magnitude lower in vapor pressure of the U03 over the U308. And that is the basic fuel element stability that we're going to be talking about throughout the rest of the talk.
This shows fuel loss for the clad and for bare tubes. Here at 2,600, we're talking about a half percent fuel loss for a 1,000-hour test. If you recall, in the case of the temperature distribution, there's practically nothing in terms of total quantity that was at 2,600.
The 2600 for the clad tubes is about in the range of a half percent, and for the unclad tubes in about the range of a half percent. One thing that I should have talked about in that design is the compressive loads on it and the hexagonal shape.
There's very little flow through the high temperature portion of the reactor. There undoubtedly is some flow, but that's important both from fuel loss and from water vapor corrosion. Our assumption was that the backside of the tube did not need to be coated to prevent fuel loss and or water vapor corrosion.
These in this case the clad is only on the boor So these fuel losses are coming from the whole tube and the actor is in the actual design These are going to be back to back and you're not going to have much fuel loss off the back So these are maximum fuel losses that we're talking about
This is just a plot showing the 1,000 hour test at 2,500 Fahrenheit. And this gain in weight here is not an actual fuel gain. What happens there, the tubes are fired in hydrogen. They come out in a reduced state. And then the fuel additive picks up oxygen. And if that were UO2 picking up that oxygen,
then you'd have volume instability. In this case, the density actually is increasing, so you've got excellent volume stability. That's 2500 F. This is just another one, shows the reproducibility. This particular one is at 2800 F. And there's nothing in the design that comes up into that temperature range.
Now I'd like to go into the water vapor corrosion. And what this table here outlines is uncoated VEO tubes at various dew points and at various temperatures. And in these cases, you generally get crystal deposits in the tube, which results in pressure drop increases.
And you'll notice here that a lot of these tubes have 10. Here's one with 34% to 41% increase in pressure drop. And that's going to not work. These are coated tubes with the
And here you see practically no pressure drop, 1%, 2%. There's one there that's 4.5% roughly. But it shows that these are 1,000 hour tests. They're in the range of 2,500 here. There's 2,600 down here. And some of them are negative, some positive. So the coating is protecting against the water vapor corrosion.
They're coated with these zirconia tubes. These numbers over here tell you that there's no really a deposit in the tube, no crystal formation in the tube. And here we've gone up again. Here's a 2800F test, 2700 test. So I think we can conclude that
the water vapor corrosion has been solved. Radiation effects on the en-fuel tubes are important because beryllium oxide has a hexagonal crystal structure and in general radiation causes
hexagonal crystals to expand more in the c-axis direction than in the a so if you get a lot of growth then this results generally in a significant strength decrease and the question was whether or not that radiation would decrease the strength enough and
allow the loads on the design loads. Two types of damage that we are concerned with, atom displacement damage, which if you count for the particular reactor design here, here's 1,000 hours over here, and at room temperature, or very low temperature,
you get up here where through the course of that lifetime, about six or seven times 10 to the 21st of the atoms have been moved out of position. Now, that doesn't mean they remain out of position, but they've been moved. Now, if you up the temperature to somewhere in the range of 1800 F, then you saturate here at about 10...
2 times 10 to the 18th displacements. The other thing that concerns us in beryllium is that when you absorb a neutron, you get two helium atoms. And those helium atoms can diffuse to grain boundaries, and maybe they'll diffuse out, and maybe they won't. But if they get on the grain boundaries, they can seriously lower the strength. This is a plot.
or a table listing some radiation effects in BEO. And remember I talked about the C-axis, and if you look here, this C-axis expansion here is about eight times what the A-axis was, and that continues throughout in the test. Also notice that here at
This sample here was irradiated at 1700 F. It's about half of what it was radiated at 700 F and about a third of what it was in the very low temperature true to 300 F and that Percentage wise
In terms of what that expansion would cost, it would cost the core about a half a percent, and the core, I think, was 44 inches in diameter, and if all expanded like that, it would increase it by about half an inch, which Otto's design will take care of. On the density, which again is a measure of volume expansion,
Here at 2.8 times 10 to the 20th, we're only getting about a small percentage increase in density. This is actually about two-tenths. The worst case down here for the low density is about three-quarters of a percent, and for the high density is about three-quarters of a percent, and it saturates out.
pretty well back in the range of 2.3 times 10 to the 20th and those dosage was what we expected in the unfueled portion of the reactor. A lot of reactor testing was done particularly on the fuel elements and primarily directed toward their stability under the power producing
densities that they were going to be subjected to and also for fission product retention. I was surprised myself when I added up the hours of tests. In the litter we had 88,000 hours less than the other reactors, but the total hours of tests for our BEO fuel elements of about 19,000 hours.
That was an excellent summary of the status of the ceramic fuel element at the end of the GEANP program. We've just been informed we have about 12 minutes left, and in that time, could you summarize the fabrication of the fuel element that we had just talked about slightly earlier? In the fabrication...
We start out with the UO2, mix the fuel at it. It's calcined, blended, et cetera, reacted, then mechanically ground to get a fine particle size to incorporate in the BEO. On the other side for the coating, we start out with the mixture and go on down. We bring them down here together, and we co-extrude to get the coating on the bore of the tube. Then they're centered down here, and finally they're ground.
and cut to a length and inspected. Let me say something about the grinding. With the tolerances that are required in the design, the straightness and the stack up in order to have closely nestled tubes, at the time that the project closed, we were just getting to the point
where we might have been able to extrude a hexagonal tube, keep it straight enough, and the dimensions close enough that it would not have needed grinding. I don't believe that we would have ever arrived there. I'm now in the pellet fuel business, and we've been trying for years to center the size. We still have a long way to go, so I think that for the ceramic reactor,
we would have always had the grinding. And John Mundy and those people developed straddle grinding where the grinder went down two sides of the tube at once to real fine art. And that was the way that a lot of the mock-ups that you've seen have been produced and all the in-pile tests were straddle ground tubes if they were hex. In the co-extrusion process,
This gave a lot of credit to Fred White and Al Covell who really developed this process. This is the clad cylinder. You notice that the volume of it is very small compared to the matrix. The whole thing extrudes down here, and one of the real accomplishments was this pin down here where you see this narrow pin.
To down here and then expanding the clad out against the tube wall and that really solved the core extrusion process and getting uniform coatings and a good bond Let me show you some of the temperature capabilities far beyond what can be utilized
But beryllium oxide melts at about 4470, Y2O3 at 4370, and so forth. All of these temperatures are far above what we'd ever be able to utilize. The other problem, though, is you can't just look at the melting temperatures. You have to look at the system temperatures. And the system temperatures that we're talking about here is the BEO,
the Y2O3, the zirconia, and the UO2. And in some cases, if you were to use alumina-alumina. So you're going to have some liquid in the fuel element with alumina present in the system at 2770. And in our current system, when you initially center, you're going to have liquid present at about 2870.
may exist throughout the life of the fuel element. Very small quantities, but this is what really limits you in terms of temperature capability and not the melting points of the individual oxides. In the design, the loads that are on there, you're worried about or concerned about compressive creep.
I show this chart merely to indicate some of the flexibility that you have in ceramics in decreasing or increasing certain mechanical properties. In this case, you've got a grain size here of 13 to 25 microns when it's centered at 2750. We didn't get as high a density as we normally get there. These other structures show...
higher grain size with higher centering temperatures. You can also center for a longer time and get a higher grain size. The other aspects here that is critical, and if you recall the chart where the BEO incorporating the fuel at a lower vapor pressure, if you'll notice in some of these large grains here,
the fuel additive is actually completely surrounded or encased within the grain. And that's virtually the reason for that half-order magnitude decrease in fuel loss from the VEO fuel element relative to the fuel additive itself. Here's a cross-section. These are zirconia ID coated tubes.
Notice the rounded corner here, that's the straddle grinding left this little round corner, particularly if the tube was a little small. Here this white layer is your zirconia coating to prevent the water vapor corrosion. I mentioned the effect of grain size on some of the properties.
Here's a grain size down here in the range of 12 to 15, where you've got modulus ruptures up in the 40,000, 50,000. As your grain size increases, your strength goes down. On the other hand, if you were worried about compressive creep, your compressive creep would be decreasing.
I mentioned modulus of rupture. These tubes are centered at 2800, which is below that eutectic. If you were to center above that eutectic, what you would see is a very similar shaped curve, but it would be significantly lower. Strengths on the 2850 material.
somewhere down around in the 25 to 20,000 range, almost duplicating that curve. The compressive creep, which was of concern because of those radial loads, if you look at all of it over here with temperatures up to
2800 and these loads are presumably what the design They would experience in the design they range down here from Maybe the worst case is one and a half percent At 2850 which again is far above the temperature of interest the
If you look at the 2950 centering temperature up here, you'll notice that the compressive creep over here is like 0.02, 0.08, 0.09, 0.23. And where you have centered at a much lower temperature, well below the eutectic, unfortunately I don't see any here, but believe me that the...
compressive creep rate is quite a bit higher. In fact, there's a big change here from 2950 down to 2850, which is again below the eutectic, so that you don't get the very large grains that you get up here in the higher temperature center. Now let me summarize.
While this ceramic story has been presented by me, it involves the dedicated efforts of literally hundreds of engineers, chemists, machinists, welders, draftsmen, lab techs, test groups at Oak Ridge, Idaho Falls, subcontractors, consultants from industry and academia, reproduction personnel, and last but not least, the secretaries who decoded illegible scrolls of engineers like me.
And let me say this, everybody in this room is a part of this story. Each of the above made their contribution in some way to the nuclear-powered airplane. I'm not competent to judge whether or not the power produced by the ceramic reactor was capable of meeting the Air Force mission requirements, excepting that the mass flow
The bulk temperature could be produced by the overall temperature distribution in the reactor. I am confident that the Brillium UO2, Y2O3, ZRO2, Y2O3, and the BEO material were capable of providing such performance. I believed it at the project termination, and I believe it today. That was great.
And thank you so very much for all that input. You've summarized an awful lot of material here in these two hours of tape. Thank you again.
ZrH Moderators
To do this, we might substitute the water for beryllium. And in order to get the enzyme age higher and the moderator ratio higher, we would drill holes throughout the unit, and we were going to pack zirconium hydride into that system. Now, one of the reasons we were going to pack zirconium hydride into that system was because we figured that we could operate at pretty high temperature.
we could operate a fairly high temperature with that beryllium uncladded, and we could operate into this area where it didn't begin to break off, which allowed us to go somewhere up around 1,800 degrees Fahrenheit. That's a great improvement over the liquid. It would have been a great improvement. Now, the big problem we were having at that time was, again, the...
Zirconium hydride was ground very fine because, again, it was used as flash powders for pictures. It was used as pyrophoric materials and so on. So we had to do something. So a fellow by the name of Dick Hunnell and I went up to Beverly, Massachusetts to talk with metal hydrides because we thought if we could take, because it was so finely ground, if we could have it with coarser particles and then a little bit...
other particles coming on and not quite as coarse and then the very fine particles so if we pushed it into these little holes that we have then we would go along and we could then increase the material the hydrogen content in each one of those holes now one of the reasons why we designed the guard designed
If you go back, George, to the original picture we showed, what he did is he took the various holes, which had few elements in, and I'll give you a point one of these like this. If you would focus the camera on this one here, if you want to do this, focus the camera on this. This would be the actual size that we were working on. And the few elements, as we've seen before, would fit right in between here.
I'll show you that. A few of them would fit in like that. Now there's a little play in here because there was insulation in here, but this is approximately the size. At that time we were thinking about that we would make these pieces about five inches across the flats, and so we were going to that. Now I'm going to go back a little bit.
While we were at Beverly, Massachusetts, Dick Connell and I were looking over and talking with the people of how they ground their powder and so on. They took us out into the work shed where they were making zirconium hydride. And it came in. Some of it was flat. Some of it was in rods and so on. And while we were there, they were unloading every torque. And I happened to notice that it was very strange that some of the stuff, the hydride, was fairly solid. It looked good.
And others were all spoiled and breaking up, and others were cracked very badly. So I came back, after Dick and I came back, I got to thinking about, well, why would this stuff be different? And so I talked to Jim about it, Jim McGurdy. Oh, yes. And I said, Jim, let's do a little experiment. Why don't, can you loan me your brazing furnace? Now, Jim and some of the other fellows were using brazing. They were brazing the few elements. And this is a...
This furnace was a hydrogen furnace, but it had a mouth on it about this big. We could slug the material into the hot area, and we could pull it back out in the cooling area. But it also was equipped with, you could bring pure hydrogen in it. You could also cut out the hydrogen and refill it with either argon or helium. So I said to Jim, if I could use it for a few days, let's try some experiments. So I went back in, and I went to the shop, and they made some.
little rods that might go into that little insert there and i packed them and i put them in the furnace and uh but i uh put in inert gas in it to begin with then the rods were heated up and then i started cutting the argon or helium whichever it was in this case it would be helium i started cutting it down increasing the hydrogen and we did that for a while and pulled the pieces out
after, I suppose, four or five hours, I don't know what it was, and looked at them and it looked like we thought nothing happened to them. We thought they weren't even hybrided. But they were nice and shiny and they're straight. And like a winner. Well, we didn't know it was a winner, so we had to go get the chemist to tell us whether we had a winner or not. And so we took it down and they analyzed it.
Sure enough, we had a good answer of H, which we couldn't really believe. So the next thing we did will be another picture coming up. We had to really sell the idea. So I went down to the shops and I found a piece of zirconium, and I had it made up in a shape like that, and put it back in and run through the same cycle.
We hadn't really got the cycle right down yet, but we ran through the same cycle, pulled it out, and that came out. I showed that to Calkins, and he got excited. So that started almost immediately. It became classified. And if I remember correctly, I can't remember what declassified it. One was called Lunex, and the other was Ranex. And I'm not sure which is which now. Yttrium came in, and then we called it one or the other. But anyway.
It was classified. Now, George, the next thing they had to do, Vince wanted us to, Dr. Coggins wanted us to do an in-pile test at the MITR. Now, you should have good memories of this. That brings up a lot of memories, Earl, because we had a very tight deadline. And I remember it was 4 o'clock in the afternoon. But you loaded your kids in the car. We made up this rig, remember? That's right. We made up this rig with a...
tail line what was it 20 foot long or something like that you shoved it in your station wagon you loaded your kids in and you headed for idaho we headed for idaho and we got out there with almost enough time but we really didn't have enough time but we had to prove to them that they're willing to put this dangerous material in their reactor and so not only do we have to have the rig to work perfectly and remember and part of this is welding very thin stainless
Only Joe North could really do that real well. There were some real good craftsmen that went into this rig. But then here we are all ready to go, and you've got to convince them it's safe to put in the reactor. Yeah, well, I tell you, I remember we had a deadline. I think it was 4 o'clock. That's right. And we had to get it in. We met with the safety group. Well, it was imprinted in their minds that zirconium hydride was pyrophoric.
everybody knew it was used as flash powder and cameras and so on so we uh we talked and we showed them graphs and they all looked at us and finally the time was coming and i was getting exasperated and i said to one guy i said call up the welding department have them to get a settling torch ready and let's take this sample we had a sample there let's take this sample down and i'll play the settling torch on it to show you it won't explode that shut them back for a minute
So as a result, they didn't want to see that or not. They felt their own safety. They said, go ahead. This was very close to the time. We might have had 30 seconds left at that point, Earl. So we had a real close call there. So we got down there. And then everybody was working on it, putting it in the reactor. People were doing this. People were doing that. And you pretty soon, you came up and said, Earl, are you sure that they've got that tubes hooked up right?
And I looked over and I said, George, you're right. They don't. They got the water going into the hydride. Wow. So you saved the day on that. So we changed it around real quick. And they started up the reactor. And there was no explosions. We were measuring the output gases and so on. Nothing happened.
for the long period of time that we had it in. I think we had it in two or three weeks for the cycle. So that got things going. So we had all of these people working, and that was one of the things that we had in our organization that I'd like to speak about.
There was no one really trying to keep anything to themselves. Once we finished the job, it went over to somebody else, and they would come in and talk to you if you wanted to, but they had their job, and they did it well. And then the other people coming in the manufacturing relied upon the development people, and they did their job well. And it was a good working organization, and I think that's why this organization has stayed together on our POPSI programs and all of that, because everybody worked very well together.
Very good. Now let's go on a little bit, George. And we finished up. We started making very large pieces of zirconium. And another thing happened about this time. The people from the Tell Memorial Institute that came down, Stan Papaki, a guy by the name of Hodge, and another guy by the name of Chuck Boyd. I don't know if you recall him or not.
But anyway, they came down, and they had a program that they were hydrating, not hydrating, but cladding. They thought they could clad metallic fuel elements, and that was just the whole problem. They wanted to come down and talk to us about cladding metallic fuel elements. And they had a little furnace about this high, and it had a tube in it that's about that big, and they were heating it up.
and they would put gas pressure on it as it was heated up, and that would cause the cladding to come down on the fuel element. So they stopped in the office one day, and we were talking, and I thought, hey, this is not a bad idea. Maybe we could do this and clad it. And this is where Carl came aboard about this time, and we started working on, we had their unit, and we set some samples up. We had a contract with them. We set samples up to be tested.
And then we decided to do our own job here. Now, Carl, would you like to tell us about this furnace? Because you're very familiar with it. Well, this was a furnace set up in one of our labs. And essentially, it was a large muffle with a very thick wall lowered into a furnace in which we could control the temperature using the furnace. And the pressure was set up.
such that it would have a large surface area on the inside. We would heat the furnace up, and when it approached over 1,000 degrees Fahrenheit, we would apply over 1,000 pounds per square inch, and this would compress the cladding over the hydride, and if all went well and we had no leaks, then...
We would cool the furnace down and take the head off, and we would have a bonded, clad hydride. You said no leaks. No. I remember one day we had a leak. We had a lot of problems with this because we were learning, and we were very new in the...
the pressure bonding business, and we had probably the largest setup in the U.S. at the time. We did at that time. And we started out with a retort, I think it was of 304 stainless steel. And as we heated and cooled the furnace with various runs, the inside, the stainless steel started
from a straight down situation and after a while it flaked off such that it was getting narrower and narrower and narrower and then when we applied the pressure then it would at one point in time it would bulge out down at the bottom and in order to get the the muffle out of the furnace we actually had to to pull it out with I think we had a forklift and we pulled it
out of the top and it almost extruded it out from the furnace to get out that that that uh piece of steel was was very hot i remember one time we were working away and i think it was bill nagger bob simmons and probably you and uh how buildings were all in there at the same time when we were pumping this thing up to a thousand pounds and all was that started hissing and we all went out the door at the same time no one paid any attention to anybody
That hissing was hydrogen? No, we were using the inert gases. Then we moved from that furnace onto the more sophisticated ones that were developed then. And these were the big autoclaves. Now the difference was that this was internal heating. And outside walls were cooled. And instead of 1,000 pounds, we went up to as high as 10,000 pounds. And the instrumentation will be shown on the next slide.
in a little bit, and I believe this is Ernie Bartel, which was one of the guys that was doing a lot of the work on that. And so this was the instrumentation that was used in this particular one. And this is how it stayed up. But a lot of it was down in the hot part, was down in the concrete, and it was surrounded by concrete. It was a concrete pit. Now we'll move on to some of the other work that we had.
We were studying at that time. And this is some of the work that we did, that you particularly did on this, Carl. Can and seal it off, leak test it, and then have it pressure bonded. And if successful, we would have not only an oxidation-resistant cladding capable of high temperature, but also we would have a good bond, a good metallurgical bond between the hydride and...
the external cladding oh that is good this you can see you can see the cladding on the outside the brighter appearing material and how it is bonded to the the hydride itself very good all right now we're going to continue on after we had this we're going to go into some of the larger pieces that we formed and these are extruded and after this this is pretty well proven by 1958
There's an overlap in the genealogy that we were talking about on the materials here because some of it was going on back and forth. But in about 1950, oh, I'd say 55, we begin to come up with extrusions of this. These extrusions, this material, this is zirconium, and it was extruded.
And in the links, as I said on the first one that I showed you, it was big, about this size. And they were about, they were somewhere between 36, the final was it would be cut to 36 inches. But we were extruding them out at about 4 feet because we needed to crop off the ends and the extrusions weren't complete, but we could cut.
In between, we could cut off this excess material and send it back and have it re-melted. So we could cut that and we got a real good extrusion, high density. So that was one of the things. Now, this also had two purposes that I find out. And if you notice that was the, Carl, if you remember, we were having a problem when we first started out. That the inside clad was fine. You pulled the vacuum on the inside. We could get it clad because it was expanding out into the inside.
diameter but the outside we were having a problem with because what we were getting was wrinkles and so we decided to put some dimples in here which went along very well because they were using those later on for to put the hollows for the control rods those are planned wrinkles those were well those were planned that was that was stretched that was stretched the skin tightly over the material
And that was when you begin to come up with the other materials on that. Now we're going to go into, here's one that was used in one of the HTRE reactors. And this was, now some of the fellows were up busy and they were doing this. This was in the work that we were doing on HTRE 2. It was the first time that the hydride had been put into a HTRE.
And if you'll see the stem on there where you evacuated the material and where it was pulled down. But in HTRE two, it still didn't come up. We didn't have the high-pressure system at that time, and so it doesn't show as tight a skin. And that was where we used two different types of cladding, if you remember. We had about a .15-mil moly in there to retain hydrogen.
And then we over cladded it with iron crumb aluminum. Earl, as I remember, the HTRE was heat transfer reactor experiment. That's the name of it. HTRE. HTRE.
5.1.4 Yttrium hydride moderators
This is the second session on moderator materials, and we have with us today Bob Simmons. Bob joined the AMP program in the spring of 1955 after graduating from Iowa State University at Ames, Iowa. First of all, Bob, I want to welcome you for driving through this miserable weather to get here to help us to work on the history of the AMP project. It's my pleasure. Over here we have Carl Blolander.
a graduate from the University of Cincinnati, and Carl will be talking about some of the research and developing of the cladding material. When you joined us, Bob, in 1955, we were deeply involved in the research and development of zirconium hydride, but we were also looking for a material that could operate at higher temperatures and hold a higher hydrogen content. By 1954,
The initial work has strongly indicated that yttrium and cerium were better, had a higher affinity for zirconium, a higher affinity for hydrogen and zirconium. I'd like to take a moment here to show you a series of curves and show you what the stability of the zirconium hydride. This line here represents zirconium. It starts out at a much higher N sub H, which is the ability to hold hydrogen.
But after about, when we drop down to about 1,200 degrees Fahrenheit, the hydrogen content begins to slope off and gradually goes down to where, operating at 2,400 degrees, it's just practically nil. When we reached the point we were able to get some cereal, or rather, first of all, we checked some calcium, and the calcium holds up very well, but drops off extremely rapidly.
as you get here. Now, when I say drop off, what does that mean? That means it will build a tremendous hydrogen pressure above the metal, and if it's cladded or something like that, it will cause it to, it will definitely cause it to blister. When we developed, found some cereal, we tested that, and we found out that the curve run out pretty straight out of this area, and then it began to drop off quite rapidly.
We were able to get, and Bob will be discussing this just a few seconds, we were able to get a little bit of yttrium, and when we tested that, we found out that the curve remained very flat until we reached about 2,200 degrees Fahrenheit, and then it began to slowly slope off. But even here at 2,400 degrees Fahrenheit, we still had a very high N sub H, much higher than we had for yttrium.
So when we got to that point and found out that the yttrium was probably the preferable material to use, this is when Bob was with us and being a graduate from Ames, Iowa, we then assigned him to that job. So Robert, I'm going to turn it over to you and start talking about the technology and the separation of yttrium.
Right. Well, the rare earths are 14 metallic elements which have atomic numbers 58 through 71. If you look at the periodic table, you'll notice where Earl is pointing. That's a series of 14 elements.
that are known as the rare earths. There's a fascinating story about how they got their name, but I won't bother you with that. Those 14 elements form what is known as the lanthanide contraction, and it's a phenomena of the periodic table. It makes yttrium, which is atomic number 39, if you'll point that out.
It's above, up in the table, 39, right next to zirconium. Okay, I'm trying to get it right now. 39? Yeah. Right here, right here. That lanthanide contraction meant that the atomic radius of those 14 elements contracted slightly through the whole series. And in so doing, it meant that yttrium had an atomic radius that fell between erbium and holmium.
in that series. And therefore, yttrium was quite similar to the heavy rare earths, the heavier ones, and it occurred with the rare earths in nature. In fact, from about 1794, when the first rare earths were discovered in Ytterby, Sweden.
There was very little known about the individual reverse anitrium because of the extreme similarity in chemical characteristics. They all have a plus three valence in their chemical compounds. Extreme difficulty in separating them, requiring exhaustive fractional precipitations and crystallizations, even distillations.
That was the case up until about 1944. At that time, as you all know, the atomic age dawned, and the scientists in the Manhattan Project realized that the individual rivers had to be separated and studied, and that included atrium, because they were born as fishing products.
both in the bomb and in nuclear reactors fortunately about this time there was a man on the project on the manhattan project that was interested in the rivers he was interested in separating them his name was frank spedding and he was a professor at iowa state college at that time it's now called university frank
invented an ion exchange technique that permitted the separation of the individual rivers and yttrium. It also permitted the separation of lanthanum. I forgot to mention lanthanum always occurs with the rivers also. So that was the situation up to the time that Earl was talking about where we got our first yttrium metal. It was a little...
little ingot, about a pound in size, and that material was hydrated and the values of the n sub h that was shown was quite plain. It showed the advantage of yttrium hydride as an ANP moderator.
It also permitted the measurement of the thermal neutron absorption cross-section, which is a value that a moderator should have a low value in that so that it does not absorb thermal neutrons. And yttrium of that purity at that time was measured around one to two barns, which was satisfactory. However,
The material at that time in 1954 was quite dirty. The microstructure had a lot of metallics. It was very poorly behaved. It had very poor behavior with respect to oxidation. It even oxidizes at room temperature. And also the workability was very bad. A lot of cracking.
So at that time, GE and the Atomic Energy Commission and the Air Force came to the conclusion that they should initiate a national yttrium program. And that program was to develop production methods for yttrium and to determine its properties suitable for a moderator.
I think the success of that program can be best seen in a photograph that I used in a talk at Denver in 1959, and it's shown here. It shows one of our secretaries, Marilyn Ladd. She's holding in her hand one of the little billets that was made in 1954, and she's kneeling beside.
80 to 100 pound ingots of yttrium metal that was produced in Ames Lab during the period 1955 to 1960. Actually, most of those were produced in 1959. Intermediate alloy reduction method that was developed. It was better known as the sponge process. And it also gave high purity.
yttrium metal, which had a cross-section consistently between 1.2 and 1.4 barns. And that value was consistent with ANP moderator requirements at that time. Let's jump into the Ames laboratory. When we first started out, of course, we did not have the capability at GE to do this.
And so we had to search out the professionals in this particular area. And, of course, we went to Ames, Iowa. And I would like you to name some of the personalities that worked hard on it because it proved to be pretty much a complete unit within itself. And you, of course, was the liaison person tying the work in with GE there. And we, of course, were taking the engineering information.
and coming up with the designs and so on that we could, and the purity of the metal. So will you dwell on the personalities of Ames? Yes. First of all, let me say that I think that, personally, that all of the people that worked on yttrium at Ames Lab did an excellent job. And I can't mention all those names. But I will pick out the ones that I dealt with.
exclusively on the atrium development. And the first one, of course, is Frank Spedding. This man was what we call the pioneer of uranium because he developed the uranium reduction process at Ames Lab, furnished uranium for Oak Ridge.
purification during the Manhattan Project. And later on, as I said, he became interested in the rare earths and all of his work as director of the Ames Laboratory was in that area for quite a considerable time. In fact, he became so well known in that field that he won the Nichols Medal.
for excellence in chemistry in 1952. And there's a picture of him receiving the medal that I happened to run across out of the CNE news at that time, and I wanted to save it. Frank was an excellent, excellent scientist. The next fellow that I dealt with was Jack Powell.
Jack was instrumental in the development of the ethylene diamine tetraacetic acid process for an exchange that I mentioned previously. The next guy was Norm Carlson. Norm was the inventor of the intermediate alloy sponge process for the production of yttrium metal.
Okay, now we want to talk about the winning of the metal from the product of the oxide. Yeah. So we will pick that up. I mentioned Norm Carlson. Norm, he invented this process. It was the only place that we really got a large-scale production of high-purity yttrium metal was at Ames.
And this process was one in which, if you look at the flow chart that I tried to show here, I hope you can read it. It starts with, Ames started with xenotime. That's a mineral that's high in yttrium, but it's got all the rest of the rare earths in it. Remember I talked about the minerals, the occurrence with all the rare earths. Well, it's treated with sulfuric acid.
that essentially digests the mineral and separates what they call a yttrium concentrate. The yttrium concentrate then was taken and loaded onto ion exchange columns, and you had an ion exchange separation of the yttrium. And now it's not shown on there, but yttrium coming out as an EDTA complex.
was precipitated with oxalic acid and so formed an oxalate. The oxalate then was fired at around 950 C to Y203. The Y203 then was converted in Fred Schmidt's rotary kiln to YF3, the fluoride.
The fluoride was mixed with calcium chloride, magnesium, and calcium, and then loaded into a bomb, a retort, heated, and that's where your reduction occurred. You got reduction by calcium, and in the presence of magnesium, you formed an alloy, about a 20% magnesium interim alloy.
and then the calcium chloride was there as a slag promoter. In other words, the calcium fluoride that was formed as a result of the reduction was entered into the slag with the calcium chloride, and it could be separated then from the alloy easily. The alloy was then demagged.
That means just heat it under vacuum and pull the magnesium off. That left you a sponge of yttrium metal. Now that sponge was compressed into electrodes and then consumable art melted under a vacuum to give you those large ingots. And these are the, this is the product, the metal products of that process. That's the...
Sponge down below, that's your pressed electrode. You can see it was welded together, goes into an arc holder. What's not shown there between that little consumable electrode and that big ingot, there was a couple of things happened. They'd melt several of those small consumable electrodes into bigger ones, and then they'd mount the bigger ones and remelt those under vacuum until you finally worked up.
to about an 80 or 100-pounding. That essentially was the Ames process, and it was producing, well, it produced 10 tons of yttrium in the period 1957-1959, and it was used also, same process, at Crane Metallurgical in Chicago.
And they produced, in the latter part of the program, they produced around 9,000 pounds of yttrium. Now, I should say one thing here. That is, just before the demise of A&P, Norm Carlson wasn't satisfied with the workability of the, even the yttrium metal produced in 1959. That is, the ability to extrude it.
Because some of that material in 1959 was difficult to extrude. They had to go to 1350 Fahrenheit at nuclear metals to do it. So he decided he's going to make a small batch of yttrium using a variation on that process flow sheet. Principally, that was taking molten yttrium fluoride, which he already had, and stirring that.
magnesium yttrium alloy with it in a titanium crucible with a titanium stirrer. He got a marvelously ductile yttrium metal out of that. He got 400 pounds of it. Unfortunately, about all that happened to that was that it was sent out to Hunter Douglas in Riverside, California and was extruded and it was successful at 500 degrees Fahrenheit and I think
it probably would have squirted out at a lower temperature even than that. Okay, Carl, I'll pick you up on this, on a little bit of the fabrication after this technology that we have worked out in the zirconium was transferred over to the fabrication of the yttrium hydride. So would you give us a few words on some of the things in the transfer of the technology?
to making different, testing to making different specimens? We started out with small rectangular specimens shown in the middle of the photograph, in the middle of the photograph. They were quite, quite small. We would clad them with iron chrome aluminum type alloys. We would affix a
a tube on the end of the can and then have that sealed off so that the yttrium hydride was, in essence, a vacuum. At that point, we would put these canned yttrium hydride specimens into a pressure bonding apparatus, heat it up to above 1,000 degrees.
and about 1,000 psi, pounds per square inch. And the bond was then formed between the outer sections of the yttrium hydride. And after it was cooled down, we would take it out of the retort, and then it would be
fully pressure bonded and the the cladding was affixed and we were able to at that point go on to more uh extensive curve shapes uh now we'll move back to uh uh to uh you bob and uh well first of all i would like before i move to you is uh give uh talk about a little bit of some of the work it was put together uh and uh out of this work
The AEC did sponsor several books. This is the one book was put out by, can you get a picture of that? And this is one book was put out by and put together by the University of Denver, where they took the work that we did at Evendale and correlated it with all the other information on high-rights.
The authors of this book was Bill Miller and Jim Blackridge, who headed up the laboratory, George Littlewitz and Charlie McGee. Now, Charlie McGee was one of our own people at GEAMP, but later on transferred as an instructor at the University of Denver. And finally, I would mention Chuck London, which also did a tremendous amount of work.
Bob, there was another book put together on the rare earths. Would you like to touch on the basis on that? Yes, this is the book. Carl's got one. It was a book of the symposium given in 1959 in Chicago on the rare earths. And there were, I don't know, about 10 of us.
At least. That gave papers there. Carl, I think yours was on... Mechanical fabrication. Mechanical fabrication. Mine was the mechanical property. I remember that. There are all you young fellows there. That's us. That's you. Back in 1959. I'll point them out for you. Name them off and I'll point to them. This is Carl and there's Carl there.
John Lockman. This is you, Bob. Yeah, that's me. And we have Ray Letter, Coy Huffine, Joe Williams, and John Collins. And that's where that information was given, and that was the source of the book here. That's right. Now, what were the spinoffs, and what do you think was accomplished out of all this work?
we went over into nuclear materials propulsion operation, NMPO, instead of ANP, right? Right. And at that time, we were using Y2O3, yttrium oxide, to stabilize the high-temperature form of zirconium oxide. We were also using it to stabilize UO2 in a BEO fuel element, which was the primary ceramic fuel element.
in the ceramic A&P reactor. There was a number of things that's happened through the years. For instance, in 1962, there was neodymium dope glass for highly efficient laser guides, the guides being the tubes by which laser beams can be focused down.
and they were highly efficient with neodymium in it. Coatings of lanthanum oxide and tungsten oxide were put onto tungsten, and these behaved at high temperatures without oxidation, which is quite a feat. In the jet engine era, there was a...
additions of cerium and lanthanum and gadolinium oxides to the nickel-based superalloys, for instance, that were used in jet engines. And they were put in there primarily to prevent what they call hot corrosion. And hot corrosion was caused by sulfides, which were always present. I think the most fascinating one that I've noticed as they...
In the period of 1989 to 19, well, 1988 to 89, there's a Y, yttrium barium copper oxide that was formulated and found to have, be superconducting at about 93 degrees Kelvin. Now that means, what that means is that this material will conduct electricity with
with no resistance in liquid nitrogen temperatures. Liquid nitrogen at 73 K, which means now you've got a means of using perhaps magnetic levitation to move vehicles, practically speaking, because you can use liquid nitrogen to foster it.
So I think that's going to be a place where improvement is going to be coming. And it is an yttrium compound. A fascinating yttrium use that I've noticed in the 90s. And it's rather peculiar because it concerns yttrium-90. It's an isotope. Yttrium-90 is a soft beta producer.
Beta ray is a soft way of destroying a tumor. But the problem with that, you can inject it into the site or you can put it in hepatitis. There's no way to target the tumor. Somebody found out that there was a way to bond monoclonal antibodies to that yttrium-90. Bind it chemically.
Then what happened is that when they put it into the tumor site, on the tumor they have what they call an antigen that's always present on the surface. And that monoclonal antibody recognizes that antigen, and it attaches itself to it. So you see what happens? At trim 90, the soft beta rays are right down on the tumor, no matter whether you inject it, wherever you inject it, in that vicinity.
So you might say that that's a real magic bullet. I think in closing, one thing I want to say. Several years ago, I picked up an article that said, whatever happened to the airplane nuclear aircraft reactor? Well, the reactor is only on paper.
HTRE 3 is probably mothballed. But what we saw was a tremendous technology transfer, both as I said in technical information and in scientific talent, and it benefited this nation and other nations.
5.1.5 Shield, Control Rods, Braze
Well, Jim, this is one of the last sessions that we have on materials. We've gone through all the different parts of the building of a reactor. These parts were not considered, when we first started out, as a high priority. But we picked them up later on. The first high priority was the fuel element, correct? And then the moderator. And then we followed with the moderator. And that took a development of about a couple, two or three years.
But as it turned out, the creativity that was necessary to bring these into reality was almost as great as that for the fuel elements and moderators. The structures and configurations we have are entirely new. And 10 years before the project started, we just couldn't do it.
because we did not have the technology to do these rather exotic things. Yeah, I was thinking this morning as I was reading over some of this stuff. When we first started, really, very little technology. And this, actually, we had to explore the full periodic table, almost from one end to the other, because some of the requirements were the required light elements.
for neutron, thermalizing the neutron. And we went all the way up to the heavy elements to collect the gammas. And we also had to go into the rare earths to find out those elements that had high cross-sections but were stable. So it took us quite a while to analyze. And every time we went into a new area, it had to be proven by the other groups of people that did the testing. Yttrium was essentially unknown.
when we started out, and its nuclear cross-section was excellent. So with a fair amount of effort, we managed to get 10 grams of yttrium, and the purpose was to see if it had any effect on stainless steel as a strengthener. And as it turned out, it had no strength whatsoever advantage. But later on, we found that this was the optimum hydride.
And the finding of the yttrium hydride really made possible the high temperature reactor. And incidentally, the AEC effort that went into providing the thousands of pounds of yttrium also had as a byproduct large amounts of the rare earth elements, which were hitherto unknown, such as gadolinium and europium.
which then turned out to be valuable for control rods. And some of those are still being used today. And they are, European in particular. Yeah, and some of the other. One of the other things that we had to do, you would think at the time we were operating and trying to put the fuel arms together with stainless steel, that there would have certainly been enough brazes available to use.
But we found out that we couldn't use any of the common brazes. Well, there were commercial brazes available for brazing the heat-resistant alloys, iron base and nickel base. But they all contained silicon. And we found, after a relatively short time, that the silicon content of the braze would react with the UO2 in the core.
and liberate elemental uranium, which could then diffuse to the surface of the cladding and provide a leakage of fission products. So we had to go ahead and develop a braze that did not have this property. And we turned our attention to germanium. Right, I'm sorry. Well, that's all right. I have little snips now. Germanium is a member of the silicon family and has a lot of the properties.
of silicon but unlike silicon it has a low affinity for oxygen and as a result it is non-reactive with ul2 and additionally the oxide of germanium is reducible by hydrogen so that it is possible to heat treat braze structures in hydrogen without picking up any oxide such as is the case with the silicon containing brazes yeah
I think we had a patent on that. There were patents on it. Additionally, the strength of the germanium brazes was very good as compared to the silicon brazes. When they were tested, we found out that the strength was quite high. Now, we had a number of other things going into the reactor assembly, as I said before.
we really did not give these the high priority until we got ready to uh and the engineers come up with a what we call the high temperature reactor which was brought in by the fuel elements which was brought in by the development of the zirconium high right and the interim high right and that that eliminated all the pumping and all the other things and gives a completely air-cooled reactor so we will begin to talk about now some of the other things that we had to get into and uh
Just for the beginning to help out, we had the front, what we call the front plug, we had the reactor, and then we had a tremendous amount of shielding going around, and then we had the rear plug. Now, one of the interesting things about it, why we wanted to get the reactor as small as we could, is because any time we increased the diameter of this reactor, the shielding weight went up. A closer look at the wavy wall design. In the shielding, the...
And the first objective is to moderate the neutrons down to where they can be absorbed and then incorporate the neutron absorbent materials, typically boron. But as our work proceeded, we abandoned boron and went to some other materials. And then finally, the heavy shielding to absorb gammas at the backside.
which is sort of interesting because we went from one end of the periodic chart to the other end of the periodic chart to stop the radiation. Well, the low end, you get the moderating effects, and the high end, you get the absorption effects. And then we go into, well, now we have a general cutaway of the major shielding. And again, that was a...
We went to the light elements in this case. Well, now we have a general cutaway of the major shielding. And again, we went to the light elements in this case, which were forced into the lithium hydride. And we went from lithium, because it was part of the lithium. This lithium-6 has the high cross-section.
And the hydrogen did the moderating. So we had some absorption. So it was a combination of moderation and absorption. Moderation by the hydrogen and absorption by the lithium. But in between that, in between the shielding that we had, we had this layer of beryllium for the reflector. And again...
We were thinking ahead, if I remember correctly, we were thinking ahead, and if the beryllium didn't work, we came up with a zirconium beryllite, which was quite brittle, but we did have that. We did work on it. We tested it in a reactor. I remember going and testing it in the LTI. Yeah, that was an advantage, and that was one of the compounds that this effort produced.
But basically, I think, in the end, beryllium was the material that was most satisfied. The machinist had worked, got that down pretty good. Yeah. Yeah, they had that worked out. But I wanted to go into the, a little more into the lithium hydride now. We had to use a number of ways of trying to get this in a shielding design. And one of the...
Well, first of all, let's talk about the methods that we worked. We did some cold pressing. Cold pressing, hot pressing, and melting. Melting. And lithium hydride is really a very nice material. It has a very nice melting point at a relatively low temperature. And in the molten state, it does not decompose.
be held as a quiet liquid without any real trouble. The only real problem is that on cooling there is a considerable shrinkage and for that reason special solidification processes had to be developed essentially cooling directionally so that as the bottom went solid
the top would remain liquid and continue to feed material into the shrinking structure. Well, that's its own cooling. We started out, it cools it up. Well, yeah, that is one cooling. It just keeps going up, and as it cooled, you got to the top, and then you put a pipe on it. You had a pipe, and then you just cut it off and seal it, and that was it. Another advantage, lithium hydride is reactive with the moisture content of air, so...
and it is reactive with air, so it was necessary to clad it with oxidation-resistant alloys such as stainless or nickel-based alloys. And luckily lithium hydride is non-reactive with these metals, so you can actually melt lithium hydride in a stainless steel container without any problem whatsoever.
which is sort of unusual because most of the hydrides particularly start degassing or losing hydrogen as they go into the melting phase. Yes. And that was a hydride that we come up with. Well, this is the only hydride that you can melt without losing the hydrogen. Right. So that was a big step forward in cutting down the size of a reactor.
the weight, more not the size, but the weight. And then the next thing, of course, we had to do would be the testing of it. And we did test it. We have one design here where it was put together and tested. I think there's a 45-inch part of the shield we had never done, John. We had a 45-inch part of the shield, 45, not inches, but degree.
basically one quarter of it. And that was built by the Budd Company, was that correct? I think it was. Budd built that. They went to Oak Ridge and they did some testing there. And I believe Wright-Patterson did the thermal cycling on it. I'm not sure, but I think they did. Yeah, the main problem with lithium hydride is heat is going to be developed in it, and there is a heat removal problem. There is also a high
low thermal conductivity and thermal expansion, so there are problems of cracking. So it became important to come up with structures in which there could be a dissipation of heat and also a resolution of the problem of the structure being integral under conditions of thermal cycling.
To achieve that objective, various honeycombs and wire structures were incorporated within the lithium hydride, and these provided a means of reinforcing the structure under conditions of thermal cycling.
also were a means by which heat dissipation was no longer a problem. That is a way of traveling out to the surface. These systems were air-cooled, and so there was a possibility of getting the air, the bram air coming in and cooling the system, so we did that to require any other work. One of the last areas that we moved to, that we went through a series of things, was the control rods.
And at the beginning, of course, the first test reactors that we run was primarily with boron carbide and aluminum matrix. And then as the temperatures got, like at the HTRE-1, we went to the boron carbide with the stainless steel matrix. And that was cladded with stainless steel.
And then finally, as we got to the final reactors where we're getting ready to really test reactors for the test actual running of a nuclear aircraft, we went to one of the rare earths that we had come out as a product of the interim separation. Well, here again, the problem was temperature.
boron carbide aluminum matrix type of element was well known in the industry and essentially almost readily available. But unfortunately, the temperature limitations of that material are limited by the melting point of aluminum, which is relatively low. Additionally, if you remove the aluminum and go over to a stainless steel type of
structure with a stainless steel cladding you get the oxidation resistance but here again it turns out that boron carbide is reactive with the iron base alloys at such temperatures so it was really then there are no other refractory metals of boron that could be used for this purpose so
That's when our attention was turned to the rare earth oxides that were nuclear poisons, and europium oxide in particular. And here we have a material that is stable at all temperatures to the melting point of the claddings. And it was ideal for operating to the limits of the oxidation resistance of the claddings.
Being a ceramic, the problem of thermal expansion and thermal cracking was an issue, and the final product that was used was actually a 60-40 mixture of europium oxide and nickel, a ceramite-type structure that resembled the fuel element.
where here again the nickel is serving to stabilize the structure against thermal cracking and also providing a means for the heat to be brought to the surface for dissipation by air. Now we studied several techniques of making this fuel element, I mean this fuel element that they control right.
We did some swaging studies on it. We had a large extrusion press, and we extruded some. And finally, I think we found out the most effective way was another development we had, which was the hot isostatic gas pressure bonding. And this illustrates a little bit of one of the products was gas pressure bonded. And in between each segment was the europium nickel mixture.
And it had been centered to a reasonably high density. And then we had little metal slots in between here, which were actually, after it was completed, they were cut, x-rated cut, so that they could produce little segments of fuel elements. And then these were laced together with ink canal, I believe, thin strips of ink canal.
so that if they got hot at the thermal expansion, if it was uneven, the Iconel strips, of course, would allow it so it wouldn't hang up in the reactor. And this is a final picture with the poison sections in between there. There's one little thing that I might add here, and it's really of first importance.
the methods of consolidation such as the hot gas isostatic pressing. This came into being, it was started at Battelle as a laboratory type of operation and we picked it up and essentially transformed it into a commercial application and it turned out to be ideal for
consolidating cladding and fuel elements for making moderator components and cladding them. But also here in the control rods, it turned out to be the ideal way for joining together the cladding and the matrix so you had a continuous metallurgical structure. Just as I've had John Collins told me after the program was ceased and there was very few people there, the one area was the busiest.
was the high-temperature gas autoclaving bonding, and Joe Hopkins was doing that afterwards. I think Ernie Bartell used to do it, and then Ernie left, and Joe was doing it. There is another thing that I think is worthy of mentioning. While we did use the brazing extensively, the electron beam welding came into...
in the latter part of the work. It was a tool that was used routinely for accomplishing the most beautiful welds and providing structures that 10 years earlier would have been impossible to achieve.
Yeah, we could probably ask John whether or not that that's what the ultrasonic drilling begin with that come along to the Electro yeah, the electrochemical drilling and electrical discharge drilling those are when you're working with materials such as tungsten and tantalum and iridium various types of ceramics and
combinations of these. When this project started there was no machine shop that could have produced these things. That's right. But during this period all of these new methods of machining came into being and really contributed to the success of the effort in producing the things that had to be produced. Yeah, I think our machine shop was on the forefront of all machining at that time. They got caught with a lot of elements that they
didn't know about and they got caught with a lot of high temperature materials that were very hard and ceramics that had to be machined and so on.
5.3.1 High temperature materials
When the GE Aircraft Nuclear Propulsion Program was canceled in 1961, the U.S. Atomic Energy Commission awarded GE a follow-on program to focus on high-temperature evaluations of nuclear materials in support of the AEC's established or planned nuclear projects, especially those associated with the liquid metal fast breeder reactor.
This new program led to the name of GE ANPD being changed to GE NMPO, Nuclear Materials and Propulsion Operation, and later changed to GE NSP, Nuclear Systems Programs. In my discussion, I have selected a few mechanical property evaluations that have attracted special attention and that have been the subjects of formal publications.
The first of these involves measurements of creep rupture behavior of refractory metals to 5400 Fahrenheit. Special recognition should go to the late Pete Flagella for his very capable direction and management of this particular material study. All the metals involved in this high temperature measurement are shown in the first slide. With all these tests,
They were performed in either hydrogen or argon using tungsten rod furnaces that evolved from our years of A&P experience with high temperature sources. One illustration of the type of data. Before I go into that, I wanted to focus on the first slide, which shows the four refractory metals that were evaluated.
along with three refractory metal alloys. One example of the type of data generated is provided in the second slide, based on tests of tungsten. These stress rupture results cover the temperature range from 1,600 degrees C to 3,000 degrees C. That is 5,400 degrees Fahrenheit.
and they identify fairly parallel isotherms. If, for example, the 100-hour stress rupture strength is desired as a function of temperature, that information can be read right off this particular graph, this graph of stress versus rupture time. A similar plot has been prepared based on steady-state creep rate.
So a plot of stress versus this creep rate has been made available. Similar plots were made available for all the metals that I mentioned in a previous slide. One other very special materials evaluation involved low-cycle fatigue testing of 304, 316, and 348 stainless steel in the temperature range of 800 to 1500 degrees Fahrenheit.
This type of materials test was not mentioned when the GENMPO program was first formulated by the U.S. Atomic Energy Commission, but their extensive studies of nuclear component failures, and I might say non-nuclear component failures as well, suggested that such failures were probably due to a fatigue mechanism. As a result, GENMPO was approached in 1964
to include this special testing in the NMPO effort with the objective of determining the best experimental procedure to use, acquiring the necessary test facilities, and, of course, to perform the test in sufficient detail to serve the objectives of the program. The technology of this special effort was assigned to Ray Stentz, Tom Slott, and Jim Burling.
test procedures, it was decided to perform strain-controlled evaluations using push-pull loading of inductively heated hourglass-shaped specimens and to use an elastic hinge extensometer for diametral strain measurements. This detailed assessment also led to the conclusion that the commercially available fatigue test machines, and there weren't too many of these at the time,
were not capable of achieving the type of fatigue testing evaluations that were planned for this program. It was decided, therefore, to direct the NMPO efforts to the design and development of special closed-loop servo-controlled machines that were uniquely suited for low-cycle fatigue testing.
I wanted to show a slide that exhibits the hourglass-shaped specimen that was employed with the diametral extensometer in place. This extensometer is positioned at the minimum diameter point of the specimen. And that diameter, by the way, was 0.25 inches for the most part. Diameters above and below that were evaluated.
The general test procedure involved the quarter-inch diameter specimen. This is the elastic hinge here, and any displacement here, which is caused by loading. For example, if you load in tension, the diameter decreases, and then when you reverse the load and go into compression, the diameter increases. The amount of change of the diameter divided by the diameter itself gives the value for the
This value of strain is picked up electronically at the end of the elastic-hinge extensometer, which is six inches beyond the hinge itself, and that's fed into the control system, which we'll talk about later, and allows the system to keep repeating the amount of strain that was set for the particular test.
In preparing for a test at elevated temperatures, before you could put the specimen on test, we would put an induction coil to surround the specimen. We would slip that coil right down over the specimen, and then instrument the specimen with a chromolymal thermocouple right near the minimum diameter point, and then the specimen would be ready for insertion into the test machine.
I wanted to list a few of the problem areas that had to be solved before an adequate test machine and a reliable test machine could be put on test. I've had some lines here to divide the thing into three different parts and as we'll see later each induction or each fatigue machine had three particular
One was the induction heating system and the thermocouple attachment I've already mentioned. And then the second part contained the holding fixture, the specimen holding fixture, the load frame, and the hydraulic actuator. This provided the cyclic push-pull loading to the specimen. And then the third part of the fatigue machine is the control console. This controls the temperature to the set limit.
It controls the rate of cycling. Usually that was 20 cycles a minute. It controls the strain range that was imposed, and it controls the cyclic shape. And all the data recording system, all the data recording equipment is part of the control console. This gives load readouts, strain readouts, et cetera.
Once these problems were solved, the systems were built in-house and proved to possess the reliability desired. This was a major development for this program. Another major development involved the conception, design, and fabrication of the analog strain computer. This device accepted instantaneous signals of the diametral strain and the axial load.
and in conjunction with elastic modulus and Poisson's ratio values, provided an instantaneous computed value for the axial strain component. Thus, for isotropic materials, a method was provided to allow low cycle fatigue tests to be performed in computed axial strain control when a diametral strain measurement was being made. This had never been done before at elevated temperatures.
This strained computer was described extensively in an issue of the Journal of the American Society of Testing and Materials and was viewed as such a valuable contribution to the fatigue testing field that it won the 1971 ASTM Templin Award for the authors Tom Slott, Ray Stentz, and Jim Burling. A final slide shows an example of the type of
fatigue data generated in this program. This is a plot of the total strain range and it's plotted against the cycles to failure. All the data for the three different stainless steels are combined on one plot to illustrate at least one point. These are the data at 800 Fahrenheit, 430 centigrade, and these are the data
at 1500 Fahrenheit, 816 centigrade. The magnitude of the temperature effect there is very prominent. As a matter of fact, it amounts to a factor of 20 from the low to the high temperature. About 15 of the fatigue machines developed on this program were built in-house to serve the request for additional data of this type.
Some of these machines were assigned to the radiation effects program that we'll be hearing about a little bit later. We are now going to continue our discussion of low-cycle fatigue in a visit to the Martest Incorporated Fatigue Laboratory. After GENMPO and GENSP were canceled, several engineers associated with the fatigue testing effort formed a company
Load Cycle Fatigue
called Martest Incorporated, to provide fatigue testing services to various customers. This company now has some 50 fatigue machines in operation. To complete our coverage of the low-cycle fatigue project, I'm going to turn the discussion over now to Ray Stentz. Ray, welcome to the project. Thank you. Ray was...
involved with this fatigue project since 1964 and is the inventor of the analog strain computer that i spoke about earlier in my presentation ray is one of the engineers who left ge to form a private fatigue testing company
and he's going to review in general terms the features of the low cycle fatigue machine, right? Specially designed machines have been developed.
to subject materials to low cycle fatigue exposures. This is a typical low cycle fatigue machine. It consists of an induction generator, a load frame. The load frame applies the forces to the specimen, and a control console, which accepts the signals from the load frame and controls the strain and the temperature. I'm going to cut. I'm going to move in so we can get these things a little closer.
electrical currents through a copper coil which surrounds the specimen. This in turn induces currents in the specimen which heats it to the desired temperature. Now the load frame is the device which originally secures the specimen at the top. It's connected to this platen. The bottom of the specimen is connected to a hydraulic actuator which moves up and down to supply the strains and the forces to the specimen.
contains instrumentation so that we can measure the temperature, we can determine the force on the specimen, and we can determine the strain of the specimen by putting an extensometer on the gauge length of the specimen. The control console, which is this device, accepts the signals from the load frame from the instrumentation. It compares these signals constantly with what's required.
and it makes corrections as necessary by sending electrical signals to the load frame and to the induction generator. This particular load frame in a strain control test would constantly control the strain amplitude throughout the test until the specimen breaks and separates into two pieces. The number of cycles required to do this is a measure of the fatigue life.
of the specimen this particular fatigue life is consistent with only the temperature that we ran the test at the strain range we used and the frequency of loading other useful information is also obtained beside the fatigue life the instrumentation provides a record of the material response
This is used by material engineers and designers to determine the cyclic, elastic, and plastic response of the material and the forces required to maintain the strain range throughout the test. This information eventually goes to a database which is used to predict low cycle fatigue behavior of actual components. That's all.
We can get an actual setup of the machine that's running. This is an actual fatigue test in operation.
As you can see, the extensometer is coming in from the right on the specimen. The specimen is encircled with the induction coil, and the actuator is going up and down to impart the strain on the specimen. The motion is very small, and you probably can't see that. If we switch over to the console...
Can you see this part here? Yeah. OK. This needle moving indicates that an instantaneous correction has taken place, controlling the strain on the specimen. This recorder indicates the force and the strain that's present on the specimen. Every cycle is recorded.
In addition to this, we periodically record what we refer to as hysteresis loops, which was a measure of the strain and the force in each cycle. These numbers indicate the cycle at which that recording has taken place.
High temperature material properties
Now that we've finished talking about the high-temperature mechanical property evaluations, I'm going to turn the discussion over now to Alan Feith, who is in charge of our high-temperature property study that dealt with the physical properties of metals and ceramics. So now I'd like to introduce Alan Feith. Alan? Thank you, Joe.
In the aircraft, nuclear reactors were in the range for which most of the properties were known or readily obtainable with existing equipment. The lower power requirements of the reactor allowed the use of aluminum and alloy steels for structural elements. After the termination of the ANPD project, the Atomic Energy Commission requested that we determine the thermophysical and mechanical properties of materials which could be used in advanced reactors operating at considerably
higher temperatures this led us to the development of equipment and methods to accomplish these tasks in some cases we measured properties of materials at or near the melting point in essence we built upon the earlier work and extended the knowledge of material properties most of the studies were presented at technical conferences and published in journals it was our good fortune that we had access to
furnaces which could operate at the temperatures of interest. Our job was to develop experimental methods which could be adapted to these furnaces. We also had to be aware of peculiar things which could happen at high temperatures. Our specimens as well as the tungsten heating element of the furnace had to be protected from oxidation by use of very dry inert gas such as helium. The specific properties which we measured were
thermal conductivity, enthalpy, thermal expansion, electrical resistivity, and thermal diffusivity. My co-workers were Ray Hine, Art Loskamp, and the late Mike Fensel, as well as Carol Johnstone. The furnace used in
For thermal conductivity, electrical resistivity, and thermal expansion was a tungsten tube furnace shown in the figure here. The tungsten tube furnace is here. This particular picture has the apparatus for thermal conductivity, which was a radio heat flow apparatus. The furnace was brought up to temperature, held at a steady state.
And then the center HTRE was energized to provide a thermal gradient in the specimen. The concept was simple. We had to design the specimen thermal guards to keep the center specimen isolated and the center HTRE.
The most difficult thing was measuring temperatures to a degree of accuracy required. In our work, we actually measured thermal conductivity of ceramics, which were very low, and metals, which were considerably higher, a ratio of about 50 to 1 from the high thermal conductivity to the lower thermal conductivity.
In the next picture, I want to show a few examples of the things which we were measured. We measured beryllium oxide from 1,000 to 2,200 degrees centigrade, and yttria from 1,000 to 2,400 degrees centigrade. Next slide.
Molybdenum was one of the higher conductivity materials and required that we measured thermal gradients to the hundredth of a degree centigrade. We also experimented in measuring tungsten with different volume percents of uranium dioxide. You will notice that uranium dioxide has a very low thermal conductivity compared to tungsten.
And these are various percentages of UO2. In addition to measuring the properties, our customers required that we understand the reasons for the behavior of the material and its effect on a property.
For instance, heat is conducted through solid materials by several mechanisms. One of the components is the lattice conductivity, which is similar to the travel of sound through solids and moves at the speed of sound in the particular material. Heat also is carried by electrons, so the resistance of the material has some effect on the total conductivity. Some material studies are dielectric.
Insulators at normal temperatures and conductors at elevated temperatures. This also has a bearing. Thermal radiation contributes as well. Transparency can change with temperature, and heat can be radiated across pores in materials whose density is lower than theoretical. Hence, a correction for porosity needs to be applied. It has been observed that phase changes affect the flow of electrons.
and the electrical resistivity and thermal conductivity as well. Fitting property data to mathematical equations was another requirement. In many cases, power series equations could be used. However, in the case of thermal conductivity, some of the components were exponential in nature, and the power series could not be used. At the time of this work, that proved to be a challenge to the most studious computer analysts available.
Data from these studies and those discussed below were published and disseminated at various conferences. The next property I want to discuss briefly is enthalpy. And in the picture shown now, we show the apparatus. The specimen was heated in a tungsten rod furnace contained in a capsule. When it reached the proper temperature of study,
the container specimen was dropped into a calorimeter. The temperature rise of the calorimeter then determined the heat capacity or enthalpy of the material in study. Several pictures now are shown of the enthalpy of these materials. The picture before us now is that of unalloyed tantalum.
We measured that from 1,000 degrees centigrade to just under 2,400 degrees centigrade. We are now looking at a tantalum tungsten alloy. You'll see it's still linear and there's not much change in enthalpy. And here we have two other materials which are also tantalum alloys.
Again, the significance of displaying these pictures is the fact that we were able to make these measurements over temperatures from 1,000 degrees centigrade to 2,400 degrees centigrade. We show here that when we do an alloy material, we can also calculate by a weight percentage the enthalpy of alloys from the
known values for the elements. This was molybdenum and rhenium. We turn now to the measurement of thermal expansion. The furnace use was similar to that used in thermal conductivity, and we adapted a specimen holder and parallel telescopes to read
the change in length of the specimen as it was heated again we had to attain the constant temperature to allow us to make the measurement of the length with the parallel telescopes we're now going to show several slides of the materials which are measured we measured many metals in our thermal expansion studies we looked at the
Alloys of rhenium and tungsten. The chart before us now shows the thermal expansion of rhenium, tungsten 25%, rhenium, and tungsten below it. Measuring electrical resistivity, we used a four-probe method. We did this on numerous materials.
was the same as for thermal conductivity. The specimen was used in place of the center HTRE. And this is a schematic of the apparatus used and the specimen here. It was a rather simple specimen to use and still required high precision and measurements. Typical of our measurements of
Electrical resistivity is that shown here for molybdenum. The last property I want to talk about is thermal diffusivity. Thermal diffusivity is a rate of heat passing through a solid material. We could do this with very small specimens.
The specimen used in this particular apparatus was less than the size of a dime. We used a laser to give us a very rapid heat rise on one side of the surface and then watched the effect of that with an infrared detector on the opposite side.
on an oscilloscope and photographed. Analysis of the photograph allowed us to determine the thermal diffusivity. The fact that we could do this on a very small sample allowed us to use that method for determining the effects of radiation on various materials.
Unfortunately, we do not have pictures of the apparatus, nor do we have data to demonstrate. The work in measuring these several properties extended the knowledge of thermophysical properties at that time and probably has been useful in many applications.
5.3.2 Radiation Effects
My discussion represents a review of the GE NMPO Radiation Effects Project that was directed by the late John Motiff. Radiation effects are actually changes in the internal structure of materials that occur once they are exposed in a nuclear reactor to high levels of neutron radiation. These changes have led to the term radiation damage.
because of the detrimental effects on mechanical properties as well as physical and electrical properties that have been observed. A serious loss of ductility, termed radiation embrittlement, has been cited as a principal problem area. In John's project, he organized a detailed approach to evaluating radiation effects on potential metals and alloys
for use as fuel cladding materials in the LMFBR, the liquid metal fast breeder reactor. His plans were to not only measure the magnitude of the radiation effects on properties, but also to employ metallography and electron microscopy to aid in explaining just why these changes were taking place. His principal investigators on this effort were Fred Kingsbury, Joe Beeler, Bob Brough,
and the late Jack Smith. In one of John's reports, he listed atom displacements and transmutations as the two processes by which crystalline solids are damaged by the neutron bombardment. And this type of neutron bombardment, as we know, is present
in all reactor environments. Talking about atom displacements for the moment, I've pictured here a simple body-centered cubic structure, which is typical of tungsten, molybdenum, tantalum, and several other refractory metals, with an atom at each corner of the cube and one atom positioned in the very center of the cube.
In response to the neutron damage, several important things can occur. For example, if we pick this atom of tungsten, for example, and it is involved in an elastic collision with a fast neutron, it can either strain the position of that atom, or it can knock it right out of that position and have it located in an interstitial position.
This creates a vacancy here in the lattice. So when we talk about atom displacements, that's the kind of collision that we're talking about. And it can either strain the lattice or it can produce a vacancy which has an effect on the mechanical properties. The next slide is a brief...
discussion of transmutations. And I've picked the boron transmutation, which is an N-alpha reaction, so that when a neutron is absorbed by the boron atom, this is the symbol for the neutron, it emits an alpha particle, and the boron splits into a lithium-7 atom, and it gives off
a helium particle. Another transmutation which will come up in subsequent discussion is when a tungsten atom undergoes an N-gamma reaction. It is converted from tungsten-187 to tungsten-188, and that changes to rhenium in a certain period of time, and eventually
That changes to osmium. Now all these are 188s up here to indicate that nothing happens to the atomic weight. So they're the two reactions we'll be hearing about in subsequent discussions. Now the next slide is to give some indication of just how extensive John's program was.
all these materials, all these refractory metals and these refractory metal alloys. And he studied quite a few nickel-based alloys. This A286 is an austenitic stainless, Inconel 800, and a couple of stainless steels, and another nickel alloy, Inconel 625.
The next slide indicates another facet of the program, because the general theory then was that you can't measure radiation damage effects by measurement of just one property after the irradiation, and you should include at least four. Well, John included stress rupture and creep.
tensile tests, electrical resistivity, hot hardness, low cycle fatigue, substructure studies, and damage calculations. And prior to the program, prior to the radiation program, all these materials were accumulated. All the necessary test specimens were
machined, and then they were all sent in various groups to be exposed to the neutron radiations, either in the Oak Ridge reactor or the engineering test reactor in Idaho. And the range of the fast fluences, fast neutron fluences that was used went from five times
5.9 times 10 to the 17th to 1.4 times 10 to the 21st NVT. This is an expression of the neutron density within the reactor. The radiation temperatures, and this, by the way, has an important influence on just how the materials behave after irradiation. It's the temperature at which they're exposed to the neutrons.
The temperature range was from reactor ambient, which is just about room temperature, to 1300 degrees C. And now I would like to show a few graphs depicting several interesting behavior patterns reported in John's program. This first graph is based on stress rupture data for A286 tested at 1350 degrees Fahrenheit in the irradiated and
unirradiated condition. These are the results for unirradiated material tested at 1350 degrees Fahrenheit. This chart gives the stress in kilograms per square millimeter and the rupture time in hours along the abscissa. Notice the drastic reduction
the rupture life when exposed to 2.5 times 10 to the 19th neutrons per square centimeter. This is a typical behavior pattern that has been reported for these kinds of materials. Notice now the reverse effect as we see here for an exposure to 8 times 10 to the
This is for tungsten tested at 2,000 degrees Fahrenheit. And the important observation that was made is that for the unirradiated material, the stress rupture behavior is given by this line. And for the irradiation to 8 times 10 to the 19th neutrons per square centimeter,
rupture life is increased substantially. John stated his belief that transmutation products enhance the mechanical properties of the refractory metals by a solution-hardening mechanism. In this case, tungsten transmutes to rhenium, and parts per million quantities of rhenium that are formed go into solution to strengthen
the material. And finally, I wanted to show an illustration of the effect of irradiation temperature on mechanical behavior. You remember I said earlier that the temperature at which the irradiation is performed has an important effect on the properties.
This observation is shown here where the unirradiated material has a creep response. This is percent elongation as a function of time. So this is a creep curve of unirradiated material. All tests, by the way, are of a molybdenum tested at 750 degrees C at a stress of 18 kilograms per square millimeter.
to 1.6 times 10 to the 20th at room temperature, the creep curve increases noticeably but not significantly. When the irradiation temperature is increased to 700 degrees C and 1,000 degrees C, the increase in the creep strength is quite substantial.
I am now going to turn the discussion over to Bob Brown, who was in charge of the electron microscope laboratory in Building D and worked extensively on the radiation effects program. Bob? Yeah, I would like to talk a little bit about some of our electron microscopy studies that we did on irradiated materials. And maybe first I should explain a little bit about what an electron microscope is.
In a way, it's similar to an optical microscope, a typical light microscope, except instead of using optical light, we use electrons, a beam of electrons, and we focus them with electromagnetic coils instead of glass lenses. The short wavelength of the electrons allows us to magnify things to much higher magnifications than we can with a light microscope. It also allows us to image much, much smaller detail than we can see with a light microscope.
To do studies on irradiated materials, irradiated metals, refractory metals in particular, which I'm going to talk about, you first have to prepare the sample so that it is transparent to the beam of electrons. This means we must have a very, very thin metal section. So in order to do that, we first slice or saw a very thin wafer from the specimen, and then we electrochemically thin it until a small hole breaks through.
At that point, we have a specimen which is transparent to the electron beam, and we put that in the electron microscope and scan the region around the hole where it is vanishingly thin. That is the electron transparent region. So now, to give a few examples of some of the work that we did under the program, we studied refractory metals, and in particular, I'm going to start off with tungsten.
I have a series of diagrams here, a series of pictures showing irradiation damage to tungsten metal. The first picture on the left-hand side shows a rather featureless background of unirradiated material. This would be the normal tungsten prior to irradiation. You can see a few dislocation tangles, a dislocation network coming down through here.
Dislocations are really not anything on an atomic scale, but they are due to a misalignment of atoms in the crystal lattice. Most solid materials, metals, ceramics, and so forth, have the atoms arranged in a regular crystallographic arrangement. It might be cubic, hexagonal, something like that. When those blocks are slightly misaligned with respect to each other,
The disalignment boundary then is a dislocation. It's a region where there's a slight strain in the lattice and therefore we see an image in the electron microscope which we call a dislocation. So that's what these boundaries are or these tangles of dislocations are that you see here. This would be called a sub-grain boundary. In the next picture, we see an example of tungsten which has been irradiated to a low neutron dosage.
Again, we see some of these dislocation tangles indicative of the sub-grain boundaries, but we see a lot of tiny black spots superimposed in the matrix. Formerly there was nothing there, now we're seeing many, many tiny black dots. These are diffraction contrast images of defect areas caused by
Either interstitial atoms, which have been knocked out of their normal site and have now lodged in between the regular positions, or vacancies left behind when the atoms were knocked out of their original site. You can imagine this occurring when you put an extra atom into the lattice, or maybe an extra marble in a pile of glass marbles. The lattice has to stretch, it has to give, it has to expand to accommodate that. So you have distortion there.
Likewise, the site from which the atom was knocked out collapses a little bit. You also get distortion there. So we can have two different kinds of defects caused by bombarding with neutrons, radiation damage. They can be interstitial atoms or they can be vacancies left behind. And the little black dots that we're seeing here are the visible image of those defects, those radiation-produced defect clusters.
Finally, on the right-hand side, in the third picture, we see a piece of tungsten which has been irradiated to a high neutron dose, and the damage is extreme. The small black dots have now coalesced into much larger black dots. In some cases, we can even see the beginnings of little loops forming where they have coalesced into larger clusters.
Again, at this point we don't know whether they are vacancy clusters or interstitial clusters, although we can do experiments in the electron microscope to determine whether they are vacancy or interstitial in nature. You can imagine that clusters like this in a metal then have the effect of creating barriers to
movement of grain boundaries or sub-grain boundaries or dislocations that occurs when a metal is stressed or put under some kind of a strain or tested. This hardens the metal and causes embrittlement, which of course is a detrimental factor. So in the next picture, we're going to see what happens when we anneal these radiation damaged structures. Now we have a picture where
We have annealed a piece of tungsten which had been irradiated to a high neutron dosage. It formerly had many of these black dot defect clusters in it. Now we see that we have pretty well cleaned up the background. We don't see much there, but instead we're seeing some very large, well-defined loops. In this case, they happen to be labeled A, B, C, and D because they are lying on four distinct crystallographic planes in the tungsten.
the cubic lattice of the tungsten material. The drawing up in the right-hand corner is an attempt to show those four planes. It doesn't show up too well in a flat perspective like this, but those loops are actually lying on four planes within the tungsten specimen. By doing these rather sophisticated tilting and diffraction contrast experiments in the electron microscope, we were able to determine that these are planes of vacancies, not interstitials.
This occurs because vacancies are less mobile in the lattice than interstitials. When an irradiated specimen like this is annealed, heated to a high temperature, the interstitials tend to migrate out to the grain boundaries and disappear, or if they run into a vacancy, the two are mutually annihilated.
But the net result is you wind up with excess vacancies, and these vacancies then tend to cluster into the loops that we see here. Lattice forces cause them to coalesce into planar loops, which lie on these single crystallographic planes. To visualize it, you can think of maybe taking a stack of papers.
and cutting a circle out of one and stacking them back together again the lattice is then going to collapse around that vacancy loop so that's what we see in those cases i don't have a good picture to show but if we further do experiments on refractory metals like tungsten and molybdenum at high temperature where the irradiation is actually carried out at high temperature then we actually saw these vacancy loops coalescing into vacancy
Clusters or bubbles almost look like bubbles in the microstructure. Those too act as barriers to dislocation motion during working of the metal and cause hardening of the material. Finally, I'd like to finish up with a little discussion of kind of a special project that we did studying some single crystal vanadium.
In this case, we started with vanadium rods, which had been specially grown in a favored orientation, so we knew what we were dealing with going into the electron microscope. These were irradiated, then slices were taken, and our little specimens prepared for the electron microscope. And the accompanying diagram now shows what we saw when we put these into the electron microscope. This is a composite picture showing...
two electron micrographs, it shows a diffraction pattern, the electron diffraction pattern up in the top, and a drawing down in the bottom. First if we talk about the two electron micrographs on the top, we can see first of all this bullseye over here very plainly showing up, kind of a white center, a darkened inner circle, and finally a larger outer circle.
That is actually a cross-section of a damaged shell. We also see in the same diagram, coming down diagonally across, a series or an envelope of damage, dark inner region, a slightly lighter outer region. I should mention that the matrix itself is surrounded by these black dot defect clusters that we talked about before. So there's damage throughout. We have enhanced damage in these shell regions.
The left-hand picture shows a very good example of a transverse section of a damaged shell where we have a core running down through the center, an inner damaged shell which is very heavily damaged, highly populated with black dots, an outer shell which is again highly damaged but not as high as the inner shell, and finally the matrix on the outside. What this really consists of can be shown in the diagram down here at the bottom.
where we see that we discovered that in this vanadium metal, this single crystal vanadium material we were working with, were occasional impurity fibers of vanadium boride. And boron is an interesting material when you put it in a nuclear reactor because it will take up, will absorb a neutron and become unstable, and then it will undergo a transmutation reaction. It will actually split.
because it's unstable, and throw out an alpha particle and a lithium ion. That's what's going on here. We are bombarding this little vanadium boride core with neutrons. It then absorbs the neutron. It becomes unstable. It splits, and the ejected lithium ions have a certain range.
a rather short range they go out and cause damage along the way which causes this inner damage shell the alpha particles or helium ions can travel a little farther they create damage to a larger distance as you can see in the outer shell here so we're actually seeing a cylindrical damage shell which surrounds these tiny thin vanadium boride fibers which were embedded in there
On later annealing, I don't have a picture to show, but on later annealing of these structures, we were able to see lithium vanadate crystals located in the inner shell and helium bubbles located in the outer shell, just what you would predict from this transmutation reaction.
Finally, one of these micrographs was used as the cover of an article in the Journal of Applied Physics. The article appeared in the journal, and they actually chose one of our electron micrographs to be the cover photo, which was kind of a unique experience for us. Thank you very much.
5.3.3 Chemistry
I'd like to introduce Jim to the group, and he will be part of this section on the analytical chemistry. Good afternoon, Jim. Good afternoon. Would you explain to me who this gentleman is in the picture here? Well, I was taken just a few years ago, actually about 40 years ago, and that's me when I was much younger.
I didn't have the color hair I have today, which I now dye to fit my age. Oh, you dye to fit your age. Of course. It is actually black. I may have learned that, but I need something to put on top. I think Ronald Reagan and I have the same genes. Okay. That's right. Jim, in the analytical group, I know it was a fairly large group, but who are some of the members that you can recall right now? Just so much you don't want to overlook anyone, but you had such a large group, I'm sure.
that you can, let's recall some. Well, I joined a group in 1956 when Al Rebo was in charge. At that time, we had a number of different laboratories. Dick Kupel ran the spectrographic laboratory. Al Schaefer ran the inorganic wet laboratory. Bob Stewart was in charge of the radiochemistry laboratory. And I was in charge of the analytical research group.
There were also Ruth Menke, Warren Davis, Alex Rosenberg, Bob Williams, and a number of other people. All right. I have another picture on. To begin the next part of your discussion, I have another picture on to start us out. Would you explain this particular picture here? This is a picture of Alex Rosenberg and myself.
That instrument that Alex is operating is a gas chromatograph. It was an instrument that we developed for the determination of impurities in neon. When we were given the job of analyzing neon, I went around and looked at the commercial instrumentation that was available, and the best instrumentation available had a sensitivity of only about 100 parts per million.
And our job was to develop a technique capable of measuring impurities at a part per billion, which we were able to do. This is the instrument made up of a number of commercial components. And we used amplifiers, flow controllers, and things of that nature. And we received the General Electric Technical Excellence Award for that development. Right.
You had that development, and there were a number of other analytical techniques at which you worked on, which I think deserves to be mentioned. I will put the ones that have been identified, and I'd like for you to enlarge on those and explain each right now. One of the problems that people in the ceramics group were having was that
when they made a fuel element, a uranium oxide fuel element, at high temperatures, the uranium oxide would volatilize and result in a substoichiometric uranium dioxide, which, although stable at high temperature, when it cooled to room temperature, dissociated into uranium oxide and uranium metal. When this was clad and put in a reactor,
The uranium metal would migrate along the grain boundaries, react with the cladding and cause the cladding to rupture. And our job was to determine the amount of free uranium metal in these fuel element cores and eventually when they made them out of uranium oxide and thorium oxide to determine both uranium and thorium.
None of the conventional methods— This is item three? No, this is item two. Item two. None of the conventional methods would work because the determination of a few parts per million of uranium in a matrix that had as much as 85 percent uranium, you couldn't measure uranium and tell the difference. So the technique that we came up with was to expose the fuel element.
and hydrogen, the hydrogen would migrate along the grain boundaries, react with uranium, form uranium hydride, then evacuate this system, heat the hydride up, it would dissociate, and then we would measure the hydrogen on a chromatograph. And because uranium and thorium would form hydrides at the same temperature, but they would dissociate at different temperatures, we were able to separately determine uranium and thorium metals.
and the fuel element cores. Oxygen to uranium ratios, this came about because they were having a great deal of difficulty measuring the thermal conductivity of oxygen to uranium oxide samples that were manufactured in different atomic energy installations. And this is caused by the fact that
The technique used was to oxidize uranium dioxide to U-308 and then make a calculation which would give you the initial uranium to oxygen ratio. However, the work that I had done previously in Oak Ridge taught me that U-308 was really an accident, that the uranium oxygen ratio was dependent upon a number of things, such as the previous history.
of the uranium oxide, that is whether it had been made from a precipitation with ammonia, whether it had been made from hydrolysis of the fluoride. It depended upon the vapor pressure of the oxygen such that actually the atmospheric pressure would have an effect on the final uranium ratio, uranium to oxygen ratio.
We're talking about item three. Also the length of time and the temperature and all of these things and so therefore people at different installations around the country were getting different results. Now these results under normal circumstances would have been quite satisfactory because they were within the limits of accuracy that one would expect.
However, since the values were not correlating, that is the thermal conductivity values were not correlating, we had to determine something that was more accurate. And so the method we chose was to dissolve the uranium and then convert it, the solution, the uranium and solution, convert it back to an oxide by evaporating the acid.
evaporating the rest of the liquid, and then igniting it under a preset set of conditions. In this way, we destroyed the history. The uranium in the solution didn't know where it came from. Now, for scientists like you, of course, you know that the blue ones would come from the precipitation with ammonia, and the red ones would come from if it had been precipitated as a fluoride. No, that's very enlightening, Jim.
Well, I mean, scientists really learned that the best way of distinguishing some of the atoms in the atomic species is by the color of the atoms. Right. You want to move on to four, then? The application of the double withdrawal countercurrent extraction technique to analytical chemistry. This came about when I went to a meeting in New York, and I heard a fellow by the name of Craig.
discussed the use of the double withdrawal technique to separation of materials that were very difficult to separate by extraction. And after the meeting was over, I asked him if anybody had ever tried to do this analytically, because it seemed like an excellent technique to determine elements that were very difficult to separate.
And really the foundation of chemical analysis is separation because when things are in solution by themselves, if you have a pure material, it's relatively easy to analyze them. The main problem is to separate things from other elements that would cause interference with your measurement technique. Well, I set about trying to do this, and it turned out to be a...
very interesting mathematical problem and was able to develop some mathematics that allowed allowed us to predict how this thing might work and we actually used it to separate a couple of elements of tungsten and molybdenum and we theorized that it could have been used to separate neighboring rare earths which at that time was an extremely difficult problem.
The only way I know that it had been done at that time was by spedding out at Ames, Iowa. And that paper turned out had very little application to the fuel element work, but I got more requests for that paper than anything else I had ever published. Requests came from all over the world because they were interested in it as an analytical application for the determination of things that were
Very similar, technically. All right. You had some other interesting work that you did on item five. Would you briefly touch on that, then? Well, as an analytical chemist, you find that depending upon what material that the people are working on around you,
For example, yttrium oxide, as you remember, we did a lot of work on that. You develop methods for determining a bunch of elements in yttrium oxide, and then six months later they're interested in something else, say beryllium oxide, and so you develop the same methods over and over again. And it finally seemed to me that maybe we ought to have a method that would allow you to determine an element regardless of the matrix. And I set about trying to do this and ended up...
quite a bit of success. And as a test of it Dr. Lewis Garden from Case Institute suggested that well why don't you determine it in the presence of all of the elements. So what we did is we set up a sample that contained all the elements in a periodic table except of course things like the transuranium elements or gases or things of that nature. And this was
undoubtedly the worst sample that anybody has ever tried to analyze and we applied these methods to it and they work quite well and that's what i call it some rather unusual chemical analyses well we move on now in your work you are published a number of papers and so on and let's run through the list of those now and you give us a brief description of some of the papers
here i have it'll take two sheets to put it on now these are just the papers that you published while you were with the amp program is that correct that's correct well we've already touched on some of them and as you can see the analysis of beryllium oxide came out in a number of different parts and this is the sort of thing i was mentioning before that
At this particular time, we were working on beryllium oxide and a number of these procedures I had used in the past for something else. Things that were not published was the application of these particular techniques to Bureau of Standards samples. These methods were able to be applied to Bureau of Standards samples. Matter of fact, I analyzed all of the applicable samples.
And almost without exception, I was able to get the, quote, correct answer the first time I did it. Now, I don't know whether you realize it, but the establishment of the purity, not the purity, but the, yes, the impurity content of a Bureau of Standards sample sometimes took several years. And these techniques were able to be applied immediately. And many labs.
They would have lots of labs, and I think the way, if I remember correctly, was that they would send out a sample to a number of labs, and they would take, average all the lab results together, and that become the point of interest for the standardization. I worked on some of those, and you're quite right. The big problem to set up one of these things is always to get a sample that's homogeneous.
so that when you send it around to the various laboratories, the results that you get back are indicative of the method rather than the inhomogeneity of the material. It is quite a job to establish a standard. There's no doubt about that. And that's what I was really trying to do with these methods.
6.1 Thermocouples
I see. Now, I know you and Jim McGrady did a lot of work in the beginning. I believe Jim came up with the idea of the iron chrome aluminum alloy. Can you give us some? Yeah, but first being the tungsten uranium alloy. Yeah. Can you give me a little history on that? Well, at the beginning of the beginning, which we should, it all started when Jim approached me, I think maybe it was early 1957 in the corridor outside of L-8 or 6.
said he wanted to do some phase diagram research. And he was particularly interested in some high melting point materials. And as we continued this discussion, Jim made the observation of, well, we're going to have to do some thermal analysis work, but we don't have a thermocouple that'll reach these temperatures. And so he said, well, let's push the phase diagram research aside and start with a thermocouple. And he mentioned
Let's take a look at refractory metals. And he particularly mentioned the element rhenium, which was pretty much a novelty in the business in those days and only recently discovered. So he sent me about over the program to screen through high-metaling point materials in the refractory metals group and see if there were any good combinations for a high-temperature
usable thermocouple. And so we all started with that and one of the things we had going for us immediately was the fact that in previous work the so-called gallium pool high temperature furnace had been constructed and was available so that we could reach at least the temperatures in the 4,000 degree range. So we started with available materials was another thing and of course we had
unalloyed molybdenum wire and we were able to get some pure rhenium wire and we started by calibrating this combination and this also served to check out a calibrating system and to kind of shake down a furnace and the apparatus we needed to do the work. A little later on we got into looking at other refractory metal combinations including tantalum and
and others, and we quickly came to the conclusion that our best avenue to follow was with the alloys of tungsten and rhenium. And that involved a second problem, and that was getting these alloys made into the form of wire, which we could use for the work. And we found a source that would take the
the rhenium powder that was available and fabricate various percentages of tungsten and rhenium. And as a result, we then were able to pursue these combinations and start doing some basic calibration work. And these charts we have here, these figures show some of the early work with pure tungsten versus pure rhenium and then
joining tungsten, for example, 20% rhenium with tungsten 3% rhenium and so forth. And later on, we were able to get tungsten 26% rhenium material. As it turned out, in the end, the best combination we came up with was a tungsten 26% rhenium on one leg and a tungsten 5% rhenium on the other.
And we concentrated most of the work from then on on that particular pair of wires. And these calibrations you see here go up now up to the 5,000 degree Fahrenheit range. So we were able to do this because we could convert that gallium pool furnace with tungsten heating element and eliminate the moly out of the system. And a lot of the initial work was concentrated on
getting the furnace in the condition where we could reach these kind of temperatures yeah now you uh of course you had to run the tables on there and then you had to make up the chart yeah uh the emf chart yes and um how long did that take you to do that then to get that in position where you knew that you could duplicate wire after wire and it would be at the same output well this this work
went on up through late 1959. We were pretty much in a position where we had a practical thermocouple that we could use, and we had not only a calibration curve, but we had calibration tables, which could be used to make what we felt were pretty accurate measurements, determining the stability of these materials after continual heating and cooling, the effects on the EMF curve.
of this kind of usage. And we did prove that the combination, they were very stable through many hours of use. So we had what looked like a pretty practical thermocouple inside of a couple two and a half years. Now after that, you finished that work and the end of the so-called AMP project. Has there been any further work done on this?
Well, as I indicated before when you asked me where I had gone from the AMP department, I went to Hoskins, who are an old-time manufacturer of specialty metal alloys and have made quite a mark for themselves by producing highly reliable thermocouples such as chromal allumel and other combinations that
are usable up to the 2,000 degree Fahrenheit range. So Hoskins was a logical place to extend their expertise up to scale further in the temperature, and they were highly interested in the work we had done with the tungsten uranium, and that's what my initial work was comprised of there, was to put this into a commercial product, including the marketing aspects and whatever.
And that took a couple years of work. And the important point I think here is that we did some cleaning up of the calibration figures in the tables, but basically the calibration is the same one that we developed early on. You mentioned that the Bureau of Standards is using that as a standard calibration this morning when we were talking.
Is that true from the standpoint that the calibration that you guys worked out is essentially the same as it is now? After a number of years, and we were in contact constantly with the Bureau on these developments, the decision was made to accept this calibration as the standard for this particular thermocouple combination, and that's where it lies today.
The original calibration has pretty much become the standard of the industry. Okay, you mentioned the fact that Hopkins, Hoskins rather, is making this now. Could you give me an idea of what the market might be and where it is? Well, it has to be obviously a small market, but it's a very lucrative market in terms of profit.
in my travels around i've seen it in use all over europe certainly and as well as the far east the japanese are particularly interested in it and and are one of the principal customers for hoskins material they are involved in a lot of high temperature research over there as well as for example in the manufacture of artificial diamonds
and places like Coby Steel. They use the thermocouple then to measure the temperature in the diamond matrix or what? In autoclave, the autoclaving part of that production system. As well, the Chinese, strange enough, are also manufacturing now this thermocouple combination in Beijing. In fact, I don't...
I haven't followed this lately, but I believe they are trying to market it through Canada. Well, that's the route coming through Canada. Well, yeah. At any rate, it's kind of a worldwide thing. And, of course, in the early days, the Russians as well were doing a lot of work with this. But they have never taken it into a, that I know of, into a commercial venture. Uh-huh.
Well that pretty well sums up the major part of the thermocouple. What we thought we'd do today is look at some pictures we have in the shop showing some of the
6.2 Shops and Labs
components and John would comment upon them as we show the pictures. The first picture we have is the graphite shop showing the general area there. The lighting in that shop had to be explosive proof and so did the motors on all the machines and the vacuum system in there took care of the dust.
so that workers would not have to breathe that. In general, that's about the extent of that. They made graphite dyes and components for the laboratories and manufacturing fuel. Well, they also worked on other ceramics other than graphite, didn't they, John?
They did work on some, but not anything of the nature that would have been done in the special metals group. Okay. All right. Okay. Here's a picture of the special metals shop. Right. There's a young fellow there doing an operation in a dry box.
And his name is Elmer Burns. And Elmer was in that shop for quite some time. The lower picture shows some of the filtering systems that were required to filter out all of the hazardous materials in the solutions that were used during the machining operations.
Here's a picture, John, that's interesting. I think it's a straddle grinder, isn't it? Yes. Clay Brasfield referred to this operation, and I happen to have a picture of it which shows the straddle grinding of the beryllium oxide UO2 tubes in its early stages. And it was indexed by a small wheel at the top of that long slanted piece there to get the hex shapes.
and the wheels were diamond coated and plated on the wheels, so they did a very precise job. Okay. And here's one of the BEO reflector segments, I believe, from the... Up the other way. That's it. Yeah, that's the HTRE, one beryllium metal.
reflector that went inside the core and surrounded the active part of the core. Okay, now the next one here is one that's dear to my heart. I was trying to get materials information on beryllium oxide, and I wanted to find out what, get a static modulus measurement on beryllium oxide. And I asked if
you could grind a spring for me out of BEO. And I was surprised when it happened, and what a beautiful job they did. You can tell us how it was done. Yeah, that was done with an ultrasonic, or what they call a cavitation machine, and it was machined in its entirety by that method. It was one of the unique machines that we had the capability of using on this type of material.
The beryllium oxide has such a high modulus and is so brittle, everybody was shocked when they could take a BEO spring and spring the material like that. That is really unique. Here's some other components that you people made in the shops. You might want to comment on those a little, John. Well, there's the metallic fuel element up in the upper area there, Bert, the circular parts.
They were made and formed on equipment that was made in the general shops, dyes and so forth. And then there's some ceramic parts of all types that were various generations of the development program. And a couple other items that may have interest. It looks like a clad moderator. Right, some moderator samples. A sample, yeah.
It's been evacuated and sealed, is it not? Right. That was done, and the evacuation was done and welded in an electron beam welding machine. Pulled a vacuum. And was this a fuel element sample? Yeah, I think that is just a small fuel element cartridge type for testing.
Explain this one, John. That's the general shop area where all the smaller machines were contained. And they did the general machining of most all materials except ceramics and the graphite. They made their dyes and other components, as well as much tooling for the special metals. I mean, the special.
fuel element line and in general it it covered quite a broad type of work in that area that was a big shot there's a big shot there that was only about half of the area yeah this looks like a tube sheet of some sort john contained in that general shop was a jig mill which you see in the upper picture there and that is
machining a tube sheet. It's a very accurate machine, similar to a jig, but only it's done horizontally. And that's a very tough material that that is machining there. I think that had the, it was Incon LX with a high cobalt content. Made it very tough to machine. In the lower picture, we show a planer there, which gives you a
Kind of an idea of different types of machines we had. It's small compared to some, but a very versatile machine. This looks like an EDM machine. Yeah, you got that right. That's Bob Roll up there on the EDM machine in the upper picture.
And that is electro-discharge machining. And out of the development of that type of work, they were able to do some pioneering work in the aircraft turbine business for them. And it's quite a unique piece of equipment. It erodes the material by spark discharge between generally a copper electrode.
the workpiece. On the lower picture we have a gun drilling machine and we had a requirement for that for deep hole drilling and it would do up to about three-quarters of an inch in diameter. We weren't making guns though. Okay. I've seen some of those EDM machines that could...
Drill a hole so small you can hardly see it. That's about right. Okay, here's a picture of a surface grinder. Yes, typical type of grinder that you might find in a real large shop, a big surface grinder. And that's just one of many. We had internal, external, and smaller surface grinders and honing machines. So we were well equipped for that type of work, too. There's the honing machine down at the bottom.
Now here comes a big one. This is a very unique machine. It's in a temperature-controlled room that is tightly controlled. I think about two to three degrees maybe. I can't recall for sure. And it has optical measuring systems on it. And it's a Swiss-made jig bore. It's a very unique machine.
And it's very high precision. It has circular measuring systems within, I'm trying to remember now, two seconds. And that is quite accurate. Very unique. In the same room, go ahead Bart, in that same temperature controlled room, we have a smaller jig bore.
and a precision lathe to do very high precision work that requires temperature control. And you see one of the workers, I can recall his name, Joe Finkbiner. Here? No, that's Dewey Stringer, I'm sorry. That's Dewey Stringer, yeah. And Joe Finkbiner is next, and then I don't know that third gentleman. That's Joe Finkbiner. Right. I don't remember his name. I don't recall him. We're getting old, Burke. You know.
OK, here's another. This was part of the solid moderator program. This shows one of the zirconium hydride tubes being machined. And the chips on that were very pyrophoric if they weren't controlled under water. And generally, you didn't see chips collect like this.
In any operation, they were kept collected and put under water to make sure we didn't have fire. And at each machine, we had buckets of graphite powder because that's the only thing that would extinguish the flame if it got started. And that was done in the general shops. And we did tons and tons of that stuff.
Ever have any big fires? Never had a fire. Have some little flames? No flames. Not that kind. Here's a very young John Mundy. Maybe you can comment on that and some of the other people with us. That was a bunch of eager beavers that collared me to collect their suggestion awards. Starting on the right is Les Patton.
Okay. And I don't know the next gentleman or don't remember him. And that's Pete Hosky and then Bill Bishop. And this guy, he's not there. Okay. What have we here, John? We have a sheet metal shop. It's not a very good photo, but it's showing one of the...
big punch presses, and you see some of the busybodies there. They had various equipment in the sheet metal shop. They had ring forming equipment and cylindrical, yes, okay. You can see the
cylinder rolling equipment. It's a center picture, a ring forming at the top. And they had various sizes in these types of machines. And on the bottom... Turret punch? Yeah, that's one of the turret punches they had, the Wiedemann turret punch. This looks like a very complicated piece. Yeah.
Test equipment. This is part of the CTF or core test facility. Insulation pads that went on the bottom plug and they were instrumented. They had to be shaped and sectioned and fit properly and we'll show you the picture of the part that they went on in a moment. That's showing the inside surface of those components and they were fabricated in the sheet metal shop and instrumented in the instrument shop.
By the way, the guy that does the welding on the thermocouples named Joe North, and he was about the only one that could do that consistently with a very good success rate. Joe was a very talented welder. Yeah, he had x-ray vision, I think, when he did his little needle-point things with his weld torch. He did.
This is the assembly of that unit, isn't it? Yeah. You can point out there the segments showing those pads. Those pads are down in here. They're all over that section. Yeah. And up at the top, the dome piece. This is the shield plug, right? Right. At the aft end of the reactor. Okay. Oh, the size of that? It was about...
Oh, I think that was about 48 inches or more in diameter. I can't recall specifically that body there and then the struts that are pipes that extend outward for cooling and instrumentation exit wires. Looks like some of the components that you made. Yeah, these were some of the development components that were made in some
led on to actual products and explosive forming. Are the dies, are these the dies? These are the dies. The two upper pieces there that look like halves fit inside that ring. And then this other ring that you see there fit down on the top. And the cylindrical piece there at the bottom was the starting form. The plug. Starting form. The beef explosion form. No, no, it was the starting cylinder that we started with.
And all those parts had to be in and sealed with beeswax. And that area around the dyes had to be evacuated. All the joints were sealed with beeswax. It had to be evacuated with a vacuum pump so that when the explosion took place, it did not compress air in those cavities where it was to form or it would burn through.
And that was one of the tricky parts of explosive forming. What was the explosion device? The explosive was primacord. And we were limited to 100 grams in this operation. Which way is this? This way? Let's see. No, turn it up the other way. This way. This way. Okay, this is the facility where we did the explosive forming.
And you see a charge blowing water completely up in the sky there. And it was a tank that was two-walled with sand between the two walls and filled with water. And the die was always set on a pedestal down near the bottom of it. And the charge was placed in there with wires going out the control room. And they were set off at that point.
That's what you see. This is the result, isn't it? This is the result. The result of the explosion for me. Right. One shot. Getting those rings around there. How precise could you make these things? Could you repeat? Oh, the actual result was very close to what the dye was.
and the repeatability was was very good oh good what was that particular piece do you remember i can't recall as they were looking at uh convoluted tubes for uh air airflow into the reactor and i think this was part of the development of that okay this is part of the welding capability
You see welders at work on a large fabrication, which is the torus in the upper part of the CTF facility. This is the large torus, and I think that was the exhaust part of it. And the diameter, it was in the neighborhood of, oh, I'm guessing now, somewhere about 30 feet or in that order.
I can't recall exactly the diameter of the whole donut shape. It's on the high bay. It's in the high bay on the iron construction floor there where everything could be bolted down. And that floor was used quite extensively throughout the fabrication and erection of the parts that we made in Evendale. Now, they're TIG welding? They are TIG welding. And by the way, all the welds had to be of X-ray quality.
And that was pretty difficult in many cases. And these welders were all fairly versatile in that area. Is there electron beam welding? Yes, we have, as I mentioned earlier, electron beam welding, which was done for some very, very small stuff and the canned stuff that I mentioned a while ago where you had to evacuate the containment.
and then weld it so that there was a vacuum inside. And that was done in that EB welding machine. You're telling me this is welding development? Yes, this is welding development. Jim Silphin is the operator there. Jim put together some equipment that we had bought and made an automatic
TIG welder, which wasn't available on the market at the time. He took components from a couple of machines and a wire feeder and combined them to do automatic TIG welding on a very piece of critical material that we were working with at the time. Back in the 50s, there were a lot of things that weren't available on the market that we had to...
Much innovation took place in that shop that people never knew about because of its classification, and there were not answers in the outside industry as a whole, because I was part of trying to find those answers. But they just were not available, and many times the people like grinding wheel manufacturers would say, you know more than we do.
We were out ahead of the game in many cases and didn't know it. I think there was a lot of spinoff from those programs, John. Okay. We had a clean room where all the components that had to be kept clean internally were assembled and welded in there. For instance, the core had to be taken into that room as components and sealed.
plastic and only opened up inside. The workers were required to wear nylon clothing and hats, gloves, and booties. And they had to change in a change room to get to go into that room. Through an airlock. Through an air chamber, yes. And we show Walter Goloski there.
In the top photo, assembling the shield plug, which is upside down, and down in the lower photo, we see a welder welding up one of the shield cans that went on top of the reactor, I mean, yes, the reactor, I'm sorry, and through the shield plug. That was in the clean room. How big was the clean room? The clean room was probably a good...
50 by 50 and it has a positive pressure air pressure in there so that outside dust did not come in this looks like part of a that's the lower part of the cocoon or ctf facility containment vessel is right and uh showing the uh exhaust parts or tubes there and
That whole piece was manufactured in the shop with the exception of the dished head on the bottom of the dome. That was purchased from an outside vendor, probably Ladish, I don't recall. But it's 304 stainless steel, and the sheet metal shop fabricated those tubes, and the wells were done with TIG, and again, X-ray quality.
The wells were about, if I recall, an inch and a quarter, or the wall was about an inch and a quarter thick in that lower area there. How big, what's the diameter of that? The diameter of that was in the neighborhood of, with the tube sticking out, probably 140 inches, something on that order. Wow. That's about 12 feet, huh? Yes.
Here's back in the high bay. Right. This is another picture of one of the high bay machines, the vertical boring mill. And that happens to be the instrumentation ring that was part of the CTF, where all the instrumentation fed through little tubes in the ring. And that's on the 120-inch boring mill.
boring mill that was 144 inches in diameter by 144 inches tall. So it was even bigger. We had a 25-ton crane in this area and a 30-ton crane in the other area. There weren't too many machines that big around, were there, John? No, there weren't. That 144, that was a special order, I think.
This is the shield plug in the high bay being worked on by two machines at one time, and it is not penetrating the inside of the shield plug, so therefore it can be out in an unclean area or a regular area. But you've got two machines working on that thing at one time, which is pretty unique. Yeah, that is. Well, here's the pizza oven. Yeah, here's the...
The biscuit oven, people, Bert calls it pizza. This oven was used, I might mention how big it is. It's 12 foot high, 12 foot wide, and 20 feet long. And there's a car on rails that comes out. Its capability is about 2,100 degrees Fahrenheit. And during open house for families one year, we put on a demonstration out there.
with Big Fred Bentz baking biscuits in that furnace, and he had a table set up with jams and jellies and butter so that you could enjoy them. That was quite unique. Looks like a... This is a core insert, solid moderator version of it.
And it shows that the assembly in the high bay and the instrumentation that took place out there. You see the solid moderator at the bottom there. I think Otto Worke had something to do with this thing. This is the instrumentation part above the moderator? Right. And last but not least, we couldn't do without inspection.
So we showed Mr. Hoxtel up there with his 30-inch comparator. That's Lee, wasn't it? Lee Hoxtel. Lee Hoxtel. Yeah. And he's still around. And in the lower picture, it shows the general area. And that was in a temperature-controlled room also. And it had all the necessary gauge blocks and that type of measuring instrument to support the work that we did in the shops.
John, here's a bunch of guys. Oh, my. Some of whom I still remember, so I can't. But I'm sure. Well, there's Jerry Moore in the center. I'm sure a lot of people remember him, the guy, the older gentleman. Jerry Moore. And then. There's Buck Jordan. Buck Jordan. There's John Mundy. Oh, my. John Tennefeld. Jake Keeney. Jake Keeney.
I can't remember his name, Walt. Isn't that Walt? I can't tell from here. Anyhow, hopefully some of them see their picture and remember it. Okay, this is first. This is first. Right. Okay. This is a picture of the shield plug and core.
as you come down in the transition area. There's a shield plug at the top, and then you have the transition area there. Coming from there to there, and that's where the air entered in between those two. You see the control rods coming down through there, and then the active core.
And it shows how the fuel elements would be loaded. Would be loaded up from the bottom. From the bottom. And that's in the high bay where it was assembled before shipment. And then we have a picture on the back. On the back, if you can zero in on that, you can see the torus, the small torus, and the large torus, and the shield plug and reactor assembly in the background there.
Where they're assembling that at Idaho. This is at Idaho? This is at Idaho. At Idaho, okay. But all these components you made back here in Evendale and Shetown. Quite a task. Well, John, I'll tell you, there was a lot of good machinists and manufacturing people in that organization that did a tremendous job.
Good to have you here today. Thank you. And you'll be part of the bigger picture. I hope. Thanks a lot. John, let's put up on the screen the list of employees in Manufacturing Shop. We'll quickly run through them, and we'll just show them. Here's the Manufacturing Shop supervision.
Here's you. Here's the employees of machine shop number one.
Here's the employees of machine shop number two. Here are the employees of machine shop number three.
Here's machine shop number four. John, I know I know most of these people, but I can't put the faces with the names anymore. Neither can I any longer. Here's machine shop number five.
And here's the sheet metal shop. John Carrier ran that, didn't he, John? John Fleming, John Carrier, and Loren Hudson. The difference between some of these numbers is that some are night shift and some are day shift. Okay. Like one and two are. Oh, this is another shift on shop number two. Yeah, that's correct.
OK. Here's the well shop.
This is shop number two again. That's the well shop. Shop number two of the well shop. Right. Second shift. Second shift. Okay. And finally, here's the special metal shop.
I might mention, Bert, that there were probably three or four other people in there, but I had no listing on them. Okay. Where's Elmer Burns? I didn't see Elmer's name in there. Did we miss him? He would have been in the... Special Meadows. Okay. Quite a few people in that organization. Right. And they reported to... Gus News. Gus News. Yeah, at one point. And Ida Luther-Yose. Right. Well, thanks again, John.
You're quite welcome. I enjoyed it. I really appreciate it. Thank you very much. Thank you.
6.3 Health Physics
In the materials group at the A&P, we worked with many toxic materials, and we also worked with very hazard materials. Not only did we do it in the laboratories, but it was also carried out in the machine shops that were protected, and they operated under the conditions for hazard materials. We also had a hot shop, and that was also one of the problems we had of watching materials.
Now, to keep everybody safe in the lab, in the neighborhood, we had a group called the Health Physics and... Industrial Hygiene. And what was the other name? Health Physics and Industrial Hygiene. Industrial Hygiene. Heading up that, starting in 19... You joined the group in what year? In 1952. 1952, and you worked there until you went to the A&P department, not the other department. The Jet Engine Group. The Jet Engine Group. So when did you go there? 1979.
Who was speaking there is Ralph Hopper, and we have Don Winker over there. When did you join the group, Don? I guess I joined about 61. 56. Oh, 56. 51. The thing ended there. 56. Yeah, 56. And then you worked there until how long? 30 years. And you and Ralph worked together. Could you give me also some names of the other individuals that worked with you? Let's talk this off.
I think as a group we had at one time 36 people in our group. Gordon Boyle was a health physicist. Hard Perry was an industrial hygienist. Harry Coley had charge of the laboratory operations and also community monitoring. We had a truck that we went out into the neighborhood and took air samples out in the neighborhood just to show negative data. Then there was Larry Brock who had education and training.
Pat Adams, who had the monitoring people in the laboratory, Ray Stewart, who had the safety group, and John Kale, who was one of our specialists in radiation protection, who worked for us. But let me correct you once, Earl. We didn't work with any radioactive material out in the machine shop. It was only material... No, I know that. Yeah. The radioactive material was always handled in the special metal shop, which was the L-4-5 block and so on.
And I might mention that when the... Let me give you a little history about Building D. It was built in 1943, and it was the last building built in the Evendale complex. It was actually built as an aluminum foundry. And then in 1945, the day the war ended, everybody was laid off, all the equipment in the building were mothballed.
until 1951 and at that period is when the amp contract was signed and we took over the building from 1951 to 52 and a half the building was completely gutted and as a result of that the design was made we knew we were going to be working with radioactive material so a lot of effort and thought was put in how to design the building to safely handle radioactive material
Actually, the building was broken down into about four areas, five areas, the office area, the engineering areas, the laboratory area, and the shop area, a machine shop area. The laboratory area got the most effort put into the design, mainly because we knew what we were working with, we knew we were in an urban area, and we knew that if we had a serious accident,
it would not only injure the people, our people working in the building, but would also get into the neighborhoods and maybe the jet engine plant and shut them down. So a lot of effort and design was put into the proper facilities to handle radioactive material. The lab was completely isolated in that it had its own air conditioning system, it had its own controlled exhaust system, which were
All the exhausts going out were filtered through HEPA filters. HEPA stands for High Efficiency Particulate Aerosol Filter. And we also had a controlled liquid waste drain system that any water from the laboratory area went into holding tanks in the basement. And these tanks were never emptied until we took samples, analyzed it, and approved the disposal of that water. Any known contaminated water
was put into 55 gallon drones and these were shipped out for disposal same way with any known radioactive waste and so on this was all handled special and shipped out and buried in approved burial sites that pretty well covers uh that's far more than i knew about that ralph but i would like to ask you one question what did you do with all those urine samples we used to give you we sent those to the laboratory to be analyzed and i might say that
All our air samples, all our urine samples, and I don't know of any overexposures to radiation. I don't know of anybody that suffered any radiation damage, but we did have a fairly tight program, and we've got a book of procedures that we followed. Don, have you got that book?
Yeah, I remember a lot of problems you used to give us on the things we wanted to do. This is just our procedures on safe handling of radioactive material and the very toxic materials and so on. But we followed that to a T, and all the laboratory managers and so on cooperated in such a manner that we never had any problems along those lines. Okay. Now, you want to pick up this information on the board here? One of the key things that we had to do was educate.
the people who, not necessarily the engineers, because a lot of them had a technical background. They understood radioactivity. But we had a lot of clerical help, and we had a lot of lab technicians that didn't have the education and training to understand radiation. And our biggest job, we had two fellows in a group, and their assignment was education and training.
You might turn that one around, and we'll talk about, you know, the biggest thing when you mention the word radiation is just it connotes a psychological fear, mainly because you can't see it, you can't smell it, you can't hear it, and you can't taste it. And as a result, the people who don't understand it, they're just very fearful of it. It's only when you explain all these situations,
The activity of how radioactivity works and so on and the damage that it can do and how you protect yourself is when they start feeling at ease to work with the material now The other thing the the general public they don't understand stay in radioactivity either and Then the mass media doesn't help the situation For example the Three Mile Island incident. It was a serious reactor accident, but the
The mass media, they really blew that out of proportion, and they really scared everybody about radioactivity. And that fear is still existing today, 20 years afterward. People are just scared of reactors, and reactors can be handled very safely. And we're going to have to use them one of these days. That's right. Okay, then we also would talk about what radioactivity is.
You want to move to another flip the back flip the back now All right, what radioactivity is and it's actually a process in which the certain nuclei and undergo spontaneous Disintegrations Not all materials are radioactive. It's just a very small percentage of materials are radioactive. There's very few natural occurring radioactivity I think there's about seven or eight materials that are naturally occurring radioactive
In fact, we have a radioactive isotope in our body in potassium 40. Every time we eat a banana or every time we take a potassium pill or something like that, we actually are ingesting radioactivity. And the body contains a fairly large amount of potassium. It's more important than sodium in the body. And you know doctors like to limit the amount of sodium for people who have high blood pressure and so on.
We get about 20 millirankins of exposure per year from the potassium-40 in our bodies. So we're also exposed to other natural activities, the cosmic rays that come from outer space. Nobody knows what causes them and so on, but we also get about 40 milligrams per year of radiation from those. And then we have the natural occurring materials in the Earth.
the trace amounts of uranium. As uranium decays, it has a very large progeny, a lot of what we call daughter products. It actually decays down into radium, and the first item coming off of radium is radon, and you hear a lot about radon these days and so on. But this is natural occurring. Every day when you go outside, you're breathing radon gas.
In a year's time, you would get an exposure about 30 milligrams to the radon gas. Those are essentially the natural occurring isotopes of radioactivity. But then we're talking about material that we work with. And we mainly work with material called uranium-235, which is enriched uranium. It's one of the isotopes of natural occurring uranium. And one of the disintegrations coming off of uranium
235 is alpha radiation. I should say alpha 234 comes off of uranium 234, but they can't separate them when they when they run them through the diffusion chambers to separate the various isotopes of uranium, but Don Don's got a demonstration for alpha radiation an Alpha particle is actually a helium atom has two protons and two neutrons in it and
So since it has two protons, that's two positive charges. And it's looking to neutralize these two positive charges. And it will pick up electrons in the air. So an alpha particle will only travel about two to five centimeters in the air.
a piece of paper uh we'll take a piece of paper in between there now yeah put a piece of paper and a piece of paper will shield the alpha but how does it shield it because it strips two electrons off of the paper you can uh off all the alpha you want on your skin and so on no damage to the skin it just picks up two electrons off of the skin but there's no radiation damage whatsoever the big problem with alpha radiation is if you get it internally there's where it does
20 times more damage than beta or gamma radiation and also x-rays because of the ionization effect it does to large molecules let me give you an example of what it does to the the human body what radiation does to the human body when it causes ionization of the other long molecules in the body this was a piece of plexiglass pure plexiglass that we
drilled a little hole in and we had a two curie cobalt source that we calibrated our radiation detection instruments with and it would sit in our calibration well we would raise the weld at different levels if we wanted something reading 100 milligram per hour we would raise it to a certain level if we wanted to read 10 r per hour we'd raise it to a higher level and so on so we calibrated our instruments in that manner i just wanted to show you that what the radiation had done to that same piece of plexiglass
This was a very similar source boat, but had years of exposure to the radiation. I don't know if you can get in very close and actually see that damage that's done. I might ask the question, did this become radioactive?
because it had a source in it and it's done that? And the answer to that is no. The only thing that can make material radioactive is active. If you inject a neutron into the nucleus and you disturb the balance of that nucleus, then it becomes radioactive. But I can touch this. Actually, it pulverized this material to a powder and so on. But it's not radioactive because it was strictly exposed to gamma radiation.
And this is the same effect in the human body. Here is a source boat that was exposed for two years. And you can see sort of the orange, yellowish area in the center. And the damage done is a lot worse than the damage done to this one, to a year's exposure. But if you were exposed to the gamma radiation from...
to your body and so on. This is the same type of damage that takes place. It actually ionizes or breaks down that molecule to where it becomes a toxic material. A very small amount of exposure to radiation isn't going to make any difference because the little damage that would be done in a short period of time would be repaired by the body. The body's constantly repairing itself and so on.
So what you're saying is that don't worry too much about x-rays when you go to the doctor. Because they're limited. If you get a high exposure of x-rays, then it can do a lot of damage. And a lot of times in treating cancer, they'll treat it with high doses of x-rays and so on. They're actually killing the cancer tissues and so on. But that's where they use a circular thing.
And pinpoint, but they don't damage the whole body, but they are damaging. They try to shoot at different days, they shoot at different angles. That's right. Okay, what do you want to pick up next, Ralph? Okay, so we talked a little about the alpha particle. Let's talk about the beta. The beta particle. Both the alpha and the beta particle come from the nucleus of the atom. When a neutron splits off, splits, it gives off a proton and a beta particle.
and the beta particle comes flying out with a lot of energy. There's usually gamma involved there. The gamma is the binding energy that held that nucleus together, so there's usually gamma energy coming off. The proton then stays in the atom, and it increases the atomic number of that atom by one. So if you had hydrogen and you had an hydrogen atom that split, it would become helium and so on. And if you had uranium and you...
you had a neutron split it would become neptunian one higher in the atomic number all right now we're going to back into another area well we talked about that now let's talk about some of the areas that uh a lot of work was done and uh you people played a very important part to be sure that no one uh received over exposure by the way while you're at this uh you had film badges did you not right here's an example of a film badge
A film badge is only good for measuring beta and gamma radiation. And the reason for that is the alpha, a piece of paper would shield the alpha. So the photographic film that's in this film badge, if I can get it open here, the photographic film that's in this film badge has got a paper coating on it, so it certainly wouldn't pick up the alpha radiation. Here's a picture, a piece of the film that we pull out of the...
of the film badge. Let me explain a little more now. I just want to show you the innards of this film badge. Actually, it has an open space. You can determine the energy of the radiation that the film badge is seeing because it has an open window. Can you see the open window? And then it has a
piece of plastic the same plastic as a film badge then it has a piece of cadmium and a piece of indium foil and they they stop the radiation to a point absorb the radiation to where you can determine what the energy level of the radiation is for beta and gamma radiation okay i think that pretty well covers that pretty well covers that yes uh yeah well now we can move on into here
where you monitored this operation. None of the work that we've done so far in these presentations, we have discussed the hot cells. Right. Let me talk about the hot cells. Most of the work done in the laboratory was done with powdered uranium oxide metals and so on, uranium oxides.
After we made fuel elements and so on, we would send the fuel elements to Oak Ridge to irradiate them to see how they would withstand the neutron fluxes and the heat from a reactor. And then they would ship pieces back to the Evendale, and they would be put into the hot cell. This is the front end of the hot cell where everything's done remote. These are three foot thick, high density concrete walls, and they have a view grid.
view glass made out of leaded glass that's also three three foot thick so everything is done by two remote handlers using just these two fingers so anything that you can do with two fingers you could do in a in a hot cell here is the back end of the hot cell let me show you the back end again it has three doors on it there's uh uh one two and three
One was the area, cell one was the area where we loaded up our hot cell material. The other two cells, we had, we polished the, we prepared metallography samples and then we polished them in there. Here is another picture of the back end of the hot cell where a cast coming up from Oak Ridge is loaded.
Back of the door. That's the back of the door, and it gets rolled into the hot cell. That's Stu Layton that's unloading that, and here's another picture of Stu, and we're sort of dedicating this to Stu because Stu died of a heart attack about four or five years ago, and he did all the work back in the hot cells, the main work in the hot cells. A very creative individual.
Yeah, he did design a lot of the equipment I know. It was used to help out because it was hard to handle a lot of the things. Right, right. You couldn't handle the activity in this material. Within maybe 10 minutes, you'd be dead from radioactivity. That's how hot some of this material was. And that's why you had to use the hot cells to handle the material. Now we're going to show you the best dressed man.
in the AMP program? I was in my younger days. But apparently there, what I'm monitoring is the cell before they put something in, make sure that it's clear. We always check them out before and after just to make it safe for people to be able to work in there and load them up. There's always a lot of monitoring to do.
On the other side of the operating site, we took air samples just to make sure there was nothing that could come into the air. And, of course, we took them on the hot side, too. But also for surface contamination, you know, this radioactivity, if it's a powder or dust and so on, if it gets on the floor, you can track it around. And as a control that we didn't expose anybody unnecessarily.
We were constantly taking smears on the floor to determine if any material was being tracked out. We also did this in the labs with the powdered materials also. Did I mention before that the labs were specially designed that they were under negative differential pressure with the rest of the building? That meant all the air from the rest of the building was being sucked towards the lab. So if we did have an accident to occur, that it wouldn't get away into the rest of the building.
the individual labs themselves were under negative differential pressure with respect to the hallways that they were in. So all the air was flowing from the rest of the building to the hallways to the individual labs. Then inside the labs, we would have hoods and dry boxes. They were under negative differential pressure to protect the labs themselves. And we had a very, very elaborate air conditioning control system.
that if anyone left the door open for more than 20 seconds, it would measure the differential pressure in the room and automatically shut down the air supply to maintain a differential pressure, a negative differential pressure. So these facilities were well thought out. And as I say, I don't know of anybody that got an overexposure to radioactive material. I bet I'd never heard of anyone being sick on that.
Not only did you have to take care of some of the radio activity, you also took care of us in case we had a fire, right? Right. We had a volunteer fire department. It was headed up by Ray Stewart. And there's a picture of our volunteer firemen. Twice a year we would have exercises, and we would plan on different accidents happening and so on.
Okay, now we'll move on. I think we could end it up by Don and you demonstrating the use of Geiger counters and other body counters that you may have here. This is about the most sensitive instrument that we have. This also has a speaker on it that you can hear every burp. So as you move a source close to it,
So that was one of the methods of protection, of human body protection from radiation, is distance. If the radiation falls off inversely, it's the square of the distance. So the farther you get away from the source, the less radiation you would receive. Other protective methods were time. If you had a job to do that's fairly high radioactivity, you would try to do it in a hurry. First of all, you would practice.
so you wouldn't make any mistakes dropping material or spilling material and so on if it was very hot. Then you would practice ahead of time that you could get the work done properly and so on. And then you'd get in there and do it with a minimum of time so you kept your radius and levels down. The other method was shielding. We have a piece of lead. Turn yours on.
This will create like a static. But watch when I shield it. Well, that's a... You sure that's lead, Ralph? That's lead, but that's an awful thin piece of lead. We didn't practice. We used to have that lead brick there. I also brought along some uranium ores. Here's one that's fairly hot.
Let me just show you. This probably has enough uranium in that it would be cost effective to refine the uranium. Also got some very low level amounts.
It wouldn't be very cost-effective to mine this grade and to recover it, but this is a fairly high-level one. Also a third one that we have. It's a little higher than this one, though. So those are different grades of redactivity. Also... Yeah, I think that's interesting. That dish there represents the technology that we had.
A lot of times they're using frits before the 1940s. Is that true there? Yeah, that's right. And a lot of people had those in their house and didn't realize. They thought it was pretty, but they did not realize the problem that it might have with it. Very beautiful colors. It's called fiesta ware. And if you had a serving of 12 and had these all stacked up with the saucers and the bread plates and the dinner plates and so on, you could actually get about 150 millirenken per hour.
off of these. They've since been banned by the public health service. About 1950, they banned these materials. They're still around. They are still around, but people don't realize the hazard associated with them. But that's only in the red and sort of light yellow colors. Yeah, there's a yellow. This is an orange, yeah. This is uranium salt, and the yellow is a different uranium salt.
In fact, most yellow brick in homes contain a low-level amount of uranium. It's the uranium salt. Okay, Earl, that's all. Well, we've learned an awful lot about what you did down there. I often wondered what was going on. And I thank both of you and Don for coming in today and adding this portion of your presentation to our general presentation we'll have. Okay, very good. Thanks a lot.
7.1 Critical Experiments
Hi, my name is Bob Evans.
And I'm here with Jack Simpson today to continue to review some of the activity that went on in the ANP program, aircraft nuclear propulsion, between the years of 1951 and 1961 when General Electric was in control of the program. Our specific objective today is to review a little bit about reactor physics and what it means in terms of the ANP program.
and also to review what went on at Evendale in terms of critical experiments. And Jack will explain what critical experiments are a little later on. As far as when we talk about a reactor, a nuclear reactor, as concerned with aircraft nuclear propulsion, we're talking about a reactor core that's primarily composed of uranium-235, highly enriched.
Highly enriched, in our case, is about 93.4% U-235 with probably the rest in U-238. We're talking about a reactor that's made up of this fuel, and we're talking about a reactor that's made up of a moderator, a moderator being necessary to slow down the fast neutrons that are born at fission in order that they may be absorbed by the U-235.
We're talking about a reactor that contains a reflector, a reflector being necessary to reflect those little neutrons that are trying to escape the core back into the core. We're talking about a reactor that has a control system, a system that allows us to control the level of power in the reactor by controlling the neutron flux and also to provide a means of shutting down the reactor should the reactor become uncontrollable or for some other reason by way of using safety rods.
These rods, in our case and most of ANP cases, were boron. And boron, of course, is a high absorber of neutrons, especially thermal neutrons. We're talking about a reactor that has to have its fuel elements cooled. When the reactor is fissioning at high rates, it's producing a lot of heat. This heat has to be removed or the reactor will melt down.
In the case of AMP reactors, we're talking about air cooling. And this air cooling is provided by a jet engine. And of course, we're talking about sensors that we use to monitor the level of neutron flux and also to tell us what the period of the reactor is and other things concerning temperature, et cetera. What does a reactor do? OK.
That's relatively easy to explain. U-235 fissions, and it fissions by capturing a thermal neutron. Thermal neutron in the process of this capture splits the nucleus of U-235 into two parts. These parts have relatively high kinetic energy. This energy is absorbed in the fuel element proper and is translated into heat.
The moderator is essential in order to slow the neutron down. Of course, the reflector is trying to keep the neutrons from escaping, and the airflow cools the reactor fuel elements. Critical experiment is an experiment that's set up to try to simulate an actual reactor design with similar materials and similar or exactly the same fuel.
And the idea here is to try to produce the power distribution and the flux levels that you'd expect in a full-size reactor, but at a much lower power. We're talking of powers levels of 1 to 10 watts, whereas our full-size nuclear reactors are probably going to run at anywhere from 20 megawatts up to 100 megawatts, depending on the design.
In a critical experiment, we want to provide means of changing these configurations. If the reactor designer hasn't designed things exactly correctly, we want to be able to manipulate those configurations in terms of fuel and moderator and structural materials in order to see what the effect on the power distribution and power generation is. Of course, we want to control the power level with our control rods, neutron flux being
a significant factor. We want to measure the neutron flux in the reactor in order to determine exactly what the power level is. Of course, we want to measure the flux distribution to find out where the power is being generated. The purpose of all this, of course, is to verify design calculations to make sure that when we go and actually build a reactor and put all the time and effort and money into doing this, it's going to behave the way we expect it to behave.
And also, when we operate our full-size reactor, we're going to have to know how effective the control rods will be, and the critical experiment will allow us to do that. And of course, it'll allow us to measure how effective the reactor is in terms of maintaining its capability of maintaining its power level over a period of time. And this is determined by measuring the reactivity of the system.
the so-called parameter which is called K effective, effective reactivity. If that's equal to 1, the reactor is just critical. And any values greater than 1 means that the reactor has the capability of being put on a positive period and have its power level increased. We talked about reflectors. What does a reflector really do? This chart gives you an idea of
what's going on. If we have a core that's bare here without any reflector, these are reflector sections here, usually beryllium or certainly beryllium in all our cases, but they possibly could be other things. A and P only use beryllium. If you don't have this reflector, you can see what the flux distribution looks like. And you can imagine that if you had a fuel element in the center of the core, you're going to get a very high power generation. If you had a fuel element out here, you get a low power generation. And that's not too good.
for trying to get a uniform air temperature out of the rear of the reactor. By putting a reflector in, you allow the neutrons that are trying to get out or reflected back. This skews up the outer part of the flux curve and, of course, gives you a much better distribution of flux across the reactor and a much flatter power distribution. We can then further compensate for the fact this is still skewed by adding, say, more fuel out in this region of the reactor to get the power level up.
We talk about reactivity in terms of what it does to the system and what it is in relation to addition of uranium-235. In general, if you compare a reactor with U-235 in it to a reactor that's made up essentially of U-238, which, of course, some of the big early piles were out at Hanford used to produce plutonium.
The Hanford piles were gigantic because they required the use of 238, which has almost no cross-section for fission. But the residual U-235 that's natural in U-238 did most of the fissioning. Since there's so little of it, it required a great big reactor to get it to fission. In terms of the effective multiplication, the capability of the reactor to generate more neutrons than it absorbs,
We look at a curve like this, which was actually a curve for a HTRE 3 reactor. We're talking about plotting pounds of uranium installed versus the effective reactivity. And the point being here, we're shooting for a design characteristic. Of course, we want it to be greater than 1 because we want the reactor to be critical and we want it to be greater than 1 to be able to generate a positive period.
The characteristic of the HTRE 3 was a multiplication factor of 0.03. This is a lot of multiplication. The desire here is to be able to override things like xenon and have the capability of operating the reactor for long periods of time. Xenon, of course, is a poison that builds up after the reactor is shut down. Jack will probably talk a little bit about this later.
When we talk about reactivity periods in relation to effective multiplication, if we're just at or near an effective multiplication of one, the reactor period, though it's possible to get a reactor period, is going to be very, very slow. And of course, as the reactor, as the effective multiplication is increased, we can get a much, much higher period. It's important to control that period so that the reactor
does not go critical at a faster rate than we want it to. We talked about cross-sections a few minutes ago. I'll give you an idea of what a cross-section means in relation from one material to the next. We talk uranium. These are all of the odd materials or fissionable materials. The even one, of course, U238, is relatively non-fissionable. And you can see why. The fissionable materials have cross-sections of over 500 barns.
particularly plutonium, whereas the uranium has a cross-section of very low down around 10 barns. You can't imagine you're going to fish in uranium with fast or thermal neutrons. Poisons, the idea being to absorb, as we mentioned, the neutrons that we want to control, there are three very prominent materials that are available to use as poisons, boron, cadmium, and europium. Some of these are very rare.
Most of the work we did was with boron, which has a cross-section for absorption of 700. But you can see that europium would be a lot better up around 4,000. Up to this point, we've just talked a little bit about reactor physics. We're now at the point where the EMP program has been brought to Evendale, and we're contemplating doing critical experiments in Evendale facilities.
And Jack, I know you were involved in this program practically from the start. Right. When we initially started in 1952, basically trying to create a reactor design, you've got to remember that in those days, computational power, as we know it today, just did not exist. Today we use...
neutron calculations, multi-hundreds of groups. In those days, a big deal was a two group, two energy groups of neutrons, calculations solved by banks of people standing around pushing or sitting around pushing mechanical calculators. That meant that the entire theoretical basis for reactor design was pretty shaky.
So as a consequence, much of the design had to be done empirically or experimentally, which is where the critical experiment function came in. The recognition was made that when we moved to Evendale, Ohio, we would have to have that facility. And correspondingly, the back end of Building D, which Bob is pointing out there, which consisted of a bunch of sand bins for a foundry which had been in place during World War II,
was converted. Part of the conversion area involved making or creating two cells in which we would install critical experiment rigs. Critical experiment rig is a device that we're going to show a few pictures of in a little while that actually contains the uranium fuel that Bob's been talking about and all the other goodies. Here's a layout of the facility in building D. There was two cells in it.
the cell one and that cell room seven. I don't know if you can read the lettering, but it's not important. It's basically the cell which is colored in the lower view. And immediately adjacent to the cells were the places where the fuel arms were put together. And also immediately below there, when you put it on area six, that was a control room from which the reactor
function was controlled in the case of water tanks we would control the pumps and rods from there in case of moving tables we control the table all and viewing the process for the most part was a closed circuit tv remember this is the middle 50s so and all the rest of the facility you see there in the chart was basically supported provided to support the activities trying to measure the power distribution within the reactor again
Computational facilities in those days were extremely limited. Only during the last two years of the A&P program did we have a computer anywhere close to what we now today recognize as a big-time computer. And so we had this facility where we constructed the reactors, exposed the fuel to
the measuring devices in order to determine the power distribution. In 1953, as I mentioned, the program, the entire A&P program was redirected. And one of the first things we needed to do was to embark on a proof of principle set of experiments, so-called heat transfer reactor experiments, which Bob has beside him there showing.
The concept basically consisted of a large tank of water perforated by 37 tubes through which air would flow. In each of these tubes, a fuel element would be located. And control rods would then be inserted in the interstices between the air coolant tubes. These control rods, as Bob mentioned, consisted of boron, boron carbide, boron oxide.
used to adjust the multiplication of the reactor, or the k. So we had to, next step was to come up with the concept of what an experimental fuel would look like, which is shown here. It's a multistage one. This isn't actually the first one, but it's the second one. And it consists of a series of 10, I believe it's 10, maybe it's more, maybe it's 15 on the HTRE, 18 on the HTRE 3.
Horizontal stages in the horizontal flow through a direction. Air went from left to right in this model. And each of those segments consisted of a series of concentric rings. These rings were fabricated out of material, which we talked about a few minutes ago, which was a nichrome clad, nichrome uranium oxide meat.
The design of the reactor was such that it could be represented, the geometry could be represented very simply in a tank of water. It so happens that at Oak Ridge there was a facility created during the early NEPA days which had a water tank in one of the cells. Arrangements were made to borrow the tank for a couple of months,
mock-up of the reactor was constructed and installed in that tank. You see here on the chart that picture of that mock-up. The 37 tubes simulated there and around it is the tank in which the water would be placed. In August, latter part of August 1954, the same facility was recreated in Cincinnati with different kinds of fuel elements.
representations, a beryllium reflector from the very beginning. Can you point it out here, Bob? Yeah. And the rods you see sticking down in the interstices there are the control rods, which are moved up and down by either motor-driven assemblies or by spring force in the case of a rapid shutdown mechanism, otherwise called scramming.
At the bottom of the tank was a 12-inch valve. It's hidden there, but when the scram function was performed, that dump valve opened and all the water ran out of the tank into the basement corridor immediately below the cell. Well, this water took a long time to prepare because of the purity required.
We basically would shut the reactor down by just scramming the rods and not dumping the water. So a lot of experiments were performed. That facility operated for about a year and a half. Jack, you remember that little flapper valve device we had? There's a big story associated with that. Oh, Lordy, yeah. One of the things we needed to do...
in order to get permission to operate a reactor in Cincinnati was to analyze all possible accidents that could befall us in operating this thing. This is so-called safety report or hazard report, whoever. You're a half-full guy or a half-empty guy. And one of the things we had to evaluate was what would happen if the...
The pump proceeded to pump the water too fast into the tank, and it overflowed into the top of the tubes and got into the inside of the tube. So if you see if that water got too high, went up over the top here, and got down inside the tube, what would happen? Well, what would happen was not nice. So we had to provide a way of abetting that. The way was determined, OK, let's put a standpipe outside the tank.
and let the standpipe be the safety valve. Well, we did, and we then failed the various water pump rate protection devices to see whether water would stop. Well, it didn't stop soon enough with the initial height on that standpipe, so we cut off an inch. Now, this was a 12-inch stainless steel tube, one inch thick, and we cut it off.
cut it off an inch at a time. I was sitting in the cell with an Air Force guy, and we'd say, okay, we'll cut to here, and somebody would come in and cut it. We'd try it again. Still wasn't right. This was done with a standard hacksaw, too, if I remember correctly. You got it. I wasn't running the hacksaw, but it was cut with a hacksaw. So finally, after a few days, we finally got it down to the level that was considered acceptable, and under these
simulated accident conditions, where the low pump kicker in failed, and we kept going with the high rate pump. So we got rid of that. And then we proceeded to load the reactor incrementally. And you're seeing here a cross section of the first reactor, the HTRE one.
37 tubes arranged in the pattern that's quite apparent here. The interstitial regions, some of them had absorber rods of boron. And one of the places there, right there, was a tube which allowed us to start the chain reaction by inserting a source. In our case, we were using a source made of polonium beryllium. Polonium is an alpha emitter.
striking beryllium creates neutrons. The neutrons were then used to initiate the chain reaction and provide a minimum level for our instruments because we needed to know exactly what the neutron level was. So as we added fuel, uranium around there, that apparent level of neutrons kept increasing and increasing or multiplying, surprisingly enough.
And that multiplication factor we referred to earlier, which was so abysmally low in the initial Oak Ridge experiment, basically tracked much better. And we achieved criticality with roughly 33 tubes loaded with fuel. Bob has got to hear now a representation. Let's put that chart up. It's gotten out of... Okay. That's the chart you were talking about, I think. Right. So, yeah, as we loaded...
fuel elements, I should say, the inverse of the multiplication, which is handier to plot than the actual multiplication itself, since that gives us a scale of 1 to 0 instead of 0 to infinity, is shown here. After loading the initial seven tubes, the center tube and the six surrounding it, we had a multiplication of about, let's see,
0.85 was about 1.2. And we used that as a basis. Then we continued on. Then we actually got a multiplication of about 3 and so forth down the line. Each time you measure, how many neutrons am I measuring at a given detector, or what's a neutron flux, compared to what it was without any fuel? What it was now was what it was with just the source. It's called a multiplication.
plotting the reciprocal of that multiplication gives us this curve, since we know that when the reactor will be self-sustaining, when the neutron flux is there and the artificial external source has been removed, which is represented by this condition here. So we march down the curve looking like that, adding each of those large black dots represents a fuel loading increment.
And our desire is to somewhere get down here and be able to extrapolate to the point where we think it's going to go critical. Right. Now, the chart has been simplified somewhat because there are actually two multiplication curves generated as we go along. One with the control rods totally withdrawn and one with the control rods totally inserted. So that as we go down, we will make sure that when we get to the point where we're critical with the rods out,
We are definitely subcritical with the rods in. It's a most desirable condition. You might add that a lot of this procedure was developed by Enrico Fermi during the first critical experiment, which was conducted at Chicago. On my birthday. On your birthday. Seriously. Was it? Yeah, absolutely. Were you there? No. It was a big stack of graphite and uranium and uranium metal.
with the control rods in, the stack was about 25 feet in diameter and about 20 feet high. But it was natural uranium, and it took a lot of uranium to get it to go critical. But they followed basically the same procedure, and those procedures essentially went down through time to the point where we did almost the same thing. Yeah, without getting into all that satirics thing that you...
It's a factor of you don't want to more than double the multiplication on any successive step. All right. It's also a factor that it's going to tell you pretty fast if you're not doing things right. Because if that curve starts diving in someplace with only a few tubes in the reactor, you know you probably haven't gotten the loading right somewhere along the way. Or you've got some bad instruments. Right. OK, here we see the
representation that was used in the critical experiment facility for the fuel element loading and for the purpose of power measurement, uranium fission power measurement. Now, I've got it over here. Now, you're looking at an assembly of six concentric tubes. These tubes are made out of 10 mil thick nichrome,
10 stages long or high, which corresponded, in our case, its reactor was on its nose, to the airflow direction, although no airflow was obviously used at these low power levels. Around each of these cylinders, we would wrap additional nichrome, thin sheets, 1 to 10 mils in thickness, and uranium, and fasten the uranium and metal.
on there with standard straps used to strap insulation on pipes turned out to be a very economical way of doing it and after each cylinder was loaded with its assembly of uranium foil or fuel elements and nichrome the six were assembled on a supporting cross member down there and then the entire assembly was inserted into one of the 37 tubes we referred to earlier
Now, in order to measure, as we talked before, one of the things we needed to know is how does the fission rate distribution vary as you move through the reactor. From a thermal design viewpoint, you want that fission rate distribution to be as uniform as possible, since that would represent a, that gets you closer to the ideal thermal efficiency of your heat generating machine.
So that became our paramount measurement need. Now, the way we would do it is you put bare metallic uranium in place on these fuel elements here. And then we would place half-inch squares of aluminum, something like this.
In intimate contact with the bare uranium metal. Literally, it looked just like that. Half-inch squares on tape and wrapped around the fuel element. Then the reactor, that foil impregnated fuel element was loaded in the reactor. The reactor was run for approximately 20 minutes.
It was very precise timing needed. And the reactor shut down the fuel and was quickly removed from the reactor. These foils stripped off and placed in a counter. Now, what had happened while they were in there, these half-inch square foils had captured fission fragments due to being the intimate contact between these aluminum foils and uranium foil.
These fission fragments were then decaying. Now, so remarkably enough, we call them catcher foils. They caught the fission products, fission fragments, resulting from the fission. The radioactivity measured in those foils was then a direct measure of the fission rate activity occurring in that place, spot immediately below the foils. Therefore, by measuring these foils throughout the reactor,
all other conditions being equal power level time and so forth the relative radioactivity of each foil would then be a measure of the uranium power distribution generated throughout the reactor with that our thermodynamics folks and our metallurgists could then determine okay in this part of the reactor we need to have so much uranium in in this particular fuel element
at this particular location. These specifications were then given to the manufacturing folks, and they would proceed to fabricate the high temperature materials using the specifications which we had developed this way. And again, I want to reemphasize that we were limited in the calculation of capabilities. That's why it requires such extreme measures to get good power distribution.
measurements in the reactor. We initially, we then embarked, since this became quite a problem in terms of controlling human exposure to the radiation, you couldn't help it, later on we developed techniques to load perhaps four to five hundred foils in a reactor for a given run.
and mechanized counting techniques, connected the counters to the beginning computers or the initial computers we had there, and that was done in order to reduce...
the human radiation exposure associated with the folks doing the experiments and removing these foils from these, because they were radioactive after removal from the reactor. Not only that, Jack, we constantly use the same fuel from one experiment to the other, and there's bound to be some residual activity. Right. There were times when we would just plain stop operating for a while, for two or three days, and let everything cool down, so that when we next...
started to load some catcher foils on the on the fuel elements the exposure would be reduced I remember we really had to take a lot of pains and we bring it up on a period and we had to turn the corner very precisely so we didn't overshoot and had a nice flat profile right the whole the idea was as I
Yeah, we had to reproduce conditions. The total exposure, the total number of watt seconds, had to be maintained or known for one exposure at one run to another, because we wanted all these measurements to be normalized in order to provide this precise power distribution measurement needed for the fuel element loading determination. OK.
Now, one of the other unknowns at the time, it turned out to be really unknown, was what effect the temperature of the moderator water would have on the reactivity or the K of the reactor. Initially, it was thought, oh, if the water gets hotter, the density is going to go down. Therefore, the reactivity will go down. Well, it didn't quite work out that way. So we had some HTREs installed in our tank there.
steam heaters with electrical support. And we started to heat the water up on an experiment one day. And after doing a lot of preliminary work, and lo and behold, we had to keep running the control rods in, which meant the reactivity kept rising. We kept going it in, running it in, running it in. And after a short period of time, we said, wait a minute.
we have to stop here we're running we had the rods too far in so we shut down and we just changed the initial conditions we added some control rods in at that were static and just the manual type manual type they just went in stayed there they didn't they didn't get moved because at no time could we ever get into a condition where our control rods could not shut down the reactor
So this gives you a picture of the change in reactivity as a function of the water temperature in the reactor tank. Going from southwest to northeast indicates the reactivity is increasing as water temperature goes up. You can't read the numbers on the grid, but that's what it's saying. I think we went up to about 150 degrees eventually. In several steps.
Right. Because of the, you can't tell it by looking at this side there, but that's a non-trivial change as a result of the water temperature. In fact, this can be a positive thing also. As time went on, we found that we were getting so much xenon buildup that one of the only ways to get the reactor started was to heat the water up. So much, that water facility, we kept it for a while.
All the experiments were needed, were performed. The power reactor was designed, built, and started to operate in Idaho at our test facility out there. So it became quite apparent that, okay, the heat transfer reactor experiment, number one, which is what we're doing, was a success. We were able to design fuel elements. We could run it at temperatures.
appropriate to the needs of the program at that time. Do you remember how many hours we put on that critical experiment? It was a lot. Oh, the critical experiment? Oh, hundreds. Hundreds. Hundreds of hours. Normally they would run about 20 minutes, I think, per pop, right? Our foil exposure was a 20-minute run, right? And we'd get maybe three of those a day. And so there were a lot of them made.
we'd served our purpose, the reactor was designed, was operating, there was no more real need for the critical experiment in Evendale. But at that time, the design requirements of the program had moved on. Water was nice to prove the proof of principle, but it was not realistic to think about that for an aircraft propulsion device. So the next consideration was, well, okay, how else can we use, can we get a hydrogen
in around the uranium. Now, the moderating function that water provides is due to the interaction of the high-energy neutrons, and Bob talked about earlier, interacting with the hydrogen present in the water. Just like billiard balls. Exactly. One of the things that was considered was, well, maybe we can hydride betel. So the idea was, well, we can hydride zirconium. Zirconium is a very good hydrogen absorber. And actually, you can get as much hydrogen per cubic centimeter in zirconium as you can in water.
Some of the other metals you can actually get considerably more. What you're seeing here is a representation of the reactor with hexagonal cells of zirconium hydride with roughly a four-inch or three-and-a-half-inch diameter hole through them. Into these holes were then placed the filaments that we just finished talking about. And that zirconium hydride hexagonal cell then took the place of the water.
Otherwise, everything else is the same. Surrounding this structure was a beryllium reflector. Okay, the mock-up that we did for that reactor was constructed in the other cell from which we did the water tank. In this one, we divided the reactor with a line passing right down the center line of the reactor such that the upper half or upper half of the cylinder was
restrained on the second floor of the cell. The lower half was mounted on a hydraulic piston similar to that which is used in garage lifts and so forth. And the reactor would be brought together remotely from the test cell or from the control room at the top of the steps there.
actually then lift, the hydraulic lift would lift the upper half off of its support pads by about a three-eighths of an inch. And the reactor, all further activities on the reactor were then conducted from the controller, as I mentioned up there. Did we have the big concrete doors in this facility? Yeah, okay, yeah, good point, Bob. The other door, the other cell which we started had the five feet, well, equivalent of five feet of concrete,
mounted on, hinged like a big bank vault door in the test cell. And that was manually operated. Well, Bob and a few other guys and myself got pretty damn, pretty tired of pulling that door open and closed all the time. So it got a motor put on it. But we had something to say about the construction of this cell. So we had a door mounted.
were created on the one wall of the cell. It was roughly 100 tons of concrete. And it was moved back and forth in which to load structures into the cell. And then we had a very small personnel door up there at the top of the steps from which our people went in and out. OK, here's a picture of a table facility, which was basically our next experiment.
It's not the exact facility, but it's close enough. It consists of a series of a whole stack, I should say, of hexagonal aluminum tubes into which we would stick uranium. Stick is a very gross word. Place. Insert. Insert. Good word. Uranium, zirconium hydride, beryllium, beryllium oxide, or whatever. Each segment of this thing would have half the reactor on it.
and one half would be fixed, the other half would be remotely closed with the other half, and we would then have a close to a critical assembly. Then the control rods, which are mounted out here, would then be removed or inserted, depending on whether we're having fuel or absorber, to bring the reactor critical. That facility, which is sort of like this one here, is one which we used.
for the ANP program initially, for the zirconium hydride program, then for the fast reactor, then for the beryllium oxide reactor program. And later on, after ANP was canceled as a side issue, that facility was then moved to Idaho, where we did fast reactor experiments under the NMPO, or follow on to ANP. And when the NMPO program was terminated, that facility was turned over to another contractor. I don't know who it was.
And they continued to use it for another 15 years. And before it was dismantled in the middle 80s, it was the longest running experimental facility in the United States. Just another example of GE quality. Absolutely. It was a real top of the line design, including the hydraulics, Bob, which you did. Right. OK, here you see a picture of the beryllium oxide
fuel elements which were later used for the latter part of the in the latter part of the program was decided that metallic clad fuel elements were just not going to cut it in terms of turbine inlet temperature for the various mission profiles which we were trying to achieve and the only way we could get it is going to a higher temperature material namely ceramic which in our case turned out to be beryllium oxide and we had to start
doing critical experiments on beryllium oxide reactors. We would have pure beryllium oxide, and we would cut up our thin uranium metal foil, stick it inside the tube, and that formed our fuel element simulation. The actual tubes, the beryllium oxide tubes, this happened to be fuel tubes, and their appearance is dark, blackish, compared to the white. Beryllium oxide is white, right. Okay.
Okay, now here's a picture of the ceramic reactor mock-up. Here's the hexagonal tube structure I alluded to earlier. Surrounding it are five-sixths of the reactors consist of beryllium oxide tubes with uranium metal inserted inside the tubes. The upper sector there, roughly one-sixth of the reactor, was actually beryllium oxide, uranium oxide fuel elements. And that was...
It was a little better geometric match to what we had to get to. And also, the large amount of metallic uranium provided a safety mechanism, which is beyond our scope of our discussion today. But anyhow, it was needed for safety. Jack, how good of a simulation? Did you compare what a simulation section was doing compared to the actual fuel tubes? How good was the simulation?
It was quite good across the radius. Within a fuel element, it was not as good as we desired because when we had one milth thick metal in there, it was more of a point absorber rather than a spatial absorber. So we needed to have the other. So this is another representation of a similar core with the beryllium oxide core, the beryllium oxide reflector.
and the places for the control rods to go basically near the reflector region. This type of thing was built up on the tables, right? Yeah, we had those. We built two of the facilities similar to the one which we just looked at. The first one was a six-foot square, and the other was an eight-foot square. This was a six-foot square, I think. Right, right. The first one was six feet.
And then the one for the XMA mock-up itself had to be, we needed an 8-foot square. Actually, it was an 8-foot cube. An XMA reactor was a much more powerful reactor in tension. I know the HTREs were around 20 megawatts. That was over 40, wasn't it? I think it was over 100. Over 100, that's right. That's right. This was over 100. Yeah, right. So now, during the latter part of the 50s and early 60s,
computational capability which we had available to us in terms of shield design, reactor design, thermodynamics, had just grown so remarkably that the critical experiments function was redirected to a large degree into some other areas. One of the areas was shield heating, which required a whole new kind of instrumentation,
we had to develop because we still had some problems in shield design, which required some experimental methods of measuring shield heating. No. So in summary, the critic experiments function these days, 1995, is very, very limited.
in as much of what you used to need the clinical experiment for, you can do with a very satisfactory amount of accuracy using calculational techniques. And there are very few reactor designs going on anyhow in the United States for any purpose whatsoever. But you've got to admit that those were pioneering days. Oh, yeah. Considering the fact that merely
five or six years prior to that where the first critical experiment ever conducted it was pretty it's pretty interesting achievements that we have we accomplished right we were we had the first automated foil counting thing uh we demonstrated in there where we could count the foils uh the power of these catcher foils which i alluded to earlier
results were put out and came out on a punch card machine, which we had in there in those days. This is the middle 50s. Punch cards were a big deal, big time. And take those punch cards down to a 650 computer and have the data reduced in a few minutes. Of course, it took you weeks to program the bloody computer. And then later on, we create a direct pipeline from our control rooms into the computer center. And when we came up, we did certain measurements, piped them directly into the computer, and this was in 1960.
And then we had an IBM 701, 704. So that's basically the history of the critical experiments at A&P, where we went from very crude stuff, as you've seen, to hooking it up with the computer, which is the modern-day savior, I guess, to some folks.
And one other thing we might mention in closing, that these were real reactors. The fact that we only operated 1 to 10 watts doesn't mean they couldn't have been operated at 20 megawatts. Of course, they wouldn't have held together very long. The capability was there. Very large heat generation capability, very small heat dissipation capability, which means you get real hot. Okay.
7.2 Shielding
This is a model of the Convair NX-2 nuclear-powered aircraft. The NX-2 was never built. Its gross weight was about 500,000 pounds and was about B-52 in size. Three XNJ-140 power plants were located inside the fuselage in a three-abreast installation. Here the power plants were readily available for remote removal.
Radiation shielding was used both around the reactors and around the crew compartment. Convair optimization studies were based on an acceptable crew dose of 0.02 rem per hour within the five-man compartment. The crew compartment was located about 100 feet forward of the engines. A large part of the radiation striking the crew compartment is straight line coming from the reactor core. However,
Radiation scattering off the airframe and the air requires that shielding surrounds the crew compartment. The engine shield was designed to reduce radiation levels so that induced radioactivity of the aircraft parts did not hamper subsequent aircraft maintenance. Also, the radiation damage did not limit the engine mounted parts.
The XNJ-140, this is a model of it, was based on the direct air cycle, where compressor discharge air was heated in the reactor core in the uranium-fueled area and went out through the turbine and the nozzle. The OD of the side shield was about nine feet. The power plant was 61,000 pounds, two-thirds of which was in the reactor shield area.
The 183,000 pound weight for all three power plants was reasonable when one considers that the 747 might take off with 250 to 300,000 pounds of jet fuel. The 140 reactor power plant was a four-bearing machine, the thrust bearing and number two bearing. The red indicates the annulus, which is the active core.
The blue shows the front shield plug, the rear shield plug, and the side shield. The compressor discharge air flows outward, then down, and becomes a plenum ahead of the reactor. The mixed air temperature of about 1800 degrees coming out of the reactor then goes into a larger plenum and goes upward and into the combustor and the
The duct configuration was similar to that of the HTRE 3 reactor, which was tested in Idaho. The effect of the ducts on radiation leakage was not as severe as once expected. Neutrons are slowed down through a series of elastic or billiard ball collisions with atomic nuclei. The lighter the nuclei that the neutrons strike, the more energy they lose per collision.
The slowing down or moderating process is vital to neutron shielding. This is because the materials for neutron absorption work best on thermal neutrons, for example, boron, cadmium, europium. The neutron shield materials in the XNJ-140 were borated beryllium within the pressure vessel and lithium hydride in the outer shield, outside the vessel.
These were chosen for thermal and mechanical requirements. Gamma rays lose their strength through collisions with heavy elements. This makes dense elements with high atomic numbers most attractive. Lead, iron, tungsten, tantalum are good. In the XNJ-140, bore-rated 304 stainless steel in the front and rear shields absorbed gammas and suppressed the thermal flux. Cooling of the front and rear shields was by compressor discharge.
The compressor further reduced the straight line radiation to the crew compartment. This is a drawing of the front shield. The stainless disks are 304 stainless and one weight percent natural boron are located close to the core where they are termed shadow shielding. 550 pounds.
located there and a little parts up here. The beryllium of one weight percent natural boron discs and sectors filled most of the plug and weighed 4,100 pounds. The incandilic structure weighed about 1,000 pounds of the total 5,700 pound front shield. Secondary heating caused by gamma rays requires cooling of the shield. The coolant bleed flow is from the duct here radially downward.
into a tunnel around the shaft going to the rear end to the rear plug. This is an isometric of the rear shield plug. The primary shielding material in the rear shield was beryllium plus one weight percent natural boron. The shield structure was heat treated in canal X. A small amount of stainless steel with boron was used. The total weight was 4,400 pounds. The beryllium
Again, the larger volume. The beryllium inserts were placed inside the shaft at the aft end. They were aligned radially with the front and rear shield plugs, but not in the core region. There was 30 inches in the front and 43 inches in the rear inside the shaft. The inserts were positioned with torsional springs.
This is the 140 radial or axial section of the core. The two bundle axial loads were contained on the forward end by 12 beryllium sectors, on the aft end by aft retainer. The latter consisted of 12 Inconel-X twin tube sheet sectors. This aft retainer was internally cooled.
and externally insulated. It eliminated the need for tie rods going through the core to hold the aft retainers. The next chart is the 140 radial cross section. Red indicates the annular core and it consists of the uranium-fueled beryllium oxide tubes. The side reflector
8.5 inches thick and 90% solid with BEO hex bars. The beryllium oxide cooling tubes made up the other 10%. This reflector region also served as neutron shielding and somewhat as gamma shielding. Also, the outer reflector and the bore-rated 304 pressure pads enclosing the tube bundle radially served as a thermal shield and limited energy deposition in the outer shield.
Some neutrons were slowed down and were absorbed in the borated pads. This is an isometric of the side shield. It was outside of the pressure vessel and was cooled with ram air coming through here, going out there. It consisted of annular sectors of lithium hydride cast into 99DL stainless steel structural cans.
Lithium hydride contained a good thermal neutron absorber, lithium 6. Like boron, it suppressed thermal neutron absorption in structural materials and minimized induced radioactivity in aircraft structures and engine components. The total weight of that side shield was 16,000 pounds. The thickness of this side shield through here, the thickness of that annulus, was about 17 inches. One-third of the weight
was within structure, that 1990L material. The top forward sector is shown next. The RAM error comes through the sector and exit out to another section. The next chart
shows the internal cooling tubes that are placed inside this can before the lithium is cast into it. Shielding material was removed at the sides of the engines. This permitted an 83-inch centerline spacing
among engines, and shared shielding. At the left and right engines, shield cheeks were mounted to the airframe to minimize induced activity in the aircraft wings. On the MTR, I worked mostly on the shield design, which was actually at the time one of the big uncertainties. The nuclear reactor
design situation and physics was in pretty good shape. And where I worked there was right down there with Jimmy Lane, who was one of the big unsung heroes in the whole nuclear program. He was the guy that could calculate anything on the back of an envelope. And to get a quick
very close approximate answer to almost anything in nuclear or thermal or aero or electrical or anything and he worked directly for Alvin Weinberg and then Eugene Wigner then he's put in a few years down there and that is when Wigner and Weinberg were putting together their their work on the
physics of nuclear reactors. Now, of course, people had to be designing nuclear reactors before that book came about. There were reports that, for example, reports from the so-called Lexington Project, which was an MIT AEC project, which we'll talk about later, where they'd had academics look at physics of shielding. And these were hopelessly complex discussions of the thing.
talked to Jim Lane about, you know, how are we going to calculate the shielding, what comes out of the shield. When I got there, the shield configuration, external configuration, was already set. It was seven feet thick of concrete, period. And we had to have air ducts go through that thing. So I asked Weinberg, where did that seven feet come from?
There was no changing it at that stage of the game because they were already designing the foundations in one thing or another. He says, oh, well, Gene and I sat down one afternoon and we figured the shield should be seven feet thick. So that's how it got to be seven feet thick. Then we had to put in all kinds of penetrations, air ducts and control things and electrical conduit and et cetera, et cetera, all within this thing. So I sat down with...
with Jimmy Lane. And it turned out that on the quarter of the page, he would do a calculation that I'd been looking at these academic reports an inch thick, and you still couldn't figure out what in the world they were saying. He would just rip it off. And then I said, where's all this stuff come from? And he says, oh, that comes mostly from a guy out in Hanford by the name of Wendy, W-E-N-D-E.
And he wrote a report out there on the shielding. And I said, oh, gosh, that's how they got shielding on the Hantrid reports before they had all these academic studies. And he said, yeah, we had to have shields. Well, where did he get the stuff from? He says, well, you know, back in the 1920s in Helvetica Acta, there was a big issue devoted to the shielding of radioactive sources.
And they took radioactive cylinders and cubes and one thing or another with different field configurations around them and came up with all these very simple formulas, all starting with the inverse square law and with exponential random absorption. Beautifully simple. All that stuff really comes out of Helvetica Acta. Here's my copy of Helvetica Acta.
Lo and behold, except for problems of scattering around corners and in ducts, which they didn't get into very much, the basic shielding stuff was all done back in the 20s and very well written up, but completely lost track of by the academics who were trying to figure out the shielding later on. The mathematics they used for the simplest things, the academic stuff, was just, you just couldn't believe how they would.
and mess things up. So in any event, I then went in to calculate the scattering through ducts in one thing or another. And we got a lot of experiments running down at Oak Ridge on the reactor. The swimming pool reactor got built by that time. And so we ran the slab experiments of different kinds of shield slabs and combinations.
of ducts and one thing or another. And so we got them pretty well, the shield design figured out. But in order to make the shield work, it turned out we had to put in big clumps of iron at one place or another and thicken up the shield here and there. And it ended up being the most awful mess. And so we were bemoaning that fact. And then Ted Rockwell, who worked for Admiral Rickover,
who was a very smart guy, just put out a book on the life of Rickover, the biography of Rickover, as a matter of fact. And I said, can't we just pour some high-density concrete in there and get rid of all of this patching that we've done? And he said, well, you could use barite. It has a density of ordinary concrete is about 2.4 or something like that, and barite has a density of around 3.5. So we made believe we poured the whole shield full of barite.
and we were able to eliminate all the patchwork and everything else, and it solved the problem, except for one thing. The contractor didn't count on having concrete that weighed 50% more than ordinary concrete, so he went broke on me. Hi, I'm Mel Lapides. I was with the ANP project from 1952 to 1962.
The Air Force actually set up three organizations for conduct of the shielding work on the A&P project. General Electric A&P, of course, was responsible for the overall design development and integration of the shield into aircraft. Oak Ridge was involved with both basic measurements at
the lid tank facility and special measurements related special facilities I should say related to the unique scattering problems of the aircraft reactor. Convair had similar facilities but in contrast to a tower facility for scattering experiments they actually flew a test reactor. One of the
fundamental concepts of ANP shielding was that they really worked on two shields called the divided shield. You had the crew shield where the radiobiological effects of radiation on a multi-man crew were of major concern. Fortunately, perhaps, that crew shield was separated by as much as 100 feet from the reactor core source. The reactor in turn
had front, rear, and side shields that were independently designed to different requirements. The overall concept of the divided shield was that this was the method of getting overall weight optimization.
One of the distinguishing characteristics of the reactor shielding approach is suggested by a distributional plot of how the dose rate from either neutrons or gammas, not considering streaming immediately, varied with the azimuthal angle, where zero would be horizontal to the core centiline, would be the core centiline.
The dose builds up rather rapidly and then reaches a maximum at anywhere from 90 to 150 degrees. And what that meant was that the
carrying shielding beyond this area was of perhaps limited value because it didn't contribute to scattering. I guess this must be scattered dose rate. Integrating the two previous slides gives you an idea of the philosophy of shielding.
The crew shield, which would be over here, was the dominant parameter determining that shielding was separation distance. The front shield was what would be called the shadow shield to keep direct radiation from impinging on the crew shield.
And that front shield was carried out only until the scattering, which was the other source of radiation to the crew, was no longer of benefit. You remember the line showing the azimuthal distribution.
And so its weight optimization and benefit limit decided the size of the front shield. The rear and side shields were configured to provide radiation damage and activation limitations to facilitate maintenance.
It isn't clear to me that the designs were highly validated, except at a very basics level, but it was of interest to note that they assumed very heavily that an awful lot of items in the aircraft would be replaced.
even though that might not be normal practice in an aircraft at the time. I guess it's also worthwhile to remember that people were just starting into radiation hardening, and certainly there were very few solid state devices around at that time. I guess there are better radiation hardening systems.
systems that are more radiation resistant than semiconductors, but they certainly were not emphasized until much later in the 50s. So the trade-offs amongst all these parameters and defining the dose rate and the activation limitations for maintenance all contributed to the shield design and that
So the latter, other than radiobiological, dictated the side and rear shield design. I suspect that instrumentation somehow necessary for operation of the system may have been the most severe limit in the damage arena. We're going to shift now to some of the
mostly pictures of facilities and equipment developed within the shielding program. The first and perhaps best known was the tower shielding facility built at Oak Ridge and it was essentially a question of building a gantry tower or portable gantry tower so high and with
such a minimal amount of solid structure so that you could loft the reactor and study its effect on a crew shield without being concerned about ground or air scattering, which turned out to be no mean trick, particularly since at the time people were not
totally aware of neutron activation in the air issue referred to as the nitrogen-16 issue. They also could test mock-ups of ANP shields. This is a schematic of the tower shielding facility. These, what I guess you call lacy girders, are of the order of 200 feet high.
There's a swimming pool reactor down here and this is a hoisting system for a reactor shield that is being tested. I believe that the reactor could also be lofted as the case may be.
These guy wires are what facilitated these very open, porous girders. Earlier, more conventional bridgework designs with a gantry crane up on top proved to give too much side shielding. And this is the hoist house
and the control room is over here. Here are two views of the tower shielding facility. This one shows the actual completed system, and I think right here is a mock-up shield.
being set up for test. This is the actual reactor system, which is a swimming pool reactor normally housed in this concrete swimming pool, and this is a self-contained reactor. The requirements on this system, of course, were that it be capable of lifting both the shield and the reactor system, so we're talking
probably of the order of 10-ton capacity. The second special facility for the ANP program was the Convair Fort Worth Nuclear Aircraft Research Facility that became operational in 54 and ran through 1961. It was used for fuel fuselage activation studies, ground and flight validation of shields, and radiation damage studies.
Shown here is the reactor and crew shield mock-up that were part of the NARF test program conducted by Convair. This was, I believe, used both on the ground and in the air. This picture shows the
a crew shield mock-up being loaded into a Convair military aircraft. The crew shield contained various detectors and early in the program the reactor was simulated by various
gamma sources and I believe neutron sources, and then ultimately by the operating reactor shown previously. I believe it was in this series of testing that they truly discovered the magnitude of the scattering problem, including the
nitrogen-16 issue, and of course they were able to get information about activation and radiation damage by having sensors spread out throughout the plane. These schematics show the aircraft with the reactor and crew shield in place.
And there's an instrument capsule between the two and additional instrument locations all along the capsule. That's a top view and you can see a side view here. It never... somebody, not I, knows how they manage to protect the pilot during these flights. This...
shows the actual configuration of the crew shield where the operating pilot actually stood. He must have had great faith in shielding calculations and or they ran the reactor up at power gradually.
7.3 Mechanics
My name is Art Ross. The facility that was done for using a chemical combustor to test the X-39 engine. In this there were large ducts connected with bellows. And here I'll just show you a quick picture which will be amplified later on for you.
In that picture, there were two ducts that leave the X-39 engine and go to chemical combustors where they pick up the heat. These have expansion joints which are large bellows, and you'll be able to see that better in the next picture. And every time the rods were tensioned, the big duct would cock out of line, and they kept...
re-tensioning and re-positioning, and every time it cocked. And finally, someone asked me to look into it. And it was a problem that I had come across in looking through theses I would use for my master's degree, I believe. And it was described by an aircraft company, Boeing in Seattle, and they were asking
suggesting it is a thesis problem. And what happens is it's a fluid column buckling in which columns supported by bellows or other expansion joints would buckle as if they were loaded axially, even though if the system is under internal pressure, even though there's no axial load in the system. A very unique and really hard to believe for most stress analysts problem.
Well, we looked at this problem, and this turned out to be a case of fluid column buckling. And so we didn't have to reposition anything. We just put some chocks in that just held it, because they're really not much load to cock it. And that solved the problem for that. And as opposed to metallic components, ceramic components have no ductility, essentially have no ductility at room temperature.
and therefore are very susceptible to thermal stress analysis. So when we started to design that element, thermal stresses became a very critical item in the design of the reactor, which was all ceramic. One problem, we had beryllium oxide, as I remember, I think a beryllium oxide reflector plates in the reactor.
And we did experiments with these. They were non-nuclear experiments of the setup of the reactor. And when we opened it up, every beryllium oxide reflector plate had cracked. The normal two-dimensional analysis we had done for these
would not have predicted cracking at all in the direction they were at, wouldn't have predicted cracking, and certainly wouldn't have predicted cracking at the locations. And we looked at it as an end effect problem, and using some probably analysis that Gabe Harvey of General Electric Company in Schenectady, who I got to know very well later on, had worked on end effect problems, and
I utilize a mechanical analog to analyze it, and sure enough, a short, rather than infinitely long, reactor plate would have resulted in high stresses, higher than predicted by two-dimensional analysis, and in a direction different. And so that showed that the three-dimensional analysis would predict cracking.
And also, luckily, it showed that all one had to do was use plates half as long, each plate half as long as the other, and the problem disappeared. So we did that, and there was no, needed no redesign, because there were many plates stacked end-to-end, so we just had twice as many plates stacked end-to-end. Each plate in itself was cheaper, since the shorter plates were much cheaper to make than long plates, so we changed to short plates, and lo and behold, there was no problem.
The last problem I'd like to describe is one that I don't think had been solved by the time I left, and this involved the ceramic fuel element cylinders themselves. Our present finite element analyses could have analyzed that in detail because they can answer, can solve nonlinear problems and three-dimensional problems. But at that time, Howard and Eagle and I
wrote an analysis of the stresses in the ceramic fuel elements using a finite element program, not a finite element program, a finite difference program, and as a matter of fact presented that at a GE research division symposium. During operation, I mentioned before, during operation, any solid piece of matter
internally generating heat will have to have stresses locked into it because the inner portions expand greater than the outer portions. It's like the core of a ball expanding against the outer surface of the ball. Nuclear energy is unique in that pattern that you cannot find a non-zero stress distribution.
zero stress temperature distribution. The problem that we had at the time I left was that the ceramic element could be brought up the power and run the power quite successfully. And if run for a long enough time, however, the stresses would relax and at temperature there'd be no stress, which sounds great.
and is great, but the problem is that a reactor has to be shut down at some time. When the reactor shuts down and it comes back, tries to get back to its original shape, the zero stress condition gets stresses reappearing almost opposite in sign. So during operation, for instance, you would normally have compressive stresses near the center.
When you cool it down, you get tensile stresses near the center. At operation, stresses are not important because they will yield and creep out and the material is not brittle. But when it cools down, the stresses reappear, tensile stresses on the inside, and now they reappear on a cold and brittle ceramic tube.
One of the members of Art Ross's Applied Mechanics Group is Howard Eagle, and we're pleased to have him with us today. Howard, why don't you tell us a few things that you fellows were up to? Okay, thank you, Otto. As Art Ross has mentioned, safety of a nuclear-powered engine was paramount, and if any failures were to occur, they should do so in a fail-safe mode. One question raised was,
What would happen in the unlikely event of the airplane crashing into a granite mountain? We all know that airplanes don't crash, but what if? In such a crash, would there be time to sense the impact and scram the reactor? That is, could we activate and insert the control rods? Or would the reactor as a critical mass blow up the mountain and surrounding area?
Since we were the resident dynamic experts, we were given the challenge. Working with Convair structural design and configuration data, we developed an analytical dynamic model of the aircraft and the nuclear power plant, albeit it was a simplified model as the computing capacity of that day was very limited. At the time, it served its purpose.
We saw the initial velocity impact problem which indicated the reactor control system would be able to sense a deceleration wave prior to the reactor itself crashing into the mountain. Sometime later, we constructed an even more simplified model using an equivalent thin-walled cylinder
to represent the fuselage portion of the plane and a solid mass for the reactor component. We took the model to the D Street facility in the Missile Space Division at that time in Philadelphia and used their shock tube to conduct a test. Tony Coppa, at that time working in MSD,
was working on dynamic buckling of shells and was happy to help us and perform the test perform the test the test performed consisted of tony firing a sabo from the shock tube into our freely suspended model high speed pictures were taken and accelerometer data recorded
This test agreed with an analysis we had made of the specimen and provided some credence to an airplane crash analysis. At the time the airplane crash analysis was carried out, there were of course skeptics that a crash could be sensed and a system activated, much less predicted through analysis.
Let's fast forward the clock some 25 years and think of something common in our everyday life. There's something everyone comes in contact, I don't think they come in contact with, but is aware of. That's the automobile inflatable airbag. Today cars are analyzed using finite element dynamic models. Cars complete with dummies are crashed into abutments to demonstrate the safety of the
airbag concept and I'm sure we've all seen films of such tests demonstrating that impacts could be sensed and such control systems activated. I mentioned earlier that there was also widespread significance of our dynamic analysis methodology and that occurred in the technology transfer of the programs that were developed in
anp and i'm sure this happened not only in structural mechanics but in other areas thermo and thermodynamics etc after transferring to philadelphia at the at program termination frank krieger and i developed enhanced versions of the dynamic computer codes which we had developed and added new capability programs for stiffness generation
eigenvector eigenvalue or the or the modes and frequency computation and response analysis to sinusoidal random and transient loading loadings became the cornerstone for dynamic analysis in both the re-entry systems division and the space division well into the 1980s
The other divisions also in the company benefited from this work dynamic analysis program development through the medium of the stress and vibration workshops which were started in 1965 era when GE introduced the GE 635 computer that which later became the Honeywell 6000 and we shared between the different components of the company.
all the techniques that had been developed and these were among them. The relatively brief history of the ANP program saw tremendous advances in the area of structural mechanics computations.
See Also
On this page