Chapter 9

From Development into Qualification: Flight Tests

(July - December 1960)

In mid-1960, NASA and its Space Task Group hoped soon to begin launching a major qualification flight test for Project Mercury every six weeks. If all went well, these tests of the operational vehicles should permit a man to ride into space before the end of the year. But if Mercury's developmental experience to date was any guide, troubles could be expected to pyramid and might require more than six months to correct. Since the ultimate goal of Project Mercury was to achieve man-in-orbit rather than merely a sounding-rocket ride by a man into space, the Task Group would be running concurrent flight tests with the Little Joe, the Mercury-Redstone, and the Mercury-Atlas combinations. But attention and impetus were focused on the accomplishment of manned orbital circumnavigation.

NASA Administrator T. Keith Glennan sent a memorandum to his Director of Space Flight Programs, Abe Silverstein, on July 11, 1960, prompting him to make every effort to put forward to November the launch of MR-3, long designated the first manned suborbital flight. If that was not possible, Glennan urged Silverstein to hold fast the schedule for the first manned launching before the end of the year. Silverstein replied that the manned event had just been reset for the week of December 5. By mid-August 1960 the most realistic estimate of the earliest possible man- launching changed the program management plans once again and reset the MR-3 launching for mid-January 1961. As late as October 1960, this optimism prevailed while work on capsule No. 7 for MR-3 proceeded "somewhat better than expected."1

Having once called the Army's stillborn Project Adam a "circus stunt" because it proposed little more than shooting a man into space, Hugh L. Dryden, Deputy Administrator of NASA, had himself set a precedent for the criticisms of those influential scientists who came to regard Project Mercury as more of an exhibition than a demonstration. During 1959 few had raised their voices against NASA's plans and STG's development program for a manned satellite. But during this election year of 1960, many citizens scrutinized - and Eisenhower even established a commission to study - all national policies, goals, and ideals. [264] This White House-sanctioned introspection led to some criticism, not entirely constructive, of the civilian space agency, which all too often was equated with Project Mercury.2

Most Americans appeared to approve Mercury as a potentially stupendous adventure, and many Congressmen anxiously hoped that NASA would mobilize the Nation's vaunted technological know-how to put the first man above the atmosphere. Although Dryden, George M. Low, and other NASA officials recently had warned repeatedly that the Russians could and likely would achieve manned space flight first, no one in NASA seemed to wonder whether the Soviets would send men on ballistic suborbital missions before committing a man to orbital flight. Most citizens seemed to confuse their feelings of hurt pride with loss of prestige and were reluctant to accept Eisenhower's difficult rationalization that America should abjure any "space race" with Soviet Russia. But NASA followed Eisenhower's leadership in this matter and reinforced the official attitude by insisting that Mercury was an "R and D" program whose pace could not be forced.3

Glennan in his public statements appeared torn between the pressures of public sentiment expressed through Congress and the news media, on one hand, and the demands of loyalty to the Chief Executive and to technological realism, on the other. Aware of the Nation's late start in rocket propulsion development and yet of its amazingly rapid achievement of a workable ICBM, Glennan knew that the United States still did not have the weight-lifting prowess to join an avowed contest with the U.S.S.R. But Glennan also shared the aerospace community's satisfaction on May 20, 1960, when the Atlas first flew higher than 1,000 miles and over 9014 miles downrange from Cape Canaveral into the Indian Ocean. By this time the Thor and Jupiter intermediate-range missiles were operationally deployed abroad. The Titan ICBM, in spite of some developmental failures, was emerging into a second-generation intercontinental missile.4

Mercury still was only a fractional part of NASA's total space effort, but publicity and public interest had reinforced each other until the manned program clearly had become the most promising hope of "beating" the Russians into space. When the Soviets orbited Korabl Sputnik II on August 19 and the next day recovered two dogs, Strelka and Belka, from it, grounds for complacency among Americans evaporated.5 National phobias, stimulated by partisan criticism of the alleged "missile gap," were further distorted by technological chauvinism with respect to Soviet accomplishments in space. Popular attitudes were exacerbated after the "spirit of Camp David" was destroyed by the U-2 incident and after Khrushchev used the U-2 affair to destroy the summit conference in Paris.

Speculations on high policy and international relations were not the business of the field workers on the Mercury program. But as citizens they could not avoid being aware of some wondrous possibilities for the historic significance of their work. Both landlubbers and space lovers could find many excellent reasons to think that the ICBM and nuclear warheads might possibly become plowshares of peace rather than tools of terror if directed toward the exploration of space. [265] Peaceful coexistence and even international cooperation might be force-fed by the exorbitant economics of the competition to put men into orbit. Whatever one's particular brand of concern, there were motives aplenty to work on Project Mercury.

Toward the end of June 1960, the Space Task Group took another hard look at the status of Project Mercury. Having formalized three separate series of engineering inspections and tests - progressing from development through qualification into reliability phases - STG faced with increased confidence some criticism from technical associates. It felt it could gauge accurately the soft spots in the major systems for Mercury. Of the 17 nominal systems for the capsule, all but five or six by June were reported finished with qualification tests and almost done with reliability testing. The major unfinished items were the reaction control system, pyrotechnics, the retrograde and posigrade rockets, and the satellite clock.

Capsule system tests had revealed that certain pressure regulators, solenoid and relief valves, and thrust chambers for the reaction controls using corrosive hydrogen peroxide were going to be troublesome when operating in a high vacuum. On the other hand, the environmental system was progressing better than expected, with only five components still unqualified: the emergency oxygen bottle, a pressure reducer assembly, the odor and carbon dioxide absorber, a high-pressure oxygen transducer, and a suit-circuit water separator. The abort sensing and implementation system (ASIS) for the Atlas was 95 percent qualified, but its counterpart for the Redstone was not.6

The communications and tracking network faced four outstanding problems: no one had much experience with Atlas guidance and tracking at long ranges and low elevation angles; the reliability of the high-speed data links was unknown; capsule antenna patterns were erratic enough to make radar acquisition problematic; and control procedures and techniques as yet were untried.

Astronaut training, the Task Group believed, was virtually complete for disorientation, tumbling, and familiarization with high levels of carbon dioxide absorption. Adaptation to weightlessness and lectures on space sciences were 90 percent complete, but training in navigation and communications (at reduced pressures and with high heating, noise, and vibration rates) was less than a third finished. The training of ground crews in procedures for preparing, launching, and monitoring an astronaut in flight had only just begun. And NASA's planning for recovery operations in the summer of 1960 was grandiose, asking "virtually for the deployment of the whole Atlantic fleet." This requirement came down abruptly after NASA met with the Navy at the Pentagon and was shown that fleet operations of this scope might cost more than the entire Mercury program.7

The climax of the debate over reliability analyses came in early summer 1960, when NASA Headquarters decided to issue an independent contract with McDonnell for making assurance doubly sure. Associate Administrator Richard E. Horner and his deputy, Nicholas E. Golovin, the mathematical systems analyst who had come to NASA from the Advanced Research Projects Agency, achieved [266] their first point on June 9, 1960, when a separate contract with McDonnell was signed for a reliability study of all Mercury capsule systems. Estimated to cost $52,892 with a fixed fee of $3,323 and planned to be administered by the Bureau of Naval Weapons representatives in St. Louis, this small contract was designed to provide Horner's office with the data it needed to analyze and evaluate the reliability efforts and achievements of McDonnell, of all 10 capsule subcontractors, of some 200 suppliers, and indirectly of STG's reliability monitoring and mission planning.8

Golovin's approach to a reliability prediction program was unusual to both the Space Task Group and to many of his professional colleagues. It reversed the common procedure of beginning with parts analysis and proceeding to the whole system. Golovin had recently explained his theoretical point of view before the American Society for Quality Control, citing other missile program precedents for inverting the crucial problem: "start with a definition of failure for the system, and then work back through subsystems and components to the data on parts failures." Glennan and Horner had approved this approach as an aid to fulfilling their desires for better "confidence coefficients" before accepting the readiness of the capsule for unmanned and manned suborbital and three-orbit missions. This kind of systems analysis used deduction and fully exploited "numbers game" techniques and data processing machines to check on the inductive systems engineering of STG and McDonnell. The experimentalists at the working levels, and many of the engineering managers, including STG's Director, Robert R. Gilruth, believed they saw a worthless expenditure of effort in this innovation.9

NASA Headquarters saw STG dragging its feet on this issue by the end of June. Glennan therefore tried another tack. He wrote directly to James S. McDonnell, shortly after a personal visit and briefing at the factory:

As you know, during the last month there have been a number of discussions between my Office of Reliability and Systems Analysis and various members of your staff on the problem of Mercury capsule system reliability. These talks were the result of my having directed the Office of Reliability and Systems Analysis to prepare for me an objective quantitative evaluation of the anticipated mission and flight safety reliability of the Mercury capsule system. It has now been brought to my attention that discussions have not yet resulted in mutual agreement on getting this job seriously underway.

I would appreciate it if you would give the matter your personal attention and have your staff responsively consider providing, as promptly as possible, the information detailed in the enclosed "Proposed Work Statements for McDonnell on Mercury Capsule System Reliability."

If you foresee any serious problems in this connection, I would appreciate your bringing them directly to my attention, and I will be glad to set up a meeting in Washington to reach a full meeting of minds.10

The work statements enclosed in this letter, prepared by Golovin's assistants Landis S. Gephart, William Wolman, and Catherine D. Hock, called for precisely defined reliability definitions, assumptions, diagrams, equations, and estimates of each [267] subsystem design, together with all available test data from every source. The basic reasons for requesting this information were to allow NASA "to review and evaluate the techniques and the data employed by McDonnell" in its reliability report (No. 7007) issued almost a year earlier, and "to update and upgrade the reliability predictions and probability equations" for mission success in the light of uneven changes of component parts supplied to McDonnell:
With all its subcontractors, McDonnell has established a reliability requirement for each major equipment. This requirement has been expressed either as a mean time between failures for a continuously operating device or as a probability of success for a single shot device, and has been incorporated as a firm contractual requirement in the appropriate McDonnell Specification Control Drawing. McDonnell also recognizes that "a requirement without a test to demonstrate compliance with it is meaningless." Accordingly, McDonnell has specified a variety of tests aimed at demonstration of the reliability requirements imposed on its subcontractors.
Golovin and associates wanted to examine all test plans and test results on every Mercury capsule component from pre-installation acceptance through systems, compatibility, qualification, and life tests. In short, they wanted virtually a whole library of files at McDonnell opened for their inspection promptly, within two weeks if possible. This was not quite possible, but the founder of McDonnell Aircraft did reply personally to Administrator Glennan in mid-July:
I am happy to inform you that our company started work on 9 June 1960, the same day on which Dr. William Wolman made his first specific request, even though this request was only verbal [sic]. Our company is now at work on every one of the programs therein outlined even though we still have no contractual authorization for any of it.

* * * * * *

We are in full accord with providing as fast as humanly possible (without diluting other Project Mercury effort) whatever work is desired by NASA to assist in the reliability evaluations of Project Mercury. . . .11

A few days later Golovin's group, having requested Silverstein to show STG how invidious was its prejudice against the "numbers game," journeyed down to Langley Field and briefed the Task Group on how Headquarters proposed to raise quality by quantitative methods. Reliability goals for each major capsule system, progressive analyses, and periodic reviews, plus a new order of simulated mission-testing stringency, were proposed and accepted by STG. Since the last major reliability meeting at Headquarters on February 29, 1960, had been so acrimonious, STG was surprised to find how little difference there now appeared to be between Golovin's approach to reliability and its own. On July 21, Paul E. Purser logged this note for Gilruth: "Spent most of the day in the meeting with Dr. Golovin, et al. They sounded fairly reasonable. If we had held such a meeting several months ago, there would have been a lot less misunderstanding."12

Shortly after this rapprochement, Horner resigned from NASA to go to industry, [268] Golovin resigned later to join the President's Science Advisory Committee staff, and Gephart and Hock obtained an expansion of the McDonnell reliability contract to cover the astronaut's task description and performance evaluation. Glennan meanwhile pressured Silverstein, who pressured Gilruth, to do something formal about taking into account contemporary mathematical techniques used in missile programs to enhance managerial confidence in reliability, hence in readiness before a launch. Gilruth in turn gave the job to John C. French, who proceeded to organize a "reliability and quality assurance office" in the Space Task Group. There was special significance in the word "assurance," because STG had by no means capitulated to the statistical approach nor to the mathematicians' belief in the efficacy of reliability prediction.13

Had the qualification flight tests actually started earlier, perhaps much of the debate over what to expect from Mercury launches would have been obviated. But while still standing on the threshold of the major flight test program after almost two years of virtually simultaneous work on detailed design, engineering, and manufacturing, the Mercury spacecraft developers had to talk out some of these difficulties before they could call for a vote. Far more significant than the formal reliability program in the long run were the test philosophy, test programs, and the test work in "space chambers" that could more realistically simulate the hot/cold vacuum of the exospheric environment.14 To move in that direction required a move toward the "spaceport" at Cape Canaveral, Florida.

1 Memos, George M. Low to Abe Silverstein, "Information for Program Management Plan Meeting," Oct. 6, 1960; T. Keith Glennan to Silverstein, July 11, 1960; Silverstein to Glennan, "MR-3 Launch Date," July 16, 1960; Silverstein to Robert R. Gilruth, "MR-3 Launch Schedule," July 25, 1960; Walter C. Williams to NASA Hq. International Programs Office, "Monthly Summary of Project Mercury Activities," Aug. 8, 1960. Warren J. North, "History of Mercury Schedules: Earliest Possible Manned Flights," chart, Aug. 13, 1960. See also Abe Silverstein, "Progress in Space Flight," Astronautics, V (Nov. 1960), 24-25, 140-142.

2 For the Report of the President's Commission on National Goals, see Henry M. Wriston, et al., Goals for Americans (Englewood Cliffs, N.J., 1960). Note especially the section by Warren Weaver, pp. 101-124, on "A Great Age for Science." Cf. J. L. Penick, Jr., et al., eds., The Politics of American Science: 1939 to the Present (Chicago, 1965), 221.

3 House Subcommittee of the Committee on Appropriations, 86 Cong., 1 sess. (1959), National Aeronautics and Space Administration Appropriations, testimony of Hugh L. Dryden, 15; House Committee on Science and Astronautics, 86 Cong., 2 sess. (1960), Review of the Space Program, Part II, testimony of George M. Low, Feb. 16, 1960, 761. For an excellent analysis of political positions and public opinion on American space policy (1957-1963) as a whole, see Vernon Van Dyke, Pride and Power: The Rationale of the Space Program (Urbana, Ill., 1964). Eisenhower's position is described on pp. 82-83.

4 Glennan's introspection on the role of international competition was best expressed in an address at a Yale University symposium on Oct. 7, 1960. See also letter Glennan to Eugene M. Emme, Oct. 19, 1965. For an overview of Air Force programs, see Ernest G. Schwiebert, "USAF's Ballistic Missiles - 1954-1964: A Concise History," Air Force and Space Digest, XLVII (May 1964), 51-166, later published as A History of the U.S. Air Force Ballistic Missiles (New York, 1965).

5 The best open monograph comparing Soviet and American space accomplishments is Charles S. Sheldon II, "The Challenge of International Competition," paper, third American Inst. of Aeronautics and Astronautics/NASA Manned Space Flight Meeting, Houston, Nov. 6, 1964, revised and reprinted in Senate Committee on Aeronautical and Space Sciences, 89 Cong., 1 sess. (1965), International Cooperation and Organization for Outer Space, Appendix A, 427-477.

6 "Project Mercury Discussion," brochure, STG, June 20, 1960. See also memo, Dieter Grau to Dir., Guidance and Control Div., Marshall Space Flight Center, "Unsatisfactory Condition on MR Abort Sensing System," Oct. 11, 1960; minutes, "Resume of Mercury-Redstone Panel 2 Meeting," LOD-MSFC, Aug. 24, 1960.

7 "Project Mercury Discussion," B-276, B-187, B-258, B--204; comments, William Underwood, Executive Sec. of CMLC, to Eugene M. Emme, Nov. 1, 1965; draft Ms., B. Leon Hodge, et al., "Recovery Operations Portion," for Mercury Technical History, Aug. 1963.

8 NASA Contract No. NAS-190, "Reliability Study of Mercury Capsule System," June 9, 1960, was signed by William P. Kelly, Jr., for the government and by D. P. Murray for McDonnell Aircraft Corp.

9 Nicholas F. Golovin, "An Approach to a Reliability Prediction Program," American Society for Quality Control. Transactions of 1960 Convention, San Francisco, May 25, 1960, 173. See also, memos, Silverstein to Deputy Assoc. Administrator, "Project Mercury Reliability Analysis," June 21, 1960; Golovin to Dir., Office of Space Flight Programs, "Project Mercury Reliability," June 23, 1960. "Old data and wrong ground rules gave bad figures from our standpoint," said Walter Williams in interview, Houston, Aug. 23, 1965. Silverstein and Low tended to side with the working levels on this issue: see Low, comments, Oct. 5, 1965.

10 Letter, Glennan to James S. McDonnell, Jr., June 30, 1960, with enclosure [from which next quotation is taken], "Proposed Work Statements for McDonnell on Mercury Capsule System Reliability," June 30, 1960, 2.

11 Letter, McDonnell to Glennan, July 13, 1960, NASA Central Files, Washington. Apparently Mr. McDonnell was unaware of the NASA-MAC reliability contract NAS-190.

12 Paul E. Purser, log for Gilruth, July 21, 1960; "Informal Reliability Discussion for STG by AAR Staff," July 21, 1960; Ms. notes on reliability meeting, Purser, July 21, 1960; F. John Bailey, Jr., interview, Houston, July 16, 1964. For a later statement of Headquarters' policy see Landis S. Gephart, "NASA Requirements for Reliability and Quality Assurance," in Western Space Age Industries and Engineering Exposition and Conference: NASA Day, April 27, 1962, NASA SP-4 (Washington, 1962), 49-56.

13 John C. French, interview, Houston, Aug. 3, 1964; memo, Gephart to Everett W. Quintrell, "Background Information on Astronaut's Task Description and Performance Evaluation," Aug. 31, 1960. Cf. letter, J. Y. Brown, Contract Manager, McDonnell Aircraft Corp., to W. P. Kelly, Jr., NASA Contracting Officer, Aug. 24, 1960.

Previous Next Index