Pease, Robert
Dublin Core
Title
Pease, Robert
Description
J-2 Engine
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_000033_A
Oral History Item Type Metadata
Interviewer
Bilstein, Roger E.
Interviewee
Pease, Robert
Transcription
[00:00:00] Roger Bilstein: Interview with Bob Pease, J-2 engine. Now one of these things in the design of the J-2, I get the impression that it was designed to be tested at altitude and sea level or something like that. But no, designed to be tested at sea level although it was going to be fired very high. Was this a particular problem? Was it the first engine to be designed that way at least in the liquid hydrogen technology class?
[00:00:32] Bob Pease: No, let me say that there was a particular desire to be able to fire the complete engine including the bell nozzle which was a 27.5 to 1 air ratio nozzle, under the sea level test stand conditions. The reason for doing this of course was number one to ensure that we had the ability to fire the engine with more ease, which you can do in a sea level environment than if you have to do it in an altitude chamber. The test frequency—the question was important—so the engine bell nozzle was designed to flow full at the sea level test conditions. This was important in order to get the proper heat fluxes and so forth into the chamber. In other words, the chamber was being cooled by hydrogen. The hydrogen of course picks up heat as it goes up through the wall and comes into the injector, so trying to get as near as possible the proper heat transfer conditions on the chamber under sea level firing. Another thing involved was the fact that if the nozzle didn't flow full, not only would it affect the heat transfer to the thrust chamber wall, it would also make it difficult to [tape cuts out] …Not recording?
[00:02:36] RB: Okay, we're on again. [laughs]
[00:02:38] BP: Did you miss all that or should I go?
[00:02:40] RB: Why don't you just recapitulate? I think I moved this over here because sometimes we get a lot of overtone noise from air conditioning equipment.
[00:02:48] BP: Well, I think I can shorten that down now.
[00:02:50] RB: Okay.
[00:02:52] BP: I think if I understand your question, you're saying…The basic question you asked was the engine, some reason for having the engine designed to both be fired at sea level and at altitude, and was this a first in the engine business? To recapitulate, yes, there was a reason for having the engine be able to fire both at sea level and altitude. The reason to have it fire at sea level is in order to obtain more easily the testing, numbers of tests and so forth, and to do it under an easier test environment than you have to when you're playing with the diffusers and altitude equipment. That facilitates your development effort that way. Because of this, the engine was designed to give as near as possible the proper performance on the sea level test by having the nozzle flow full. If it doesn't flow full, you have problems in measuring the thrust, and you also have a difference in heat transfer to the thrust chamber wall.
[00:04:23] BP: As it turned out, the engine…There were still differences in the way the engine performed at sea level and the way it performed on the altitude test. There were some changes in the engine power balance due to the difference in pressure ratio between sea level and altitude across the turbine areas. There was also some difference in the heat transfer to the thrust chamber wall simply by the fact that under the altitude conditions you had, I'd say, a very reduced atmospheric situation, and you weren't even transferring atmospheric heat into the thrust chamber wall. [laughs] There were some differences, both in those two areas. However, I don't know if I'm answering your question completely on this. Was this a first? I think it was a first from the standpoint of trying to get a nozzle of a higher expansion ratio to flow fully at sea level. The nozzle was a special contour which was designed in order to facilitate this. That part of it was a first. I believe on previous programs, such as the RL-10, we had to use some type of diffuser system if we were going to flow the full, complete [engine?] nozzle. So I don't know, does that…?
[00:06:07] RB: Yeah. That answers the question, I think. I’ll have to do some research on it too, try and find out some more things. Maybe I can pick up the phone sometime, just call you back.
[00:06:16] BP: Yeah, fine.
[00:06:17] RB: Now, they did use vacuum diffuser equipment in the J-2 development.
[00:06:22] BP: Yes, they did.
[00:06:23] RB: Was there a difference, say, in the diffuser equipment that Rocketdyne installed there and what was used at Tullahoma?
[00:06:31] BP: Yes.
[00:06:32] RB: Why was there a difference? I mean, how did the difference occur and what was the intent? Why do some of the testing at Susanna and why do some at Tullahoma?
[00:06:41] BP: Well, what amounts to this is that when the program started, it was recognized that there still had to be an altitude system type test to verify the engine behavior on their altitude. Even though the engine was designed such that it could be tested at sea level, it was still recognized there would be some differences when testing in an altitude environment. From a very important beginning program outset, the steam operated diffuser was planned at Santa Susanna to evaluate the engine under vacuum operating conditions. The steam operated diffuser operated in such a manner as that steam was supplied during the engine start transient, which pulled the capsule pressure down to something in the neighborhood of 80,000 feet. When the engine started, the diffuser would operate in a manner of…It would be activated by the exhaust gases of the nozzle in order then to maintain the capsule vacuum pressure under main stage operating conditions. When the engine shut down, the steam would come on again to try to maintain the altitude conditions. Well, when this system was actually put into operation, it was discovered that the steam system was unable to maintain the 80,000 foot altitude during the start transient. Part of the reason for this was the fact that when the diffuser and steam system were designed, they didn't anticipate the amount of hydrogen fuel lead that was going to be produced by the engine. Hydrogen being densities and so forth it is, this fuel lead very quickly saturated the altitude producing capabilities of the steam system such that the altitude would drop down to less than 20,000 feet during the start.
[00:09:12] RB: If you're shooting for 80, that's not a very good condition. [laughs]
[00:09:16] BP: Right. Not only that, they also found that there was considerable steam blowback and turbulence that came back and got into the nozzle area, which also fouled up the heat transfer inputs to the chamber one thing or another. Between the pressure ratio across the oxidizer turbine and the steam blowback, it was felt that our altitude simulation was less than optimal. From that standpoint, we then made an investigation to find out if the Tullahoma facility would be available in order to run some additional series of altitude tests. In fact, along with this investigation, we also investigated constructing another altitude facility at Santa Susanna, which would have been operated by another type of steam system, one which would have kept the altitudes up in the neighborhood of 90 to 100,000 feet at all times. Rocketdyne had successfully demonstrated what they called a hyper flow steam system using basically an H-1 engine to generate steam. Instead of having a steam accumulator, which had limited capacities, they were going to fire essentially a rocket engine with water injection, which would have generated a very large steam capacity instantaneously upon demand for keeping the altitude up. That system looked very good and would have cost us, however, several million dollars to put that in at Santa Susanna. In the process of looking at all this, it became apparent that it would be more cost effective to activate the Tullahoma stand, the J-4 cell at Tullahoma for this purpose. That's how the Tullahoma program came into being. Once we got into the J-4 cell, we did find that we did learn some further knowledge as to the engine starting characteristics that wasn't apparent at the testing at Santa Susanna, either under the altitude or the sea level testing.
[00:11:55] RB: So you had to do a little redesign?
[00:11:580] BP: Yeah, there did cause some redesign, but it was not of a major nature, mainly orificing and balancing of the engine start primarily. The Tullahoma conditions on the transfer to the thrust chamber and so forth were further verified as being accurate. This is just a side note, when I looked at the flight data, they found that the flight data substantiated our performance results at Tullahoma. Tullahoma was a pretty good facility for that standpoint. One thing that Tullahoma didn't do, or any of the testing didn't do, was did not find what we call the ASI problem of 502.
[00:12:51] RB: Can you go into that a little bit?
[00:12:55] BP: Sure.
[00:12:57] RB: We're still running.
[00:13:02] BP: You might have already got some history on this, you may have known, we did fail an ASI line in the S-II stage and also in the S-IVB stage on flight 502. The ASI fuel line was later discovered to be a fatigue failure of the line caused by flow vibration, or its flow induced vibration. Turns out that this was not discovered under sea level testing or even under altitude conditions of the Tullahoma stand because, well, number one, under sea level conditions, there's a condensation of air that gets in between the bellows section. This does two things. Number one, it's maybe a lesser effect of it, but it does provide some damping in the bellows. Secondly, it does affect the heat transfer boundary layer conditions on the inside of the flow path. When Rocketdyne ran laboratory tests and took photographs, movie photographs, and made other flow dynamics studies, they found out that the heating effect of the line actually changed the flow induced vibration characteristics. I'm not a flow vibration expert, so I'm trying to explain it in…
[00:14:49] RB: Neither am I. [both laugh]
[00:14:53] BP:…Simple terms as I understand it. In other words, apparently it changed the boundary layer conditions on the inside of the line so that the flow pattern wasn't the same. Now, why didn't this happen at Tullahoma? Well, there's a good altitude condition we were obtaining—100,000 feet or so—there's still enough nitrogen present in the cell to—this is a theory is presumed on to, at least the calculations have been made—to provide the same effect.
[00:15:35] RB: Okay, I didn't understand that then. So was nitrogen still existing in the cell to create that kind of frost buildup and have the same damping effect?
[00:15:43] BP: Well, there's both nitrogen and water, and I say nitrogen because I believe in Rocketdyne's testing they proved that frost alone would not prevent the failure. It really, really had to be the air condensation. In other words, yeah, if it were packed with ice, it's one thing, but a minor frost buildup, I think alone was not considered sufficient to prevent the failure. The way I understand it now is that they believe there was enough nitrogen condensation within the cell to have done this. Now, I guess one way of…Maybe another way of tackling the problem is that I don't really know if this is a feasible way if nitrogen could have been substituted for some other gas such as a helium purge. The reason they don't like helium is because the lighter the gas is the harder it is to pump it out.
[00:16:45] RB: But isn't that the way they finally got to fix on it though? They put it in a helium tank or something?
[00:16:51] BP: Yes, as a matter of fact, in the testing at Santa Susanna, they tested with a helium environment, vacuum environment, with [arid?] environments, and they found out that it would fail under the helium environment because the helium wouldn't condense. It would fail under the vacuum environment, the hard vacuum environment in a test small test capsule, they can pump it down quite a bit higher altitude to Tullahoma. So that's how they finally pinpointed it was by isolating these different environments. [00:17:34] BP: So it's quite an interesting investigation. I think to many of us it's sad that there are always certain problems that one cannot predict in advance and all the analytical techniques cannot even plan for. Well, now that we know about this one, we will always plan for it. I'm sure that in the space shuttle management program or other future programs, we will still continue to plan altitude type testing. Although in this particular case on J-2, it didn't find this problem.
[00:18:22] RB: You were out at Rocket...Were you the MSFC representative out there?
[00:18:27] BP: I was the resident manager for the J-2.
[00:18:32] RB: Do you remember very much about the story on the injector face and the use of Rigi-Mesh? Were you involved in that?
[00:18:42] BP: Yes.
[00:18:43] RB: Could you give me a little bit more?
[00:18:45] BP: What would you like to know about it?
[00:18:46] RB: It's my understanding that it was the Pall Corporation that came up with the Rigi-Mesh stuff in New York, and they got this out of the Atomic Energy Commission program and used it for filtering gases or something?
[00:19:00] BP: Yes, I believe that's correct. I know the material was produced for purposes of such of that nature. I guess it also had been used on the RL-10 program.
[00:19:15] RB: My question is this. You had some problems with the injector, I guess. Well, normal development problems. Where does the decision come to get into Rigi-Mesh? Does Marshall say that?
[00:19:28] BP: Well, I just happened to be intimately involved in that one. We started the J-2 program. Rocketdyne had proposed a copper ring type injector, which was typical of most previous Rocketdyne designs with the various orifice holes drilled at various angles in the copper. The Rigi-Mesh had been used successfully on the RL-10 program and had been in fact introduced to the RL-10 program by Lewis people. Before that, where they got it from—I think what you say is right—the Atomic Energy Commission may have used it for filtering devices, one way or another. But it was used in the RL-10 program as an injector plate medium in order to allow the hydrogen to filter through and partially cool the face by that means. People were concerned at that time about the problem of cooling the face and felt that Lewis people felt we might have some problems with the copper. In the early part of the J-2 program, we got back to the initial discussions on how to get it into the program. It was suggested to Rocketdyne by NASA that they include in the development program an alternate design of a Rigi-Mesh face type injector, and that they evaluate both the copper design which they proposed and this alternate Rigi-Mesh injector. Sometime downstream, there would be a decision made as to which one was best.
[00:21:27] RB: In this case, the impetus came from NASA to Rocketdyne?
[00:21:30] BP: Right, and this is part of NASA's transfer of technology, whatever you want to call it. We were aware of the material, we were aware that it had been used in the rocket engine programs, and we felt that Rocketdyne ought to look at it also. So after—I don't remember the exact time frame, but I think it was within probably about the first six to nine months—the early injector testing, they did test several copper versions and several Rigi-Mesh versions. The overall assessment of Rocketdyne came to the conclusion—and NASA agreed—that Rigi-Mesh performed not only performance-wise better in terms of the particular concentric post orifices where the hydrogen came out around each post, each LOX [oxidizer?] post in the Rigi-Mesh face. The design performed well from an Isp standpoint, and it performed well from a cooling standpoint, there was no problem with burning. We did have some problems with the copper injector on burning the rings and lands and so forth. I remember one where we had green flame coming out the chamber during one test. [both laugh] I'm sure they could have solved the copper problem, I'm sure they could have solved the cooling problem, but it just turned out that in that evaluation the Rigi-Mesh was best.
[00:23:04] RB: This is an interesting feature it seems to me of the Saturn, or NASA programs generally, that you do hear here is Marshall operating, cooperating with Lewis and with the contractor. It's the kind of technology of one center, in this case Lewis, that comes through Marshall into the Rocketdyne facility, that there’s this interchanging [inaudible]...
[00:23:23] BP: I remember that particular one because I had worked with the Lewis people on the technical committee of the source evaluation for the J-2. In that capacity I got familiar with those people and that particular subject. I was a part of convincing Rocketdyne that we should do this.
[00:23:53] RB: Now again there's a question I wanted to ask, kind of a specific one. Rocketdyne got the contract and started to work to put together an experimental engine and later fired an experimental engine. Now where the hell do you get an experimental engine? Do you just go out in the shop and start hand working an injector face, put together a few tubes? How do you do that?
[00:24:18] BP: Well, you first have to design it.
[00:24:20] RB: Yeah [laughs]. Well, what I'm getting at is there's no manufacturing lines set up yet really. The first test engine really is a handmade product. Is it accurate to say that?
[00:24:32] BP: No, no it isn't. Now the J-2, Rocketdyne designed a J-2 engine as they envisioned it in their original proposal model.
[00:24:43] RB: Oh, the computer model. I know they used a lot of computer stuff, configurations…
[00:24:47] BP: That's true, they looked at their analysis, and they sized the ducts, tubes and so forth with a proper thrust chamber size with a proper heat transfer and all that type of thing. In other words, they designed the engine from a dimensional standpoint based on all their analysis. They went out and released orders for material, which is not just raw material but vendors to make certain basic castings and forgings and so forth, and even some components which they purchased from vendors, which was either modification off the shelf items like small valves and things of that sort. Or they actually ordered components from certain vendors to their own specifications. This material comes in house, it's machined to the drawings; it's checked out to the specifications; and they put an engine together. Now they call this an experimental engine. I believe the first one was essentially called “Experimental Engine” because they had very little if any experience with the major components. One exception was the hydrogen turbopump, which was basically a modification of one that they had already built and run for the Atomic Energy Commission on the Rover program. We were starting with basically an off the shelf turbopump, which was modified somewhat for the J-2 application. The oxidizer pumps were not too unsimilar from the previous engine programs like the H-1 and the Atlas and other LOX/RP engines. The oxidizer pumps were quite similar in design to what they had used before. Some modifications such that J-2 they used separate pumps rather than two pumps run by a gearbox. Each pump had its own turbine and that was a new innovation from previous engines.
[00:27:12] BP: It is not just a handmade type of thing. It's actually analyzed, designed, hardware is released and built, and so forth. Not only that, they didn't just release one set of hardware, they released groups of hardware. I think they released maybe four or five engines worth of hardware, probably even the initial release. From the first engine firing, the experimental engine, they found out some things, which they went back and made some corrections to. The hardware was already in the pipeline. It was like they might have to, well, I can't think of a specific example that occurred on the J-2 but…Well, I think of one thing that happened, our famous 27.5 to 1 thrust chamber, which was flows full at sea level, developed a phenomenon which they hadn't predicted, and that's called side loads. During the start transient, a very eccentric load developed in the chamber, forcing the chamber over to one side. Some of the first engines they ran, they actually physically distorted the chamber, distorted the actuator systems, and things trying to hold the chamber in place, engine in place. They had to quick come up with what they call a horse collar, which was a device that came in and grabbed the chamber at the throat. It would hold the engine stable, take the load during the start, and once it was running then the thing would drop away so the engine could be gimbaled. But that was not predicted, and that's an example of a kind of a modification to find out you have to make.
[00:29:02] RB: Were they able to work out the side load problem that they were on and get rid of that horse collar thing?
[00:29:07] BP: No.
[00:29:08] RB: That stayed in the whole program?
[00:29:09] BP: Yes it did. The S-II stage and ground firing used devices to hold the chambers in place.
[00:29:14] RB: And the S-IVB then also has the same thing?
[00:29:17] BP: During ground firing.
[00:29:18] RB: During ground firing.
[00:29:19] BP: And that's ground equipment.
[00:29:20] RB: Okay, it's not flight hardware?
[00:29:21] BP: No, the side loads don't exist in flight because of the vacuum condition.
[00:29:26] RB: Okay.
[tape cuts out]
[00:29:27] RB: It takes a minute or a second to get here, to run up. Okay, I think we're probably back on here.
[00:29:35] BP: Okay, so I kind of went beyond maybe the scope of your original question.
[00:29:39] RB: No, I'm glad to find out. I didn't know about it.
[00:29:41] BP: There were any number of problems that, of course, came out of the early engine firings, which caused either modification of hardware, which had been released in the system, or certainly got cranked into the next batch of hardware design that was going to be released in the system.
[00:30:04] RB: Could you pick out a problem that you remember, especially that developed during the test and qualification program that brought about a redesign of hardware?
[00:30:14] BP: We had lots of those. [laughs] I don't know if there's any one you want me to emphasize...
[00:30:22] RB: We read the NASA press releases and the AIAA presentations by Rocketdyne people—which isn't to say anything against them, you know?—but you get the impression that this was a beautiful program. You know there must have been a problem somewhere.
[00:30:36] BP: I guess this is a matter of debate. I don't know. My own personal philosophy is that you have a development program to find out what's wrong with things and correct it. In other words, I think it's more accurate to say that happiness should be finding a failure rather than not finding a failure because if you don't find it, then you're going to be hung up sometime in a critical mission or flight or something of that sort. The J-2 program did find lots of failures and had to correct them during the program. The most difficult part of it came into play when we had to release engines for production, and we had engines in the production pipeline, and we were still finding failures. It's particularly difficult because then it means that the solution has to be found that we can crank into the production line and fix that particular piece of production hardware. Not only before it's delivered from the line, but we had cases where engines had already been delivered to the field that we had to send out modification kits to fix them. A lot of people look at this as saying, “Well gee, why didn't you find all those things before you sent those engines out to production?” Truth of the matter is that the program always needs the production hardware before it's fully developed.
[00:32:18] RB: For schedules? To meet your schedules?
[00:32:21] BP: For schedules. We have not only on the Saturn program, but on every program in the missile rocket business that I've ever looked at including the Air Force program, we have a situation of what we call concurrent development and production. General Schriever recognized that during the early IBM days. We certainly had it during the Saturn Apollo. The engine being one of the long lead items for the entire system suffers particularly because the vehicle people would like to get an operational engine cranked in their vehicle as soon as possible. Many times they'd like to have it rather than waiting to have it at some point where you install the engine down at the Cape before you launch it. They're not satisfied with that. Usually they want to crank the engine in in the early part of the vehicle build up cycle back at the vehicle manufacturing area so that the whole system, engine to vehicle, goes through the system checkouts. Maybe sometime in the future, we can promote a philosophy to give us more time in the engine area where we can install the engines at the other end of the line. [both laugh] Which just means maybe as much as a year of expert development and maturity.
[00:33:47] RB: Can you give an example? Do you remember what kind of modifications you were sending out?
[00:33:53] BP: Yeah. Well, let me give an example that hit us in the production line before we were just beginning to deliver the first flight engine for Saturn IB, which was going to Douglas with engine number fifteen, 2015 I believe. Now this engine was in the stages of final assembly when we discovered that we had burned out a section of the gas generator wall, actually put a hole in it. In fact, I believe this happened on one of the battleship engines that we had sent to Douglas Sacramento. It happened on another engine or two that were being fired in Santa Susanna. This launched an intensive development effort to solve this problem, which Rocketdyne came up with a change to the throat of the gas generator. They call it [inaudible] choke ring, which diverted the gases somewhat so they didn't impinge on the wall. So this [inaudible] choke ring then had to be worked into all the gas generators that were on the line. It meant actually cutting the gas generator, which was welded to the fuel turbine [manifold?], taking this assembly off, reworking it, re-welding everything back up again. It was quite a re-working turnaround problem. It was a modification that could be incorporated into existing hardware to solve the problem. That's an example of a problem that hit us during the manufacturing cycle.
[00:35:56] BP: One of the most serious development problems I think in terms of seriousness that we found was the fact that we had some very definite problems in the engine start sequence. During an attempted semi-qualification, which we call PFRT testing, we did discover that we did have a control problem or a sequencing problem. We started build up which actually caused the LOX valve to be forced closed such that the LOX valve had such hydraulic pressure against it it wouldn't open properly. This caused excessive back pressure since the gas generator was tapped off above the LOX valve. The gas generator LOX balance went LOX rich. We got too hot temperatures and burned out the turbine and a few other things. This did point out the fact we had a problem there.
[00:37:17] BP: We'd also been having problems in starting the engine due to the hydrogen pump stall, which in case power balance was very sensitive between the LOX and fuel turbine speed buildups. If the pressure in the chamber built up too fast because the LOX pump was putting more LOX in the main thrust chamber causing the main thrust chamber pressure to build up too fast, this caused excessive back pressure on the hydrogen side. It got ahead, you might say, of the hydrogen pump buildup, and we had that condition which forced the operating and the operating curve of the pump, which pushed it over into the stall region. The pump would stop pumping. That was called hydrogen stall.
[00:38:13] BP: There was extensive development effort launched to find improvements to the start sequence, which necessitated running a much larger number of engine system tests than we had originally planned. The only way you could work on this problem was to run the engine system. The other problem I mentioned when we burned out the gas generator, we solved the problem in the gas generator pit. In this case, there was no other way to solve the problem except to make many, many cut and tries, as it will, at the engine test itself. Rocketdyne used all of their analytical methods in the start computer, which they had developed and so forth, and tried to make refinements on that computer in order to get a better handle on things. They used all the techniques that they could. Really what it came down to is that you had to make an adjustment and then try it. The final solution had to be tested under altitude conditions to make sure that it was also valid because of the altitude affecting the start.
[00:39:37] RB: Okay, that's very good. You know, these are a good example, I think, you know, of what happened [inaudible]. Another brief question here, and then I've taken enough of your time, I think. J-2 was built for a mission of around 500 seconds, but the lifetime was about 3,750 seconds around that area.
[00:40:01] BP: Yes, 3,750 was the qualification one.
[00:40:04] RB: Okay, now why the difference? What were you doing inside those 3,750 seconds that you needed that many of when the actual mission was only 500?
[00:40:16] BP: Well, I guess that's a philosophical sort of question. It gets back down to some basic assumptions. Number one is the fact that I guess when an engine is delivered to the Cape for flight, it's already going to have a certain amount of time on it. It's not going to be zero time. There's about 750 seconds allocated for the acceptance test of the engine. Then there was another 500 seconds allocated for the vehicle test, ground test. They did fire the S-II stages and S-IVB stages on the ground for quite a while. I guess they fired all of the S-II stages. Then as a reserve, there were additional time allocated beyond that.
[00:41:19] BP: Suppose you have a vehicle problem or a reason to refire the vehicle for some reason on the ground, you might have to run another complete test. You can rationalize yourself that you might have engines at least 1,500 seconds that would still have to be qualified for flight. It turns out most of them didn't have that much time on them when they got to the Cape. More like, I say 1,500 to 2,000 seconds, you could rationalize. You might be in that range, but most of them were in the 1,000 to 1,200 range that actually turned out. All right, so then beyond that, you say, “Well, why would you need any more time?” This is strictly a matter of judgment in fact that trying to make a maturity reliability type assessment, I think the goal of 3,750 was chosen because it would also increase the confidence level upon which the engine say of some lesser time would have their flight.
[00:42:34] RB: Is this part of the manned rating philosophy too?
[00:42:38] BP: Yes. Well, it was at least in our minds at that time. The fact that we were going to fly manned in previously qualified engines, much less time. We felt that 3,750 seconds was perhaps not an unreasonable amount of time at all for a hydrogen type of an engine. I can't say that this is any magic number for a manned space flight because it wasn't. The F-1 had a qualification time of 2,250 seconds, which was felt reasonable for the conditions that you expect with a LOX/RP type engine. The fuel made a considerable difference because there were more effects on the F-1 life by using RP fuel because of coking, carbon deposits, and other other types of problems that you've got with the fuel. Hydrogen engine stays just brand shiny clean. You don't have those kind of problems, and you can expect a longer life inherently from a hydrogen engine. This was one reason why I'd say somewhat arbitrarily 3,750 was chosen to be at least some measure beyond what we have done in the LOX/RP type engines. That's about it. There are probably some other philosophical reasons that you might get from some other people. I don't know.
[00:44:14] RB: [laughs] Well, do you have any, you know, things that you'd like to say?
[00:44:16] BP: Well, just to make one more little sidelight on this whole thing…In fact, I don't know when the designer looks at 3,750 seconds, whether he designed the bearing carrier much different than whether it was 3,000 or 2,000 or what. We actually ran some R&D engines with, of course, some component replacement here and there, but we had components on R&D engines which had gone well over up to 20,000 seconds. We had several other samples at ten, 15,000, that range.
[00:45:00] RB: It gives you a pretty high confidence factor.
[00:45:02] BP: Well, yeah, as a matter of fact, it's because of this kind of data that we feel rather confident that the space shuttle main engine we can eventually attain the seven and a half hour life on mission requirements. In fact, we don't even know today that some of the J-2 components wouldn't have gone much beyond the 20,000 seconds because we just gave up at that point. Some of them weren't even worn out. If we had to today take a J-2 engine and make it a reusable type of engine to go to a long life—many missions—we have some data, which there will be some areas in the J-2 that we know are life limited, which would have to be redesigned. From a general standpoint, it did give us high confidence in the J-2 from the J-2 data that many of the hydrogen type components would run considerably longer than what we ever tested before.
[00:46:18] RB: Well, I've taken a good share of your time here.
[00:46:22] BP: Well, I understood you wanted about an hour. Anything else you have on your list that you want to go into?
[00:46:30] RB: No, some of the…There are questions that I picked up just reading odds and ends of things. One thing that is really a stupid question, but what does J-2 mean?
[00:46:43] BP: [Laughs] Doesn't mean anything particularly. The Rocketdyne had a system of numbering engines. Well, maybe it's a lack of system. We had…I can't really say where it started. Rocketdyne had an E-1 experimental engine program. They had the F-1 program. It seemed to me there was a D program somewhere that didn't live very long. I don't know what happened to some of the other numbers in between. They probably died in the proposal [mill?] somewhere. But there was H-1, of course, which was picked on H-1. I think they skipped I for obvious reasons. The next thing that came out of the bag after H-1 was the J-1.
[00:47:48] RB: Oh, J-1?
[00:47:50] BP: Rocketdyne peddled this around the country in the advanced design sales department, I'd say, for a while.
[00:48:00] RB: Was it a hydrogen engine?
[00:48:02] BP: Yeah, it was an advanced design concept really, which they called J-1. Then when NASA got serious about going into a development of a hydrogen type engine of the 200,000 pound class, then to distinguish from the version that they had been peddling around the country, they would call it the J-2. I guess basically I don't even think J-2 appeared in their proposal, but when they got down to negotiating the contract, they said, “Well, we're going to call this the J-2 since we had J-1. Does anybody object?” NASA didn't have any numbering system established for engines like the Air Force had the [XLR?] type series. We didn't have any official numbering system, so we said, “We don't know it makes any difference.” Well, J-2. Of course, we already had the H-1 being one, and the F-1 program had already been started. It seemed chronological to us to call it the J-2. [both laugh]
[00:49:13] RB: While you were out there, do you remember any incidents that struck you or remember as being humorous or funny? People out there who were really strong characters, little anecdotes about any of them?
[00:49:30] BP: I probably could if I thought about it a little bit. After all those years, we certainly had a lot of interesting meetings. [laughs] It got pretty wild at times, but let me think a minute here. Well, I really can't remember anything offhand. As I say, I'm sure I would if I thought about it a bit.
[00:50:07] RB: What about Von Braun?
[00:50:09] BP: He used to come out quite often to the West Coast and visit the Rocketdyne and review the program. He attended many of our review meetings that we had in the early days. Later on, he got…Demands were heavier on his time, he came out less often. He usually managed to come out once a year and review all of the activities at Rocketdyne.
[00:50:40] RB: Did you find those helpful, necessarily, or was it something you did because it was for his primary information benefit?
[00:50:47] BP: No, I wouldn't say…Obviously, I think that it was not only a general feeling by the NASA people, but I think the contractor, too, appreciated the opportunity to bring the center director—Von Braun, of course, being the center director at the time—up to date on what was going on. I think this helped the morale a lot because you did feel that the top management was interested in the effort. Not only that, I think that Von Braun himself was quite an inspiration to the program and any meetings that he attended. His comments and observations were—I would say—very, very helpful because it put what we were doing in one detailed part of the center—Apollo—and the whole system somewhat in more context as to really what our objective was. That was helpful. I think those types of meetings perhaps gave us more insight on where we really should be going and probably some of the day-to-day directives that we used to get. [both laugh]
[00:52:17] RB: It's too easy to put the director of File 13 down there, I suppose too.
[00:52:24] BP: My only general comment on the whole thing, I thought it was a very interesting program. I think that certainly learned a lot from the program. Much of what we learned from the program, we will attempt to apply in general to our management of future programs. In other words, not just learning what to do, but certainly a few things that we learned, I hope perhaps, not to do in a future program of this type. One thing I think that we learned, I think might be of specific interest, in looking back over the J-2 program history…We've done a lot of this in the last couple of years, going back into our history and deciding how we might possibly improve the approach, the management as a development approach in the management. This is even over and above the technical things that are obviously going into the new program. We're trying to apply this to our management approach now, the take on the shuttle engines. Just for example of this, I think we learned in looking over the J-2 program we need to have more emphasis on the component and subsystems testing. Also, we learned we need to get in earlier to testing limits of the hardware in J-2. For example, we had run the hydrogen turbopump a number of months at some fairly nominal type missions. They're within specifications, so to speak. Performance was fine, no problems. Well, I won't say no problems, but the things seemed to perform well basically. When they started to run what you might call higher limits to higher power levels, higher speeds and things of this sort, it might be on the outer fringes of the operational environment. In other words, the engine probably wouldn't run there, except maybe under some unusual conditions or what have you. We failed the turbine wheel, and found we got into a serious vibration problem with the turbine wheel, what they called a non-synchronous whirl.
[00:55:29] RB: A non-synchronous whirl?
[00:55:31] BP: That's right, but it was a critical vibration problem at a certain speed, the turbine just wouldn't stand up and developed a fatigue crack and came apart. Unfortunately, that caused us then to put some limits on the production hardware in the sense that we had to inspect turbine wheels out in the field for a certain number of engines until we got to the production model that had the new turbine wheel in it. People look back on that and say, “Now had we the first couple of weeks, we’d run that turbopump, taken that machine up to that condition, we would have found the problem sufficiently early to prevent any impact at all on the program.” This might sound a little critical of some of the things we did in the program. It doesn't bother me to talk about it because I think that a development program is a learning process in itself, and nobody can be 100% smart in developing something that's completely unknown or… You just can't predict everything. These types of things will, I think, make us a little smarter going into the new programs. On the other hand, there are probably some things about the new high-pressure designs that we think we understand today, but we'll come crawling out of the woodwork at it. [both laugh]
[tape ends]
[00:00:32] Bob Pease: No, let me say that there was a particular desire to be able to fire the complete engine including the bell nozzle which was a 27.5 to 1 air ratio nozzle, under the sea level test stand conditions. The reason for doing this of course was number one to ensure that we had the ability to fire the engine with more ease, which you can do in a sea level environment than if you have to do it in an altitude chamber. The test frequency—the question was important—so the engine bell nozzle was designed to flow full at the sea level test conditions. This was important in order to get the proper heat fluxes and so forth into the chamber. In other words, the chamber was being cooled by hydrogen. The hydrogen of course picks up heat as it goes up through the wall and comes into the injector, so trying to get as near as possible the proper heat transfer conditions on the chamber under sea level firing. Another thing involved was the fact that if the nozzle didn't flow full, not only would it affect the heat transfer to the thrust chamber wall, it would also make it difficult to [tape cuts out] …Not recording?
[00:02:36] RB: Okay, we're on again. [laughs]
[00:02:38] BP: Did you miss all that or should I go?
[00:02:40] RB: Why don't you just recapitulate? I think I moved this over here because sometimes we get a lot of overtone noise from air conditioning equipment.
[00:02:48] BP: Well, I think I can shorten that down now.
[00:02:50] RB: Okay.
[00:02:52] BP: I think if I understand your question, you're saying…The basic question you asked was the engine, some reason for having the engine designed to both be fired at sea level and at altitude, and was this a first in the engine business? To recapitulate, yes, there was a reason for having the engine be able to fire both at sea level and altitude. The reason to have it fire at sea level is in order to obtain more easily the testing, numbers of tests and so forth, and to do it under an easier test environment than you have to when you're playing with the diffusers and altitude equipment. That facilitates your development effort that way. Because of this, the engine was designed to give as near as possible the proper performance on the sea level test by having the nozzle flow full. If it doesn't flow full, you have problems in measuring the thrust, and you also have a difference in heat transfer to the thrust chamber wall.
[00:04:23] BP: As it turned out, the engine…There were still differences in the way the engine performed at sea level and the way it performed on the altitude test. There were some changes in the engine power balance due to the difference in pressure ratio between sea level and altitude across the turbine areas. There was also some difference in the heat transfer to the thrust chamber wall simply by the fact that under the altitude conditions you had, I'd say, a very reduced atmospheric situation, and you weren't even transferring atmospheric heat into the thrust chamber wall. [laughs] There were some differences, both in those two areas. However, I don't know if I'm answering your question completely on this. Was this a first? I think it was a first from the standpoint of trying to get a nozzle of a higher expansion ratio to flow fully at sea level. The nozzle was a special contour which was designed in order to facilitate this. That part of it was a first. I believe on previous programs, such as the RL-10, we had to use some type of diffuser system if we were going to flow the full, complete [engine?] nozzle. So I don't know, does that…?
[00:06:07] RB: Yeah. That answers the question, I think. I’ll have to do some research on it too, try and find out some more things. Maybe I can pick up the phone sometime, just call you back.
[00:06:16] BP: Yeah, fine.
[00:06:17] RB: Now, they did use vacuum diffuser equipment in the J-2 development.
[00:06:22] BP: Yes, they did.
[00:06:23] RB: Was there a difference, say, in the diffuser equipment that Rocketdyne installed there and what was used at Tullahoma?
[00:06:31] BP: Yes.
[00:06:32] RB: Why was there a difference? I mean, how did the difference occur and what was the intent? Why do some of the testing at Susanna and why do some at Tullahoma?
[00:06:41] BP: Well, what amounts to this is that when the program started, it was recognized that there still had to be an altitude system type test to verify the engine behavior on their altitude. Even though the engine was designed such that it could be tested at sea level, it was still recognized there would be some differences when testing in an altitude environment. From a very important beginning program outset, the steam operated diffuser was planned at Santa Susanna to evaluate the engine under vacuum operating conditions. The steam operated diffuser operated in such a manner as that steam was supplied during the engine start transient, which pulled the capsule pressure down to something in the neighborhood of 80,000 feet. When the engine started, the diffuser would operate in a manner of…It would be activated by the exhaust gases of the nozzle in order then to maintain the capsule vacuum pressure under main stage operating conditions. When the engine shut down, the steam would come on again to try to maintain the altitude conditions. Well, when this system was actually put into operation, it was discovered that the steam system was unable to maintain the 80,000 foot altitude during the start transient. Part of the reason for this was the fact that when the diffuser and steam system were designed, they didn't anticipate the amount of hydrogen fuel lead that was going to be produced by the engine. Hydrogen being densities and so forth it is, this fuel lead very quickly saturated the altitude producing capabilities of the steam system such that the altitude would drop down to less than 20,000 feet during the start.
[00:09:12] RB: If you're shooting for 80, that's not a very good condition. [laughs]
[00:09:16] BP: Right. Not only that, they also found that there was considerable steam blowback and turbulence that came back and got into the nozzle area, which also fouled up the heat transfer inputs to the chamber one thing or another. Between the pressure ratio across the oxidizer turbine and the steam blowback, it was felt that our altitude simulation was less than optimal. From that standpoint, we then made an investigation to find out if the Tullahoma facility would be available in order to run some additional series of altitude tests. In fact, along with this investigation, we also investigated constructing another altitude facility at Santa Susanna, which would have been operated by another type of steam system, one which would have kept the altitudes up in the neighborhood of 90 to 100,000 feet at all times. Rocketdyne had successfully demonstrated what they called a hyper flow steam system using basically an H-1 engine to generate steam. Instead of having a steam accumulator, which had limited capacities, they were going to fire essentially a rocket engine with water injection, which would have generated a very large steam capacity instantaneously upon demand for keeping the altitude up. That system looked very good and would have cost us, however, several million dollars to put that in at Santa Susanna. In the process of looking at all this, it became apparent that it would be more cost effective to activate the Tullahoma stand, the J-4 cell at Tullahoma for this purpose. That's how the Tullahoma program came into being. Once we got into the J-4 cell, we did find that we did learn some further knowledge as to the engine starting characteristics that wasn't apparent at the testing at Santa Susanna, either under the altitude or the sea level testing.
[00:11:55] RB: So you had to do a little redesign?
[00:11:580] BP: Yeah, there did cause some redesign, but it was not of a major nature, mainly orificing and balancing of the engine start primarily. The Tullahoma conditions on the transfer to the thrust chamber and so forth were further verified as being accurate. This is just a side note, when I looked at the flight data, they found that the flight data substantiated our performance results at Tullahoma. Tullahoma was a pretty good facility for that standpoint. One thing that Tullahoma didn't do, or any of the testing didn't do, was did not find what we call the ASI problem of 502.
[00:12:51] RB: Can you go into that a little bit?
[00:12:55] BP: Sure.
[00:12:57] RB: We're still running.
[00:13:02] BP: You might have already got some history on this, you may have known, we did fail an ASI line in the S-II stage and also in the S-IVB stage on flight 502. The ASI fuel line was later discovered to be a fatigue failure of the line caused by flow vibration, or its flow induced vibration. Turns out that this was not discovered under sea level testing or even under altitude conditions of the Tullahoma stand because, well, number one, under sea level conditions, there's a condensation of air that gets in between the bellows section. This does two things. Number one, it's maybe a lesser effect of it, but it does provide some damping in the bellows. Secondly, it does affect the heat transfer boundary layer conditions on the inside of the flow path. When Rocketdyne ran laboratory tests and took photographs, movie photographs, and made other flow dynamics studies, they found out that the heating effect of the line actually changed the flow induced vibration characteristics. I'm not a flow vibration expert, so I'm trying to explain it in…
[00:14:49] RB: Neither am I. [both laugh]
[00:14:53] BP:…Simple terms as I understand it. In other words, apparently it changed the boundary layer conditions on the inside of the line so that the flow pattern wasn't the same. Now, why didn't this happen at Tullahoma? Well, there's a good altitude condition we were obtaining—100,000 feet or so—there's still enough nitrogen present in the cell to—this is a theory is presumed on to, at least the calculations have been made—to provide the same effect.
[00:15:35] RB: Okay, I didn't understand that then. So was nitrogen still existing in the cell to create that kind of frost buildup and have the same damping effect?
[00:15:43] BP: Well, there's both nitrogen and water, and I say nitrogen because I believe in Rocketdyne's testing they proved that frost alone would not prevent the failure. It really, really had to be the air condensation. In other words, yeah, if it were packed with ice, it's one thing, but a minor frost buildup, I think alone was not considered sufficient to prevent the failure. The way I understand it now is that they believe there was enough nitrogen condensation within the cell to have done this. Now, I guess one way of…Maybe another way of tackling the problem is that I don't really know if this is a feasible way if nitrogen could have been substituted for some other gas such as a helium purge. The reason they don't like helium is because the lighter the gas is the harder it is to pump it out.
[00:16:45] RB: But isn't that the way they finally got to fix on it though? They put it in a helium tank or something?
[00:16:51] BP: Yes, as a matter of fact, in the testing at Santa Susanna, they tested with a helium environment, vacuum environment, with [arid?] environments, and they found out that it would fail under the helium environment because the helium wouldn't condense. It would fail under the vacuum environment, the hard vacuum environment in a test small test capsule, they can pump it down quite a bit higher altitude to Tullahoma. So that's how they finally pinpointed it was by isolating these different environments. [00:17:34] BP: So it's quite an interesting investigation. I think to many of us it's sad that there are always certain problems that one cannot predict in advance and all the analytical techniques cannot even plan for. Well, now that we know about this one, we will always plan for it. I'm sure that in the space shuttle management program or other future programs, we will still continue to plan altitude type testing. Although in this particular case on J-2, it didn't find this problem.
[00:18:22] RB: You were out at Rocket...Were you the MSFC representative out there?
[00:18:27] BP: I was the resident manager for the J-2.
[00:18:32] RB: Do you remember very much about the story on the injector face and the use of Rigi-Mesh? Were you involved in that?
[00:18:42] BP: Yes.
[00:18:43] RB: Could you give me a little bit more?
[00:18:45] BP: What would you like to know about it?
[00:18:46] RB: It's my understanding that it was the Pall Corporation that came up with the Rigi-Mesh stuff in New York, and they got this out of the Atomic Energy Commission program and used it for filtering gases or something?
[00:19:00] BP: Yes, I believe that's correct. I know the material was produced for purposes of such of that nature. I guess it also had been used on the RL-10 program.
[00:19:15] RB: My question is this. You had some problems with the injector, I guess. Well, normal development problems. Where does the decision come to get into Rigi-Mesh? Does Marshall say that?
[00:19:28] BP: Well, I just happened to be intimately involved in that one. We started the J-2 program. Rocketdyne had proposed a copper ring type injector, which was typical of most previous Rocketdyne designs with the various orifice holes drilled at various angles in the copper. The Rigi-Mesh had been used successfully on the RL-10 program and had been in fact introduced to the RL-10 program by Lewis people. Before that, where they got it from—I think what you say is right—the Atomic Energy Commission may have used it for filtering devices, one way or another. But it was used in the RL-10 program as an injector plate medium in order to allow the hydrogen to filter through and partially cool the face by that means. People were concerned at that time about the problem of cooling the face and felt that Lewis people felt we might have some problems with the copper. In the early part of the J-2 program, we got back to the initial discussions on how to get it into the program. It was suggested to Rocketdyne by NASA that they include in the development program an alternate design of a Rigi-Mesh face type injector, and that they evaluate both the copper design which they proposed and this alternate Rigi-Mesh injector. Sometime downstream, there would be a decision made as to which one was best.
[00:21:27] RB: In this case, the impetus came from NASA to Rocketdyne?
[00:21:30] BP: Right, and this is part of NASA's transfer of technology, whatever you want to call it. We were aware of the material, we were aware that it had been used in the rocket engine programs, and we felt that Rocketdyne ought to look at it also. So after—I don't remember the exact time frame, but I think it was within probably about the first six to nine months—the early injector testing, they did test several copper versions and several Rigi-Mesh versions. The overall assessment of Rocketdyne came to the conclusion—and NASA agreed—that Rigi-Mesh performed not only performance-wise better in terms of the particular concentric post orifices where the hydrogen came out around each post, each LOX [oxidizer?] post in the Rigi-Mesh face. The design performed well from an Isp standpoint, and it performed well from a cooling standpoint, there was no problem with burning. We did have some problems with the copper injector on burning the rings and lands and so forth. I remember one where we had green flame coming out the chamber during one test. [both laugh] I'm sure they could have solved the copper problem, I'm sure they could have solved the cooling problem, but it just turned out that in that evaluation the Rigi-Mesh was best.
[00:23:04] RB: This is an interesting feature it seems to me of the Saturn, or NASA programs generally, that you do hear here is Marshall operating, cooperating with Lewis and with the contractor. It's the kind of technology of one center, in this case Lewis, that comes through Marshall into the Rocketdyne facility, that there’s this interchanging [inaudible]...
[00:23:23] BP: I remember that particular one because I had worked with the Lewis people on the technical committee of the source evaluation for the J-2. In that capacity I got familiar with those people and that particular subject. I was a part of convincing Rocketdyne that we should do this.
[00:23:53] RB: Now again there's a question I wanted to ask, kind of a specific one. Rocketdyne got the contract and started to work to put together an experimental engine and later fired an experimental engine. Now where the hell do you get an experimental engine? Do you just go out in the shop and start hand working an injector face, put together a few tubes? How do you do that?
[00:24:18] BP: Well, you first have to design it.
[00:24:20] RB: Yeah [laughs]. Well, what I'm getting at is there's no manufacturing lines set up yet really. The first test engine really is a handmade product. Is it accurate to say that?
[00:24:32] BP: No, no it isn't. Now the J-2, Rocketdyne designed a J-2 engine as they envisioned it in their original proposal model.
[00:24:43] RB: Oh, the computer model. I know they used a lot of computer stuff, configurations…
[00:24:47] BP: That's true, they looked at their analysis, and they sized the ducts, tubes and so forth with a proper thrust chamber size with a proper heat transfer and all that type of thing. In other words, they designed the engine from a dimensional standpoint based on all their analysis. They went out and released orders for material, which is not just raw material but vendors to make certain basic castings and forgings and so forth, and even some components which they purchased from vendors, which was either modification off the shelf items like small valves and things of that sort. Or they actually ordered components from certain vendors to their own specifications. This material comes in house, it's machined to the drawings; it's checked out to the specifications; and they put an engine together. Now they call this an experimental engine. I believe the first one was essentially called “Experimental Engine” because they had very little if any experience with the major components. One exception was the hydrogen turbopump, which was basically a modification of one that they had already built and run for the Atomic Energy Commission on the Rover program. We were starting with basically an off the shelf turbopump, which was modified somewhat for the J-2 application. The oxidizer pumps were not too unsimilar from the previous engine programs like the H-1 and the Atlas and other LOX/RP engines. The oxidizer pumps were quite similar in design to what they had used before. Some modifications such that J-2 they used separate pumps rather than two pumps run by a gearbox. Each pump had its own turbine and that was a new innovation from previous engines.
[00:27:12] BP: It is not just a handmade type of thing. It's actually analyzed, designed, hardware is released and built, and so forth. Not only that, they didn't just release one set of hardware, they released groups of hardware. I think they released maybe four or five engines worth of hardware, probably even the initial release. From the first engine firing, the experimental engine, they found out some things, which they went back and made some corrections to. The hardware was already in the pipeline. It was like they might have to, well, I can't think of a specific example that occurred on the J-2 but…Well, I think of one thing that happened, our famous 27.5 to 1 thrust chamber, which was flows full at sea level, developed a phenomenon which they hadn't predicted, and that's called side loads. During the start transient, a very eccentric load developed in the chamber, forcing the chamber over to one side. Some of the first engines they ran, they actually physically distorted the chamber, distorted the actuator systems, and things trying to hold the chamber in place, engine in place. They had to quick come up with what they call a horse collar, which was a device that came in and grabbed the chamber at the throat. It would hold the engine stable, take the load during the start, and once it was running then the thing would drop away so the engine could be gimbaled. But that was not predicted, and that's an example of a kind of a modification to find out you have to make.
[00:29:02] RB: Were they able to work out the side load problem that they were on and get rid of that horse collar thing?
[00:29:07] BP: No.
[00:29:08] RB: That stayed in the whole program?
[00:29:09] BP: Yes it did. The S-II stage and ground firing used devices to hold the chambers in place.
[00:29:14] RB: And the S-IVB then also has the same thing?
[00:29:17] BP: During ground firing.
[00:29:18] RB: During ground firing.
[00:29:19] BP: And that's ground equipment.
[00:29:20] RB: Okay, it's not flight hardware?
[00:29:21] BP: No, the side loads don't exist in flight because of the vacuum condition.
[00:29:26] RB: Okay.
[tape cuts out]
[00:29:27] RB: It takes a minute or a second to get here, to run up. Okay, I think we're probably back on here.
[00:29:35] BP: Okay, so I kind of went beyond maybe the scope of your original question.
[00:29:39] RB: No, I'm glad to find out. I didn't know about it.
[00:29:41] BP: There were any number of problems that, of course, came out of the early engine firings, which caused either modification of hardware, which had been released in the system, or certainly got cranked into the next batch of hardware design that was going to be released in the system.
[00:30:04] RB: Could you pick out a problem that you remember, especially that developed during the test and qualification program that brought about a redesign of hardware?
[00:30:14] BP: We had lots of those. [laughs] I don't know if there's any one you want me to emphasize...
[00:30:22] RB: We read the NASA press releases and the AIAA presentations by Rocketdyne people—which isn't to say anything against them, you know?—but you get the impression that this was a beautiful program. You know there must have been a problem somewhere.
[00:30:36] BP: I guess this is a matter of debate. I don't know. My own personal philosophy is that you have a development program to find out what's wrong with things and correct it. In other words, I think it's more accurate to say that happiness should be finding a failure rather than not finding a failure because if you don't find it, then you're going to be hung up sometime in a critical mission or flight or something of that sort. The J-2 program did find lots of failures and had to correct them during the program. The most difficult part of it came into play when we had to release engines for production, and we had engines in the production pipeline, and we were still finding failures. It's particularly difficult because then it means that the solution has to be found that we can crank into the production line and fix that particular piece of production hardware. Not only before it's delivered from the line, but we had cases where engines had already been delivered to the field that we had to send out modification kits to fix them. A lot of people look at this as saying, “Well gee, why didn't you find all those things before you sent those engines out to production?” Truth of the matter is that the program always needs the production hardware before it's fully developed.
[00:32:18] RB: For schedules? To meet your schedules?
[00:32:21] BP: For schedules. We have not only on the Saturn program, but on every program in the missile rocket business that I've ever looked at including the Air Force program, we have a situation of what we call concurrent development and production. General Schriever recognized that during the early IBM days. We certainly had it during the Saturn Apollo. The engine being one of the long lead items for the entire system suffers particularly because the vehicle people would like to get an operational engine cranked in their vehicle as soon as possible. Many times they'd like to have it rather than waiting to have it at some point where you install the engine down at the Cape before you launch it. They're not satisfied with that. Usually they want to crank the engine in in the early part of the vehicle build up cycle back at the vehicle manufacturing area so that the whole system, engine to vehicle, goes through the system checkouts. Maybe sometime in the future, we can promote a philosophy to give us more time in the engine area where we can install the engines at the other end of the line. [both laugh] Which just means maybe as much as a year of expert development and maturity.
[00:33:47] RB: Can you give an example? Do you remember what kind of modifications you were sending out?
[00:33:53] BP: Yeah. Well, let me give an example that hit us in the production line before we were just beginning to deliver the first flight engine for Saturn IB, which was going to Douglas with engine number fifteen, 2015 I believe. Now this engine was in the stages of final assembly when we discovered that we had burned out a section of the gas generator wall, actually put a hole in it. In fact, I believe this happened on one of the battleship engines that we had sent to Douglas Sacramento. It happened on another engine or two that were being fired in Santa Susanna. This launched an intensive development effort to solve this problem, which Rocketdyne came up with a change to the throat of the gas generator. They call it [inaudible] choke ring, which diverted the gases somewhat so they didn't impinge on the wall. So this [inaudible] choke ring then had to be worked into all the gas generators that were on the line. It meant actually cutting the gas generator, which was welded to the fuel turbine [manifold?], taking this assembly off, reworking it, re-welding everything back up again. It was quite a re-working turnaround problem. It was a modification that could be incorporated into existing hardware to solve the problem. That's an example of a problem that hit us during the manufacturing cycle.
[00:35:56] BP: One of the most serious development problems I think in terms of seriousness that we found was the fact that we had some very definite problems in the engine start sequence. During an attempted semi-qualification, which we call PFRT testing, we did discover that we did have a control problem or a sequencing problem. We started build up which actually caused the LOX valve to be forced closed such that the LOX valve had such hydraulic pressure against it it wouldn't open properly. This caused excessive back pressure since the gas generator was tapped off above the LOX valve. The gas generator LOX balance went LOX rich. We got too hot temperatures and burned out the turbine and a few other things. This did point out the fact we had a problem there.
[00:37:17] BP: We'd also been having problems in starting the engine due to the hydrogen pump stall, which in case power balance was very sensitive between the LOX and fuel turbine speed buildups. If the pressure in the chamber built up too fast because the LOX pump was putting more LOX in the main thrust chamber causing the main thrust chamber pressure to build up too fast, this caused excessive back pressure on the hydrogen side. It got ahead, you might say, of the hydrogen pump buildup, and we had that condition which forced the operating and the operating curve of the pump, which pushed it over into the stall region. The pump would stop pumping. That was called hydrogen stall.
[00:38:13] BP: There was extensive development effort launched to find improvements to the start sequence, which necessitated running a much larger number of engine system tests than we had originally planned. The only way you could work on this problem was to run the engine system. The other problem I mentioned when we burned out the gas generator, we solved the problem in the gas generator pit. In this case, there was no other way to solve the problem except to make many, many cut and tries, as it will, at the engine test itself. Rocketdyne used all of their analytical methods in the start computer, which they had developed and so forth, and tried to make refinements on that computer in order to get a better handle on things. They used all the techniques that they could. Really what it came down to is that you had to make an adjustment and then try it. The final solution had to be tested under altitude conditions to make sure that it was also valid because of the altitude affecting the start.
[00:39:37] RB: Okay, that's very good. You know, these are a good example, I think, you know, of what happened [inaudible]. Another brief question here, and then I've taken enough of your time, I think. J-2 was built for a mission of around 500 seconds, but the lifetime was about 3,750 seconds around that area.
[00:40:01] BP: Yes, 3,750 was the qualification one.
[00:40:04] RB: Okay, now why the difference? What were you doing inside those 3,750 seconds that you needed that many of when the actual mission was only 500?
[00:40:16] BP: Well, I guess that's a philosophical sort of question. It gets back down to some basic assumptions. Number one is the fact that I guess when an engine is delivered to the Cape for flight, it's already going to have a certain amount of time on it. It's not going to be zero time. There's about 750 seconds allocated for the acceptance test of the engine. Then there was another 500 seconds allocated for the vehicle test, ground test. They did fire the S-II stages and S-IVB stages on the ground for quite a while. I guess they fired all of the S-II stages. Then as a reserve, there were additional time allocated beyond that.
[00:41:19] BP: Suppose you have a vehicle problem or a reason to refire the vehicle for some reason on the ground, you might have to run another complete test. You can rationalize yourself that you might have engines at least 1,500 seconds that would still have to be qualified for flight. It turns out most of them didn't have that much time on them when they got to the Cape. More like, I say 1,500 to 2,000 seconds, you could rationalize. You might be in that range, but most of them were in the 1,000 to 1,200 range that actually turned out. All right, so then beyond that, you say, “Well, why would you need any more time?” This is strictly a matter of judgment in fact that trying to make a maturity reliability type assessment, I think the goal of 3,750 was chosen because it would also increase the confidence level upon which the engine say of some lesser time would have their flight.
[00:42:34] RB: Is this part of the manned rating philosophy too?
[00:42:38] BP: Yes. Well, it was at least in our minds at that time. The fact that we were going to fly manned in previously qualified engines, much less time. We felt that 3,750 seconds was perhaps not an unreasonable amount of time at all for a hydrogen type of an engine. I can't say that this is any magic number for a manned space flight because it wasn't. The F-1 had a qualification time of 2,250 seconds, which was felt reasonable for the conditions that you expect with a LOX/RP type engine. The fuel made a considerable difference because there were more effects on the F-1 life by using RP fuel because of coking, carbon deposits, and other other types of problems that you've got with the fuel. Hydrogen engine stays just brand shiny clean. You don't have those kind of problems, and you can expect a longer life inherently from a hydrogen engine. This was one reason why I'd say somewhat arbitrarily 3,750 was chosen to be at least some measure beyond what we have done in the LOX/RP type engines. That's about it. There are probably some other philosophical reasons that you might get from some other people. I don't know.
[00:44:14] RB: [laughs] Well, do you have any, you know, things that you'd like to say?
[00:44:16] BP: Well, just to make one more little sidelight on this whole thing…In fact, I don't know when the designer looks at 3,750 seconds, whether he designed the bearing carrier much different than whether it was 3,000 or 2,000 or what. We actually ran some R&D engines with, of course, some component replacement here and there, but we had components on R&D engines which had gone well over up to 20,000 seconds. We had several other samples at ten, 15,000, that range.
[00:45:00] RB: It gives you a pretty high confidence factor.
[00:45:02] BP: Well, yeah, as a matter of fact, it's because of this kind of data that we feel rather confident that the space shuttle main engine we can eventually attain the seven and a half hour life on mission requirements. In fact, we don't even know today that some of the J-2 components wouldn't have gone much beyond the 20,000 seconds because we just gave up at that point. Some of them weren't even worn out. If we had to today take a J-2 engine and make it a reusable type of engine to go to a long life—many missions—we have some data, which there will be some areas in the J-2 that we know are life limited, which would have to be redesigned. From a general standpoint, it did give us high confidence in the J-2 from the J-2 data that many of the hydrogen type components would run considerably longer than what we ever tested before.
[00:46:18] RB: Well, I've taken a good share of your time here.
[00:46:22] BP: Well, I understood you wanted about an hour. Anything else you have on your list that you want to go into?
[00:46:30] RB: No, some of the…There are questions that I picked up just reading odds and ends of things. One thing that is really a stupid question, but what does J-2 mean?
[00:46:43] BP: [Laughs] Doesn't mean anything particularly. The Rocketdyne had a system of numbering engines. Well, maybe it's a lack of system. We had…I can't really say where it started. Rocketdyne had an E-1 experimental engine program. They had the F-1 program. It seemed to me there was a D program somewhere that didn't live very long. I don't know what happened to some of the other numbers in between. They probably died in the proposal [mill?] somewhere. But there was H-1, of course, which was picked on H-1. I think they skipped I for obvious reasons. The next thing that came out of the bag after H-1 was the J-1.
[00:47:48] RB: Oh, J-1?
[00:47:50] BP: Rocketdyne peddled this around the country in the advanced design sales department, I'd say, for a while.
[00:48:00] RB: Was it a hydrogen engine?
[00:48:02] BP: Yeah, it was an advanced design concept really, which they called J-1. Then when NASA got serious about going into a development of a hydrogen type engine of the 200,000 pound class, then to distinguish from the version that they had been peddling around the country, they would call it the J-2. I guess basically I don't even think J-2 appeared in their proposal, but when they got down to negotiating the contract, they said, “Well, we're going to call this the J-2 since we had J-1. Does anybody object?” NASA didn't have any numbering system established for engines like the Air Force had the [XLR?] type series. We didn't have any official numbering system, so we said, “We don't know it makes any difference.” Well, J-2. Of course, we already had the H-1 being one, and the F-1 program had already been started. It seemed chronological to us to call it the J-2. [both laugh]
[00:49:13] RB: While you were out there, do you remember any incidents that struck you or remember as being humorous or funny? People out there who were really strong characters, little anecdotes about any of them?
[00:49:30] BP: I probably could if I thought about it a little bit. After all those years, we certainly had a lot of interesting meetings. [laughs] It got pretty wild at times, but let me think a minute here. Well, I really can't remember anything offhand. As I say, I'm sure I would if I thought about it a bit.
[00:50:07] RB: What about Von Braun?
[00:50:09] BP: He used to come out quite often to the West Coast and visit the Rocketdyne and review the program. He attended many of our review meetings that we had in the early days. Later on, he got…Demands were heavier on his time, he came out less often. He usually managed to come out once a year and review all of the activities at Rocketdyne.
[00:50:40] RB: Did you find those helpful, necessarily, or was it something you did because it was for his primary information benefit?
[00:50:47] BP: No, I wouldn't say…Obviously, I think that it was not only a general feeling by the NASA people, but I think the contractor, too, appreciated the opportunity to bring the center director—Von Braun, of course, being the center director at the time—up to date on what was going on. I think this helped the morale a lot because you did feel that the top management was interested in the effort. Not only that, I think that Von Braun himself was quite an inspiration to the program and any meetings that he attended. His comments and observations were—I would say—very, very helpful because it put what we were doing in one detailed part of the center—Apollo—and the whole system somewhat in more context as to really what our objective was. That was helpful. I think those types of meetings perhaps gave us more insight on where we really should be going and probably some of the day-to-day directives that we used to get. [both laugh]
[00:52:17] RB: It's too easy to put the director of File 13 down there, I suppose too.
[00:52:24] BP: My only general comment on the whole thing, I thought it was a very interesting program. I think that certainly learned a lot from the program. Much of what we learned from the program, we will attempt to apply in general to our management of future programs. In other words, not just learning what to do, but certainly a few things that we learned, I hope perhaps, not to do in a future program of this type. One thing I think that we learned, I think might be of specific interest, in looking back over the J-2 program history…We've done a lot of this in the last couple of years, going back into our history and deciding how we might possibly improve the approach, the management as a development approach in the management. This is even over and above the technical things that are obviously going into the new program. We're trying to apply this to our management approach now, the take on the shuttle engines. Just for example of this, I think we learned in looking over the J-2 program we need to have more emphasis on the component and subsystems testing. Also, we learned we need to get in earlier to testing limits of the hardware in J-2. For example, we had run the hydrogen turbopump a number of months at some fairly nominal type missions. They're within specifications, so to speak. Performance was fine, no problems. Well, I won't say no problems, but the things seemed to perform well basically. When they started to run what you might call higher limits to higher power levels, higher speeds and things of this sort, it might be on the outer fringes of the operational environment. In other words, the engine probably wouldn't run there, except maybe under some unusual conditions or what have you. We failed the turbine wheel, and found we got into a serious vibration problem with the turbine wheel, what they called a non-synchronous whirl.
[00:55:29] RB: A non-synchronous whirl?
[00:55:31] BP: That's right, but it was a critical vibration problem at a certain speed, the turbine just wouldn't stand up and developed a fatigue crack and came apart. Unfortunately, that caused us then to put some limits on the production hardware in the sense that we had to inspect turbine wheels out in the field for a certain number of engines until we got to the production model that had the new turbine wheel in it. People look back on that and say, “Now had we the first couple of weeks, we’d run that turbopump, taken that machine up to that condition, we would have found the problem sufficiently early to prevent any impact at all on the program.” This might sound a little critical of some of the things we did in the program. It doesn't bother me to talk about it because I think that a development program is a learning process in itself, and nobody can be 100% smart in developing something that's completely unknown or… You just can't predict everything. These types of things will, I think, make us a little smarter going into the new programs. On the other hand, there are probably some things about the new high-pressure designs that we think we understand today, but we'll come crawling out of the woodwork at it. [both laugh]
[tape ends]
Duration
0:57:30
Files
Collection
Citation
“Pease, Robert,” The UAH Archives and Special Collections, accessed July 9, 2026, https://oralhistory.uah.edu/items/show/630.