Return to Flight: Richard H. Truly
and the Recovery from the
by John A. Logsdon
Seventy-three seconds after its 11:37 a.m. liftoff on September 29, 1988, those watching the launch of the Space Shuttle Discovery and its five-man crew breathed a collective sigh of relief. Discovery had passed the point in its mission at which, on January 28, 1986, thirty-two months earlier, Challenger had exploded, killing its seven-person crew and bringing the U.S. civilian space program to an abrupt halt.2 After almost three years without a launch of the Space Shuttle,3 the United States had returned to flight.
Presiding over the return-to-flight effort for all but one of those thirty-two months was Rear Admiral Richard H. Truly, United States Navy. Truly was named Associate Administrator for Space Flight of the National Aeronautics and Space Administration (NASA) on February 20, 1986. In that position, he was responsible not only for overseeing the process of returning the Space Shuttle to flight, but also for broader policy issues such as whether the Challenger would be replaced by a new orbiter, what role the shuttle would play in launching future commercial and national security payloads, and what mixture of expendable and shuttle launches NASA would use to launch its own missions. He served as the link between the many entities external to NASA —the White House, Congress, external advisory panels, the aerospace industry, the media, and the general public— with conflicting interests in the shuttle's return to flight. In addition, he had the tasks of restructuring the way NASA managed the Space Shuttle program and restoring the badly shaken morale of the NASA-industry shuttle team.
The citation on the 1988 Collier Trophy presented to Admiral Richard H. Truly read: "for outstanding leadership in the direction of the recovery of the nation's manned space program." This essay recounts the managerial and technological challenges of the retum-to-flight effort, with particular attention to Richard Truly's role in it. However, as Truly himself
346 RETURN TO FLIGHT
|It's thumbs up for the shuttle as the STS-26 Discovery crew celebrate their return to Earth with Vice President George Bush. The orbiter completed a successful four-day mission with a perfect touch down on October 3, 1988, on Rogers Dry Lake Runway 17. In this picture, from left to right, are: Mission Specialist David C. Hilmers, Commander Frederick H. (Rick) Hauck, Vice President George Bush, Pilot Richard O. Covey, and Mission Specialists George D. Nelson and John M. Lounge. (NASA photo no. 88-H-497).|
FROM ENGINEERING SCIENCE TO BIG SCIENCE 347
recognized, the recovery program was a comprehensive team effort;4
as the first post-Challenger flight approached, he sent a memorandum
to the "NASA Space Shuttle Team," saying:
As I reflect over the challenges presented to us, and our responses to them, my overriding emotion is one of pride in association. You —the men and women who compose and support this unique organization— should take great pride in having renewed the foundation for a stronger, safer American space program. I am proud to have been a part of this effort; I am proud to have witnessed your extraordinary accomplishments.5
Immediate Post-Accident Events6
When Truly was named NASA Associate Administrator for Space Flight, he told the press that in the three weeks since the Challenger accident he had not "had one moment" to review information about the mishap.7 Whether he realized it or not, Truly was entering a very chaotic situation. At the time of the accident, NASA Administrator James Beggs was on a leave of absence to deal with a Federal indictment unrelated to his NASA duties. (Beggs was later completely exonerated of any wrong doing and even received a letter of apology from the Attorney General for being mistakenly indicted.) Acting as Administrator was NASA Deputy Administrator William Graham, a physicist with close ties to conservative White House staff members but no experience in civilian space matters prior to being proposed for the NASA job. A few weeks earlier, Graham had been named Deputy Administrator, a White House political appointment, over the objections of Beggs and other senior staff at NASA; in his short time on the job he had remained largely isolated from career NASA employees. When Challenger exploded, NASA was thus bereft of experienced and trusted leadership.
Graham was in Washington when the accident occurred. Later in the day, be flew to the Kennedy Space Center with Vice President George Bush and Senators John Glenn and Jake Garn. The latter three flew back to Washington after consoling the families of Challenger crew members and meeting with the Shuttle launch team. Graham stayed behind; in a series of phone calls to the White House during the night, a decision was made to have the President appoint an external review commission to oversee the accident investigation. Although Graham had been briefed by his NASA staff on how the investigation after the 1967 Apollo 1 fire had been handled, he apparently did not argue that the NASA Mishap Investigation Board, set up immediately after the accident, should continue to lead the inquiry.
This naming of an external review panel was in marked contrast to what had happened nineteen years earlier, on January 27, 1967. When he learned that a fire during a launch pad test had killed the three Apollo 1 astronauts, NASA Administrator James Webb immediately notified President Lyndon Johnson, and told him that NASA was best qualified to conduct the accident investigation. Webb later that evening told his associates that
348 RETURN TO FLIGHT
"this is an event that we have to control .... We will conduct the investigation. We will get answers. There will be no holds barred. We'll issue a report that can stand up to scrutiny by anybody." Meeting with the President the next day, Webb told him "They're calling for investigations.... A lot of people think it's a real issue for the future, and that you ought to have a presidential commission to be clear of all influences." But, argued Webb, "NASA is the best organization [to do the investigation]."8 Johnson concurred in Webb's approach; NASA had already selected the initial members of the accident review panel, and they set to work immediately. Certainly there were external reviews of the Apollo fire, particularly by NASA's congressional oversight committees. However, their starting point was the NASA-led investigation.
By not even attempting to retain control of the Challenger accident inquiry at the start, NASA found itself subject to searching external scrutiny and criticism, and the space agency had to share decision-making power during the return-to-flight effort with a variety of external advisory groups overseeing its actions. Dealing with, on one hand, the desire to get the Shuttle back into operation as quickly as possible and, on the other, the recommendations of advisory groups who gave overriding priority to safety concerns and organizational restructuring, was one of Richard Truly's greatest challenges between February 1986 and September 1988. This was particularly the case as the accident investigation quickly changed from one focused on the technical causes of the Challenger mishap to one broadly concerned with NASA's organization and decision-making procedures.
On February 3, President Ronald Reagan announced that the investigation would be carried out by a thirteen-person panel chaired by former Secretary of State William P. Rogers; the group quickly became known as the Rogers Commission. Reagan asked the Commission to "review the circumstances surrounding the accident, determine the probable cause or causes, recommend corrective action, and report back to me within 120 days."9
Within a few days after the accident, NASA investigators had pinpointed a rupture in a field joint10 of the shuttle's right Solid Rocket Motor(SRM) as the proximate cause of the Challenger explosion. As the Rogers Commission began its work, there appeared to be little controversy on this issue. However, in a closed meeting at the Kennedy Space Center on February 14, Commission members were "visibly disturbed" to learn that engineers from the firm that manufactured the SRM, Morton Thiokol Inc., had the night before recommended against launching Challenger in the cold temperatures predicted for the next morning; that their managers, at the apparent urging of NASA officials from the Marshall Space Flight Center, had overruled their recommendation; and that more senior NASA managers responsible for the launch commit decision were unaware of this contentious interaction. This was a "turning point" in the investigation; the Commission immediately went into executive session. It decided that the NASA team working with the Commission should not include any individual who had been involved in the decision to launch Challenger. It decided to broaden the scope of its investigation to include NASA's management practices, Center-Headquarters relationships, and the chain of command for launch decisions-in effect, shifting the focus of the inquiry from a technical failure to NASA itself. At the end of its executive session, the Commission issued a damning statement suggesting that NASA's "decision-making process may have been flawed."11
FROM ENGINEERING SCIENCE TO BIG SCIENCE 349
This indictment of shuttle management provided the backdrop against which Richard Truly would work in succeeding months. As the Rogers Commission tried to fix responsibility for the "flawed" decision to launch Challenger, the agency was rampant with internal conflicts and finger-pointing. The New York Times reported on its front page that the Marshall Space Flight Center, the key organization for diagnosing and fixing the SRM problem, was "seething with resentment, hostility, depression, and exhaustion."12Aviation Week described the U.S. space program as being "in a crisis situation."13 Truly remarked in his first press conference "I have a lot to do"; he was certainly not overstating the situation.
Truly Takes Charge
While he may have been unfamiliar with the details of the Challenger mishap, Richard Truly was no stranger to the space agency; he had been a NASA astronaut from 1969 to 1983, had piloted several of the early unpowered tests of the shuttle, and had flown as pilot on the second shuttle mission in November 1981 and as commander of the eighth shuttle mission in August-September 1983. He left NASA on October 1, 1983, to become the first head of the Naval Space Command; it was from that position that he returned to NASA to assume control of the Office of Space Flight. Truly was an engineering graduate of the Georgia Institute of Technology and an experienced Naval aviator. To most, the combination of his technical background and astronaut experience —and his absence from NASA for the period preceding the accident— made him well qualified to head the return-to-flight effort.
Truly spent his first weeks as Associate Administrator becoming familiar with the situation he had inherited, organizing his immediate office, and establishing a close working relationship with the Rogers Commission. As soon as he entered office, Truly became chair of the "STS 51-L Data and Design Analysis Task Force,"14 which had been set up by Acting Administrator Graham to provide NASA support to the Rogers Commission. One of Truly's crucial early decisions was to bring in J. R. Thompson as vice-chair and day-to-day head of this task force; in effect, this put Thompson in charge of NASAs part in the accident investigation. Like Truly, Thompson had been a long-time NASA employee, but had been in another job in the years preceding the Challenger mishap.15 Other members of the task force were astronaut Robert Crippen; Col. Nathan Lindsay, Commander, Eastern Space and Missile Center; Joseph Kerwin, Director, Space and Life Sciences, Johnson Space Center; Walter Williams, Special Assistant to the NASA Administrator; and the leaders and deputies of the six task force teams on development and production, pre-launch activities, accident analysis, mission planning and operations, search, recovery and reconstruction, and photo and television support that had been set up to parallel the organization of the Rogers Commission investigation. The task force in turn drew on all relevant resources of NASA.
Between intensive task force efforts during March and April 1986 and the equally intense activities of the fifteen-person investigative staff of the Rogers Commission (plus a parallel investigation by the staff of the Committee on Science and Technology of the
350 RETURN TO FLIGHT
House of Representatives),16 it was unlikely that any aspect of the accident would go unexamined. This was especially the case given the constant media scrutiny of the investigation.
By the end of March, Richard Truly was ready to go public with his return-to-flight
strategy. In a March 24 memorandum which he later described as a "turning
point" in the recovery effort,17 he
listed the "actions required prior to the next flight":
Truly reviewed his strategy before an audience of over 1,000 at the
Johnson Space Center; his remarks were televised to other NASA Centers.
He argued that "the business of flying in space is a bold business. We
cannot print enough money to make it totally risk-free. But we are certainly
going to correct any mistakes we may have made in the past, and we are
going to get it going again just as soon as we can under these guidelines."
The New York Times reported that "his upbeat words appeared to be meant
to lift spirits at the beleaguered agency and to turn the staff's eyes
forward to the shuttle's future. . . ."19
In just over a month after taking office, and well in advance of any recommendations from the Rogers Commission and the Congress, Richard Truly had set out the general outlines of the strategy he would follow over the following two and one half years. However, that it would take that long to return the Space Shuttle to flight was likely inconceivable to him and his associates at the end of March 1986. NASA planning at the time called for at worst an 18-month delay to July 1987 in launching the next shuttle. Left to its own devices, it is possible that NASA and its industrial contractors could have met this schedule. NASA was no longer a free agent, however; the Challenger accident and the resulting external scrutiny of NASAs decisions had changed the agency's freedom of
FROM ENGINEERING SCIENCE TO BIG SCIENCE 351
action forever. Over the coming months, Truly would have the almost impossible task of balancing the pressure to fly as soon as possible in order to get crucial national security and scientific payloads into space while convincing the agency's watchdogs that a return to flight was adequately safe. It was not to be an easy assignment.
Trying to Get Flying Soon
As mentioned earlier, it was clear within a few days of the accident that the direct cause of the mishap had been a failure in the joint between two segments of one of the shuttle's two solid rocket motors. That failure was in turn quickly traced to the failure of the "O-rings" designed to prevent the escape, through the joint, of the hot gasses generated during SRM firing. On March 11, Acting NASA Administrator Graham told a congressional committee that a redesign of the SRM joint and seals would be needed, and estimated the cost of the redesign at $350 million.20
Responsibility within NASA for overseeing the SRM lay with the Marshall Space Flight Center in Huntsville, Alabama. On March 25, Truly, acting on his memorandum of the previous day, announced the creation of a Solid Rocket Motor Team "to recommend and oversee the implementation of a plan to requalify the solid rocket motor (SRM) for flight, including the generation of design concepts, analysis of the design, planning of test programs and analysis of results, and any other initiatives necessary to certify flight readiness." The following day, Truly named James Kingsbury, Director of Science and Engineering at Marshall, to head the team on an interim basis .21
Within a few days, Kingsbury told The New York Times that he believed a redesigned solid rocket motor could be ready for flight within twelve months, and would not require ordering substantial new hardware. "We can use everything we have, and just modify it," he told the Times. In particular (though it was not publicly acknowledged at the time), NASA hoped to be able to use 72 steel casings for the SRM that had been ordered six months before the Challenger accident. As would become evident in the course of the accident investigation, NASA had been aware for some time of problems with the original design of the field joint; these casings had been planned to accommodate a new joint design incorporating a "capture fixture" that had been suggested as an improvement on the original joint design as early as 1981. 22
In its eagerness to get started on the return-to-flight process, NASA appeared to be getting ahead of the findings and recommendations of the Rogers Commission, which was not scheduled to report to the President until early June. For example, Truly had said on March 25 that it was probably infeasible to add a crew escape pod to the shuttle orbiter, but "certainly if the Presidential Commission concludes we should do that, we will do it."23 Particularly troubling to the Commission was the speed with which a redesign of the SRM field joint was being proposed. On May 7, the Orlando Sentinel, in an article headlined "Red Flags Fly Over Joint Redesign," reported that "engineers redesigning the shuttle's flawed booster joint will submit a preliminary plan to NASA today, but members of the Challenger
352 RETURN TO FLIGHT
Commission say the agency is moving too fast on the project and could repeat its mistakes." Some Commission members, the article claimed, "are so concerned about Marshall botching the redesign that they want an independent panel of experts to approve the new joint."24
NASA had little choice but to respond to the Commission's concerns, particularly once they had become public; the agency in the wake of the Challenger accident had lost the ability to act counter to those reviewing it from the outside. The commission's concerns were communicated in a private meeting with NASA's top officials, and a response followed quickly. On May 9, Truly announced that James Kingsbury would be replaced as head of the solid rocket motor redesign team by John Thomas, who had been Spacelab Program Office manager at Marshall before being assigned to the 51-L Data and Design Analysis Task Force in March. This was a switch that had been in the works for some time, but it may have been accelerated by Kingsbury's bullish approach to SRM redesign. Truly also announced that "an independent group of senior experts will be formed to oversee the motor redesign" and that this group would be involved in all phases of the redesign effort, "will report directly to the Administrator of NASA, and will thoroughly review and integrate the findings and recommendations" of the Rogers Commission in carrying out its responsibilities.25 The interactions between this external panel, which was appointed by the National Research Council (NRC) in June, and NASA during the redesign and testing of the SRM would be a key determinant of the pace of the return-to-flight process.
On May 12, Richard Truly got a new boss. James Beggs had long since resigned as NASA Administrator. The White House, in March, had nominated James C. Fletcher as his replacement. NASA Administrator from 1971-1977, the period during which the Space Shuttle had been approved and developed, Fletcher was quite familiar with the program. It took two months for Fletcher's nomination to be approved by the Senate. After being sworn in by Vice President Bush, Fletcher told the press that, if necessary changes to make the shuttle safe were not completed by the July 1987 target date for the next launch, "we just won't fly."26
In effect, any chance of a next launch before early 1988 had vanished with NASA's acceptance of the oversight role of an external advisory group, though it took several months before the agency fully recognized that reality. If there had been any prior doubt, it was now clear that the recommendations of the Rogers Commission, due out in early June, would be the defining context for NASA's return-to-flight effort, at least in the public mind. It was clear, moreover, that those recommendations would go well beyond the need for a redesign of the SRM to many other suggestions on how the Space Shuttle should be operated and managed; The New York Times commented that, with such a broad set of recommendations combined with White House and congressional pressure for full compliance with them, "the complexity of NASA's [and thus Richard Truly's] task appears to have been greatly magnified."27
FROM ENGINEERING SCIENCE TO BIG SCIENCE 353
The Rogers Commission Report
The Presidential Commission on the Space Shuttle Challenger Accident
(the official name of the Rogers Commission) submitted its report to President
Ronald Reagan on Friday, June 6; the report was released to the public
the following Monday. The over 200-page document, which contained detailed
assessments of the causes of the accident and of NASA's overall failings
related to the mishap, culminated in nine recommendations. Among them were:
Recommendation I - "The faulty Solid Rocket Motor joint and seal must be changed. This could be a new design eliminating the joint or a redesign of the current joint and seal." Also, "the Administrator of NASA should request the National Research Council to form an independent Solid Rocket Motor design oversight committee to implement the Commission's design recommendations and oversee the design effort. "
Recommendation II - "The Shuttle Program Structure should be reviewed. " Also, "NASA should encourage the transition of qualified astronauts into agency management Positions. "28
Recommendation III - "NASA and the primary shuttle contractors should review all Criticality 1, 1R, 2, and 2R items and hazard analyses. "
Recommendation IV - "NASA should establish an Office of Safety, Reliability and Quality Assurance to be headed by an Associate Administrator, reporting directly to the NASA Administrator."
Recommendation VI - "NASA must take actions to improve landing safety. The tire, brake and nosewheel system must be improved. "
Recommendation VII - "Make all efforts to provide a crew escape system for use during controlled gliding flight. "
Recommendation VIII - "The nation's reliance on the shuttle as its principal space launch capability created a relentless pressure on NASA to increase the flight rate ... NASA must establish a flight rate that is consistent with its resources. "29
In carrying out its mandate, the Rogers Commission had inter-viewed more than 160 people and held more than 35 formal investigative sessions, generating more than 12,000 pages of transcripts. The full-time staff grew to 43, plus some 140 part-time support specialists. In the end, the report toned down any strong criticism of NASA's overall performance and responsiveness; such a harsh approach had been proposed by Commissioner Richard Feynman.30 Rather, the report's recommendations were followed by a conciliatory "concluding thought": "the Commission urges that NASA continue to receive the support of the Administration and the nation.... The findings and recommendations presented in this
354 RETURN TO FLIGHT
report are intended to contribute to the future NASA successes that the nation both expects and requires as the twenty-first century approaches."31
On June 13, President Ronald Reagan directed NASA Administrator Fletcher to implement the Rogers commission recommendations "as soon as possible," and asked for a report within thirty days on a plan for doing so.32 NASA's response came on July 14; Administrator Fletcher told the President that "NASA agrees with the [Rogers Commission] recommendations and is vigorously implementing them." On June 20, in a memorandum to Richard Truly, Fletcher said that he would take direct responsibility for implementing recommendation IV on a new safety organization to replace what the Rogers Commission had characterized as NASA's "silent safety program."33 Fletcher told Truly that "the Office of Space Flight is directed to take the action for all other Commission recommendations." Fletcher asked him to "status me on your progress on a weekly basis."34
While submitting its report to the President, NASA released a schedule for the return-to-flight effort that slipped the earliest possible date for the first launch by 6-8 months, to early 1988. Administrator Fletcher noted that some within and outside of NASA were urging that the three remaining Space Shuttles be returned to flight immediately, with constraints on the conditions under which they could be launched, but that, although he was "uneasy" and disappointed "about the additional delay," in view of the large visibility of the accident ... when we start flying again we want to make sure that it is really safe."35
Implementing the recommendations of the Rogers Commission, and modifying them when justified, would occupy much of the time of Richard Truly and his Space Shuttle team for the next twenty-six months. They worked in the glare of constant congressional and media scrutiny and outside reviews of their actions. There was little margin for error in their task. This was in marked contrast to the situation in the months following the Apollo accident, where, after one round of congressional hearings on the NASA accident report, the space agency made the required technical and management fixes without anyone looking over its shoulder. Indeed, NASA in August 1968 even secretly made a decision to send the second post-accident mission, Apollo 8, around the moon. This decision came before the modified Apollo capsule had been tested on the October 1968 Apollo 7 flight.
Fixing the Solid Rocket Motor
As mentioned earlier, a Solid Rocket Motor Team based at Marshall (but including personnel from other NASA centers, particularly Johnson), and led since May by John Thomas, had gotten an early start on SRM redesign. Sharing leadership with Thomas was Royce Mitchell, another Marshall engineer. Working with the NASA team was a parallel group of engineers from the SRM manufacturer, Morton Thiokol.
This group was headed by Allan J. McDonald, who had been one of those vociferously opposing the launch of Challenger on the night of January 27. McDonald's testimony to
FROM ENGINEERING SCIENCE TO BIG SCIENCE 355
the Rogers Commission about the events of that night had brought him much positive media attention. Following that testimony, however, Morton Thiokol had reassigned McDonald and another senior engineer who had opposed the launch, Roger Boisjoly, to jobs not related to the SRM. Congressional outrage at such a reassignment and NASA pressure had led the firm to restore McDonald to a central role in the SRM effort.36
The Marshall and Morton Thiokol teams played the central role in developing an approach to SRM redesign and testing; from late 1986, the team worked out of temporary quarters near the Morton Thiokol facility in Brigham City, Utah, north of Salt Lake City. The SRM redesign effort received two overall directives from Truly's office: most fundamentally, "to provide a solid rocket motor that is safe to fly," and, secondarily, "to minimize the impact of the schedule by using existing hardware if it can be done without compromising safety."37
Thomas revealed on July 2 that the redesign effort was focusing on two alternatives for fixing the field joint, both of them based on using the previously ordered castings.38 On August 12, he announced an over-all plan for SRM redesign, which included not only changes in the field joint but also fixes to the SRM nozzle-to-case joint and to the nozzle itself. The redesign proposed for the field joint incorporated the capture feature that had been discussed since before the Challenger accident, added a third O-ring, and made other modifications.39
NASA's plan was controversial. For example, the front page of The New York Times, on September 23, reported "rising concerns that it [NASA] may be discarding more reliable designs in an effort to save time and hundreds of millions of dollars."40
Among those with reservations about the path NASA was taking were members of the NRC Panel on Technical Evaluation of NASA's Redesign of the Space Shuttle Solid Rocket Booster. This was the external review group that had been established in June at the urging of the Rogers Commission; the eleven-man Panel was chaired by H. Guyford Stever, a highly respected engineer who had been Director of the National Science Foundation and Science Adviser to President Gerald Ford.
The Stever Panel's first report was submitted to James Fletcher on August 1. It acknowledged that, of the factors driving SRM redesign, "safety is the prime consideration," but that "the critical national need for the launch capability of the shuttle makes time a close second." The Panel expressed early concern that the test program for the redesigned motor "meets only a minimal requirement."41
Over the next two years, the Stever panel would keep constant pressure on NASA to explore alternative designs and to conduct an extensive test program.42 The panel's next report was submitted on October 10, after NASA had announced its choice for the redesign of the field joint. The Panel gave only a tepid endorsement to NASA's plans, noting that "if this approach is successful, i.e., if the test program succeeds and the level
356 RETURN TO FLIGHT
of safety is judged acceptable, the shuttle flight program can resume at the earliest time." The Panel expressed some skepticism about the likelihood of such success, however, urging that "NASA maintain a program to explore and develop original, possibly quite different designs ... for the contingency that the baseline design may not offer sufficiently good performance and margin of safety." It noted that if the design competition had not been constrained by the desire to use the previously-ordered castings, "we believe that more basic alternatives to the basic design would probably be preferred once thoroughly analyzed." The Panel also told NASA "we believe that the planned test program requires significant augmentation with additional facilities and tests."43
NASA, after spirited internal debate, concluded that the panel's suggestions were well-founded, and added a number of partial and full-scale tests to its plans. On October 16, NASA also announced that it would follow the Panel's recommendation and build a second facility for full-scale tests of the SRM.44 NASA did get Panel endorsement of its decision not to follow one of the Rogers Commission recommendations. At the urging of member Joseph Sutter of Boeing Aircraft, the Commission had suggested that the redesigned SRM be tested in a vertical position, since that was thought to more closely simulate the various conditions during actual SRM use. Constructing a stand for such a test would have cost twenty million dollars and added at least a year to the time before the next shuttle launch. Both the NASA Marshall team under John Thomas and Allan McDonald at Morton Thiokol argued that a horizontal test could be conducted in a way that better simulated flight stresses than would a vertical test. The Stever Panel concurred "that horizontal testing can be appropriate."45
Between 1986 and August 1988, the NASA-Morton Thiokol team conducted a test program that included eighteen full-scale but "short burn" tests of SRM joints; seventy-six tests of subscale motors; fourteen SRM assembly tests; and five full-duration tests of the redesigned SRM. Flaws in SRM insulation and seals in joint areas were deliberately introduced in a number of tests; particularly severe flaws were created for the last full-scale SRM firing before return to flight, in August 1988. 46
The test program did not always go smoothly, and on occasion produced results that forced the team to revise their baseline design. As a result, the date for the first launch slipped twice from a February 1988 target, to June 1988 and then to the August-September period. Early subscale tests convinced the team to stay with the original O-ring material, rather than introduce a substitute. The first full-scale firing was delayed from February to May 1987. The redesigned joint was first tested in a subscale firing in early August 1987; the full scale test came on August 30. (Richard Truly's reaction to the successful test was "a couple of grins.") 47 December 23 test of the new design at temperatures close to those at the time of the Challenger launch was at first called a success, but a few days later engineers discovered that the redesigned outer boot ring at the junction between the SRM nozzle and the rest of the motor had failed.48 After this test, even though it had not identified the specific cause of the failure, in order to save time the redesign team abandoned the new design and returned to one that was a modification of the pre-Challenger design and had performed well in the August test. A successful fourth full-scale test on the
FROM ENGINEERING SCIENCE TO BIG SCIENCE 357
new test stand that had been suggested by the Stever Panel came in June 1988; it simulated the bending, vibrations, and other stresses of an actual liftoff.
The final full-scale test came on August 18; it was the most demanding and controversial of the series. The need for such a test, introducing the "worst credible" flaw, had been urged on NASA by the Stever Panel as "essential."49 The redesign team used a putty knife and shoelaces, among other means, to introduce holes in the primary SRM seals; these flaws allowed the seepage of gases in order to check whether backup seals would actually work. Such deliberately induced major flaws were unprecedented in the history of solid rockets, and "months of internal debate" within NASA and Morton Thiokol had preceded Richard Truly's decision to accept the NRC recommendation and approve the politically very risky $20 million test. (If there had been a failure during the test, NASA certainly could not have launched Discovery a month later, even though the test motor contained flaws well beyond anything likely to appear in Discovery's SRMs.) Although there were some within NASA who favored the test, most did not; that Truly approved it suggests the power the Stever Panel had over the character and pace of the return-to-flight effort.50
As the test ended, Allan McDonald and Royce Mitchell, the NASA engineer who had shared leadership of the SRM redesign effort with John Thomas, leapt on the still smoking booster to check for joint failure. There was no evidence of it. In the crowd watching the test, Truly shouted "we did it!"51
A few weeks later, a Morton Thiokol spokesman announced that the test had been "as near perfect ... as you can imagine."52 With that outcome, NASA judged the redesigned SRM ready for use. In its September 9 report to the NASA Administrator, the Stever Panel concurred, noting that "risks remain.... Whether the level of risk is acceptable is a matter that NASA must judge. Based on the Panel's assessment and observations . . ., we have no basis for objection to the current launch schedule for STS-26." 53
To its great relief, NASA was now felt both technically and politically ready to return the Space Shuttle to flight. Successfully redesigning the solid rocket motor had been the "long pole in the tent" of the return-to-flight effort; with the muted endorsement by the Stever Panel of the redesign effort, the last obstacle to an initial post-Challenger flight had been removed.
One person close to the program suggested that the redesign and testing work between early 1986 and August 1988 "exceeded, by four or five times, the amount of work put into original motor work in the mid-1970s." 54 While Richard Truly was necessarily removed from the day-to-day engineering details of the enterprise, he at its outset focused efforts on only those redesign activities that were mandatory for requalifying the SRM for use on the first post-accident flight, and resisted pressures from many fronts to introduce changes, including new designs, additional tests, and different contractors, that would
358 RETURN TO FLIGHT
have delayed resumption of shuttle flights even more. 55 Truly defended the NASA-Morton Thiokol effort to a sometimes hostile Congress. He accepted the risk that the proposed "minimum necessary change" approach to redesign would not be successful, and authorized ordering SRMs incorporating the baseline design changes for the first post-Challenger flights at the time the redesign reviews were completed, but before major tests of the redesign had begun. If there had been a major design failure in the test program, NASA would have had to go back to square one, and those SRMs redesigned or scrapped.56 When the pre-launch test program concluded with the August 18 success, Richard Truly had reason to be excited.
A New Management Structure
Putting a new management structure in place was second in importance to redesigning the SRM as a prerequisite to clearing the Space Shuttle for its return to flight. Richard Truly made a reassessment of the entire shuttle program management structure the first item in his return-to-flight strategy in March 1986, and the Rogers Commission listed such a review as its second recommendation. In May 1986, newly reinstalled NASA Administrator Fletcher had charged the former manager of the Apollo program, retired General Samuel Phillips, with conducting an overall review of NASA organization and management. On June 25, Truly directed astronaut Robert Crippen to form a fact-finding group specifically responsible for assessing the National Space Transportation System (NSTS) management structure.
A first step in reforming program management was the departure or transfer of a number of those who had been in key management positions at the time of the Challenger accident. By October 1986, there were new directors at the Johnson, Marshall, and Kennedy Centers, and several other individuals at Marshall who participated in the decision to launch Challenger had left NASA.
The Crippen group submitted its findings in August. They were consistent with the views of the Phillips review, and so on November 5, after extensive consultations within NASA, Truly announced a new shuttle management structure.57Aviation Week described it as "resembling that of the Apollo program, with the aim of preventing communication deficiencies that contributed to the Challenger accident."58
The key management change was moving lead responsibility for the shuttle from the Johnson Space Center to NASA Headquarters in Washington. Arnold Aldrich, who had
FROM ENGINEERING SCIENCE TO BIG SCIENCE 359
been NSTS manager in Houston, was asked by Truly to come to Washington as Director, NSTS —in effect, the single director of the Space Shuttle Program, with all shuttle-related activities at the Johnson, Marshall, and Kennedy Centers reporting to him. He in turn would report directly to Truly. Aldrich, who was the only top-level shuttle manager who retained his position after the Challenger accident, would have two deputy directors, one for the NSTS Program based at Johnson, and one for NSTS Operations, based at Kennedy. Richard Kohrs was named to the first deputy position; Robert Crippen, the second. The Director, NSTS would have "approval authority for toplevel program requirements, critical hardware waivers, and for budget authorization adjustments .... "59
Truly in his memorandum also noted that "a key element in the ultimate success of the Office of Space Flight is a revitalization of the OSF Management Council."60 This body included the Associate Administrator for Space Flight and the Directors of Johnson, Kennedy, and Marshall (and the much smaller National Space Technology Laboratories). It had not been very active in the pre-Challenger period. This top-level group, lead by Truly, began to meet on a monthly basis, and served as the forum for overseeing the return-to-flight effort in the months following. Its meetings were described as "free-wheeling, no-holds-barred," at which "programme issues are flushed into the open and relentlessly pursued to resolution."61
A secondary aspect of the Rogers Commission recommendation on management changes was that "NASA should encourage the transition of qualified astronauts into agency management positions." Richard Truly was himself a former astronaut, and it might have been expected that implementing this recommendation would have been a straightforward matter.
The reality turned out to be somewhat different. In the wake of the Challenger accident, the public discovered that the image of the astronaut corps was very much at odds with reality, and that the group was racked with "longstanding strains and resentments," and with "low morale, internal divisions, and a management style that uses flight assignments as a tool to suppress discussion and dissent."62 Chief astronaut John Young, who had commanded the first shuttle mission, was particularly critical of NASA's approach to flight safety. 63
Truly's first challenge, then, was rebuilding a positive attitude among his former astronaut colleagues. He met with them privately in March 1986, and made sure that Crippen considered astronaut views as he reviewed shuttle program management. He was not totally successful; some in the astronaut office believed he was too ambitious in trying to return the shuttle to flight by February 1988, and was planning on too many launches per year once the shuttle was back in operation. They were critical of the measured pace of the recovery effort, given a launch target only sixteen months in the future, pointing out that after the Apollo 1 fire, the command module was redesigned in only eighteen months and suggesting that "management has either got to cut back what they want to do before restarting flights, or get a 'tiger team' approach to pick up momentum."64
By July 1987, NASA noted that "ten current or former astronauts hold key agency management positions."65 One of them had been Rick Hauck, who served from August 1986 to January 1987 as NASA's Associate Administrator for External Relations before he returned to Houston to train for the STS-26 mission. It was rather well known that Hauck was likely to command the first post-accident shuttle flight; he was thus a convincing spokesman for the safety
360 RETURN TO FLIGHT
aspects of the return-to-flight effort. Other astronauts brought into management positions had "some difficulties in adjusting to the realities of bureaucratic life," but felt that "their presence had made a difference, pointing with pride to influence on key policy issues."66
Other Changes to the Shuttle
Even before the Rogers Commission submitted its report, Richard Truly made one key decision related to reducing the risks of future shuttle operation. Some in NASA, even before the accident, were concerned about the wisdom of using a modified Centaur rocket, fueled by highly combustible liquid hydrogen, as an upper stage to carry satellites from the shuttle's payload bay to other orbits. Among the payloads for which the Centaur was to be used were two solar system exploration missions, Ulysses to explore the Sun's polar regions and Galileo to orbit Jupiter; several classified Department of Defense payloads were also scheduled to employ the Centaur upper stage.
A combination of congressional pressure and the more stringent safety criteria being applied to the shuttle after the accident led to a NASA reassessment of Centaur. Although over $700 million had already been spent on modifying the Centaur for shuttle use, and its unavailability would cause major delays in the solar system exploration program, Truly recommended cancelling the Shuttle Centaur program. Administrator Fletcher agreed and announced the decision on June 19, 1986. 67
Another key decision was to terminate planning for launching the shuttle into polar orbit from the Vandenberg Air Force Base in California. This decision meant that the very expensive Shuttle Launch Complex 6 at Vandenberg would be mothballed and that the number of overall Department of Defense (DoD) flights on the shuttle reduced (DOD would use a Titan IV expendable launch vehicle for payloads originally scheduled for a shuttle launch from Vandenberg). This decision reduced overall schedule pressure on a four-orbiter shuttle fleet, and eliminated the need for a lighter, filament-wound SRM case.68
The third recommendation of the Rogers Commission had directed NASA and its industrial partners to review, in terms of safety and mission success, all Criticality 1, 1R, 2, and 2R items and hazard analyses. Richard Truly had called for an even more extensive risk review in his March 1986 return-to-flight strategy. The Rogers Commission had also separately recommended a series of actions to improve landing safety.
That the shuttle had been flying with a number of less-than-optimum
systems and components was well known to those close to the program, but
the pressures of maintaining an ambitious launch schedule and budget constraints
had blocked any extensive review and upgrading of the shuttle before the
accident. When it became clear that the shuttle would be grounded for some
time, Arnold Aldrich, at the time still in charge of the shuttle program
at the Johnson Space Center, had on March 13, 1986, initiated a comprehensive
review aimed at identifying possible shuttle upgrades. By the end of May,
this review had identified "44 potentially [critically] flawed components
of the space shuttle ... that may have to be fixed before shuttle flights
The conduct of a comprehensive Shuttle Failure Modes and Criticality Analysis and the audit of the resulting Criticality 1 and 2 items recommended by the Rogers Commission was an extensive and complex process. In its July 1986 report on implementation of the
FROM ENGINEERING SCIENCE TO BIG SCIENCE 361
Rogers Commission recommendation, NASA indicated that "the overall reevaluation is planned to occur incrementally and is scheduled to continue through mid-1987. "70 By the time Discovery was ready for launch, the list of Criticality 1 items had grown from the 617 items at the time of Challenger to 1,568; each of those items had to pass particularly rigorous review before Discovery was cleared for flight. The number of Criticality 1R items had also grown dramatically, from 787 to 2,106 .71
Similar to his situation with respect to SRM redesign, Richard Truly found an external review committee assessing NASA's actions with respect to risk assessment and management. The National Research Council created a Committee on Shuttle Criticality and Hazards Analysis Audit in September 1986; the Committee was chaired by retired Air Force General Alton Slay. In its initial report, submitted to James Fletcher on January 13, 1987, the Slay Committee noted that it had "been favorably impressed by the dedicated effort and extremely beneficial results obtained thus far." The Committee raised a point that recurred throughout its work, that "the present decision-making process within NASA ... appears to be based on the judgment of experienced practitioners and has received very little contribution from quantitative analysis." The Committee also questioned the timing of the risk review in terms of incorporating any resulting design changes in the shuttle before its scheduled return to flight (then February 1988), noting that there may not be "time to incorporate any substantial design changes that may be indicated by the outcome" of the review.72
The Slay Committee continued its work throughout 1987 and submitted its final report to Administrator Fletcher in January 1988, although the report was not made public for two months. While generally positive in tone, it criticized NASA's risk assessment activities as still too "fragmented" and "subjective," and for not taking advantage of widely used quantitative techniques such as probabilistic risk assessment. 73 But, most important to Richard Truly and his associates, the Committee found "absolutely no show-stoppers" from a risk assessment perspective in terms of NASA's return-to-flight plans.74
Richard Truly had relieved much of the pressure of implementing the separate Rogers Commission recommendation on improving landing safety by mandating in his March 24, 1986, return-to-flight strategy that the first flight would land on one of the extremely long runways at Edwards Air Force Base in the California desert. In its 1987 report to the President, NASA said that it had identified several design improvements "to improve the margins of safety for the landing/deceleration system. Some of these improvements are modifications to existing designs and will be completed prior to the next flight." But, added NASA, improvements involving more extensive design changes would have to be certified for flight and then introduced "later in the program."75
In fact, this was the philosophy followed for almost all design changes to the shuttle in the aftermath of the Challenger accident which were not related to SRM redesign. The first post-accident shuttle flight was launched as soon as possible after the requalification
362 RETURN TO FLIGHT
of the SRM for flight; the introduction of other redesigned shuttle elements as a result of the risk reviews or of Arnold Aldrich's examination of desirable shuttle improvements did not have significant influence on the shuttle launch schedule. However, the post-Challenger reviews did have other important impacts, both before and after return to flight. The system was overall much safer and reliable on September 29, 1988, than it had been in the 1981-1986 period. The shuttle's main engines were upgraded, its brakes improved, and the valves in the orbiter that controlled the flow of fuel to the orbiter's engines modified to prevent accidental closure. But the result was a "shuttle in transition"; "the hard truth," said Aldrich, " is that the really major changes take years."76
Adding an Escape System
As a former astronaut, Richard Truly gave particular, personal attention to the Rogers Commission recommendation that an escape system be added to the shuttle to allow its crew to leave the vehicle in an emergency while it was in controlled gliding flight (i.e., after the SRMs had finished firing and been jettisoned and the shuttle's main engines shut down). In fact, a search for a viable escape system had begun in March 1986; as the search progressed astronaut Bryan O'Connor played a key role in assessing various options. Alternatives considered included ejection seats, "tractor rocket" extraction of seated crew members, bottom bail out, and tractor rocket extraction through the side hatch. All but the last alternative were eliminated by the end of 1986, but in its July 1987 report to the President on how it was implementing the Rogers Commission recommendations, NASA said that a decision to implement the side hatch, rocket-powered escape approach "had not been made." 77
NASA in December 1986 had in fact made a tentative decision to go forward with this approach, if it could be shown satisfactory in tests and installed in time for the next launch.78 By September 1987, due to delays in the testing program and the possibility that an adequate supply of parts for the system might not be available on a timely basis, NASA began to consider a simpler alternative-one using a telescoping metal pole extending nine feet beyond the shuttle escape hatch. In an emergency, crew members would attach themselves to the pole and slide away from the shuttle orbiter's wing before they parachuted to Earth.79
Based on tests of the two systems, Truly in April 1988 selected the pole escape approach. This was perhaps the last major pre-launch choice stemming from a Rogers Commission recommendation. One factor in the decision was avoiding the additional risks created by installing the pyrotechnic tractor rockets in the shuttle cabin; also, the STS-26 crew preferred the pole system. The escape system could be used only with the shuttle in controlled flight at a less than 20,000 foot altitude, with landing on a primary or emergency runway impossible. (Whether in an emergency to push the shuttle's main engines beyond their design limits to enable the orbiter to reach a trans-Atlantic abort site, or to bail out was a controversial issue up almost to the time of the Discovery launch. Astronauts and mission controllers favored a bail out option, but they were overruled by Truly who wanted to avoid losing another orbiter in an ocean ditching.)80 Bailing out of the shuttle was considered far preferable to trying to survive a water landing; one individual responsible for the escape system commented, "the orbiter doesn't survive ditching very well."81
FROM ENGINEERING SCIENCE TO BIG SCIENCE 363
Setting a Flight Rate
The Rogers Commission had identified "the relentless pressure to increase the flight rate" as a major contributing factor to the Challenger accident. Though not directly related to getting the shuttle ready for its first post-accident flight, determining the appropriate schedule for shuttle launches after the STS returned to flight occupied much of the time of Richard Truly and his staff at NASA Headquarters while the shuttle was grounded.
A first consideration was what payloads the shuttle would carry as the
launch rate was reduced; it was clear that critical national security payloads
would have first priority. After a series of intense debates within the
Reagan administration-over NASA's objections-the President announced on
August 15, 1986, that, except in situations where there were overriding
national security, foreign policy or other reasons, the shuttle would no
longer be used to launch commercial communication satellites.82
This decision and plans for its implementation announced two months later
removed a major category of payloads from the shuttle manifest; prior to
the accident, eleven of the twenty-four earlier shuttle missions had carried
one or more commercial communication satellites.
In October 1986, NASA released a shuttle launch schedule that called for a buildup to fourteen or sixteen launches per year, four years after the STS returned to flight, and after a replacement orbiter had entered service.83 This was more ambitious than the launch rate thought reasonable by yet another National Research Council review committee. At the request of NASA's House Appropriations Subcommittee, the NRC created a panel to carry out a "post-Challenger assessment of Space Shuttle flight rates and utilization." In its October 1986 report, the panel concluded that with a four-orbiter fleet NASA could sustain a launch rate of eleven to thirteen launches per year, but only if there were significant improvements in various aspects of the shuttle program. Without such improvements, the panel estimated, the maximum rate was eight to ten launches per year. The panel noted that only "under special conditions" might the launch rate surge to fifteen launches per year.84
Balancing the desire to get flying again on a regular basis, the pressure to launch critical national security and scientific payloads as soon as possible, and the need to ensure continued safe and reliable operation of the Space Shuttle was a constant challenge for Richard Truly. He recognized that "we will always have to treat it [the shuttle] like an R&D test program, even many years into the future. I don't think calling it operational fooled anybody within the program.... It was a signal to the public that shouldn't have been sent and I'm sorry it was."85 Media watchdogs were quick to report perceptions that NASA was "putting schedule over safety."86 But, as Truly had said on many occasions, "the only way to operate the shuttle with zero risk is to keep it on the ground." That was not his intent.
Return to Flight
The Space Shuttle Discovery was rolled out from the Vehicle Assembly Building to launch pad 39B on July 4, 1988; as a morale-boosting measure, throughout the day Kennedy Space Center workers and their families were allowed to drive around the pad.
364 RETURN TO FLIGHT
There were no waivers (permissions to launch even though specifications were not met) on any hardware element, and an internal NASA committee had found a "positive change in attitude" with respect to safety considerations and a "healthy redundancy of safety reviews and oversights." The group found no safety issues that would adversely affect the launch of STS-26, then set for September 6. 87
There were a few minor delays before the shuttle was ready for launch,
however, slipping the launch date to late September. A 21.8-second Flight
Readiness Firing of Discovery's main engines was conducted on August
10, and a two-day Flight Readiness Review in early September. The final
launch date of September 29 was set when it was determined that Hurricane
Gilbert would not affect operations at mission control at the Johnson Space
When Discovery roared off of the launch pad after a 98 minute weather-caused delay on the morning of September 29, a great weight was lifted off of not just Richard Truly, but the whole NASA organization. Truly would later say that "the time when the Space Shuttle did not fly was time well spent by NASA. When we look back at 1986-1988, we will see it as a time when NASA and the country took an unwanted, but necessary, breather in the space program. During this time, we took a hard look at ourselves and at what we hoped to accomplish in space. What we saw was solid. Some things needed changing and changes were made. It was a time of introspection, not without pain, but mostly it was a time when we rechartered our course and rededicated ourselves to space exploration."88
Richard Truly brought a perhaps unique set of attributes to his job as NASA Associate Administrator for Space Flight. Though admitting frustration at the inefficiencies of the political process and impatience with the need to testify so frequently to Congress and to participate in frequent executive branch meetings, he was skilled at charting a course through the political process. His status as a former astronaut gained him credibility on Capitol Hill and with the public, and legitimacy within the space flight community inside and outside of NASA. He was able to gain the support of the many external groups overseeing the accident recovery effort for most, if not all, of NASA's actions and decisions, He had enough technical background to understand the issues under debate during the recovery process. He surrounded himself with a team as committed as he was to the shuttle as the centerpiece of the U.S. effort in space.
As he reflected on his experience a few months before the shuttle returned to flight, Richard Truly suggested that during the preceding months "the high and low points have been very high and very low" and that "there have been great frustrations," particularly in dealing with the criticisms of NASA and its employees. He admitted that NASA deserved "some" of the criticism, and so his approach "has simply been to try to build a team that will win our credibility back." The high point in his experience during the return-to-flight effort, according to Truly, "has simply been watching this team come back together."89
Richard Truly accepted the 1988 Collier Trophy on behalf of all those in government and industry that had participated in the return-to-flight effort. It was an honor well earned.