Managing the Design Phase

In the year following the dry-workshop decision, Skylab moved beyond the bounds of Apollo Applications. Although much of the hardware and many of the managerial practices retained the Apollo stamp, the program took on a new dimension. The name Skylab, adopted in February 1970, signified the change of outlook: officials no longer viewed the program simply as a means to use leftover Apollo hardware. Increasingly, it was seen as America’s first space station-and perhaps the only one for many years. Several factors contributed to this change. Apollo 11’s success allowed NASA officials to give more attention to Skylab, while the Saturn V’s greater lift permitted program engineers to expand their plans and make the workshop a better laboratory and home. The program also took on increased importance as it slowly became apparent that Congress would not fund a space station during the 1970s.

MOVING OUT OF APOLLO’S SHADOW

George Mueller’s integrated plan of May 1969 listed Apollo and Skylab as NASA’s first manned programs of the 1970s. The agency hoped to move out in two general directions: on one avenue Apollo led to further lunar exploration and the possibility of a lunar base; a second route to Earth-orbital operations began with two Saturn-V workshops and proceeded to a permanent, manned space station with a low-cost Shuttle. Major milestones for the decade included:

1972-Earth-orbital operations with Saturn-V-launched workshop

1973-Start of post-Apollo lunar exploration

1974-Suborbital flight tests of Shuttle

-Launch of second Saturn-V workshop

1975-Initial space station operations

-Orbital Shuttle flights

1976-Lunar-orbit station

-Full Shuttle operations

Sometime in the 1980s or 1990s, NASA would establish bases in Earth orbit and on the lunar surface and would land men on Mars.¹

Paine was anxious to win approval for this ambitious plan in 1969 while public enthusiasm was high. The Space Task Group, a body established by President Nixon to consider America’s future space program, provided the administrator an excellent sounding board. In meetings that summer, Paine promoted a manned Mars mission as NASA’s next major objective. The task group’s September report, America’s Next Decades in Space, recommended a balanced manned and unmanned space capability and listed three possible programs leading to a manned landing on Mars before the 21st century. The most ambitious option called for a 50-man, Earth-orbiting station in 1980 and the first Mars flight three years later. Funding would reach $8 billion annually by 1976. The least ambitious option cost half that amount and delayed the Mars expedition until the 1990s. The group’s chairman, Vice President Spiro Agnew, endorsed the Martian goal enthusiastically, but elsewhere the proposal fell on barren soil. Opposition appeared in Congress and the press, and the Nixon administration approved less than three-quarters of NASA’s proposed $4.5 billion budget for FY 1971. That was one-half billion dollars less than the previous year’s appropriation and brought NASA to its lowest level of funding in nine years. On 13 January 1970 Paine briefed the press on the impact of the reduction: NASA would suspend production of the Saturn V, cancel Apollo 20, delay the initial workshop flight until late 1972, and postpone Apollo 18 and 19 until 1974.²

The following month NASA renamed its Apollo Applications Program. A widespread dissatisfaction with the acronym AAP^* had prompted Paine to seek a new name shortly after the dry-workshop decision. A committee considered nearly 100 names ranging from Socrates to LSD and recommended 8, 4 from mythology and 4 from American history. Mueller forwarded the recommendations to NASA’s Project Designation Committee with the comment that a name change “could enhance the public’s identification with the program and hopefully provide a more manageable term for everyday use.” The committee passed over the recommendations and selected, instead, a name submitted by Lt. Col. Donald Steelman, an Air Force officer on duty with NASA in 1968. Skylab, a contraction for “laboratory in the sky,” met both of Mueller’s objectives as the name was quickly accepted within and outside NASA.³

During 1970 Paine continued to press for an expansive space program despite the lack of support from Congress or White House. By June he had to concede that at least one more Apollo mission would be eliminated and that there was no possibility of further Saturn production. Paine hoped that Skylab could fly as early as mid-1972. His main concern was to have “a major mission of new significance” by 1976, something more than just another Skylab, but he was clearly out of step with the Nixon administration. NASA’s interim operating budget, made public on 2 September, provided only $3.27 billion. Two more Apollo missions fell by the wayside; the program would end in June 1972. Skylab was supposed to lift off five months later. Paine resigned on 15 September 1970.⁴

The task of defending NASA’s budget fell to George Low, the acting administrator. In October Edward David, the president’s science adviser, asked Low to evaluate the relative priorities of Apollo and Skylab in the light of further possible cutbacks. Low defended both programs, saying that “to reduce or constrain the scientific returns from Apollo by dropping one or more missions would involve very great losses.” But canceling Skylab was even less palatable: “On balance, the weight of evidence seems to favor Skylab over Apollo if a choice must be made.” The scientific returns from the single Skylab mission would probably exceed those from an additional lunar landing. America had already benefited from its Apollo investment, whereas canceling Skylab would provide no return. Finally, Skylab could lead to more new options with less risk than Apollo.⁵

David was asking Low to consider reductions in an already austere budget. In a period of 6% inflation, NASA had sought a modest increase to $3.7 billion. The Office of Management and Budget had countered with a $3.3-billion offer, which forced large reductions in the Space Shuttle and nuclear engine programs. Neither Apollo nor Skylab suffered serious cuts; their combined loss of $50 million amounted to less than 5% of the requested amount. Nevertheless, the loss could be absorbed only by slowing the pace of operations. The Office of Manned Space Flight set new launch dates of December 1972 and March 1973 for Apollo 17 and Skylab respectively. When Kennedy Space Center indicated that such closely spaced launches would require overtime, Skylab was moved back another month. The budget decision in late 1970 marked the last major change in Skylab’s schedule. Thereafter the program moved steadily toward launch.⁶

A SECOND SKYLAB

A second Skylab, under consideration since mid-1969, was a principal casualty of the 1970 budget deliberations. Shortly after the wet-to-dry switch, Charles Mathews suggested that the center program offices begin investigating artificial gravity for a second workshop; the information gained thereby would prove valuable in planning for a permanent space station. In September Mueller’s office broadened the study by asking the offices of space science and advanced research to propose other experiment payloads. Guidelines for a follow-on workshop, prepared in November, listed several options-a year-long occupation of a workshop similar to the first Skylab by four three-man crews, the addition of artificial gravity, substitution of a stellar telescope for the ATM , and a more complex group of earth-resource sensors. The additional logistical support and the new experiments would be accomplished with as little change as possible to the workshop’s basic configuration. Since the first Skylab’s backup hardware would become the second workshop, no major changes could be made on the hardware until near the end of the first missions. The committee set a series of milestones for subsequent studies: a preliminary report on 20 January 1970 to support congressional hearings, a work statement by July, and a preliminary design review in early 1971.⁷

The definition of new experiments continued into the new year. On 7 March, Dale D. Myers, George Mueller’s successor,^* reviewed the progress of preliminary studies with his staff. The group concluded that definition of a stellar telescope had advanced far enough for present needs and that major emphasis in studies should go to artificial gravity and to payloads “providing tangible benefits of general public interest.” After the meeting, Schneider asked his center program offices to provide cost estimates for three possible missions: a repeat of the first Skylab, a yearlong mission with advanced solar instruments but no major changes to the cluster, and the same configuration with advanced earth-resource instruments in place of the telescope mount.⁸

Answers from the centers conflicted. Houston wanted a &m commitment to a more sophisticated station, even if it meant delaying the first Skylab. Huntsville, fearing that a major commitment to a follow-on Skylab would jeopardize the present program, argued that a year-long mission was impossible without major hardware changes and that artificial gravity would double or triple costs. The most that NASA could afford, in Huntsville’s opinion,wasacombinedearthresources-solar astronomy mission of eight months’ duration. Both centers’ views were aired at the April meeting of the Manned Space Flight Management Council, along with Schneider’s proposals for further work. The council approved additional studies of Skylab II configurations and directed the committee on artificial gravity to present its findings by early May.⁹

Skylab II studies proceeded that summer in preparation for the FY 1972 budget discussions. Payload weight soon became a serious problem, whose solution might require modifying the second stage of the Saturn rocket. The cost outlook was more disturbing-estimates ranged from $1.32 billion to more than $1.5 billion. Schneider had discussed a second Skylab with officials from the Office of Management and Budget on 31 July and knew money would not come easily. After another review on 31 August, he informed Myers that Skylab II studies had provided sufficient data for planning purposes. Further steps awaited a funding decision.¹⁰

The decision that fall went against Skylab 11. There was some question about its utility; unless the agency made expensive modifications for artificial gravity, the mission would essentially duplicate Skylab I. NASA management found that funding another workshop dictated either a much larger budget or lengthy delays in the Space Shuttle. Although there was strong support for a second Skylab in the House space committee, the Nixon administration was unwilling to underwrite the costs, and NASA did not wish to jeopardize its future programs.¹¹

MANAGMENT TOOLS

During the summer of 1969, the program manager had his hands full managing the first Skylab. From Schneider’s point of view, research scientists moved in a world different from that of engineers. He found it difficult to convince them “that you really need the hardware six months before flight.” In defense of the scientists, they were probably influenced by their Apollo Applications experience, when schedules had slipped from month to month, allowing almost indefinite time to improve their instruments. Those improvements contributed to the rising costs of developing the experiments, a frequent subject in Schneider’s correspondence. Changes to the experimental instruments also made it impossible to freeze interfaces between experiments and spacecraft,” with further damage to budgets and schedules.¹²

Indeed, interface control was one of Skylab’s biggest problems. Aerospace engineers used interface to describe the common boundary between parts of a space vehicle, such as an electrical or pneumatic connection or a physical fit. Thousands of interfaces on Skylab required close supervision to ensure compatible connections. The Skylab program offices managed these interfaces with procedures developed for Apollo: interface control documents and intercenter interface panels.

Interface control documents provided design requirements and criteria for every interface, describing the parameters and constraints under which the common parts functioned. When the interface concerned two items designed by the same center, a level B document applied. If the interface involved two or more centers, a level A document was in order and an intercenter panel assumed responsibility. Following the program manager’s approval of a document, each center was responsible for implementing its side of the interface. Huntsville had the additional responsibility of examining both sides of flight hardware interfaces for overall compatibility, while Kennedy Space Center performed a similar role where flight hardware joined ground support equipment. Marshall, with support from Martin Marietta, scheduled and tracked interface control documents and kept the master file. In cases where program managers could not agree on panel action, the matter went to Headquarters for resolution.¹³

The elaborate system had bogged down in 1968 and had threatened to delay Apollo; a similar situation troubled Schneider two years later. At a meeting in July 1970, he noted that incomplete interface control documents were delaying the design of “various Skylab modules and many experiments.” Schneider asked Project Integration Director Thomas Hanes to review the status of all documents and recommend ways to eliminate the bottleneck. Little headway was made over the next two months, causing Schneider to direct his program managers to simplify their procedures and get their contractors more directly involved. Hanes’s office would work with the centers in developing a better scheduling and tracking system. Shortly thereafter, the centers joined forces in an Interface Working Group; meeting biweekly, the group cleared most of the backlog by early 1971.¹⁴

Intercenter panels dealt with Skylab interfaces that involved more than one center. Early in Apollo, Gilruth and von Braun had organized panels to exchange ideas and formalize agreements between Huntsville and Houston. When the three centers (Kennedy Space Center joined the arrangement in 1963) approved a solution, the panels would document the agreement. Huntsville found the panels to its liking; in December 1963, von Braun called them “the only effective medium of working out technical problems . . . which cut across center lines.” Houston was less enthusiastic. By September 1966 Samuel Phillips, the Apollo program director in Washington, wanted to eliminate them completely. He probably disliked the panels’ independence from Headquarters and may have feared that the groups were not properly documenting all of Apollo’s interfaces. Nevertheless, in March 1967 Charles Mathews established a panel system for Skylab. His initial order covered four areas where the centers worked together frequently: mechanical, electrical, instrumentation and communications, and mission evaluation. Interfaces on launch operations equipment were to be handled by the Apollo panel for the time being. Two weeks later Mathews added three more panels: mission requirements, systems integration, and systems safety.¹⁵

By August 1969 there was no question at Headquarters about the need for intercenter panels; with the number of interfaces on Skylab there had to be some formal means of tying the centers’ work together. But realignment of some center responsibilities in late 1968 had raised a number of questions about panel relationships and Schneider hoped to resolve them. Huntsville wanted to discontinue the practice of cochairmen in certain key areas and let the responsible center direct panel activities. Houston had suggested doing away with the System Integration Panel since it duplicated the baseline configuration meetings held by Headquarters. There was also support to upgrade guidance and control activities-currently a subpanel of mission requirements-to an independent panel. At a meeting of 5 August, officials decided against wholesale changes in the panel system; instead, the Systems Integration Panel was deleted and a panel for planning tests was added.¹⁶

Interfaces were part of the larger problem of configuration control. Configuration referred to the various characteristics of hardware: size, weight, shape, connecting points, and power requirements. During the design phase, engineers made frequent configuration changes, many of which affected other parts. The Apollo 13 accident provided a classic example of a breakdown in configuration control. In 1965, engineers had increased the power used to pressurize an oxygen tank without changing the protective thermostatic switches on the tank’s heater. During normal operations the error caused no problem; but an unusual operation, aimed at correcting a different problem some days before launch in 1970, applied the higher voltage long enough to weld the switches shut and damage some insulation. In space, the tanks exploded with near-fatal consequences.¹⁷

Apollo and Skylab officials attempted to avoid such errors through a series of configuration control boards. These groups evaluated changes to an approved design at one of four levels, depending on the impact of the modification. Level 4 modifications affected neither weight nor performance, such as changing the screws on an instrument from brass to nickel alloy. Level 3 boards dealt with modifications that might affect the schedule or cost of a particular experiment or module but would not affect other hardware; at these levels the centers improved many experiments without Headquarters approval. A level 2 change affected other major hardware and required the approval of the center program manager or his representative. A good example of such a change resulted from a Huntsville inspection by von Braun. Shown plans for a vacuum pump on the lower-body negative pressure device, von Braun took strong exception: “Right through that wall you’ve got the greatest vacuum in the universe.” Engineers initiated a level 2 change to drill a hole through the workshop wall. When such changes were approved by a level 2 board, the decision was transmitted to the Headquarters office for review. Level 1 actions, requiring Schneider’s approval, involved changes to hardware, software, or facilities that might result in inability to meet the operations plan and mission objectives; changes that affected milestones; and changes in excess of $500,000 or that would double the agreed-on cost of an experiment.¹⁸

Interface control and configuration documents were an important part of the documents system that Skylab inherited from Apollo. Paper work had characterized major projects of the post-World War II era, and Apollo was no exception; indeed, observers facetiously suggested that NASA was trying to reach the moon on stacks of paper. The Skylab Program Office used three types of document to direct the activities of the center program offices. “Skylab Program Specifications” established major functional and performance standards for program hardware. For example, the August 1969 edition set the probability of crew safety at a level comparable to Apollo, with spacecraft parts and systems designed to work 995 times out of 1000 and the reliability of the workshop and launch vehicle put at 0.995 and 0.990 respectively.^* “Skylab Program Work Authorizations” identified center responsibilities for more than 50 major end items, among them the one-g spacecraft trainer (Houston) and a workshop engineering mockup (Huntsville). A second list in the authorization document identified over 130 mission milestones, deadlines for specific actions. “Mission Directives” provided detailed statements on objectives, flight plans, space vehicle configurations, experiments, and center responsibilities.¹⁹

When the paper threatened to drown the program, Schneider asked his managers to review all requirements in the light of three questions. What is the minimum information needed to meet general program responsibilities? What information do you need to meet specific technical responsibilities? What information do you believe other offices will expect you to have available? The Headquarters office undertook a similar review of the documentation requirements it levied against the centers. In spite of NASA’s intentions, many participants-particularly scientists were appalled by the amount of red tape. An investigator working on the human-vestibular experiment at the Navy Aerospace Medical Institute, on first seeing the “Experiment General Specifications,” was taken aback. He told Houston officials that the cost of his experiment would increase tenfold and suggested that NASA “build a direct line between Pensacola and Houston, to carry the carloads of paper. . . .”²⁰

Of the various management tools used in Skylab, probably the most important-certainly the most prominent-was NASA’s formal system of reviews. During Apollo, NASA had developed this system to serve as key management checkpoints during program development. The first three, occurring during the design phases, were:

1. Preliminary requirements review-a review of concepts considered and of the concept chosen to meet mission objectives;
2. Preliminary design review-an examination of the basic design conducted early in the detailed design phase;
3. Critical design review-a technical review of specifications and drawings near the conclusion of the detailed design phase.

In Skylab’s preliminary and critical design reviews, the module or experiment under review was also examined for its compatibility with other portions of the space station. The next reviews came near the end of hardware development:

4. Configuration inspection-a comparison of manufactured end items (including test equipment as well as flight hardware) with specifications, drawings, and acceptance testing;
5. Certification of flight worthiness-a determination prior to shipment from the factory that flight hardware was complete, qualified, and accompanied by supporting documentation.

Whereas the first five reviews were conducted for each stage, module, and experiment, the last two covered the entire Skylab operation:

6. Design certification review-held four months before launch to certify the spacecraft design for flight worthiness and safety and to assess the design of the launch complex, mission control center, and Manned Space Flight Network;
7. Flight readiness review-held several weeks before launch to validate the operational readiness of the total mission complex.

With these seven milestones, NASA tracked the progress of Skylab hardware from drawing board to launch site.²¹

Since Huntsville was responsible for most Skylab hardware, Lee Belew directed a majority of the reviews. He appointed review board chairmen, scheduled review dates and sites, and ensured that experiment sponsors, contractors, and other NASA offices were represented. Design review teams performed the detailed examination of blueprints, spending much of their time with “review item discrepancies,” the principal means to recommend hardware changes. If a qualified individual did not like the location of an experiment or the living arrangements of the workshop, he could submit a discrepancy report. Teams then screened the reports, combining similar ones, approving or disapproving many, and submitting others to the review board for decision. The process was fully documented with center managers maintaining the status of every document as to number, title, category, date for completion, and the individuals responsible for assigned actions. One or two individuals earned a certain notoriety with the center offices by recommending large numbers of changes.²²

The changeover to the dry workshop touched off extensive reevaluations at McDonnell Douglas plants and in Huntsville. By December 1969 the process had advanced sufficiently to warrant a preliminary design review of the cluster systems. Several hundred NASA and contractor representatives divided into groups to examine requirements for and possible changes to the various systems. Three days of discussion disclosed a number of significant items. Whereas Huntsville and McDonnell Douglas had assumed the astronauts would enter the cluster in vented pressure suits, Houston was planning a “shirtsleeve” entry. The Manned Spacecraft Center also objected to the layout of the telescope mount’s control and display console, since astronauts could not monitor it and the panel for the structural transition section simultaneously.^** Another problem stemmed from the decision to incline Skylab’s orbit 50' from the equator so as to accommodate earth-resource experiments. The change posed problems for engineers working on the thermal control system. To maintain compartment temperatures within the comfort zone when in sunshine, the workshop would have to give off more heat than had been planned. Modifications for this purpose, however, increased the heating requirements during nighttime periods beyond the available power. A decision was postponed pending more detailed studies.²³

A number of other questions were discussed, but in retrospect the most important decision concerned the electrical power system. From the wet-workshop days, two separate electrical systems had evolved; one of them served the lunar module and the telescope mount. With the elimination of the lunar module, two independent systems no longer made sense, but the “minimum change” dictum in July discouraged any immediate alterations. At the December review, a proposal to combine the separate systems was approved in turn by level 3 and 2 configuration control boards. After weighing increased cost and complexity against the greater probability of mission success, Schneider approved the change. It would develop that, after the accident during launch, this decision would save the mission.²⁴

The cluster systems review generated a number of actions over the next few months, among them a detailed study of the power and thermal systems, reorientation and relocation of the ATM’s display panel, modifications of the multiple docking adapter including the retention of a side port for emergency docking, and a thorough study of the ATM’s computer software. The work proceeded under a tight schedule which received attention when Schneider and Belew met with the airlock team in St. Louis on 10-11 December. Schneider was particularly worried about the short time between critical design reviews and the delivery of flight hardware. If major problems arose at the reviews, contractors would probably not meet their delivery dates. Accordingly, Schneider wanted all personnel involved in a design to review and critique their areas of responsibility regularly. Managers were to stress content “rather than extensive formal preparation of presentation material.”²⁵

In January, Schneider pressed Belew to hold a series of reviews the following month, much like the December meeting in Huntsville. The program director was concerned that “failure mode and effects analyses”^*** were lagging and would delay the rest of the design work. He considered reviews in this area mandatory, while follow-up reviews on the electrical power, environmental control, and attitude-control systems were highly desirable. Belew did not share Schneider’s concern about work on failure modes. Although formal documentation was usually not available, Huntsville’s designers and analysts were working closely together, and Belew had taken steps to have the failure mode documents available 90 days before the critical design reviews. As for the other reviews, Belew wanted to avoid “large, relatively inefficient reviews which would in fact impede much activity which is already planned.”²⁶

Belew preferred to use monthly crew-station reviews, agreed to by the center managers in December. In these meetings, astronauts walked through mockups of flight hardware to ensure that the design met operational requirements. Attendance was held to a minimum; NASA and contractor representatives had sufficient rank to make immediate decisions on matters not involving large cost or schedule delays. The next meeting of a configuration control board then confirmed their decisions. When members of the review team disagreed, they could appeal to the board. However, review teams were encouraged to resolve matters among themselves. The reviews used engineering mockups at each contractor plant, and each mockup included appropriate interfaces. (Thus the airlock mockup in St. Louis had a workshop hatch and adjacent portions of the docking adapter.) Belew thought crew-station reviews provided a “more continuous effort of responsible parties, concentrated nearer the working level.” Judging by Belew’s weekly reports in early 1970, Skylab was one review after another. At contractor plants in St. Louis, Denver, and Los Angeles, teams of 40 to 50 engineers and astronauts participated in crew-station reviews on the major modules. In between these meetings, smaller groups coordinated daily changes.²⁷

Reviews of 70 Skylab experiments were an additional burden for the program offices. Managers were required to certify each review as to completeness and adequacy of documentation within 60 days of completion. Despite attempts to tailor the reviews to the importance of the experiment, based on crew safety and mission success, the centers fell behind schedule. In July Schneider took the managers to task for 27 uncertified reviews.²⁸

Design work climaxed in mid-1970 with critical design reviews of Skylab’s principal hardware. Each lasted nearly a week and involved upwards of 300 NASA and contractor engineers. Review boards considered an average of 200 discrepancies on each module and although most of the proposals were minor, collectively the changes could delay Skylab’s launch by several months.²⁹

THE PROBLEM OF CHANGES

Changes posed the biggest problem for Skylab managers during the first two years of program development. At the time of the dry-workshop decision, Headquarters had decreed “minimum change.” The restriction was short lived, however; by October 1969 a dozen major changes were under consideration, among them a 120-day mission for the final crew, an earth-resource package of experiments, an orbit inclined 50° from the equator, operation of the solar telescopes in an unmanned mode, and the addition of a teleprinter. That month Schneider approved a series of physical modifications to the workshop including the addition of a side access door and a window, the reversal of the “floor” equipment to the new, hard “ceiling,” and a new wardroom combining the sleep compartment with the food management area. At Huntsville, the center responsible for keeping all of that hardware on schedule, Belew protested the extent of the changes, stating that they constituted a new workshop mission. He estimated the changes would delay the schedule by six months and add $100 million to the costs.³⁰

Indeed, Schneider had not given the centers much slack. Two weeks after the dry-workshop decision, he announced a working schedule with a flight-readiness target of March 1972. By setting his deadline four months ahead of the official launch date, Schneider sought to ensure against unforeseen problems. Huntsville’s reaction in August was positive; the working schedule appeared feasible with the possible exception of the solar telescope mount. Houston officials were less sanguine. Kenneth S. Kleinknecht,^* who would soon replace Thompson as Skylab manager, noted that “AAP schedules are fluid and are being established before full definition of either the workshop or the CSM.” He saw no slack left in the schedule for problems or changes and concluded, “with such an approach, schedules cannot be met.” In December-before the changes had been fully assessed-Belew reported that his contractors were under an “extremely tight schedule.” The centers gained breathing room in January 1970 when FY 1971 budget cuts forced a four-month slip in the working schedule; but by May, Houston was pushing for further delay and some items were three to four months behind schedule.³¹

As design work proceeded, NASA officials debated the merits of further changes. On 27 March 1970-shortly after a major decision to modify the urine processing-Dale Myers announced that Skylab could accept no more experiments, since “hardware development activities have reached the stage and maturity where any significant additions or modifications will cause a schedule slip.” In May, however, Houston sought further changes in habitability aspects of the workshop. This brought loud protests from Huntsville and led to a major program review 7-8 July. The review team approved many of the proposed changes, while reaffirming the launch date of July 1972. The director of Marshall wrote Headquarters that the new changes would eliminate all slack from Skylab’s schedule. If modifications continued, Huntsville would be unable to maintain either schedule or budget. The following month, he urged Gilruth to assist him in reducing program changes since the limitations of the Skylab systems “are now being reached, or in some cases nearly exceeded.”³²

Correspondence between Belew and Schneider that summer pointed up the problem of funding, which the changes exacerbated. On 17 July Belew indicated that Huntsville would need more money if the center was to maintain the schedule. Schneider replied that there were no unallocated Skylab resources, nor was it prudent to expect more. He asked Belew to devise a way of meeting his program objectives within present resources. The plan was to include specific manpower restrictions for major contractors and in-house personnel. From the subsequent review, Belew concluded that the Skylab schedule and resources were, indeed, incompatible; Marshall needed $285 million in FY 1971 funds, nearly $50 million more than the intended allocation. Meanwhile, Schneider had found an additional $25 million for Huntsville, halving Belew’s deficit. The Huntsville manager proposed to spread the shortfall among all his major projects, bringing each down about 10% below desired funding. This looked all right until early October, when McDonnell Douglas reported that its allocation would delay workshop delivery by two months, removing all the schedule margin from the official launch date. On 7 October Belew reported that unless NASA controlled changes more stringently, it would not make a 1972 launch “at any price.” During a teleconference on the 13th, Schneider added $12 million to Huntsville’s funds so that Belew could speed up his contractor’s work. (The sum eventually came from Houston’s allocation.)³³

Scheduling pressures eased in September 1970 when Schneider dropped the idea of a working launch date, set four months ahead of the official schedule. At Houston, Kleinknecht was particularly pleased by the end of the two-schedule policy:

When people know that they’re working to a schedule that nobody expects to make, you can’t keep them motivated and people start playing games with the schedule, too. . . .The only way to run a program is to have a do-able schedule; it can be ambitious, [but it must be] one that everybody can focus on and feel that if he does his part of the job we will remain on schedule.

Schneider attempted to retain some cushion by scheduling hardware into the Cape three months before the required date.³⁴

The critical design reviews recommended many small modifications, but few large changes were proposed after the fall of 1970. As Schneider noted on 15 December: “The flexibility to incorporate changes without impacting the launch date and critical program resources has passed and each proposed change has to be considered on the basis of Skylab systems impact and how each change can impact other aspects of the total Skylab program.” Although Huntsville had opposed many of the proposed changes in 1969 and 1970-largely because of the impact on schedules and cost-after the mission the consensus was that the changes had enhanced the program well beyond their cost.³⁵

THE PROBLEM OF REENTRY

One change that had been debated and ruled out was providing for controlling the reentry of the orbital cluster when it finally came back to earth. At nearly 75,000 kilograms, Skylab would be the heaviest object ever placed into orbit, and its high orbital inclination would take it over most of the earth’s surface. The eventual reentry of the workshop-or large pieces of it exposed a problem of a magnitude that NASA had not previously had to face. For years the hazard of falling space junk-spent booster stages, spacecraft, or satellites-had existed, and treaties spelled out the responsibility of spacefaring nations for injury or damage caused by their vehicles. Starting in late 1962 the manned spaceflight centers and their contractors had studied the survival of earth-orbiting vehicles and means of predicting their impact points or controlling their reentry. Prediction was difficult, and providing for controlled reentry imposed severe weight penalties. All the studies, however, indicated such a small probability of human injury that NASA management accepted the risk, in spite of White House and State Department fears of possible diplomatic repercussions. Some measures were taken for payloads that seemed to create abnormal hazards. The unmanned spacecraft used on the test flight of Gemini-Titan 1, and the 17,590-kilogram payload of SA-5 were both modified structurally so that they would break up into small pieces on striking the atmosphere.³⁶

No such solution was possible for Skylab, however, and early in 1970 Administrator Thomas Paine called for a review of the reentry hazard and an assessment of possible engineering changes to minimize it. The resulting study considered the S-II booster stage, the four segments of the payload shroud, and the orbital workshop, concluding that there was 1 chance in 55 that a fragment of Skylab would strike someone.³⁷

As for countermeasures, the only sure solution was to add retrorockets and control systems so that ground controllers could bring the fragments down in a preselected location-preferably a wide stretch of ocean. For the S-II stage, the study group calculated, such systems would weigh about 9000 kilograms and would cost perhaps $10 million; for the workshop the weight penalties were similar and the costs even higher. The added weight of these systems would severely tax the attitude-control and electrical power systems, requiring extensive redesign and adding months to the schedule.³⁸

The study group concluded that NASA should accept the rather small risk, which was somewhat less than that expected from all other sources-meteorites and space junk already in orbit^*-during Skylab’s expected lifetime. The cost of reducing the risk by 50% was extremely high. The group recommended, however, that criteria for acceptable risk should be established early in future programs, so that planning and development could incorporate them.³⁹

These conclusions were corroborated in all important respects later in the year by a study performed for Marshall by Lockheed Missiles and Space Company. Lockheed’s experts concluded that 306 pieces of the Skylab cluster, totaling 22,600 kilograms, would survive reentry. The largest piece would be the film vault, as big as a large executive desk and weighing as much as a compact car. Lockheed’s study did not assign a significantly higher risk figure than previous studies, however.⁴⁰

In late November 1970 Dale Myers forwarded formal recommendations to Acting Administrator George Low. These largely agreed with the conclusions reached 11 months earlier-namely, that the risk was small enough to be accepted in view of the weight and cost penalties imposed by redesign. Low accepted Myers’s recommendations and ordered the Office of Manned Space Flight to work with the Office of Public Affairs and the Office of International Affairs to develop a plan for the public affairs aspects of the Skylab reentry problem.⁴¹

The first phase of program development ended in late 1970 with the completion of design work. In 16 months Skylab program offices had defined relations with Apollo, organized management tools, steered the cluster through its design phase, decided what to do about the reentry problem, and begun preparation for tests. Skylab’s appearance and objectives had undergone considerable modification, but the period of major change was over. Ahead lay hardware fabrication and tests.