Belew, Leland
Dublin Core
Title
Belew, Leland
Description
This interview also contains the end of the Ferrell/Lewis interview. Interview with Leland Belew begins at 00:07:20.
The interview centers on the development of the Saturn rocket propulsion systems, with engineers and historians discussing the technical, managerial, and historical decisions that shaped the Apollo program. Early conversation compares liquid- and solid-fueled rockets, emphasizing the higher efficiency of liquid hydrogen engines despite their greater complexity and the safety challenges encountered during early RL-10 development. Lee Belew explains that the RL-10 evolved from earlier Pratt & Whitney engine concepts to meet specific Saturn requirements, while the F-1 engine was developed ahead of a finalized launch vehicle because engine development is the longest-lead element of any rocket program. He describes how clustering engines, adopting the "all-up" testing philosophy, extensive ground testing, and systems engineering enabled NASA to meet the lunar landing deadline. Belew also highlights breakthroughs in solving F-1 combustion instability, advances in hydrogen-oxygen propulsion, and innovations such as Rigi-Mesh injector cooling, arguing that these developments transformed rocket engine design from an uncertain art into a more predictable engineering discipline. The discussion concludes with reflections on payload growth, design conservatism, technology transfer among contractors, and Belew's pride that the Saturn-Apollo program successfully fulfilled the United States' commitment to land humans on the Moon within a decade.
The interview centers on the development of the Saturn rocket propulsion systems, with engineers and historians discussing the technical, managerial, and historical decisions that shaped the Apollo program. Early conversation compares liquid- and solid-fueled rockets, emphasizing the higher efficiency of liquid hydrogen engines despite their greater complexity and the safety challenges encountered during early RL-10 development. Lee Belew explains that the RL-10 evolved from earlier Pratt & Whitney engine concepts to meet specific Saturn requirements, while the F-1 engine was developed ahead of a finalized launch vehicle because engine development is the longest-lead element of any rocket program. He describes how clustering engines, adopting the "all-up" testing philosophy, extensive ground testing, and systems engineering enabled NASA to meet the lunar landing deadline. Belew also highlights breakthroughs in solving F-1 combustion instability, advances in hydrogen-oxygen propulsion, and innovations such as Rigi-Mesh injector cooling, arguing that these developments transformed rocket engine design from an uncertain art into a more predictable engineering discipline. The discussion concludes with reflections on payload growth, design conservatism, technology transfer among contractors, and Belew's pride that the Saturn-Apollo program successfully fulfilled the United States' commitment to land humans on the Moon within a decade.
Source
University of Alabama in Huntsville Archives and Special Collections, Huntsville, Alabama
Rights
This material may be protected under U. S. Copyright Law (Title 17, U.S. Code) which governs the making of photocopies or reproductions of copyrighted materials. You may use the digitized material for private study, scholarship, or research. Though the University of Alabama in Huntsville Archives and Special Collections has physical ownership of the material in its collections, in some cases we may not own the copyright to the material. It is the patron's obligation to determine and satisfy copyright restrictions when publishing or otherwise distributing materials found in our collections.
Format
.MP4
Language
en
Type
Interviews
Audio
Identifier
ohc_stnv_000004_A
Oral History Item Type Metadata
Interviewer
Bilstein, Roger E.
Interviewee
Belew, Leland
Transcription
[00:00:00] Tom Lewis: Essentially you got another engine, and you get another propellant [batch?] or whatever it is. Anytime you launch a solid, you're essentially launching a new engine. This is my own personal evaluation of it. I'm sure solid people could give you a...
[00:00:23] Toon Ferrell: I'm not arguing about no rotating machinery. No cryogenics.
[00:00:31] TL: No purge.
[00:00:35] TF: That’s right. It's strictly the people in command at the time, what they preferred.
[00:00:40] Roger Bilstein: Von Braun was a liquid man.
[00:00:42] TF: Everything at Peenemuende was liquid, right?
[00:00:45] RB: Yeah. They’ve finally gotten into some solids finally on the shuttle, at least the boosters in some way.
[00:00:54] TL: But that wasn’t von Braun down there.
[00:00:56] RB: Uh-huh, yeah. Wasn't there a barrier to—I've got it [down?] here—about 260 ISP barrier on solids for a long time. That's about as good as they could get. You got into whether J-2, 425, something like that?
[00:01:11] TF: Yeah, we get up to 426 nominal as much back then. We've gone up to 429. Yes, they can't compete ISP-wise.
[00:01:23] RB: That'd be another reason. At least one of the design considerations.
[00:01:28] TF: Well, you don't take advantage of ISP until you get out of the gravitational range. You can still boost it to pretty high altitude with a solid and then come in with a liquid.
[00:01:40] RB: So liquid is much better at altitude? Okay, I've got a question on the H-1. Maybe you can help me, maybe not. Why was there a difference in the turbine exhaust disposal? Those outboard engines had a special thing around them.
[00:02:08] TL: Actually, LH2 is a little more exotic. You've got a lot more combustion problems. You've got more handling problems. You've got more of a disposal problem. But it sure does make it easier to cool the chamber. Like when they fired the J-2—probably the J-2 too—when they fired the RL-10, the longer they fired it, the cooler the outside of the engine got. It just coated up with frost. You can imagine 5,500 degrees inside this thing and frost on the outside, you know, [that far away?]
[00:02:44] TF: The ecologists like it. The exhaust products are pure water.
[00:02:48] RB: So no problem theoretically.
[00:02:50] TL: No problem. And you ran into interesting problems, of course. When I first started working on the RL-10, LH2 was a new state of the art as far as [inaudible]. A lot of funny things happened. A lot of interesting occurrences. When they first started out, everybody was super cautious. No matter how cautious you are, things still happen. You find out you know that, gee whiz, that this thing's not quite as dangerous as we thought. You tend to relax a little bit, and then you get into a bad problem.
[00:03:28] TL: They had a guy on one of the portable dewars that they brought fuel in from New Jersey up to the start the test facility. He was up on the carrier, and lightning struck [the vent?], and set it on fire. Here he is straddling this 4,000 gallon LH2, and damn thing's on fire. It didn't take him long to get off of there. But nothing happened because the thing was venting. It was just a little plume of flame coming out. The burning rate wasn't such that it went on down in there, so they just closed the vent. No problem.
[00:04:13] TL: But they did have a serious problem one time with Pratt & Whitney while I was there. See, whenever they ran a turbopump, all they did was run their fuel through a [inaudible]. They dumped it in the burn pit. They weren't really getting anything out because all they were doing was turning the pump and then burning it. They thought, “Well, it sure is awfully wasteful. Maybe we could collect some of that LH2 and save it.” They thought, “Well, we'll give it a try.” They put up a little test rig there and had a great big tub like sitting on a scale. They were going to see how much they were going to catch. It was a very humid, overcast, cloudy day. They’re dumping this liquid hydrogen into this tank. Usually what happens whenever you have any vapors from liquid hydrogen, it immediately disperses. It goes flying off into the sky, you know, because it’s very light. It really spreads out. Well, this day, it was a little humid and overcast and everything. Instead of doing that, all these fumes just kind of flowed over this tank and along the ground. It flowed under the door into a motor room, electrical motors. They got a spark, and that whole damn thing ignited. There was guys blowing down the parking area. The asphalt aprons were driving around like bowling balls. One guy got burned quite badly. He was in the hospital for two or three months.
[00:05:37] RB: Was this on Centaur operations or was it on the S-IV operations?
[00:05:42] TL: I don't remember. It was just the RL-10 development program. I don't know what application it was on.
[00:05:48] RB: The thing about the RL-10 that as a layman really kind of amazes me is the fact that they can get enough gas pressure just out of recirculation through the chamber to run the pump. Because on the J-2, it took a gas generator starting that thing by pressure. Yet on the RL-10, they were able to get enough pressure just by running it through the regenerative tube. Why couldn't they use that for the J-2?
[00:06:15] TF: I imagine the size of the turbine to turn is massive that you've got to move.
[00:06:23] RB: And also you've got to restart problem in space. I suppose that would affect the recirculation.
[00:06:27] TF: Well, you need a start tank for space.
[00:06:34] RB: Okay, well actually that kind of takes care of my questions.
[00:06:36] TF: You have a couple of H-1s that weren't answered.
[00:06:40] RB: There's a guy I can still get a hold of to do those. I think there’s a guy named Dave Christensen. I don't know if you know him or not. He did a lot of early work on the S-3D and H-1. I’m sure he can…
[00:06:49] TF: D.W. Westrope is the equivalent of Dick Rogers on this engine.
[00:06:56] RB: On the H-1?
[00:06:58] TF: Yeah.
[00:07:03] RB: Yeah. I was going to call [Catalgo?] then about the…
[00:07:06] TF: About the brazing?
[00:07:08] RB: Yeah.
[00:07:11] TF: I haven't answered your question on this. I think I have. I think it's [inaudible] problems.
[00:07:16] RB: Oh look, I'll just read this over I think then for my own information.
[tape stops and restarts]
[00:07:20] RB: Interview with Mr. Belew. I had a couple of questions. I don't mean to put it straight on your memory, but there's a couple of things I can't find readily. Back in the beginning of the RL-10, or the liquid hydrogen engine program, Pratt & Whitney was fooling around with the LR-119, which is the second generation. They finally wound up going back into the first generation LR-115 before they finally settled on the RL-10. I was wondering if you could remember why they gave up on the LR-119 and went back to the LR-115. Some sort of development problem, but I can't find any note, anything much more specific than that.
[00:08:14] John Stuart Beltz: Do you suppose, Roger, where they couldn't get it to run full duration, it was shutting down?
[00:08:20] RB: Well, it could be. I really haven't found it out.
[00:08:24] JSB: This is when they were going to use four of them on the S-IV?
[00:08:31] Lee Belew: We used four.
[00:08:32] JSB: And then they had to go back to the six?
[00:08:34] RB: And then they went back to six, but that was with the RL-10, I think.
[00:08:37] LB: I don't know that those definitions—115, 119, RL-10—are that significant. I don't recall that we sacrificed any, I would say, advancements in the development of an RL series engine when we dropped the 119 and went to the RL-10. As I recall, the RL-10 was taking all the knowledge from the 115 plus the engineering studies and work on the advanced nineteen. Pulling those out and putting them into an RL-10 that had an application. RL-10 was built for a specific application, the Saturn family of launch vehicles. I think when you have an application that can be defined and specified, you can get right to work and build a product and do a damn good job. As I recall it, we didn't have any concerns about the RL series, per se, and we didn't have a direct application for what they called the RL-19 at the time because what we had an application for was specific. That was just redefined as RL-10. Of course, when we started the RL-10, it didn't have a complete update as far as performance and all that, that we finally wound up with as we flew it, and then as we now see it on the Centaur vehicle. It did become improved even after we first began the development of the RL-10. I don't believe that you'll find any major significance in that nomenclature because when you asked me, it didn't ring a bell at all on why not the RL-19.
[00:10:40] RB: Well, this is in one of Akens' chronologies. I really haven't found anything else. It was just a comment made on there. I can't find anything. Rod Stewart, we talked to him, he couldn't remember anything about it either. Another thing, and one of the reports that he wrote, I think it must have been Stewart or somebody, made the comment that there was no use for the F-1 in 1959, and that this was not unusual in developing engines. I wonder if you could elaborate a little bit on that, about the development of engines and that they are sometimes begun before stages actually come into being, etc.
[00:11:22] JSB: [inaudible] the Air Force [inaudible] their original development.
[00:11:26] LB: Well, the F-1 had its beginning as far as some work relative to the technology of large rocket engines. That was, I guess, in ‘59, early ‘59. And then, see, Sputnik went up…Let's see, when was that?
[00:11:48] RB: Sputnik actually went up in ‘56. ‘57 [inaudible].
[00:11:51] LB: ‘56, ‘57. Then this whole business of space began to take hold as a new endeavor for this country and, of course, the Russians, and other interests.
[00:12:05] JSB: You were the ABMA then, weren't you?
[00:12:07] LB: Yes. There was a decision, I guess it was a final decision, at the presidential level, in the Congress, I guess it was a congressional decision, that we as a country should develop a very large rocket engine. That large rocket engine was defined as what became the F-1. It was that kind of decision that brought the F-1 into being. That program was transferred from whatever technology basic phase A, B, you know, study-type work, some very slight hardware work. It was transferred officially, I think that was in ‘60, to us. It was a Rocketdyne contracted for effort by, I guess that was the Air Force at the time. We picked it up a little, just a little before we became a part of the NASA organization. We picked it up as a rocket engine development program. That was complemented by…There were studies on launch vehicles, and those studies bracketed the size of this engine or complementary to the studies on large vehicles that were being made.
[tape cuts out]
[00:14:06] LB: It was a Saturn launch vehicle specified. The logic of doing that has a lot of history in that it's known that the rocket engine development cycle is the longest lead item in the development of a launch vehicle or in the development of an aircraft. Just take the engines that go into our airplanes. Those usually have some lead on the complete definition of an airplane.
Usually you'll say, “Okay, I'll build an airplane based on having knowledge of an engine development that's going on.” That's often the case. But we didn't do this out of…We didn't start the Saturn engine development out of figuring as to what it took and the way of time and all.
[00:15:01] LB: The other factor on, I would say, desensitizing engine development in the dark relative to the launch vehicle is the fact that it became evident that engines could be clustered and used in numbers rather than a single engine for a single stage. That sort of takes away from the sensitive sizing of an engine relative to a certain vehicle. Now there are certain limits on that. The facts are that the F-1 did become a clustered engine before we finally wound up with the Saturn V design and built it. We had several studies and had all kinds of numbers as far as clustering and certain lunar missions depending on how you conducted missions.
[00:15:55] JSB: Was there some thought of using the F-1 on the Saturn I and IB [inaudible]?
[00:16:01] LB: Well, those considerations, of course, were made in arriving at final decisions. You go through almost all of the possibilities. As you go through them, you begin to sort them out and come down to a final kind of configuration. When you do everything in series, like your flight as a single engine on a sort of an S-1B type stage, before you put it into a cluster, you can add enough time that you’ll never get your final product out. That ties back to having enough experience out of the ‘50s, you know, out of the development of large vehicles for the military, DOD and all, and then out of some of the uses of those kinds of components for space work. You begin to build the knowledge and confidence that says, “Okay, we take some bolder steps, like we go all up.” That was one of the contributions I would say that Dr. Miller had to the Apollo program: take a bold step and go all up. You launch it—a Saturn V—with everything on it. Instead of launching the first stage with nothing above it, except that means when you finally get to the first and second, there's nothing but damage above those and so on up. You shove the whole thing together, launch it all at one time, at least up to the point of the payload.
[00:17:39.] RB: Do you think it would have been impossible to have made the manned lunar landing within a decade had you not gone all up, if you'd made it by going in the building block [inaudible]?
[00:17:52] LB: We were within a year, weren't we?
[00:17:55] RB: Yeah.
[00:17:56] LB: Now that year isn't a lot of margin compared to when we began almost ten years earlier.
[00:18:05] JSB: Would you say something about the initial opposition at MSFC to the all up concept, [inaudible]?
[00:18:13] LB: Well, I think opposition is probably not the right word. I would say that we had to go through an inward process of, I'd say, which consisted primarily of systems engineering in our own minds of what that really meant. That does have a bearing on the kind of test program that is conducted on the ground, on your hardware. With all those factors thrown together where we did have a test program that then did give us the assurance that we were committed to that kind of a system now. For someone to come off, you know, to come into the system and say, “Okay, we're going to go all up,” that's one thing. But to say that and not allow a comprehensive test program to give you the kind of background you need, that would be one thing compared to structuring a test program that allows you to do more on the ground before you go all up.
[00:19:28] LB: I think that we went through the logical sequence of engineering analyses before we said, “Okay, we will commit.” That's just the way you run a program. I think that, except those kinds of, I would say, rather dramatic changes in approach without thinking and engineering your way through it would be a rather dangerous thing to do. Now, compared to the beginning, early ‘50s on large launch vehicles for the DOD application, we accounted for a certain amount of our first launches as part of the development of the vehicle. We would expect that we would have some problems in the subsystems of the launch vehicle and maybe lose several flights in the series of developing that vehicle. Now, that completely changed in this Apollo thing, which really said we've got to do better on the use of our experience—which we did—and we have to put together a doggone well thought through program that’s systems engineered and then backed up with the right test to build that confidence.
[00:20:49] RB: Is that because it was so expensive or because it was manned?
[00:20:52] LB: Well, it's what you would want to do on anything, not exclusively tied to launch vehicles or space stuff. If you had to kind of back down the experience, it's just a logical thing to do with any kind of product. You want to do the most you can with the data. That’s engineering analysis, [inaudible] most useful tools that we have, and then the testing of the systems on the ground and where the biggest cost is, try to minimize that effort. The biggest cost in this would be building several of those just for proving the hardware by flight. We minimize that effort by doing a whole lot more with the knowledge we have. It requires a much higher level of engineering talent, especially oriented toward systems engineering talent, when you take those kinds of steps. It was the right kind of thing to do looking back even though we had some reservations in the beginning.
[00:22:03] JSB: Were there technologies that came about in the ‘60s that made that possible, like automatic checkouts?
[00:22:09] LB: Well, I think the technologies were basically there. I think the better use of that capability contributed an awful lot to that. Certainly the kind of check out gear that one has, that we had at that time, had a great deal to do with, yes. When you run a check out and everything is hard lined and everything has to do with man being completely, in every piece of data extracted, that begins to get rather undoable when you talk about Apollo type vehicle—the Saturn and Apollo spacecraft on top of it. You just have to exploit the automatic check out. At the beginning of the ‘60s, I think one would have to say that basic technology was all there. It just took a challenge like the Apollo-Saturn to put that to use. When you commit to Apollo-Saturn in a decade, you doggone well have to put that to use. There's no choice. You don't make it if you don't. So I guess the answer is yes to your question.
[00:23:35] RB: What about the engine programs? Can you say that liquid hydrogen caused you more problems than the LOX/RP-1? Do you have to make a [inaudible] that there are two different kinds of engines and systems involved?
[00:23:52] LB: Well, I would say that you can go all the way back on engine development. At one time, oxygen was called secret fuel. You can go that far probably because it was less understood than other propellants like kerosene or alcohol. It just has to do with one's understanding of what he's working with. Liquid hydrogen/liquid oxygen is a combination which is the propellant in the RL-10 engine and also the J-2 engine.
[00:24:35] LB: At the beginning, there were concerns, yes, because the combination is very combustible. When exposed to the normal atmosphere, it does evaporate. You don't see the run like you do the rest of the kerosene. To the average common guy that works these systems, the technicians and all, they're less known about that combination. I think that had a bit to do with it.
The actual development of an engine has to consider lots of things like the flow and cooling characteristics, combustibility, and susceptibility of these various propellants as far as rough combustion, which is a very major part of the development on the engine on the side of the F-1. So all those were less known than the propellants we'd been dealing with. The propellants we were dealing with primarily at the time in our work were basically kerosene and oxygen. Of course, we'd use alcohol in the Redstone stuff. Then the acids, of course, were used in many of the other systems, but none up until the RL-10 had really used hydrogen and oxygen. So it was just a matter of dealing with something less known by the fact that we hadn't had the time and a number of different systems exposed to it.
00:26:21] LB: Looking at the way it came out, I would say that from my personal experience, hydrogen/oxygen is probably the best combination that we have in rocket engines today as far as ease of developing a system that performs well. The normal things one fears, that's rough combustion and cooling aspects of this type of hardware, does your high performance. It gives a bit of problem to containing it because of the relatively lightweight, pretty large tanks. That was certainly overcome very readily by the fact that something like the S-II stage developed a very large tank, which was very successful. Basically, hydrogen and oxygen engines have been very reliable in comparison to what we termed or categorized as conventional propellants up to that point.
[00:27:31] RB: We were talking to Douglas engineers when we went out to Los Angeles about a year and a half ago. Of course, it was Douglas, but they kept emphasizing that the S-IV, for example, was just an awful lot different than the Centaur. One engineer commented that we had to fight to keep the S-IV from just being another Centaur. Do you see that there are that much differences in the systems, the Centaur and the S-IV?
[00:28:01] LB: Well, the engines are essentially the same. In fact, we delivered engines to the two vehicles out of the same line at the time with some very slight modifications. The difference has to do with the system, the total system tankage and all. There are differences there.
[00:28:21] RB: Internal insulation and so on?
[00:28:23] LB: Well, I would say sensitivity to weight is more critical on a Centaur vehicle than on the S-IV. The S-IV was, in my own assessment, a more conservative approach, not so sensitive to weight as the Centaur. So from a vehicle point of view, yes, quite different. From the engine and what flows through it, basically no difference.
[00:28:56] JSB: [inaudible] primarily the pressure stabilization in the Centaur as opposed to the fact that the S-IVB just stayed up there?
[00:29:02] LB: Well, the booster on the Centaur is pressure stabilized. The Centaur stays pressure stabilized. I thought you were talking about the Centaur second stage, it had the RL-10 in it. It has less conservative design because of the sensitivity to weight than the S-IV. The S-IV, we just built it kind of conservatively. When you do that, you can come out with the test program that's more predictable, I would say more predictable. Hopefully…
[tape cuts out]
[00:29:47] LB: [repeated conversation about Centaur second stage] Hopefully, in the end, it will be cheaper than if you had pushed everything to the ultimate as far as sophistication. That's one thing we found in the area of our Skylab program. Where we can, we've been very conservative on design in order to minimize testing. You can test your test program down, that's the biggest part of your cost.
[00:30:52] JSB: using the rule of thumb, the test program costs about a third of the overall program.
[00:31:05] LB: Well, it depends on where you start if you build facilities and all as we've done in Apollo, that’s probably right. If you maximize the use of those that have been built, like we've done in Skylab, your test program will probably runs a little more than that. We don't have to build new facilities and all that. The test program is probably going to run a little more than that.
That has passed as well as analysis and all that to go as a confident to the test program.
[00:31:49] RB: Are there really big major differences between RL-10 and J-2? Is one essentially better or more efficient than the other? I realize that J-2 came out later so they were able to utilize the technology.
[00:32:09] LB: Well, there are some different concepts that apply to the two as far as the basic principle of design. I think they're designed for applications that you can't compare too easily. I think the record's pretty comparable, one to the other. We really haven't had any major problems on either of these in flight. I'd say that they would compare pretty equal as far as a good, reliable piece of hardware.
[00:33:04] RB: Did you develop any affections for a piece of hardware? You know, you like one better than the other. Does that ever happen? One engine seems to be a sweeter engine than the other?
[00:33:15] LB: Well, I think that you probably develop more respect for some. The F-1, the early years of that development, certainly developed a healthy respect for its capability to act up in the area of rough combustion. We put a major effort into solving the rough combustion problem. As such, developed techniques that I feel have contributed greatly to that technology. We took steps in that development cycle that hadn't been done before and did it in such a way that we were able to develop models. Those models I feel have and will contribute to any similar kind of engine development in the future.
[00:34:14] RB: As I recall there was some comment that when the F-1 combustion stability problem came up, there was a real problem because research on that area had stopped back in the ‘50s when it came up. It was never really solved. So when it came to the F-1, there was really a gap there. You almost had to really go into the basic R&D program. Do you recall that?
[00:34:34] LB: Well, yes. Research on combustion has been something that's rather broadly studied by several universities like Purdue. There's one. Of course, that kind of effort is done basically on small-scale hardware. You get certain types of inputs that give you the basis to do basic engine or injector design. We found on going from the H-1 size to the F-1 size that scale factor doesn't always apply. There's a certain amount of the data from the research technology that just isn't available to you. You get into the scale that your hardware is built in, your actual scale. It just doesn't scale up. In the development of the F1, we did produce that kind of data. I think it's been done in a way that people now have the confidence to think about a new engine program that they can produce it. They can produce it to be a reliable system and not fear that you have random combustion phenomena that causes you to go rough in flight. If you look back at the history of engine development and see the results of flights, there was quite a bit of problem in flights, especially in the development of flights where rough combustion was a major factor.
[00:36:29] RB: Prior to the Saturn?
[00:36:30]] LB: That was in the ‘50s. This came about in the early 60s. It's the logical evolution of knowledge one accumulates. When you go through the kind of thing we went through in the ‘50s in the development of vehicles and engines as part of it. Much of that was done on a rather crash basis. You go about it in one way. You go about it in building [you cut and try?] a lot. When you get into an Apollo kind of vehicle, you don't allocate four, five, or six launches to just seeing that thing's going to run smooth. I get back to what I said initially: you're going to do the most you can on the ground. If you can't do that, you really don't progress to the point you can afford these kind of programs.
[00:37:33] RB: Would you say that the F-1 combustion instability was really one of the key epochs in the development of the Apollo-Saturn launch vehicle?
[00:37:46] LB: Well, it could have been one of the things that would have kept you from having a vehicle like this. It's not by chance that you get that kind of a system built by applying one heck of a lot of know-how that's been built up by having accumulated that through past experiences. It's your ability to bring that to bear on your present—present being the F-1 in this case—program. That really gets you down to what about the confidence that's been gained in that field if we continue to have such vehicles as Apollo. It's very, very difficult to come by. One thing that one has to always remember, we shouldn't lose that confidence. That confidence is vested in, I'd say, not a heck of a lot of people. It's documented, but sometimes there's more to it than just what's documented.
[00:38:52] RB: Some guy out there that doesn't have to feel that way.
[00:38:55] LB: Well, that's a part of it. There's less of that black magic stuff now than there was. There used to be a lot of it way back there. There's a heck of a lot less of it than before. I feel that most of us have been in that field. It's a fairly straightforward design development cycle now to build a high-performance system engine.
[00:39:25] JSB: [Inaudible]. Would you say something about the E-1 program, which was originally thought up for the Saturn I, clustering four E-1s?
[00:39:36] LB: That was another large hydrogen-oxygen engine development, or at least technology-sponsored program, that would have been an engine for a next series of launch vehicles if there had been a need for something larger than the Saturn series, or if we had found that it's better to use single engines instead of clustered engines. Of course, our success in clustering engines kind of took that one out as far as having to have to replace the clustered J-2 engine. However...
[00:40:23] JSB: Are you talking about the M-1 now?
[00:40:25] LB: Oh, I'm talking about the M-1.
[00:40:27] JSB: Good, I wanted to hear about the M-1.
[00:40:29] LB: The M-1 is the one I’m talking about. I'm sorry. So that thing began at the time before we launched the cluster of J-2s and before that cluster concept had been fully proven. Had large numbers of launch vehicles been in the national inventory of products, it may have been economical to substitute one engine in place of five. One doesn't know, it might have been. But that work took its place, you know, it went so far, and there are certain other factors that said, “Well, that one isn't one that's worthy of continuing,” so it was dropped. Again, you know the development cycle, you know the time it takes, and if you project that there may be in the decade some need, sometimes you have to put a certain investment into technology or into the very beginning of the development cycle if you ever want to exercise the option.
[00:41:49] JSB: Well, can't you say the same thing as the reactor in flight programming and the NERVA engine? Many of the early studies, going back to about the ‘60s, you can see that the—well, the one to one—is designed for optimal nuclear upper stage. Then by about ‘62, the program is being continued, but not very much money has been put into it. It's obvious, it’s not going to be used on the Apollo program, and that probably drops out about [‘66?]. We do want to say something about it, that, you know...
[00:42:22] LB: I'm not…I haven't been close really to the NERVA engine development. As you know that development has been conducted out of a different segment of NASA. I think anyone in this kind of business has followed it, they know that it's been carried on at a level that sort of says “Here's what can be done if there's ever a need.” There were a lot of people, associated studies that said “Okay, if there's ever a need, here's what can be done with Saturn type hardware, accommodating flight of that kind of system.” So yes, that's an investment in the case certain events that take place that says that's what we want to fly. I think NERVA was carried on probably—probably rightfully—because it is pushing the start of the art. It's something that's not so conventional as our liquid propulsion engines. It was carried on through a longer period of time to a certain point, what you could do if there was a need to do it. You also knew the things you had to design around and order for it. I really don't have any strong contribution to that I could make to the [inaudible] family of vehicles and its potential future.
[00:44:15] RB: One of the things that John and I have been interested in is where some of the technology filters in. We've been talking to Dave Christensen a little bit about the origins and incorporation of Rigi-Mesh and RL-10 and the J-2 injectors. Do you remember from then how that came about, the sources of that technology?
[00:44:38] LB: The sources are dictated by the problem that one finds when putting designs to hardware and testing hardware. In the design of an engine, one of the biggest problems we've had—have always had—is the burning of the face of an injector. It overheats, and you burn a hole, and you don't have an engine anymore. The skill and the design of injectors in that doggone of the injector is provided the kind of cooling that will disallow the heat of combustion from burning the face of it, which means burning holes in it, and then you have an explosion. Anyone knows you can do that if you brute force utilize your propellant for a film cooling, which also takes away from your performance. You know, you can cool the devil out of it by just brute force, but in all my experience, that always degraded or lowered your performance. You can't tolerate throwing your performance away, just to keep the injector cool.
[00:46:02] JSB: Is this degrading in performance just called general cooling of the combustion chamber?
[00:46:05] LB: Say if you have an injector plate, and you distribute your propellants through it and through orifices or there’s all kinds of systems used. If you're not careful, you'll have back flows, and you'll burn through it and have a problem, especially around the periphery. If you want, you can drill holes around the periphery to squirt right across that thing, throw a lot of liquid, so to speak, on it and cool it. Same in the middle. But that particular propellant—it would be probably your fuel instead of your oxidizer—would not burn efficiently. You don't get the performance, your [inaudible] specific out of it. It just lowers your performance. In turn, you’re washing through your nozzle unburned propellant. Really, that’s what happens. It's like the bid on ecology. Your automobile, as they say, put too many hydrocarbons out into the streets and the air. You want to burn it all. That’s the same thing.
[00:47:18] LB: At any rate, through the development of the hydrogen-oxygen engine, they found that in order to get performance, they had to do something other than what we had conventionally been able to do on the oxygen-kerosene engine as far as cooling. The hydrogen and oxygen being as it flows much more readily through a mesh. You can't squirt kerosene through there very easily whereas you can put oxygen or hydrogen through it, hydrogen especially. It is cold. Temperature-wise it is favorable. Just by them getting into tests and trying things, they did come up with a Rigi-Mesh as a structural member and a flow passage for a certain percent of the propellant through it. It cooled it, and it distributed it in a way that they could work their design out where they had a satisfactory ratio of oxygen compared to the fuel. They were able to make a design that yielded a very high performance. They kept their performance. They got their cooling from it. They gained their structural strength also from that Rigi-Mesh. It's very…It’s metal, you know? It's a very good piece of structure for the injector if anchored right against posts and stuff like that.
[00:49:06] RB: Was that a fairly significant breakthrough in terms of LOX/LH2 engines?
[00:49:11] LB: Oh yeah. Yeah, that was [inaudible] J-2.
[00:49:17] RB: Did you have a couple getting Rocketdyne to take Rigi-Mesh? Since it had been used on Pratt & Whitney engines?
[00:49:24] LB: No. These companies tend to pretty well come in on similar things. It's not a closed environment. The producers and all that are there, they're pretty doggone dynamic environment that you find these companies in. There’s common denominator all across the face of this business: the government, the DODs, the NASA types and all that, and the suppliers and all. It doesn't serve anyone's purpose to let people find out everything in series. One year and five years later another might pick it up and at the same time you're having to throw money out the doggone window to let each guy do his complete thing. It's something that's extremely successful, who’s one to fight it? No, that really hasn't been a major factor in my experience in any of this engine development. The people, you know, transfer their own knowledge by moving around a little bit too. You find people that would work one place and then they’d work at another.
[00:51:01] RB: When a lot of it was done under government contracts anyway, it becomes…It's not the proprietary thing, it’s a government [inaudible]....
[00:51:07] LB: A lot of it is.
[00:51:11] JSB: Presumably the supplier Rigi-Mesh, which is the Pall Company, would be visiting all these people anyway.
[00:51:16] LB: They'll try to get their product in as much as they can. I guess I couldn’t sit here and say who thought about what first. Even with the research and the literature, you could have a hard time saying, “Now that's where it showed up first.”
[00:51:33] JSB: That’s one of our biggest problems.
[00:51:35] LB: I'd say that to do that would be…Well, you could do that and probably raise a lot of interest and debates. Things turn to a great extent if you write a history, say “This guy did that first and another guy did something else first.” That'd be hard to…Some guy, someone in the middle might say, “I thought about that too.” Who knows? [laughs] They wouldn't serve too much of a purpose in my mind.
[00:52:12] RB: Well, John, do you have any more questions?
[00:52:14] JSB: I’d like to ask one thing about the institution of the fifth engine between the C-4 and C-5 configurations. Is deciding to use the anchoring engine in the center to give the extra capability for whatever the [inaudible] or Houston might do for payload?
[00:52:32] LB: On Saturn V?
[00:52:35] JSB: Mm-hmm. Putting that one in the center.
[00:52:37] LB: Your vehicle is sized by the payload, by the mission. Oftentimes you begin a development—a design for a development—based on what someone commits he can do on the top. He says, “I can do this kind of a design and come out with a certain weight and still conduct a certain mission.” You may start that way, but then a few months later you may come back and say, “Whoops, I got a little growth here.” If you want a weight reduction program that early in the cycle, you can pretty well say you can't make it with weight. My experience is there's no more expensive kind of development than a weight reduction program. That's about the most expensive thing you can get into because that just connotates [sic] that you're going to have to do a fairly sophisticated redesign, and one hell of a lot of proof testing, and then you're hanging on too marginal a safety factor. That's kind of not the way to go. If you can accommodate the problem by increasing performance of the booster, that's probably your best bet. Most conservative, most [inaudible] forward, cost effective thing you can do is increased performance of your boosters.
[00:54:10] JSB: How much appears in the literature that Marshall anticipated the payload growth by the engineers?
[00:54:16] LB: Yeah, well, of course, you don't get into a program blindly, build a vehicle without looking up what you're committed to do. When you look at what you're committed to do, you look at the other fellows' calculations and assumptions and all that, and you make your own assessment. You very well may build in a little more conservatism based on your own knowledge of what it takes. We didn't begin that in the dark. We've been through that payload growth thing since the early ‘50s. Always it's been a case of having increased performance. That wasn't unique to just Saturn.
[00:55:09] JSB: The first S-IC, was it 7.5 million pounds, they uprated that somewhat now haven’t they, for wider missions to take care of the rover?
[00:55:21] LB: I think that as you mature and as you understand by experience what you have, you can always commit more of what is reserved than you could to begin with as you don't understand the detail that you have for say ten or twelve months. You begin to know that thing doggone well, so you're able to commit a much, much more narrow margin between calculated and [rationals?]l. Anytime there's a need for additional payload and you have a mature vehicle, you can always go in there and see what you can increase in the way of additional capability. After you mature, you know whether it's there or not. A lot of the capability is based on safety factors of your system. By actual use and testing, you begin to know how good those safety factors are. They can be off one point, say, a tenth of a unit, and increase your payload significantly. The very first launch you probably can't commit to that. Maybe after ten or twelve, you say, “Well, that's an acceptable”—not risk but—”It's an acceptable situation based on all the facts.” No difference in performance of any vehicle, automobiles included.
[00:57:12] JSB: What gave you your greatest thrill in the Saturn program as you look back on it?
[00:57:23] LB: Well, I would say the fact that in the total Saturn-Apollo program, the fact that we were, in a period of about less than a decade, able to go from a certain approach to a rather complete, different approach on the whole development business cycle and operational cycle, and find that at the end we were able to uphold a national commitment.
[tape ends]
[00:00:23] Toon Ferrell: I'm not arguing about no rotating machinery. No cryogenics.
[00:00:31] TL: No purge.
[00:00:35] TF: That’s right. It's strictly the people in command at the time, what they preferred.
[00:00:40] Roger Bilstein: Von Braun was a liquid man.
[00:00:42] TF: Everything at Peenemuende was liquid, right?
[00:00:45] RB: Yeah. They’ve finally gotten into some solids finally on the shuttle, at least the boosters in some way.
[00:00:54] TL: But that wasn’t von Braun down there.
[00:00:56] RB: Uh-huh, yeah. Wasn't there a barrier to—I've got it [down?] here—about 260 ISP barrier on solids for a long time. That's about as good as they could get. You got into whether J-2, 425, something like that?
[00:01:11] TF: Yeah, we get up to 426 nominal as much back then. We've gone up to 429. Yes, they can't compete ISP-wise.
[00:01:23] RB: That'd be another reason. At least one of the design considerations.
[00:01:28] TF: Well, you don't take advantage of ISP until you get out of the gravitational range. You can still boost it to pretty high altitude with a solid and then come in with a liquid.
[00:01:40] RB: So liquid is much better at altitude? Okay, I've got a question on the H-1. Maybe you can help me, maybe not. Why was there a difference in the turbine exhaust disposal? Those outboard engines had a special thing around them.
[00:02:08] TL: Actually, LH2 is a little more exotic. You've got a lot more combustion problems. You've got more handling problems. You've got more of a disposal problem. But it sure does make it easier to cool the chamber. Like when they fired the J-2—probably the J-2 too—when they fired the RL-10, the longer they fired it, the cooler the outside of the engine got. It just coated up with frost. You can imagine 5,500 degrees inside this thing and frost on the outside, you know, [that far away?]
[00:02:44] TF: The ecologists like it. The exhaust products are pure water.
[00:02:48] RB: So no problem theoretically.
[00:02:50] TL: No problem. And you ran into interesting problems, of course. When I first started working on the RL-10, LH2 was a new state of the art as far as [inaudible]. A lot of funny things happened. A lot of interesting occurrences. When they first started out, everybody was super cautious. No matter how cautious you are, things still happen. You find out you know that, gee whiz, that this thing's not quite as dangerous as we thought. You tend to relax a little bit, and then you get into a bad problem.
[00:03:28] TL: They had a guy on one of the portable dewars that they brought fuel in from New Jersey up to the start the test facility. He was up on the carrier, and lightning struck [the vent?], and set it on fire. Here he is straddling this 4,000 gallon LH2, and damn thing's on fire. It didn't take him long to get off of there. But nothing happened because the thing was venting. It was just a little plume of flame coming out. The burning rate wasn't such that it went on down in there, so they just closed the vent. No problem.
[00:04:13] TL: But they did have a serious problem one time with Pratt & Whitney while I was there. See, whenever they ran a turbopump, all they did was run their fuel through a [inaudible]. They dumped it in the burn pit. They weren't really getting anything out because all they were doing was turning the pump and then burning it. They thought, “Well, it sure is awfully wasteful. Maybe we could collect some of that LH2 and save it.” They thought, “Well, we'll give it a try.” They put up a little test rig there and had a great big tub like sitting on a scale. They were going to see how much they were going to catch. It was a very humid, overcast, cloudy day. They’re dumping this liquid hydrogen into this tank. Usually what happens whenever you have any vapors from liquid hydrogen, it immediately disperses. It goes flying off into the sky, you know, because it’s very light. It really spreads out. Well, this day, it was a little humid and overcast and everything. Instead of doing that, all these fumes just kind of flowed over this tank and along the ground. It flowed under the door into a motor room, electrical motors. They got a spark, and that whole damn thing ignited. There was guys blowing down the parking area. The asphalt aprons were driving around like bowling balls. One guy got burned quite badly. He was in the hospital for two or three months.
[00:05:37] RB: Was this on Centaur operations or was it on the S-IV operations?
[00:05:42] TL: I don't remember. It was just the RL-10 development program. I don't know what application it was on.
[00:05:48] RB: The thing about the RL-10 that as a layman really kind of amazes me is the fact that they can get enough gas pressure just out of recirculation through the chamber to run the pump. Because on the J-2, it took a gas generator starting that thing by pressure. Yet on the RL-10, they were able to get enough pressure just by running it through the regenerative tube. Why couldn't they use that for the J-2?
[00:06:15] TF: I imagine the size of the turbine to turn is massive that you've got to move.
[00:06:23] RB: And also you've got to restart problem in space. I suppose that would affect the recirculation.
[00:06:27] TF: Well, you need a start tank for space.
[00:06:34] RB: Okay, well actually that kind of takes care of my questions.
[00:06:36] TF: You have a couple of H-1s that weren't answered.
[00:06:40] RB: There's a guy I can still get a hold of to do those. I think there’s a guy named Dave Christensen. I don't know if you know him or not. He did a lot of early work on the S-3D and H-1. I’m sure he can…
[00:06:49] TF: D.W. Westrope is the equivalent of Dick Rogers on this engine.
[00:06:56] RB: On the H-1?
[00:06:58] TF: Yeah.
[00:07:03] RB: Yeah. I was going to call [Catalgo?] then about the…
[00:07:06] TF: About the brazing?
[00:07:08] RB: Yeah.
[00:07:11] TF: I haven't answered your question on this. I think I have. I think it's [inaudible] problems.
[00:07:16] RB: Oh look, I'll just read this over I think then for my own information.
[tape stops and restarts]
[00:07:20] RB: Interview with Mr. Belew. I had a couple of questions. I don't mean to put it straight on your memory, but there's a couple of things I can't find readily. Back in the beginning of the RL-10, or the liquid hydrogen engine program, Pratt & Whitney was fooling around with the LR-119, which is the second generation. They finally wound up going back into the first generation LR-115 before they finally settled on the RL-10. I was wondering if you could remember why they gave up on the LR-119 and went back to the LR-115. Some sort of development problem, but I can't find any note, anything much more specific than that.
[00:08:14] John Stuart Beltz: Do you suppose, Roger, where they couldn't get it to run full duration, it was shutting down?
[00:08:20] RB: Well, it could be. I really haven't found it out.
[00:08:24] JSB: This is when they were going to use four of them on the S-IV?
[00:08:31] Lee Belew: We used four.
[00:08:32] JSB: And then they had to go back to the six?
[00:08:34] RB: And then they went back to six, but that was with the RL-10, I think.
[00:08:37] LB: I don't know that those definitions—115, 119, RL-10—are that significant. I don't recall that we sacrificed any, I would say, advancements in the development of an RL series engine when we dropped the 119 and went to the RL-10. As I recall, the RL-10 was taking all the knowledge from the 115 plus the engineering studies and work on the advanced nineteen. Pulling those out and putting them into an RL-10 that had an application. RL-10 was built for a specific application, the Saturn family of launch vehicles. I think when you have an application that can be defined and specified, you can get right to work and build a product and do a damn good job. As I recall it, we didn't have any concerns about the RL series, per se, and we didn't have a direct application for what they called the RL-19 at the time because what we had an application for was specific. That was just redefined as RL-10. Of course, when we started the RL-10, it didn't have a complete update as far as performance and all that, that we finally wound up with as we flew it, and then as we now see it on the Centaur vehicle. It did become improved even after we first began the development of the RL-10. I don't believe that you'll find any major significance in that nomenclature because when you asked me, it didn't ring a bell at all on why not the RL-19.
[00:10:40] RB: Well, this is in one of Akens' chronologies. I really haven't found anything else. It was just a comment made on there. I can't find anything. Rod Stewart, we talked to him, he couldn't remember anything about it either. Another thing, and one of the reports that he wrote, I think it must have been Stewart or somebody, made the comment that there was no use for the F-1 in 1959, and that this was not unusual in developing engines. I wonder if you could elaborate a little bit on that, about the development of engines and that they are sometimes begun before stages actually come into being, etc.
[00:11:22] JSB: [inaudible] the Air Force [inaudible] their original development.
[00:11:26] LB: Well, the F-1 had its beginning as far as some work relative to the technology of large rocket engines. That was, I guess, in ‘59, early ‘59. And then, see, Sputnik went up…Let's see, when was that?
[00:11:48] RB: Sputnik actually went up in ‘56. ‘57 [inaudible].
[00:11:51] LB: ‘56, ‘57. Then this whole business of space began to take hold as a new endeavor for this country and, of course, the Russians, and other interests.
[00:12:05] JSB: You were the ABMA then, weren't you?
[00:12:07] LB: Yes. There was a decision, I guess it was a final decision, at the presidential level, in the Congress, I guess it was a congressional decision, that we as a country should develop a very large rocket engine. That large rocket engine was defined as what became the F-1. It was that kind of decision that brought the F-1 into being. That program was transferred from whatever technology basic phase A, B, you know, study-type work, some very slight hardware work. It was transferred officially, I think that was in ‘60, to us. It was a Rocketdyne contracted for effort by, I guess that was the Air Force at the time. We picked it up a little, just a little before we became a part of the NASA organization. We picked it up as a rocket engine development program. That was complemented by…There were studies on launch vehicles, and those studies bracketed the size of this engine or complementary to the studies on large vehicles that were being made.
[tape cuts out]
[00:14:06] LB: It was a Saturn launch vehicle specified. The logic of doing that has a lot of history in that it's known that the rocket engine development cycle is the longest lead item in the development of a launch vehicle or in the development of an aircraft. Just take the engines that go into our airplanes. Those usually have some lead on the complete definition of an airplane.
Usually you'll say, “Okay, I'll build an airplane based on having knowledge of an engine development that's going on.” That's often the case. But we didn't do this out of…We didn't start the Saturn engine development out of figuring as to what it took and the way of time and all.
[00:15:01] LB: The other factor on, I would say, desensitizing engine development in the dark relative to the launch vehicle is the fact that it became evident that engines could be clustered and used in numbers rather than a single engine for a single stage. That sort of takes away from the sensitive sizing of an engine relative to a certain vehicle. Now there are certain limits on that. The facts are that the F-1 did become a clustered engine before we finally wound up with the Saturn V design and built it. We had several studies and had all kinds of numbers as far as clustering and certain lunar missions depending on how you conducted missions.
[00:15:55] JSB: Was there some thought of using the F-1 on the Saturn I and IB [inaudible]?
[00:16:01] LB: Well, those considerations, of course, were made in arriving at final decisions. You go through almost all of the possibilities. As you go through them, you begin to sort them out and come down to a final kind of configuration. When you do everything in series, like your flight as a single engine on a sort of an S-1B type stage, before you put it into a cluster, you can add enough time that you’ll never get your final product out. That ties back to having enough experience out of the ‘50s, you know, out of the development of large vehicles for the military, DOD and all, and then out of some of the uses of those kinds of components for space work. You begin to build the knowledge and confidence that says, “Okay, we take some bolder steps, like we go all up.” That was one of the contributions I would say that Dr. Miller had to the Apollo program: take a bold step and go all up. You launch it—a Saturn V—with everything on it. Instead of launching the first stage with nothing above it, except that means when you finally get to the first and second, there's nothing but damage above those and so on up. You shove the whole thing together, launch it all at one time, at least up to the point of the payload.
[00:17:39.] RB: Do you think it would have been impossible to have made the manned lunar landing within a decade had you not gone all up, if you'd made it by going in the building block [inaudible]?
[00:17:52] LB: We were within a year, weren't we?
[00:17:55] RB: Yeah.
[00:17:56] LB: Now that year isn't a lot of margin compared to when we began almost ten years earlier.
[00:18:05] JSB: Would you say something about the initial opposition at MSFC to the all up concept, [inaudible]?
[00:18:13] LB: Well, I think opposition is probably not the right word. I would say that we had to go through an inward process of, I'd say, which consisted primarily of systems engineering in our own minds of what that really meant. That does have a bearing on the kind of test program that is conducted on the ground, on your hardware. With all those factors thrown together where we did have a test program that then did give us the assurance that we were committed to that kind of a system now. For someone to come off, you know, to come into the system and say, “Okay, we're going to go all up,” that's one thing. But to say that and not allow a comprehensive test program to give you the kind of background you need, that would be one thing compared to structuring a test program that allows you to do more on the ground before you go all up.
[00:19:28] LB: I think that we went through the logical sequence of engineering analyses before we said, “Okay, we will commit.” That's just the way you run a program. I think that, except those kinds of, I would say, rather dramatic changes in approach without thinking and engineering your way through it would be a rather dangerous thing to do. Now, compared to the beginning, early ‘50s on large launch vehicles for the DOD application, we accounted for a certain amount of our first launches as part of the development of the vehicle. We would expect that we would have some problems in the subsystems of the launch vehicle and maybe lose several flights in the series of developing that vehicle. Now, that completely changed in this Apollo thing, which really said we've got to do better on the use of our experience—which we did—and we have to put together a doggone well thought through program that’s systems engineered and then backed up with the right test to build that confidence.
[00:20:49] RB: Is that because it was so expensive or because it was manned?
[00:20:52] LB: Well, it's what you would want to do on anything, not exclusively tied to launch vehicles or space stuff. If you had to kind of back down the experience, it's just a logical thing to do with any kind of product. You want to do the most you can with the data. That’s engineering analysis, [inaudible] most useful tools that we have, and then the testing of the systems on the ground and where the biggest cost is, try to minimize that effort. The biggest cost in this would be building several of those just for proving the hardware by flight. We minimize that effort by doing a whole lot more with the knowledge we have. It requires a much higher level of engineering talent, especially oriented toward systems engineering talent, when you take those kinds of steps. It was the right kind of thing to do looking back even though we had some reservations in the beginning.
[00:22:03] JSB: Were there technologies that came about in the ‘60s that made that possible, like automatic checkouts?
[00:22:09] LB: Well, I think the technologies were basically there. I think the better use of that capability contributed an awful lot to that. Certainly the kind of check out gear that one has, that we had at that time, had a great deal to do with, yes. When you run a check out and everything is hard lined and everything has to do with man being completely, in every piece of data extracted, that begins to get rather undoable when you talk about Apollo type vehicle—the Saturn and Apollo spacecraft on top of it. You just have to exploit the automatic check out. At the beginning of the ‘60s, I think one would have to say that basic technology was all there. It just took a challenge like the Apollo-Saturn to put that to use. When you commit to Apollo-Saturn in a decade, you doggone well have to put that to use. There's no choice. You don't make it if you don't. So I guess the answer is yes to your question.
[00:23:35] RB: What about the engine programs? Can you say that liquid hydrogen caused you more problems than the LOX/RP-1? Do you have to make a [inaudible] that there are two different kinds of engines and systems involved?
[00:23:52] LB: Well, I would say that you can go all the way back on engine development. At one time, oxygen was called secret fuel. You can go that far probably because it was less understood than other propellants like kerosene or alcohol. It just has to do with one's understanding of what he's working with. Liquid hydrogen/liquid oxygen is a combination which is the propellant in the RL-10 engine and also the J-2 engine.
[00:24:35] LB: At the beginning, there were concerns, yes, because the combination is very combustible. When exposed to the normal atmosphere, it does evaporate. You don't see the run like you do the rest of the kerosene. To the average common guy that works these systems, the technicians and all, they're less known about that combination. I think that had a bit to do with it.
The actual development of an engine has to consider lots of things like the flow and cooling characteristics, combustibility, and susceptibility of these various propellants as far as rough combustion, which is a very major part of the development on the engine on the side of the F-1. So all those were less known than the propellants we'd been dealing with. The propellants we were dealing with primarily at the time in our work were basically kerosene and oxygen. Of course, we'd use alcohol in the Redstone stuff. Then the acids, of course, were used in many of the other systems, but none up until the RL-10 had really used hydrogen and oxygen. So it was just a matter of dealing with something less known by the fact that we hadn't had the time and a number of different systems exposed to it.
00:26:21] LB: Looking at the way it came out, I would say that from my personal experience, hydrogen/oxygen is probably the best combination that we have in rocket engines today as far as ease of developing a system that performs well. The normal things one fears, that's rough combustion and cooling aspects of this type of hardware, does your high performance. It gives a bit of problem to containing it because of the relatively lightweight, pretty large tanks. That was certainly overcome very readily by the fact that something like the S-II stage developed a very large tank, which was very successful. Basically, hydrogen and oxygen engines have been very reliable in comparison to what we termed or categorized as conventional propellants up to that point.
[00:27:31] RB: We were talking to Douglas engineers when we went out to Los Angeles about a year and a half ago. Of course, it was Douglas, but they kept emphasizing that the S-IV, for example, was just an awful lot different than the Centaur. One engineer commented that we had to fight to keep the S-IV from just being another Centaur. Do you see that there are that much differences in the systems, the Centaur and the S-IV?
[00:28:01] LB: Well, the engines are essentially the same. In fact, we delivered engines to the two vehicles out of the same line at the time with some very slight modifications. The difference has to do with the system, the total system tankage and all. There are differences there.
[00:28:21] RB: Internal insulation and so on?
[00:28:23] LB: Well, I would say sensitivity to weight is more critical on a Centaur vehicle than on the S-IV. The S-IV was, in my own assessment, a more conservative approach, not so sensitive to weight as the Centaur. So from a vehicle point of view, yes, quite different. From the engine and what flows through it, basically no difference.
[00:28:56] JSB: [inaudible] primarily the pressure stabilization in the Centaur as opposed to the fact that the S-IVB just stayed up there?
[00:29:02] LB: Well, the booster on the Centaur is pressure stabilized. The Centaur stays pressure stabilized. I thought you were talking about the Centaur second stage, it had the RL-10 in it. It has less conservative design because of the sensitivity to weight than the S-IV. The S-IV, we just built it kind of conservatively. When you do that, you can come out with the test program that's more predictable, I would say more predictable. Hopefully…
[tape cuts out]
[00:29:47] LB: [repeated conversation about Centaur second stage] Hopefully, in the end, it will be cheaper than if you had pushed everything to the ultimate as far as sophistication. That's one thing we found in the area of our Skylab program. Where we can, we've been very conservative on design in order to minimize testing. You can test your test program down, that's the biggest part of your cost.
[00:30:52] JSB: using the rule of thumb, the test program costs about a third of the overall program.
[00:31:05] LB: Well, it depends on where you start if you build facilities and all as we've done in Apollo, that’s probably right. If you maximize the use of those that have been built, like we've done in Skylab, your test program will probably runs a little more than that. We don't have to build new facilities and all that. The test program is probably going to run a little more than that.
That has passed as well as analysis and all that to go as a confident to the test program.
[00:31:49] RB: Are there really big major differences between RL-10 and J-2? Is one essentially better or more efficient than the other? I realize that J-2 came out later so they were able to utilize the technology.
[00:32:09] LB: Well, there are some different concepts that apply to the two as far as the basic principle of design. I think they're designed for applications that you can't compare too easily. I think the record's pretty comparable, one to the other. We really haven't had any major problems on either of these in flight. I'd say that they would compare pretty equal as far as a good, reliable piece of hardware.
[00:33:04] RB: Did you develop any affections for a piece of hardware? You know, you like one better than the other. Does that ever happen? One engine seems to be a sweeter engine than the other?
[00:33:15] LB: Well, I think that you probably develop more respect for some. The F-1, the early years of that development, certainly developed a healthy respect for its capability to act up in the area of rough combustion. We put a major effort into solving the rough combustion problem. As such, developed techniques that I feel have contributed greatly to that technology. We took steps in that development cycle that hadn't been done before and did it in such a way that we were able to develop models. Those models I feel have and will contribute to any similar kind of engine development in the future.
[00:34:14] RB: As I recall there was some comment that when the F-1 combustion stability problem came up, there was a real problem because research on that area had stopped back in the ‘50s when it came up. It was never really solved. So when it came to the F-1, there was really a gap there. You almost had to really go into the basic R&D program. Do you recall that?
[00:34:34] LB: Well, yes. Research on combustion has been something that's rather broadly studied by several universities like Purdue. There's one. Of course, that kind of effort is done basically on small-scale hardware. You get certain types of inputs that give you the basis to do basic engine or injector design. We found on going from the H-1 size to the F-1 size that scale factor doesn't always apply. There's a certain amount of the data from the research technology that just isn't available to you. You get into the scale that your hardware is built in, your actual scale. It just doesn't scale up. In the development of the F1, we did produce that kind of data. I think it's been done in a way that people now have the confidence to think about a new engine program that they can produce it. They can produce it to be a reliable system and not fear that you have random combustion phenomena that causes you to go rough in flight. If you look back at the history of engine development and see the results of flights, there was quite a bit of problem in flights, especially in the development of flights where rough combustion was a major factor.
[00:36:29] RB: Prior to the Saturn?
[00:36:30]] LB: That was in the ‘50s. This came about in the early 60s. It's the logical evolution of knowledge one accumulates. When you go through the kind of thing we went through in the ‘50s in the development of vehicles and engines as part of it. Much of that was done on a rather crash basis. You go about it in one way. You go about it in building [you cut and try?] a lot. When you get into an Apollo kind of vehicle, you don't allocate four, five, or six launches to just seeing that thing's going to run smooth. I get back to what I said initially: you're going to do the most you can on the ground. If you can't do that, you really don't progress to the point you can afford these kind of programs.
[00:37:33] RB: Would you say that the F-1 combustion instability was really one of the key epochs in the development of the Apollo-Saturn launch vehicle?
[00:37:46] LB: Well, it could have been one of the things that would have kept you from having a vehicle like this. It's not by chance that you get that kind of a system built by applying one heck of a lot of know-how that's been built up by having accumulated that through past experiences. It's your ability to bring that to bear on your present—present being the F-1 in this case—program. That really gets you down to what about the confidence that's been gained in that field if we continue to have such vehicles as Apollo. It's very, very difficult to come by. One thing that one has to always remember, we shouldn't lose that confidence. That confidence is vested in, I'd say, not a heck of a lot of people. It's documented, but sometimes there's more to it than just what's documented.
[00:38:52] RB: Some guy out there that doesn't have to feel that way.
[00:38:55] LB: Well, that's a part of it. There's less of that black magic stuff now than there was. There used to be a lot of it way back there. There's a heck of a lot less of it than before. I feel that most of us have been in that field. It's a fairly straightforward design development cycle now to build a high-performance system engine.
[00:39:25] JSB: [Inaudible]. Would you say something about the E-1 program, which was originally thought up for the Saturn I, clustering four E-1s?
[00:39:36] LB: That was another large hydrogen-oxygen engine development, or at least technology-sponsored program, that would have been an engine for a next series of launch vehicles if there had been a need for something larger than the Saturn series, or if we had found that it's better to use single engines instead of clustered engines. Of course, our success in clustering engines kind of took that one out as far as having to have to replace the clustered J-2 engine. However...
[00:40:23] JSB: Are you talking about the M-1 now?
[00:40:25] LB: Oh, I'm talking about the M-1.
[00:40:27] JSB: Good, I wanted to hear about the M-1.
[00:40:29] LB: The M-1 is the one I’m talking about. I'm sorry. So that thing began at the time before we launched the cluster of J-2s and before that cluster concept had been fully proven. Had large numbers of launch vehicles been in the national inventory of products, it may have been economical to substitute one engine in place of five. One doesn't know, it might have been. But that work took its place, you know, it went so far, and there are certain other factors that said, “Well, that one isn't one that's worthy of continuing,” so it was dropped. Again, you know the development cycle, you know the time it takes, and if you project that there may be in the decade some need, sometimes you have to put a certain investment into technology or into the very beginning of the development cycle if you ever want to exercise the option.
[00:41:49] JSB: Well, can't you say the same thing as the reactor in flight programming and the NERVA engine? Many of the early studies, going back to about the ‘60s, you can see that the—well, the one to one—is designed for optimal nuclear upper stage. Then by about ‘62, the program is being continued, but not very much money has been put into it. It's obvious, it’s not going to be used on the Apollo program, and that probably drops out about [‘66?]. We do want to say something about it, that, you know...
[00:42:22] LB: I'm not…I haven't been close really to the NERVA engine development. As you know that development has been conducted out of a different segment of NASA. I think anyone in this kind of business has followed it, they know that it's been carried on at a level that sort of says “Here's what can be done if there's ever a need.” There were a lot of people, associated studies that said “Okay, if there's ever a need, here's what can be done with Saturn type hardware, accommodating flight of that kind of system.” So yes, that's an investment in the case certain events that take place that says that's what we want to fly. I think NERVA was carried on probably—probably rightfully—because it is pushing the start of the art. It's something that's not so conventional as our liquid propulsion engines. It was carried on through a longer period of time to a certain point, what you could do if there was a need to do it. You also knew the things you had to design around and order for it. I really don't have any strong contribution to that I could make to the [inaudible] family of vehicles and its potential future.
[00:44:15] RB: One of the things that John and I have been interested in is where some of the technology filters in. We've been talking to Dave Christensen a little bit about the origins and incorporation of Rigi-Mesh and RL-10 and the J-2 injectors. Do you remember from then how that came about, the sources of that technology?
[00:44:38] LB: The sources are dictated by the problem that one finds when putting designs to hardware and testing hardware. In the design of an engine, one of the biggest problems we've had—have always had—is the burning of the face of an injector. It overheats, and you burn a hole, and you don't have an engine anymore. The skill and the design of injectors in that doggone of the injector is provided the kind of cooling that will disallow the heat of combustion from burning the face of it, which means burning holes in it, and then you have an explosion. Anyone knows you can do that if you brute force utilize your propellant for a film cooling, which also takes away from your performance. You know, you can cool the devil out of it by just brute force, but in all my experience, that always degraded or lowered your performance. You can't tolerate throwing your performance away, just to keep the injector cool.
[00:46:02] JSB: Is this degrading in performance just called general cooling of the combustion chamber?
[00:46:05] LB: Say if you have an injector plate, and you distribute your propellants through it and through orifices or there’s all kinds of systems used. If you're not careful, you'll have back flows, and you'll burn through it and have a problem, especially around the periphery. If you want, you can drill holes around the periphery to squirt right across that thing, throw a lot of liquid, so to speak, on it and cool it. Same in the middle. But that particular propellant—it would be probably your fuel instead of your oxidizer—would not burn efficiently. You don't get the performance, your [inaudible] specific out of it. It just lowers your performance. In turn, you’re washing through your nozzle unburned propellant. Really, that’s what happens. It's like the bid on ecology. Your automobile, as they say, put too many hydrocarbons out into the streets and the air. You want to burn it all. That’s the same thing.
[00:47:18] LB: At any rate, through the development of the hydrogen-oxygen engine, they found that in order to get performance, they had to do something other than what we had conventionally been able to do on the oxygen-kerosene engine as far as cooling. The hydrogen and oxygen being as it flows much more readily through a mesh. You can't squirt kerosene through there very easily whereas you can put oxygen or hydrogen through it, hydrogen especially. It is cold. Temperature-wise it is favorable. Just by them getting into tests and trying things, they did come up with a Rigi-Mesh as a structural member and a flow passage for a certain percent of the propellant through it. It cooled it, and it distributed it in a way that they could work their design out where they had a satisfactory ratio of oxygen compared to the fuel. They were able to make a design that yielded a very high performance. They kept their performance. They got their cooling from it. They gained their structural strength also from that Rigi-Mesh. It's very…It’s metal, you know? It's a very good piece of structure for the injector if anchored right against posts and stuff like that.
[00:49:06] RB: Was that a fairly significant breakthrough in terms of LOX/LH2 engines?
[00:49:11] LB: Oh yeah. Yeah, that was [inaudible] J-2.
[00:49:17] RB: Did you have a couple getting Rocketdyne to take Rigi-Mesh? Since it had been used on Pratt & Whitney engines?
[00:49:24] LB: No. These companies tend to pretty well come in on similar things. It's not a closed environment. The producers and all that are there, they're pretty doggone dynamic environment that you find these companies in. There’s common denominator all across the face of this business: the government, the DODs, the NASA types and all that, and the suppliers and all. It doesn't serve anyone's purpose to let people find out everything in series. One year and five years later another might pick it up and at the same time you're having to throw money out the doggone window to let each guy do his complete thing. It's something that's extremely successful, who’s one to fight it? No, that really hasn't been a major factor in my experience in any of this engine development. The people, you know, transfer their own knowledge by moving around a little bit too. You find people that would work one place and then they’d work at another.
[00:51:01] RB: When a lot of it was done under government contracts anyway, it becomes…It's not the proprietary thing, it’s a government [inaudible]....
[00:51:07] LB: A lot of it is.
[00:51:11] JSB: Presumably the supplier Rigi-Mesh, which is the Pall Company, would be visiting all these people anyway.
[00:51:16] LB: They'll try to get their product in as much as they can. I guess I couldn’t sit here and say who thought about what first. Even with the research and the literature, you could have a hard time saying, “Now that's where it showed up first.”
[00:51:33] JSB: That’s one of our biggest problems.
[00:51:35] LB: I'd say that to do that would be…Well, you could do that and probably raise a lot of interest and debates. Things turn to a great extent if you write a history, say “This guy did that first and another guy did something else first.” That'd be hard to…Some guy, someone in the middle might say, “I thought about that too.” Who knows? [laughs] They wouldn't serve too much of a purpose in my mind.
[00:52:12] RB: Well, John, do you have any more questions?
[00:52:14] JSB: I’d like to ask one thing about the institution of the fifth engine between the C-4 and C-5 configurations. Is deciding to use the anchoring engine in the center to give the extra capability for whatever the [inaudible] or Houston might do for payload?
[00:52:32] LB: On Saturn V?
[00:52:35] JSB: Mm-hmm. Putting that one in the center.
[00:52:37] LB: Your vehicle is sized by the payload, by the mission. Oftentimes you begin a development—a design for a development—based on what someone commits he can do on the top. He says, “I can do this kind of a design and come out with a certain weight and still conduct a certain mission.” You may start that way, but then a few months later you may come back and say, “Whoops, I got a little growth here.” If you want a weight reduction program that early in the cycle, you can pretty well say you can't make it with weight. My experience is there's no more expensive kind of development than a weight reduction program. That's about the most expensive thing you can get into because that just connotates [sic] that you're going to have to do a fairly sophisticated redesign, and one hell of a lot of proof testing, and then you're hanging on too marginal a safety factor. That's kind of not the way to go. If you can accommodate the problem by increasing performance of the booster, that's probably your best bet. Most conservative, most [inaudible] forward, cost effective thing you can do is increased performance of your boosters.
[00:54:10] JSB: How much appears in the literature that Marshall anticipated the payload growth by the engineers?
[00:54:16] LB: Yeah, well, of course, you don't get into a program blindly, build a vehicle without looking up what you're committed to do. When you look at what you're committed to do, you look at the other fellows' calculations and assumptions and all that, and you make your own assessment. You very well may build in a little more conservatism based on your own knowledge of what it takes. We didn't begin that in the dark. We've been through that payload growth thing since the early ‘50s. Always it's been a case of having increased performance. That wasn't unique to just Saturn.
[00:55:09] JSB: The first S-IC, was it 7.5 million pounds, they uprated that somewhat now haven’t they, for wider missions to take care of the rover?
[00:55:21] LB: I think that as you mature and as you understand by experience what you have, you can always commit more of what is reserved than you could to begin with as you don't understand the detail that you have for say ten or twelve months. You begin to know that thing doggone well, so you're able to commit a much, much more narrow margin between calculated and [rationals?]l. Anytime there's a need for additional payload and you have a mature vehicle, you can always go in there and see what you can increase in the way of additional capability. After you mature, you know whether it's there or not. A lot of the capability is based on safety factors of your system. By actual use and testing, you begin to know how good those safety factors are. They can be off one point, say, a tenth of a unit, and increase your payload significantly. The very first launch you probably can't commit to that. Maybe after ten or twelve, you say, “Well, that's an acceptable”—not risk but—”It's an acceptable situation based on all the facts.” No difference in performance of any vehicle, automobiles included.
[00:57:12] JSB: What gave you your greatest thrill in the Saturn program as you look back on it?
[00:57:23] LB: Well, I would say the fact that in the total Saturn-Apollo program, the fact that we were, in a period of about less than a decade, able to go from a certain approach to a rather complete, different approach on the whole development business cycle and operational cycle, and find that at the end we were able to uphold a national commitment.
[tape ends]
Duration
0:58:31
Files
Collection
Citation
“Belew, Leland,” The UAH Archives and Special Collections, accessed August 24, 2026, https://oralhistory.uah.edu/items/show/560.