Moving to the Launch Site

Although the rank and file of the Space Task Group were barely aware of the new liaison between NASA Headquarters and McDonnell reliability experts, the quest for quality control at the working level was entering a new phase. In the early summer of 1960, about 50 men from STG established residence in Florida. John F. Yardley, along with about 80 McDonnell technicians, set up shop in mobile-home trailers around Hangar S, in the industrial area within the fences [269] of Cape Canaveral. By the end of the year the number of technicians working on the capsules for preflight checkout at the Cape had grown to more than 400, most of the increase made up by contract personnel.¹⁶

At the McDonnell factory in St. Louis, peak employment on Mercury systems had reached 880 in April 1960. After that, there was a gradual decline in Mercury production workers as Yardley's field team increased to 120 by summer's end. Because STG had called for the first four capsules from McDonnell's production line before they were entirely finished, the maximum of 427 workers on the factory floor in May 1960 declined with the buildup of preflight polishing activities at the Cape. Yardley and his crews soon became the center of attention for unofficial helpers and kibitzers from other organizations and contractors, many of whom were glad to provide materials and tools that were urgently needed and in short supply among McDonnell people at the Cape.¹⁷

Yardley, his assistants at the Cape - E. F. Peters and Robert L. Foster - and other working engineers knew little about the separate reliability contract between NASA Headquarters and McDonnell. Walter F. Burke, Logan T. MacMillan, and the quality manager, N. E. Covinsky, did know that this extra business was coming to their company through separate channels, but they and their production engineers were so busy trying to make each capsule work properly that they too could see little sense in the "numbers game." Each system and subsystem seemed to have its own personality. But to guard against overemphasizing these individual idiosyncrasies, capsule No. 10 was set aside as the standard test article at McDonnell. As preflight checkouts at the Cape uncovered more and more unique difficulties, the need for still more stringent quality control was made plain.

No one recognized this more than Yardley, who in the summer of 1960 urged his company to institute a new order of reliability tests. He did not insist on statistical performance data, but he did enjoin improvement of environmental-chamber reliability testing of components. Robert L. Seat, McDonnell's senior test engineer, was pressed by Silverstein in Washington, by Lewis R. Fisher of STG, and by Yardley from the Cape, as well as by the burgeoning number of test requests between McDonnell departments, to prepare specifications for an exhaustive environmental reliability testing program. On September 26, 1960, the project to flight-test a man in orbit was supplemented by an authorization to ground-test the capsule in a simulated mission through physical environments in a "space chamber." This simulated orbital test program gradually became known as "Project Orbit."¹⁸

The reaction control system on capsule No. 2 was giving Yardley headaches. In general the power and sequential systems on all capsules were full of "glitches," or minute transient voltages from inexplicable origins. Surely more problems could be expected from space operations. So the simulated mission test program, designed specifically to detect unknown anomalies arising from four and a half hours of continuous operation in a vacuum alternately hot and cold, like "day" and "night" for the manned satellite, was welcomed by all hands. [270] Unfortunately it would take six months to build, install, test, and modify the new space chamber test facility at the McDonnell plant. Several smaller, less sophisticated "man-rating" vacuum chambers had already been used but none was capable of simulating the extremes of orbital conditions.

Prelaunch preparations at the launch site began in June 1960 with an understanding between STG and McDonnell that some rework would be performed there in addition to extravagant preflight checkout tests, but the extent of the last-minute work to be performed and the number of discrepancies to be corrected became so great that "preflight checkout" quickly came to be a misnomer. Under Preston at the Cape, John J. Williams eventually came to head the "Preflight Operations" division, instead of being simply "checkout" crew chief. Paul C. Donnelly, Archibald E. Morse, Jr., A. Martin Eiband, Walter J. Kapryan, and Jacob C. Moser gradually became involved with wholesale systems engineering as the thoroughgoing checkouts in Hangar S expanded.

Gilruth laid down the law "for what is perhaps the most important single requirement in our programs: that designs, procedures, and schedules must have the flexibility to absorb a steady stream of change generated by a continually increasing understanding of space problems." This policy of correcting every discovered deficiency and of modifying each spacecraft down to the finish line at launch time was what Gilruth meant by an "R and D" program; it sacrificed cost and schedules if necessary in the interest of quality or reliability as the experimentalists understood it.¹⁹

Through August 1960 "space chamber" ground testing for Mercury had consisted primarily of the capsule systems tests for integration and compatibility in a relatively mild vacuum and of the manned environmental control system tests simulating an altitude of 40,000 feet. McDonnell had detected many design deficiencies in these test programs. Now early development failures, arising from unanticipated interactions between parts and components and from errors in estimating the effects of environmental extremes, became most troublesome.

At St. Louis in mid-August, the "Development Engineering Inspection," a milestone meeting comparable to the Mockup Review, brought together for three days all the chief actors and participants in the hardware work on the capsule. Walter C. Williams and Kenneth S. Kleinknecht were eager to institute this old Air Force custom - the "D.E.I.," as they called it - as a basic check on systems integration and configuration control. When on August 18 the 30 members of the NASA inspection team departed, they were well assured that the Mercury capsule on display (No. 7) was safe for manned flight, but only for a suborbital mission. Orbital flight would require a higher order of precautions for reliability. "Project Orbit," taking advantage of recent advances in vacuum technology, promised to pioneer this new dimension in development engineering by bringing the space climate down to Earth. Capsule No. 10 was specifically set aside in September for environmental chamber testing at McDonnell for orbital conditions.²⁰

[271] While the Tenney Engineering Company of Union, New Jersey, was building the new vacuum chamber for man-rated environmental testing of the capsule at the Cape, and while McDonnell engineers were moving in to augment STG's preflight checkout group there, one NASA operations expert transferred back to tidewater Virginia to help Gilruth and French formulate policy and establish STG's competence to judge reliability and flight safety issues. F. John Bailey, Jr., was Gilruth's choice for the man most likely to reconcile the differences between reliability based on experience and on expertise. Bailey believed an engineer needed 15 or 20 years' experience in any specialty to be a proper judge of the state of his art; he also appreciated the value of mathematical models in the redesign stages of technological evolution. But he quickly became convinced, particularly by studying the carefully balanced engineering compromises between efforts to make the boosters perfect and to perfect the escape system, that Mercury dependability could hardly be improved except by flight testing.²¹

Everyone recognized dangers in the pragmatic experimental approach to pilotless spacecraft research, but each calculated the risks differently. Silverstein and the new Associate Administrator, Robert C. Seamans, Jr., who succeeded Horner at this post on September 1, 1960, were among those at Headquarters who justly feared that overemphasis on the uniqueness of each production capsule and on STG's policy of continuous rework might lead to so many "quick fixes" that a pyramid of unobtrusive changes could cover up the truth about whatever might go wrong.²²

Perhaps the most pertinent of these difficulties with systems integration derived from NASA bench tests of the reaction control system. The manufacturer of the RCS, Bell Aerosystems Company, ran its qualification test program from August through October 1960 and reported all phases of the testing satisfactorily completed. Subsequent tests by McDonnell, STG, other NASA engineers, the preflight teams at the Cape, and eventually by the workers on Project Orbit revealed innumerable electrochemical and electromechanical problems in simulated environments that required small changes here and there and eventually everywhere. The thrust chambers, metering orifices, solenoid valves, expulsion bladder, and relief valves each presented developmental flaws that were "solved" more often by improvisations than by scientific redesign. Karl F. Greil, a thermodynamicist who was working for Grand Central Rocket Company in 1960 to perfect the escape pyrotechnology for Mercury, joined STG and its reaction controls test team in 1961 and tried in vain to apply the same perfectionistic standards to this vastly more complicated and inherently less reliable system of moving parts:

This is the irony: the results that counted in Mercury's RCS were due to changes of the screen, heat barrier, and orifices, all of which were made upon simple first thought. On the other hand, the large amount of experimentation on the valve resulted merely in the assurance that nothing needed to be changed so far as valve design was concerned. This irony, that the simple approach [272] did the entire job while the sophisticated approach merely resulted in an "Amen", is indeed worthy of reflection, because it has in store both a risk and a lesson: a lesson because there is so much glamor cast on sophisticated pretense and so much disregard for the profane causes of all kinds of trouble; a risk because the simple remedy which did the job once without ever having become clear just how it really worked, such success without perspiration is likely to remain confined to its own historical case. But having established a precedent, it is bound to seduce us into relying on it, if it is not even bound to become a myth and a dogma.²³

If Project Mercury were on the verge of technological bankruptcy, as some critics claimed, the problem was that man was still land-locked by inadequate boosters. The Redstone for Mercury was still not man-rated. The first Mercury-Atlas flight on July 29, 1960, not only did not qualify anything, it seemed actually to have disqualified an indispensable part of Mercury. It cast everything into doubt.

¹⁴ See Low, comments, Oct. 5, 1965; Aleck C. Bond and Maxime A. Faget, Ms., "Technologies of Manned Space Systems," Chap. 14, "The Role of Ground Testing in Manned Spacecraft Programs," 167 - 177.

¹⁵ Mercury operations were governed by the agreement "Overall Plan - Department of Defense Support for Project Mercury Operations," Jan. 15, 1960, but some of the difficulties in working out specific operational procedures at the Cape may be found in Francis E. Jarrett, Jr., and Robert A. Lindemann, "Historical Origins of NASA's Launch Operations Center," Kennedy Space Center, Historical Monograph No. 1, Cocoa Beach, Fla., Oct. 1964, 69-76. See also letters, Henry N. Moore to distribution, AFMTC, "NASA Organizational Changes at AMR and PMR," June 27, 1960; and Kurt H. Debus to G. J. Weber, MAC, July 28, 1960; and Ms. paper, anonymous, "Responsibilities and Procedures for AMR Support of Project Mercury," ca. Aug. 1, 1960.

¹⁶ J. F. Shields, personnel study chart, MSC Florida Operations, Jan. 4, 1964; G. Merritt Preston, interview, Cape Kennedy, April 29, 1964; George F. Killmer, Jr., interview, Houston, Sept. 14, 1965. See also Ms. paper, Gilbert B. North, "Development and Checkout of the NASA Mercury Capsule," McDonnell Aircraft Corp. [ca. Sept. 1960].

¹⁷ John F. Yardley, William Dubusker, interviews, St. Louis, Aug. 31, Sept. 1, 1964. See also Ms. paper, H. H. Leutjen, "Ground Checkout and Launch Procedures for Man-in-Space Operations," McDonnell Aircraft Corp. [ca. Aug. 1961]; McDonnell Aircraft Corp. interoffice memo, J. T. Dale to W. F. Burke, "Mercury-Redstone Panel I Meeting at MSFC on 11 August 1960," Aug. 16, 1960. Field procurement was standardized to some extent by memo, Harold G. Collins to all Mercury Hangar S personnel, "Procurement Procedures on Contract NAS 5-59," Sept. 8, 1960.

¹⁸ Project Orbit is not to be confused with Project Orbiter; see p. 29. Yardley, Robert L. Seat, interviews, St. Louis, Sept. 1, 1964; memo, Lewis R. Fisher for Project Director, "Proposal for Environmental Qualification Test of Mercury Capsule," June 21, 1960; A. E. Wilkes, "Proposal for Full Scale Simulated Mission Test, Orbit Phase; Immediate Capabilities," McDonnell Aircraft Corp. report No. 7730, Aug. 29, 1960; memo, Floyd W. Fults to distribution, "Project Orbit Team Member's Responsibilities," McDonnell Aircraft Corp. memo No. P0-650-3, Feb. 2, 1961. For an overview of Project Orbit, see A. M. Paolini, "Evaluation of a Mercury Spacecraft in a Simulated Orbit Environment," McDonnell Aircraft Corp. preliminary report, May 29, 1962.

¹⁹ Gilruth quoted by John J. Williams and Donald M. Corcoran, "Mercury Spacecraft Pre-Launch Preparations at the Launch Site," paper, American Institute of Aeronautics and Astronautics, Space Flight Testing Conference, Cocoa Beach, Fla., March 18-20, 1963, 18. Cf. p. 28. See also draft Ms., Frank M. Crichton, "Quality Control and Inspection," for Project Mercury Technical History, July 3, 1963.

²⁰ Bond and Faget, "Technologies of Manned Space Systems"; Development Engineering Inspection Data Book, SEDR 183, McDonnell Aircraft Corp., Aug. 16, 1960. For Project Orbit, see F. W. Fultz, A. E. Wilkis, J. J. Mazzoni, et al., informal McDonnell Aircraft Corp. memos, Sept. through Dec. 1960, including preliminary McDonnell report 7869—9 [no title], June 7, 1961: all included in file by Robert A. Hermann and Joe W. Dodson, "Project Orbit notes."

²¹ Bailey interview; Mss., "Briefings, NASA-Industry Apollo Technical Conference," July 18, 1961; "Reliability and Flight Safety Problems," April 4, 1962, 8; "Reliability and Crew Safety in Manned Space Flight," Feb. 20, 1963; Bailey, "Review of Lessons Learned in the Mercury Program Relative to Spacecraft Design and Operations," paper, American Institute of Aeronautics and Astronautics Space Flight Testing Conference, Cocoa Beach, Fla., March 18, 1963.

²² Abe Silverstein, interview, Cleveland, May 1, 1964; House Committee on Science and Astronautics, 87 Cong., 1 sess. (1961), Fourth Semiannual Report of the National Aeronautics and Space Administration, 208-209.

²³ Ms., Karl F. Greil, for Mercury Tech. History, "History of the Reaction Control System," July 1962, 160 - 161. Cf. 145, 67-68 [English trans. by L. S. S.]. See also another draft Ms. by Norman B. Farmer, "Instrumentation," June 27, 1963, for some discussion of equally acute problems.

²⁴ See R. D. Korando, "Mercury Capsule No. 7: Configuration Specification (Mercury-Redstone No. 3) ," Report 603-7, McDonnell Aircraft Corp., Aug. 1, 1960, revised Nov. 10, 1960; R. F. Mackey, "General Flight Plan: Atlas Boosters for Project Mercury," Aerospace Corp. report AS-60-0000-0036, Sept. 1960.