Moon Lander: How We Developed the Apollo Lunar Module (21 page)

I had one meeting with them in which, in response to pressure from Shea in Houston, I challenged the schedule status and manpower estimates of all LM Engineering, including Drafting. They unrolled dozens of detailed estimating sheets broken down to individual drawings and went through them page by page while Krier kept up a high-pitched skirl of justification and Chandler blew cigar smoke over my head. By the time they were finished I was glad to escape with an estimate no higher than their original position. After that I relied on Rathke to deal with them—he understood them better and, if necessary, could blow cigar smoke back at Chandler.

The LM Release Records Group was led by Al Caramanica, a bright young graduate from Drafting. He had the task of transforming the group to carry out far more complex functions than it had in the past. Release records on aircraft programs was a mundane “go-fer” and record-keeping activity, which
obtained and recorded the necessary approvals on drawings from the analytical engineering groups, such as Stress, Thermo, and Loads, and released them to Manufacturing and to the customer. They also issued and recorded all the engineering orders, which were changes affecting released drawings. With the advent of rigorous configuration management on LM, the Release Group became the implementer and enforcers of the configuration control system. They handled each drawing and EO as required by NASA’s regulations, making sure that CCB approval was obtained before releasing a controlled drawing. The Release Group became a major resource for the CCB, providing the manpower needed to enforce configuration control discipline in LM Engineering and working closely with the Quality Control (QC) Department, which provided surveillance and enforcement elsewhere on the LM program.

The ultimate purpose of the engineering drawings was to provide design information to LM Manufacturing from which they could make and assemble the spacecraft and its GSE. Manufacturing was under heavy schedule pressure from program management and needed to have the drawings released to them in sequences that would permit efficient buildup of production activities in the shops. Close cooperation between Engineering and Manufacturing was imperative. Frank Messina, manager of LM Manufacturing and a seasoned veteran of aircraft production since World War II, and Bill Rathke had worked closely with each other on several airplane programs; they both knew what was needed. A joint Engineering-Manufacturing production planning committee was established to coordinate Engineering’s drawing releases with Manufacturing’s needs. Initially headed by Bill Craft of Engineering and Bill Bruning of Manufacturing, this group met daily, comparing Engineering’s plans and ability to release drawings with Manufacturing’s requirements and preferences on the sequences of releases. The result was a “negotiated” release schedule that gave Manufacturing the best available sequences and kept them informed in detail as to what was coming “down the pike.”

In especially tight schedule situations, in which days and even hours counted, we modified our basic drawing-release system, taking calculated risks to save time. We permitted the use of advance prints, which had not yet been analyzed and approved by Stress, Loads, or other analytical engineering groups, on time critical drawings. We told Manufacturing what the risks were and what aspects of the drawing might change after the analytical review, letting them decide whether the time saved was worth the risk of change. I was never comfortable with that subterfuge and tried to limit its use.

Unfortunately, the natural order in which design drawings are produced is opposite from Manufacturing’s needs. A designer visualized the whole unit or assembly he was designing, producing sketches and an assembly drawing showing how the device fits together. From this the individual detailed parts making up the assembly were pulled out and drawn up, complete with the dimensions, material specifications, and instructions that allowed the shop to
make them. Manufacturing naturally wanted the detailed parts drawings first, so they could obtain the raw materials and purchase components and then fabricate the parts. With all the parts in hand they needed the assembly drawings to put it all together, although Manufacturing Engineering needed the assembly drawings early to plan and design the assembly tooling. The production planning committee did their best to accommodate these conflicting priorities.

As 1965 stretched into 1966, there was no relief in the unending schedule pressure on the LM program. Because LM was the last major Apollo program element to be defined and placed under contract, Grumman started a year behind everyone else. NASA’s George Mueller and Joe Shea concentrated special attention on Grumman’s ability to meet schedules, and they were not at all satisfied with our performance.

Shortly before the M-5 mockup review, Gavin, Mullaney, and I were summoned to Houston for a review of LM management’s performance by Joe Shea and Bill Rector. Shea told us that NASA planned to have the first Earth orbital flight test of the CSM on a Saturn 1C booster in February 1966, and they wanted LM’s first flight to be in February 1967. Given Grumman’s performance so far there was no hope of achieving that goal. Grumman was demonstrating a 0.8 “schedule slip ratio”: every five weeks our schedule promises slipped by four weeks.

“Your management performance in other areas has been no better,” Shea said in an exasperated voice. “Your cost estimates have already doubled since negotiations and are still rising. The LM’s weight grows weekly with no sign of leveling off. By every measure, your performance is extremely bad. We had hoped Grumman would outperform North American—it looks like you are even worse. You are the people we’re counting on to make Grumman perform. If you can’t do it, we’ll have to ask Towl and Titterton to find someone who can.” I shuddered inwardly at this bare threat. Shea was at the end of his patience with us.

Shea recounted a list of serious deficiencies in Grumman’s management of the LM program. He said we had encountered unnecessary problems through management ineffectiveness. Examples were late procurements, slow and incomplete GSE identification and delivery, and sluggish staffing buildup. He faulted our subcontractor relationships, traceable to fractionalized divisions of responsibility within Grumman, which prevented us from executing a “total package procurement” approach and establishing teamwork with our major subcontractors. Most grievous of all, Shea accused Grumman of failure to understand or accept NASA’s Apollo program philosophy. He cited many examples of this, such as our recurring disagreements over reliability: whereas NASA’s approach was careful design and thorough ground testing to disclose weaknesses and verify fixes, Grumman, based on its major GNC arguments with MIT and other debates, seemed wedded to a
statistical approach. We seemed to dispute NASA’s selected approaches to black-box maintenance (ground only), qualification test requirements (test to failure not always necessary), and manned-flight prerequisites (based primarily on ground test program). We made little effort to find common-usage equipment from North American Aviation in areas like GSE, reaction control, and communication, despite repeated NASA prodding.

Finally Shea faulted us for an arrogant attitude: “Grumman is a proud organization and you are proud people. Yet your box score on technical decisions has not been good.” He found Grumman reluctant to admit to a problem without having the solution in hand, which we expected NASA to rubber stamp. He cited, for example, our switch to riveted construction on the LM’s front face without telling NASA—for months they thought it was all welded. Similarly, Grumman was far along with a two-tank ascent propulsion design before NASA knew we were considering it—the same when we went to three fuel cells from two. In every case NASA found that Grumman had failed to perform analyses that were necessary to determine whether these changes should be made. He urged us to talk out technical problems with NASA as they arose so they could jointly contribute to the solution.

It was a humiliating dressing down, but Shea ended the meeting on a positive note, outlining a series of steps to correct these shortcomings and reestablish a team relationship between NASA and Grumman. These consisted of a series of meetings in each of the recurrent problem areas where the differences in philosophy and approach would be thoroughly aired and resolved.
³

I was shaken by that meeting. I felt that much of it was directed at me personally and most of the problems lay in my areas of responsibility. Some of the perceived differences with NASA did not even reflect my own views; for example, I strongly favored the pragmatic design-and-test approach to reliability over statistics, and I was a major proponent of NASA’s decision to eliminate in-flight maintenance. That meant that I had not been effective in explaining and enforcing my own concepts within LM Engineering at Grumman. The indictment of our poor schedule performance was true, as I saw week after week at the project meetings, where despite the explanations and excuses, the bottom line result remained terrible. As for arrogance, I tried hard to avoid it myself and never failed to reprimand my people when I heard them referring to NASA or others in a demeaning manner. I felt we were newcomers to manned spaceflight and had no basis for a superiority complex.

After the M-5 review, Rathke, Carbee, Whitaker, and I bore in on the drawing production problems. In the weekly project meetings and my daily technical staff meetings we looked for constraints to drawing output and took actions to eliminate them. A new wave of Engineering staff expansion was approved and implemented. Although we had been satisfied at our former staffing levels, Shea’s review of our engineering shortcomings showed that we had much more work to do. The glamour of the Apollo program helped us
attract talented engineers from throughout the country. We expanded into Plants 35, 5, and 30, and we applied overtime and added job shoppers. The production planning committee intensified its efforts to support Manufacturing; this primarily meant developing credibility to Engineering’s promise dates.

There was progress from all these efforts but it was slow and not steady. Unforeseen problems and complications confounded our attempts at orderly recovery, and increased definition and understanding of the program resulted in longer lists of end items to be delivered. It was like fractal geometry—the closer one viewed the task ahead, the more detailed tasks were visible. At times the amount of work ahead seemed unassailable; the pile was always growing. There was a mounting sense of frustration within Engineering: the schedule pressure was unrelenting, and no matter how much we did, it was never enough. Gradually our drawing output climbed, from fifty a week to one hundred, then two hundred and still climbing. We needed more than four hundred drawings a week to meet the program’s requirements.

NASA’s unhappiness with our performance attracted Grumman corporate management’s attention. We began seeing more of Senior Vice President George Titterton, Chief Technical Engineer Grant Hedrick, and Director of Flight Test Corwin H. (“Corky”) Meyer.

Hedrick was a gifted engineer who had joined Grumman as a stress analyst during World War II when the bridge designing he was doing for consulting engineers Parsons and Brinkerhof was suspended for the duration. He impressed Bill Schwendler and Dick Hutton with his keen ability to analyze aircraft structural designs and his careful application of basic engineering principles. His wartime structural design work on the Tigercat, Bearcat, and Albatross earned him the post of chief of structural engineering in 1947. In that position he succeeded Bill Schwendler as the unofficial “chief blacksmith of the Grumman Ironworks,” the guarantor of the structural integrity of Grumman’s airplanes. He later became chief technical engineer, in charge of aerodynamics, structural analysis, and structural design. Widely respected throughout the aerospace industry, Grant Hedrick was “Mr. Engineering” within Grumman.
⁴

Hedrick was of medium height and solidly built, with sandy hair and wise-looking eyes behind rimless gold eyeglasses. A highly ranked amateur tennis player, he was trim and athletic. He grew up in Fayetteville, Arkansas, and graduated from the University of Arkansas with a degree in civil engineering. His country background showed in his practical, no-nonsense approach to engineering in which he concentrated on the facts and avoided unfounded theories or speculation. He was demanding of his subordinates, insisting that they study their problems carefully and perform thorough analyses of alternatives before coming to conclusions. Those who did not meet his strict standards seldom got a second chance. The structural engineering community at Grumman looked up to him as their leader and the arbitrator of safe airplane
design. I did not know him very well, and at first I was somewhat intimidated by his reputation and his crisp manner of cross-examination.

Hedrick’s increased involvement with LM Engineering was quite helpful. He set up a regular weekly meeting with the Engineering managers and section heads that I attended with any LM Engineering support I required. Both sides could select agenda items for discussion. These were useful meetings aimed at insuring that LM was making use of the best talent available at Grumman to solve technical problems. Hedrick had a remarkable ability to cut through complex arguments or voluminous calculations and find the hidden flaw that rendered the whole argument suspect. After NASA’s senior structural analyst, Joe Kotanchik, watched Hedrick in action at Grumman, he enlisted him to serve on an ad hoc task force assembled by Marshall, Houston, and their contractors to solve a serious problem, known by the undignified term “Pogo,” on the Saturn booster.

Pogo vibrations occurred along the long axis of the booster, when fluctuations in thrust by the rocket engines fed back into increased pressure at the pump inlet, reinforcing the magnitude of the initial disturbance. It was feared that uncontrolled Pogo vibrations could destroy the Saturn. This phenomenon only occurred in flight, where the increased thrust produced increased acceleration and hence pressure at the base of the propellant tanks. With its cause understood, it could be modeled in computer analyses. Grant Hedrick was effective in guiding NASA’s analyses toward a practical solution of the Pogo problem. In a notable meeting at Houston with Kotanchik and the structures experts from NASA, Hedrick spent almost an hour studying an array of strip chart recordings and reduced data from the latest Saturn flight test that were displayed on tables in a large conference room. Asking questions as he went, he pondered its significance. Then he returned to one set of data and pointed out numbers that appeared unreasonable and required further checking.