1966 Interview with Tom Kelly
ERTEL: This is an interview with Tom Kelly, the Apollo design engineer for Grumman, conducted at Grumman Aircraft Engr. Corp., on May 3, 1966 Bethpage, L. I., New York
KELLY: Well, I guess I've been involved in Apollo-related work as long as anybody in Grumman, actually. I started on the thing in 1960 -- April 1960. We started some in house studies at Grumman on manned space flight, and particularly manned lunar mission type activity.
As you know, at that time, there were a number of NASA funded studies, and we competed for all of them; we never won any of them but we kept conducting our own in house studies. We went down and gave our own study conclusions to the NASA people right along with everybody else; I guess we were not unique in doing this. But, the point was, we had a very active interest in house, and we just wouldn't let it die; whether it was funded, or not, we kept going with it.
So, finally, it got to the point where the Apollo competition was held; when the actual RFQ was issued, and the work statement and all that, the realization hit us what a tremendous job this thing was. And, after much soul searching and evaluation, why, the company finally decided not to go it alone on that proposal, so we ended up teaming -- teaming up with GE in a big group lashup. There was -- GE was the prime, then there was Douglas and STL, and ourselves in it with them, at that time.
Well, that was a story all in itself, but as one aside to that - the GE people decided that it was a page limited technical proposal that was being turned in for Apollo, and the scope of what you could cover in the page limited proposal is pretty restricted. So, one of the things we decided to do with GE was that we would explore some things over and above the bare requirements of the proposal, and include them as sort of extra material. In fact, they even printed them on different color paper, so if the NASA people wanted to just tear out all the pink pages, and throw them ASAP without looking at it, they could.
Well, one of the things they decided to explore was how you should actually conduct the mission, because this wasn't really covered in the -- in the Apollo RFQ at all. You were just told to propose on a particular vehicle, and it was just sort of loosely implied how the mission would be conducted. So, to come to the point, as a side operation and part of the basic GE team Apollo Program, there was a side group set up to look at various mission alternatives. And, there were five mission alternatives that were looked at, at that time. And, they just sort of drew out, among the teams, who would work on what by lots, and the Grumman team got to work on lunar orbital rendezvous, which was a very interesting choice. Because, at the end of the study period, we had to get together, and we compared what Douglas had done on their assigned mission, and what GE had done on theirs, and we all agreed that -- yeah, the way to do this mission was lunar orbital rendezvous. So, that was the recommended mission that went in as a part of that Apollo proposal. Well, as you can imagine, that sort of whetted our appetite here at Grumman, and after the Apollo proposal was turned in and done with, why, we went right ahead with these LOR studies, which were pointing to the design of something like a LEM at that time.
And, then when the decision was announced on the Apollo and it was clear that we weren't in on the CSM part of the Apollo, why, we devoted all the efforts then of our space study group, at that time, to the LEM concept. And the design concept, initially, was an extremely simple thing -- it was like a fellow sitting on a pair of handle bars, very much. But, it grew and grew, conceptually, until finally in our studies, we were -- oh, I guess, this was about April, say, of 1962, or so. By that time we were studying LEM type vehicles in a parametric form, looking at all different combinations, whether it should be a staged or unstaged vehicle; or what kind of engine complement it should have; do you really need throttleable engines, or could you use several discrete levels of engines; what kind of landing gear should you have on this thing, etc. We also, during that time, worked out all the weight tradeoff factors, which was to serve us in good stead later on, because it was obvious that was the dominating characteristic of the whole design.
So, anyway, when the RFQ for the LEM came out, why, I'd say that we were pretty well prepared. We had done an awful lot of background work and condensed it as well as we could in the RFQ. The reason that I was looking for the proposal is I wanted to show you what the vehicle looked like originally - it conceptually is the same as it is today, but, in almost every significant detail, it is different. And, there are reasons for all the differences.
I used to have a little -- oh, here is a little model of the proposal LEM; excuse me just a minute -- can I have a word with the ... Well, anyway, this is the vehicle we proposed -- it's a little dusty now.
The basic concept is the same as we now have and there are some features that are the same. Well, it was a staged vehicle, and had a throttleable engine, and it had the same overall mission requirements, I guess, in very general terms -- two men and approximately the same lunar surface stay-time.
But, from there on -- you see an up-to-date picture of the vehicle, and just about everything else is different. The location of the two hatches remains the same, we have an upper hatch and a forward hatch. But, our original concept was to allow for docking on either-hatch, so there was quite a design evolution that went to abandoning that.
I guess the most significant difference on this vehicle was the weight, This was limited to 22,000 lbs. in the proposal; we said in the proposal that -- yeah, we thought we could make 22, 000 lbs., if everything stayed exactly as defined in the RFQ; but that it was going to be an awful tight squeeze; it was really a very severely weight-limited vehicle. We pointed that out as the most significant technical problem in the program; that and the reliability of the components. As it has turned out, we were somewhat optimistic as it turned out -- we couldn't have hit 22,000 lbs. if absolutely nothing changed, and there were enough changes -- the weight has traced out, the history you see right there on the wall chart. It has been our toughest technical problem, up until now. We've got it licked now, but it has taken a real massive effort.
Well, anyway, that sort of gets me up to the proposal phase. The first thing we did with NASA was to negotiate a contract for the whole job; and there was an example of ignorance in action, if I ever saw one, at least on our part.
I think NASA was a lot smarter than we were, at that time. But neither side really understood the scope of what we were getting into, I don't think at that point. When we negotiated, we thought we were negotiating essentially to build this vehicle, or as close as we could get to it, and that really wasn't what the NASA people had in mind.
What they wanted to launch into, and what they subsequently directed us to launch into, was an extensive preliminary design phase wherein every aspect of the design was gone over again, and just from a conceptual standpoint. And, oh, that phase, I would say, lasted from -- we started in January 1963 was when we officially got under contract and we were still, grossly, preliminary designing the vehicle for at least nine months after that; swapping tanks around, changing the number of landing gear, changing the shape of the cabin, and all that sort of thing.
The gross configuration started to settle down about then; about the end of '63, and this was still in real gross terms. And, the optimization kept right on going; the definition of mission requirements got better and the definition of design features got better all through '64.
And, finally, near the end of '64, it was painfully obvious that we were at the end of our rope on the weight situation to the point of where we had to -- we had to resize the tanks again. The way this vehicle was built, the whole darn thing is wrapped around the tanks, and if you changed the tank size, you change the whole vehicle, pretty much 100 per cent. So, you see on that chart there, where the control weight takes a big jump at the end of 1964, November '64. At that point, we negotiated with NASA, a change in the control weights, simultaneously resized on the propellant tanks, and that, essentially, has been the last gross feature change that went into the vehicle.
That set up our size and we've lived with that ever since; we'll live with that for the rest of the program. Now, in this past year, it's been a rapid progression from -- I would look at that end of '64 as the final end of the preliminary design phase. It was active preliminary design all through '63, and it was semi-active preliminary design through '64, with a number of early hardware items, test models, and test articles being designed. And, then at the end of '64, we were essentially frozen on the concept, and 1965 we froze all the details of just about everything, as far as you could freeze it on paper, down to pin assignment and wire numbers.
And, now the phase we are in -- we are bringing it right up to the present -- we've progressed in our equipment and in our vehicle from paper to hardware. It's our first pass at the hardware, and seeing what the hardware problems are. So, we are cranking changes back as required by deficiencies in the hardware, and necessary changes to the hardware. But, the design is frozen to a pretty detailed level. Well, that is sort of a thumbnail sketch of where we have been as I see it -- my view point.
ERTEL: Well, I was talking to Saul Ferdman yesterday, and he brought out something that I think is a must for the history -- he said that during those early days, there were two separate configuration studies made. Do you know anything about that? Two groups working independently of one another, and worked through the problem, and came up with the same basic concept.
KELLY: Yeah, that was one of the -- that was one of the things we did right after we came on the contract. We put one group -- well, we put one group to work on the design of a vehicle that had essentially bare minimum pressurized cabin volume and as much of the equipment external, this was on the ascent stage, now - as much of the equipment external as possible. We put another group to work on a fully pressurized cabin -- a long cylindrical type thing with the crew having access to all equipment in-flight, and we did a fairly thorough study on the weight and operational advantages of a number of configurations derived from those two starting points. And they did tend to meet in the middle from a -- the most advantageous configuration tended to be neither extreme but something in the middle, which is what we have today. It's not the smallest cabin you could conceivably get, but it certainly doesn't house all the equipment by a long shot either.
ERTEL: I think it is really interesting... I don't see how you could have come up with any other type vehicle than we have now, with nothing like this happening, really.
KELLY: Oh, yeah, that was a lot of -- a lot of configuration study work. The descent stage fell into place more rapidly than the ascent stage. We went through twenty-odd descent stage configurations also and the configuration we chose showed up a pretty clear winner on the parameters we were evaluating -- weight, and operational flexibility. But the ascent stage wasn't as clear cut, and the problem we were looking at was sort of analogous to the difference between Mercury, and the Gemini capsules. Mercury capsule, why, everything was inside the pressure hull, and it was real hard for the ground crew to get at, but it was easy for the flight crew to get at, relatively. Gemini, they went just the other direction. They made just every- thing accessible from the outside; they packed the crew into a very minimum pressurized volume. So, at the time we started that study, why, it wasn't obvious which was preferable way to go for our vehicle. As we had proposed it, we had picked the middle of the road approach, but we didn't have enough backup, or justification, to show that this was clearly superior to other possible configurations. So we essentially developed that type of support - well, really, it's just coincidence that it came out with about the same answer, because we didn't try to force the answer; if we were doing the work, we wanted to come out with as objective an answer as we could.
ERTEL: How about the development of the GSE equipment; did it have much effect on -- of course, now you are just really getting into -- well into the manufacturing stage now.
KELLY: Well, the GSE has been a major part of the program right from the start. I think, looking back, it's the part that we underestimated the most, even though everybody told us -- don't underestimate the GSE -- -- even the NASA people themselves underestimated the GSE when we were all through. It's - the real problem is that to understand your needs in GSE, you have to know not only what your vehicle looks like in detail, what it contains; you have to know what your whole development and test program contains, what steps you are going to go through at every site along the way. And, this takes a large amount of detailed definition. Now, the way we set the program up, originally in engineering, has stood us in good stead; we haven't changed our basic organization. We have three engineering divisions, basically; a systems engineering, a subsystems engineering, and ground support engineering. Initially, when you first start the job, it's all systems engineering really; you are trying to define what the system is. And, from this evolves the more detailed requirements that the GSE people and flight equipment and subsystem people can work on. (We can go next door -- That's all right, I'll leave.) Okay, where were we?
ERTEL: On GSE, underestimations.
KELLY: Oh, okay, Well anyway. Systems engineering has to start off by defining the basic mission, and the basic vehicle, and then get into the major requirements of the ground systems complex. It's very difficult to get the GSE requirements pinned down in enough detail, early enough in the game. That has been our problem all the way through and I think that's -- well, I think it's just common to every program in that regard. We have made a major effort in this regard, and particularly last year, in 1965, when we were pinning down all aspects of the vehicle; we were working very hard to pin down our ground support equipment. We have over - oh, the last count I saw was 600, some odd, GSE end item types; these are different designs and, of course, there are many units in each end item and some of them are pretty complicated pieces of equipment. They get lashed together in a very complex fashion.
We've evolved a concept of GSE functional sets, to put them into a manageable grouping. A functional set is a grouping of ground support end items that functionally operates together. So, we've set up a specification and design control system that's based on the functional set concept; and we control the requirements -- system requirements -- to the functional set level, and the detailed design requirements are controlled down to the end item level. So, it's been a long hard pull on GSE, and we are just starting to see the light at the end of the tunnel, as far as getting our requirements defined, and the designs pushed out. The volume of work there is fantastic. We are putting out more total engineering drawings on the ground support equipment, than we are on the flight vehicle. And, that is saying something, because it takes a lot of engineering drawings to build a flight vehicle.
ERTEL: The interface you had with Link on the simulator gave quite a bit of input into....
KELLY: Well, yes, Link is actually a subcontractor of ours, so, essentially, we give them direction, and information, they need in order to build the simulator.
ERTEL: I was wondering if it comes from here job responsibilities-- I mentioned, as one of the best contract awards -- just like turn another shoulder, and keep on "horning it."
KELLY: Yes, we have a training equipment operation that handles this, and, well, the simulator has been another case in point where it's a real complicated device -- this simulator, very fancy, and can do a lot of things, but it has a long-lead time on getting it developed and built, and, consequently, the need for information -- firm information -- for the simulator has outstripped our ability to provide it, because we didn't even have the firm information for the flight vehicle, when we needed it for the simulator. So, we have had to work out a deal there where we just froze the information level at a particular level, and we're building the simulators to that, and we are going to upgrade and update them, with modification kit that is coming along later. But, there wasn't much alternative on that sort of thing.
ERTEL: It seems to me that you have got GSE in pretty good shape now. I think you have one installation complete at North American, and one here, and another is pretty well underway at Houston, and at the Cape.
KELLY: Well, let's put it this way -- we're in a lot better shape than we have been. I would say our requirements -- our definition of requirements -- are in real good shape. This information has come along pretty late in many cases, and, therefore, we still have schedule problems in certain areas. But, as I say, it is getting better, we can see a lot of improvement and we are putting forth a terrific effort on this part of it. Looking back at the proposal days, I think we did underestimate it, but on the other hand, I would be hard pressed to see how you can accelerate it much. The real way to accelerate the GSE is to not change anything on the vehicle. If we went right ahead -- charged right ahead and built this vehicle, our GSE would have come along a lot sooner too. But, while the vehicle we are trying to support is changing all over the map, you just can't get serious about the GSE, from a detail design standpoint.
The other thing that has changed, which I didn't mention in the history here; but there has been quite an evolution in the required development test program on this vehicle. We started out with an extremely thorough test program, I mean, you would call it methodical, and, also, idealistic. We were going to do everything on an absolutely minimum risk basis, both from a technical standpoint, and a dollar risk. And, we've had to back off on some of those ideas under the press of schedules, dollar limitations, and what have you. I still think we have a very adequate program, but it's been scaled down quite a bit from the original version, and, consequently, every time you change one of these programs, why, you change the GSE requirements, because the GSE is also tied to the particular site, and the particular way you are going to run the test.
ERTEL: I believe that there was even some LEM tests figured in the Little Joe Program at the start.
KELLY: Yes, we had some Little Joe tests originally, and, oh, one thing we did in the first year of the program was -- we had a side group working very intensively at how far -- how could you simulate the lunar landing situation on earth; what was the best way to do it; could you do it with real hardware, and with a real flight vehicle, and did it pay to do it this way, etc. We went into that for well over a year . In the beginning of the program, everyone had an intuitive feel that -- gee, we ought to be able to rig up something with a helicopter, suspending of a LEM from a helicopter, or a balloon, or something like the Langley landing rig; there ought to be someway of taking a real vehicle and flying it down to touchdown, just as though it were on the moon.
Well, we finally figured out ways to do it, all right, from a technical standpoint; but it was an extremely expensive proposition, and there was still deficiencies in the simulation. We had one rig where you would have turbojet engines that would support five-sixth of the vehicle's weight, and the LEM would fly on gimbals inside that. Any scheme that you worked out, to try to simulate the lunar gravitational field on earth, tended to be very complicated, if you did it with a real LEM. The original idea was to land a real LEM. It turned out to be extremely complicated; they all had certain technical deficiencies - you couldn't say, well, now we've got it, we're absolutely simulating the moon - it still wasn't quite right, and we could also show that, in most cases, this device couldn't do the most critical type of maneuvers. For some reasons, or other, and they differed depending on the type of scheme you had in mind. You couldn't simulate the most difficult aspects of the mission.
So the conclusion that we came up with, and I guess this was a joint Grumman-NASA conclusion, after about a year of real heavy preliminary design work on this type of flying-landing vehicle, was that it was technically feasible to produce a limited approach to a free-flight landing vehicle simulation. But, if you tried to assess it on any sort of cost effectiveness yardsticks, and compare it with what you could get with a straight fixed-base simulation, it just didn't look like it paid. You were talking hundreds of millions of dollars before you were through with this thing, and it was a dead end. So, it was washed out of the program -- but after a lot of careful thought.
And, what we have in its place, of course, is we have a pretty good fixed base simulation program, like the old Link trainer, but with a pretty good visual display, and a real good mathematical simulation of the vehicle, and we can hook certain of the flight control hardware into this rig. Our FMES (Full Mission Engineering Simulator) will be operational in about another two months, and this is our final step in the progression. We've had several different levels of engineering simulators, starting out with a simple one that we could build quickly, and working up into quite a pair of quite good ones last year, but were still limited strictly to simulation. Now, this FMES will have the capability to do a complete mission from end to end, which the others couldn't do -- the others were all pieced together. And, also, on the FMES, we can hook in the hardware -- the critical hardware in the flight control system -- the autopilot type equipment can be hooked in and actually fly with the FMES, with the computer closing the vehicle dynamic loop. So, this is what we have, and it's a good engineering tool, and is a pretty good pilot training tool, crew training tool. Of course, for crew training, this is augmented by the Link mission simulator, which is really a trainer with a lot of capability. It can simulate malfunctions of all types, etc. That was kind of an interesting historical side line, I think.
ERTEL: You don't know anybody that still has some of those early sketches, do you?
KELLY: Well, there must be some proposals around.
ERTEL: No, I mean, talking about early -- before the proposal stage; the sketches that led up to the one used in the proposal.
KELLY: Yeah, we might be able to dig up some of them. We put out a published report that came out -- oh -- three, or four, months before the competition. We put out an unsolicited study, so to speak that we showed to NASA people, and we can probably find one of them somewhere in the archives. I don't have one handy.
ERTEL: In the long run -- well, our Project Mercury history, that will come off the press about the first of September -- I think this will help us a lot in our search for material, because people will see then what kind of product we're trying for -- it's well illustrated, about 250 pictures.
KELLY: Let me just see..... (looking in desk for illustrations? ) By the way, we've been through our critical design reviews on the vehicle with NASA and that's an indication of the solidity of the design now - we've been through, essentially, a four-part series of critical design reviews that took place, I guess the first one was in December, then January, and February.
KELLY: Yeah, here's a little...
KELLY: This was the proposal -- that's right in front of the proposal. See, everything was on a smaller scale, because it was a lighter vehicle. Smaller, much smaller, cabin; had the two hatches, had a five legged gear. Everything was optimized for that particular proposal configuration.
ERTEL: I've probably got some of these down there in my files that I haven't gotten to yet...
ERTEL: Yeah. We are supposed to be the archives for all the Apollo, and Gemini. And, the last three weeks, I think, we got about 25 -- between 20 and 25 file cabinets full of stuff -- a lot of material to wade through.
KELLY: Let's go next door. Here's -- this is the descent stage, and these were five -- we studied some 20-odd configurations -- but, these were five of the principal configurations. This is the one we now have - the simple cruciform, with the tanks, and balancing locations.
ERTEL: But you said we now only have the two tanks, right?
KELLY: Oh, no, not the descent stage.
ERTEL: That's right -- that's the descent stage.
KELLY: Oh, here's what Saul was telling you about. This was the two team approach. I should have had this one when I mentioned it before. This was the two team approach - we started one group out with a -- looking at the real large type of cabin, and the other group was going to look at the small cabin, with the aft-equipment bay attached -- this was the general chronology that we tracked down to -- by the time we got down to here, we saw that they were just about the same again. We did go through a number of configurations.
ERTEL: That one will be a good one to use -- here's another.
KELLY: Oh, that's another thing I forgot to mention. We went through many landing gear configurations. Once we adopted this - this cruciform descent stage, that dictated a four-legged gear; but, there are still many possibilities as to exactly what type of four-legged gear .
We looked at lateral fold gears, radial fold gears -- lateral fold -- and finally, a cantilever configuration. This is lateral fold -- the gear folds out sideways -- one leg gets... and radial fold tripod, the two inner legs fold down themselves, and just pivots straight back. And, these are all tripods, and we compare them with the so-called cantilevered gear approach, where the primary gear strut also takes bending loads, and this was the configuration that we finally adopted.
Our proposal gear was a tripod, and it was a fixed gear -- it didn't have to fold. This gear was really a rule beater. This was a real good gear for the vehicle, if you took everything in the proposal, absolutely literally; and that was what we stated, that it was a rule beater. We put little teeny pads on it, here, that just met the bearing strength requirement in the proposal, for instance, and we made a five-legged gear, because we could just get it into the -- with the size the vehicle was turning out to be, we could just get this stiff-legged, five-legged gear into the confines of this LEM adapter, spacecraft LEM adapter, and it just fit. If we had made the four-legged gear, it wouldn't have fit, cause you have to make the gear a little longer to get adequate stability and it wouldn't have fit without folding it.
Now, obviously, it was better to have a fixed gear, than a folding gear, here, if you could possibly avoid it. But, oh, almost the first day we were on the contract, when NASA said -- well, look we can't let it just sit at that bearing pressure that we put in the proposal; we're going to change that.
Well, as soon as they changed the bearing pressure, we had to go, instead of these nine-inch pads, we had to go to the 36 inch pad, something like that. Then it wouldn't fit in without folding, and once it didn't fit in without folding you lost all the advantage of a five-legged gear, and you might just as well go to a four, even consider a three; but four was better. If you were going to have to go to a folding gear, then it is better to fold a fewer number of legs and, of course, you don't have to repeat the complex mechanism so many times. But, if you could go to a five-legged gear, or even a six-legged gear, and, by virtue of doing that, not have to fold it, then you are better off. That is why we proposed it this way, but, as soon as you change a couple of key rules, you are stuck with a folding gear. Then it pays to go back to four legs -- three legs .
ERTEL: Of course, by using the folded legs, you've got a little more space to play with than you would otherwise.
KELLY: Yes -- yes, we were able to make the whole vehicle somewhat larger, and we had to; as it became heavier, we had to make it larger. I guess that is all that's in here, on straight history.
ERTEL: How about the -- do you feel the switch over from the CPFF, to CPIF, has straightened things out, or, is it too early to tell yet, with the subs, as far as more clearly defining the general situation -- what they are going to do. I have the impression, from some people, that during the CPFF times -- now, this was not here at Grumman -- but, some of the other contracts had also undergone the same thing; that, under CPIF, things just seem to be a little better all the way around, as far as having responsibilities defined clearly, and get away from two people doing the same thing.
KELLY: Well, I don't know whether I attribute it to CPIF contracting, or not, but, I think the fact that we've been able to redefine the contracts in; much more detail has been beneficial. Now, in the process we made it CPIF, and whether that is good, or not, I wouldn't comment on it. But the fact that there was a detailed redefinition, which has resulted in scrupulously explicit responsibility delineation -- that has been very beneficial, I would say, to all parties. The whole operation is on an extremely business like basis, and there's no fooling around anywhere along the line, no room for it; it's all contractually documented. I think that has been very good; it has instilled what NASA calls "rigor" into the program. It really has. It has been a very well disciplined operation, I would say, and we are operating under an extreme degree of discipline with NASA, and we have locked our subcontractors into a similar headlock. I think it's in pretty good shape now, contractually.
ERTEL: You mean you finally started replacing "mortis" with "rigor", huh.
KELLY: Well, see, originally, we signed up on a work statement that pretty much said -- "Let's hold hands, and go to the moon. NASA, you and us, all the way" -- and now we've replaced that with a contract, and work statement, that stands about this high off the floor, and which incorporates, by reference, a stack of top technical specifications, that stands about this high off the floor. So, we've -- and the material in them is pretty good. It's really been worked over, pretty thoroughly, by both parties. So, we've got a very detailed definition of just what our job is, and we've done similar tactics with every one of our subcontractors. We really know where we stand now, it's a pretty mature operation.
ERTEL: Well, I recall, from items that I got out of the quarterly progress report, and so forth, the vast amount of things that were added to the original contract, anyway -- I mean, one responsibility after another, the plans, and everything else, were all added after the start.
KELLY: Right. Well, when we renegotiated, we renegotiated into the new contract -- the amended contract -- something like 165 contract changes, up to that point. Since that time, there has been another 60, or so, added. The ones that have been added tended to be much more detailed in nature than the previous ones that went before. The ones that went before were real big picture things -- changed the delivery dates, changed the number of units, and this sort of thing.
Here's some ancient history here -- the crew seat. We had the pilot sitting down in the original proposal, I guess because it seemed to be the thing to do. Actually, it was somebody from NASA, in the early days -- I don't know who got credit for this, it's probably somewhere in the files -- said -- -- why don't we chuck out those seats; we're only in the vehicle flying this thing for about six hours, or so, why don't we get rid of the seats, and let them fly it like a delivery truck.
Well, everybody sort of laughed at first, and then started to think that may- be that wasn't such a bad idea. So, we looked at a lot of schemes, and had one -- this one was called the bar stool, for obvious reasons. There, you just sat on it like little stools, just like a delivery truck. And we had a thing called the bird cage, where he was inside a cage like affair with webs, and we showed that to the astronauts, and they thought that all these straps and things, and particularly in the bird cage, were still too confining -- they said, gee, well, why don't you just let us stand there, we don't need anything at all.
So we laid out configurations like that, that had essentially no restraints at all, and we evaluated them in the drop test rig, where we actually dropped the pilots down, inclined in a rig, and abruptly stopped them through g force, or something, and we and the astronauts found that you needed something, and you had to have more than just hanging on with your bare hands -- mainly, because you tended to lift up off the floor. You had to have something to restrain your feet, and to keep you from being thrown sideways; you could control fore and aft motion with your arms all right, but you couldn't control sideways. So now we have a stand-up crew position, which we've had for quite sometime, but with a fairly simple restraint system - it's a window washer belt type system, that you buckle into, and it's got a couple of ropes and pulleys, and that's about all there is to it, and you buckle onto that. We also have a pair of handgrips, on the dash, they can hang onto. That's proved to be a pretty - up to now - it's proved to be pretty effective crew station for our mission.
ERTEL: How about the fuel cell change -- the fuel cell and batteries?
KELLY: Well, there's a bit of history, yeah. That was a change that we didn't agree with at the time NASA directed us to do it. I think, in retrospect now, it wasn't such a bad idea after all. The reason we were reluctant at the time was because our weight situation was critical, and there was some weight added with this change. But on the other hand, we couldn't deny that the simplicity in the battery system was very attractive, compared to the fuel cell system.
For our part of the mission, the weight penalty associated with batteries was not outlandish -- it was in the order of a couple of hundred pounds of separation weights, so it was not out of the question like it is -- I think, with the CSM, it would be out of the question to go battery there. You would be talking much more... So, I think this is a change that, if I had it to do all over again, I would say, probably, it was a pretty good idea. I didn't agree with it, at the time, mainly because I was obsessed with the weight problem.
But it has been working out pretty well -- the batteries have been performing real well, thus far, in their test program, and the other thing that has helped is that we've been able to control our growth of electrical power consumption. One of our biggest fears with the battery was at the time we were making the decision, we didn't have a very good history on being able to control the power demands, and if the power requirements had gone up appreciably, as we feared they might, we would have taken a much more substantial weight penalty.
As it turned out, we've been able to hold the line quite well on power -- we've done much better there than we have done on weight. And, as a result, the batteries have never really gotten into trouble, from a capacity viewpoint. See, the batteries are very capacity limited; you build a battery a certain size, and, man, you've only got so much in there, and that's all there is when you get rid of that. So, I think it's worked out pretty well.
ERTEL: Well, I don't know who was behind this move to change this, but certainly, after our experiences in Gemini to date, well, you know the fuel cells just aren't far enough along, yet, to be entirely trustworthy, I don't think.
KELLY: Well, of course, the counter argument to that always was - you have to have fuel cells to do the CSM part of this mission, which is still true. In other words, we have to have fuel cells developed in order to perform this mission.
ERTEL: Yeah, but they could - I mean that if something does happen, they have still got the batteries onboard... for the abort, and enough to get back.
KELLY: They don't have enough to get them all the way back; our batteries won't get them all the way back.
ERTEL: You don't think so?
KELLY: No, See, that was always the counter argument. Say, okay, yeah, great, it would be simple with batteries, but batteries are relatively inflexible -- if you powered and the man goes way up, well, you sort of had the course; and you have got to develop a fuel cell anyway to handle the translunar part of the mission. However, the batteries offer such a large reliability advantage, that it's pretty hard to argue against them, if we could get anywhere close from a weight standpoint. And, when we were finally, able to convince ourselves that ---- yeah, we were pretty close, boy, that pretty much eliminated the resistance. I think Max Faget, though, started this battery movement, and in retrospect, I would say it's pretty good sense.
ERTEL: Well, Tom, I certainly thank you for your time, I heard you say
you have got to get out of here ... End of Interview.