ORGANIZATION A N D POLICY MANAGEMENT

¶a. Management Structure Issue 1

¶Does the management of the Shuttle Program adequately defi e the lines of authority and are managers given authority commen- surate with their responsibilities?

¶186 Responses to Committee Questions, dated August 22, 1986.

166

¶Finding

¶The management of the Shuttle Program is complex and diversi- fied and it is not always clear who has authority or responsibility. NASA's "lead center" concept has resulted in placing the manage- ment of the program at JSC, one of three centers participating in the program; however, because Johnson does not have control of the other centers' resources, the NSTS program manager's author- ity to manage the program is limited and the responsibility is un- clear. Recommendation

¶NASA should restructure the Shuttle Program management to define clear lines of authority and responsibilities. This restructur- ing should take into account the special role each center must play and be especially sensitive to the need for the cooperation and sup- port of all the participants to achieve a common goal. NASA should give special consideration to moving the Program Manager to NASA Headquarters to avoid the confusion and inter-center ri- valry that result from having a large multi-center program man- aged out of one of the participating centers. Discussion

¶The line of management responsibility, authority, and account- ability for NASA programs is from the Administrator to the Asso- ciate Administrator to the Field Center Directors, who delegate im- plementation authority to a program/project manager. JSC has been designated the "lead" center for the NSTS Program and has the responsibility for systems engineering and integration, oper- ations integration, and management integration. Marshall has the responsibility for the propulsion system and Kennedy has the re- sponsibility €or launch operations.

¶The Associate Administrator for Space Flight (the Level I pro- gram manager) performs oversight over the program but doesn't have the technical staff to effectively manage the program. The NSTS Program Manager, i.e. the Level I1 manager at JSC, func- tions as a program coordinator; he is responsible for integrating the various program elements and he controls all the project inter- faces. He clearly does not control all the program elements since the individual (Level 111) Project Managers are accountable to their Center Directors who are in turn accountable to the Associate Ad- ministrator who controls the funding. For example, the Level I1 manager told the Rogers Commission that he was unaware that the SRB Project Office had procured additional Solid Rocket Motor casings to be used for testing of the joints;

Now it turns out that the budget for that kind of work does not come through my level I1 office. It is worked di- rectly between the Marshall Center and NASA Headquar- ters and there again had I been responsible for the budget for that sort of work, it would have to come through me, . . . .187

¶'87 Rogers Commission Report, Volume V, p. 1490.

167

¶The witnesses who addressed the management issues at the Com- mittee hearings had differing philosophies regarding the best possi- ble solution; however there was general agreement that the present system tended to cause confusion. There was also strong sentiment for strengthening the headquarter's role. Mr. Jesse Moore, former Associate Administrator for Space Flight, testified:

I think we need to go back and make sure we clearly define the roles of NASA headquarters, the roles of the centers in the overall management of the STS. I think we need to re-look at that kind of interaction and the kind of specific roles, responsibilities, to ensure that authority and responsibility is commensurate in terms of the role definitions for the varous levels of man- agement in NASA. I think we need to look at strengthening NASA head- quarters. I would say that in my tenure a NASA head- quarters we had a decline in staff in the Office of Space Flight. It was a decline in the number of staff, and I think we need to look at what is the proper level of staffing re- quirements to do this particular job. I also think we need to look to make sure we get as much technical expertise into the Office of Space Flight as we possibly can. 88

¶General Stafford, a former Gemini and Apollo astronaut, also stated: ". . . I guess I was never comfortable with the lead center type of management structure, after having seen how satisfactorily Apollo worked." (Note: The Apollo program was managed out of headquarters.)

¶With regard to the appropriate role of the Program Manager (Level 11),there was not a clear consensus. In discussing the Rogers Commission's recommendations, General Abrahamson stated:

. . . However, I would also like to point out that many of these recommendations have long been incorporated in NASA management procedures. The Program Manager, by definition, has the necessary authority to get the job done.19O

¶When asked if the program management should remain at JSC, Mr. Moore, who is currently Director of JSC, replied:

I think that is certainly a topic that is going to be stud- ied very, very carefully. I think there are a couple of options that can be looked at that would keep the major parts of program manage- ment that has been in operation at the Johnson Space Center at the Johnson Center. There are a lot of tools, roots and capabilities. I think, on the other hand, there should be some looks at the Office of Space Flight for finding some way to strengthen

168

the overall program management in the Office of Space Flight. And one concept might be to have a Shuttle Program Di- rector within the Office of Space Flight and working with the Level I1 program office at the Johnson Space Center. My answer is, I believe the Level I1 program office, with some strengthening, and the level I program office, with some strengthening-we can make it work and it should remain at the Johnson Space Center.

¶This was in direct contrast to the view held by John Yardley, former Associate Administrator for Space Flight. In discussing the Rogers Commission's recommendations, he stated:

The one in particular that I think I have some back- ground in that I think is not correct is they are trying to strengthen the authority and responsibility of the Program Manager at Johnson. Let me just relate what happened when I went to NASA. I hadn't been there but a couple of weeks and one of the other centers called me and said, "Hey, the Program Man- ager wants to take 15 million of my money and put it on the Orbiter." It became immediately apparent to me to have one of the center people handle the funding decisions was not going to be in the best interests of cooperative technical activity. So I pull+& 1 the final decisions on the money to Washing- ton, where 1 think they still are . . .⁹⁹ Rogers Commission Report, loc. cit. 2

¶Major Slayton made a similar observation concerning the prob- lems with having a multi-center program managed at one of the field centers:

. . . I think when you look at relationships between the centers and how the organization is structured; and you could say it could be restructured so you don't have inter- center jealousies interfering with the communications channel. A lead center concept where Level I1 is viewed by the other centers as being another center instead of having its headquarters' level is one reference I would make.

¶Major Slayton went on to say that any organization could work with the proper people:

A lot of it is in the management attitude; but again, my opinion is, you can make any organization work if you got the right people, and if you don't have the right people I don't care how you organize it, it will not work, so you still end up dealing with individuals.⁹⁹ Rogers Commission Report, loc. cit. 4

¶Is1 Ibid., p. 99.

169

¶Other witnesses also alluded to the problems with inter-center ri- valries under the current system and the break down of esprit de corps. Mr. Moore called for a new effort to re-instill the team spirit:

¶I believe a n approach to that has got to be building team work, again, to make sure-the Shuttle program involves many elements, many contractors, many NASA centers, all playing together as a team.

¶I believe we have to go back and re-instill in our people, in our participants, a team work approach . . .

¶I think the overall structure of the Shuttle program is obviously built upon people and, you know, there are humans all the way up the chain, all the way from the engineers a t the subcontractors to the engineers at the contractors, the NASA centers and so forth.

¶I think we have got to make sure that each of those participants in the program feel a dedication, feel a dedication to safety, feel a dedication to the program that they are making a valuable contribution and I think we need to do that by personal communications as well as trying to look a t our structure to make sure we have not defined something that will at least maybe encourage, tend to encourage communications breakdown.

¶General Abrahamson made the following observations with regard to changing organization:

¶It is true that when any organization is formed, it is formed to help you accomplish a particular task. By the same token, once it is there, it develops momentum and procedures and impediments sometimes to exactly what you would like to have, a dynamic and modifying organiza-

tion for the challenges of the future. This is always diffi- cult. I believe that we had a n organization that was designed for the development of the Shuttle, and when we got there, since it was only the second flight, that we had a tremen- dous change of attitude that we had to be able to create, and that was to create a n organization that would think in terms of operations of the Shuttle and overcome the flight test problems.

¶Issue 2

¶Are astronauts adequately represented in management? Finding

¶The Committee finds no evidence that astronauts are denied the opportunity to enter management if they so choose. Discussion

¶The Rogers Commission has suggested that NASA should make greater use of astronauts in management; however, the Commission report provides no basis for that recommendation. Astronauts generally have shown little interest in going into any kind of desk job, including management positions, until such time as their active flying days end. At that time, management jobs within the astronaut program become attractive alternatives to some; howev- er, opportunities in this area are naturally limited. Major Slayton expressed this very well when he testified about his experience with the Mercury Program:

170

I had the misfortune at that time of having been ground- ed due to a medical problem so I was elected to take over the management of the astronaut corps, a job I didn't par- ticularly care about, but it was the next best thing.lS7

¶Mr. Nelson asked the astronauts whether any of them felt there was a "modus operandi" within NASA that excluded either active or former astronauts from the management structure. General McDivitt stated he had seen no bias in his three years as Program manager for the Apollo Program. Mr. Nelson then asked Deke Slayton if he had ever seen any bias in NASA and Major Slayton confirmed that he too saw no evidence of bias against astronauts in management. General Abrahamson observed:

Throughout my tenure, astronauts were in key program office positions and one served as a n Assistant Associate Administrator in the Office of Space Flight. . . . I s 8

¶There was agreement among the astronauts that the astronaut office should be moved up higher in the organization. General McDivitt summed up the astronauts' position:

I think I would recommend that the Flight Crew Oper- ations Directorate be moved up to report to the Center Di- rector as well as the Flight Operations Director. I think both of those organizations are very key to flying, and having them go through another layer of man- agement before they get to the Center Director creates a filter which is not necessary or desirable for either one of them. I think it also gets them on the same level as the engi- neering arganizations within the manned spacecraft center, and gives them better access to the program.lgg b. Communication

¶Issue 1

¶Are there adequate opportunities to communicate problems within the Shuttle Program management structure? Finding

¶There are many regularly scheduled meetings and teleconfer- ences at all levels of management throughout the Shuttle Program. In addition, "special" meetings and telecons are routine. No evidence was found to support a conclusion that the system inhibited communication or that it was difficult to surface problems. Discussion

171

¶Every day at noon central time a teleconference is held among all NASA Space Shuttle Program participants. This is the daily "special" Level I1 PRCB (Program Requirements Change Board) meet,ing and includes, among others, all the managers of the vari- ous program elements, the JSC Directors of Flight Crew Oper- ations, Mission Operations, Engineering, Mission Support, SR&QA, and Space and Life Sciences. Program status, urgent problems, and program requirements are brought up at this meeting. The PRCB convenes by teleconference on alternate Fridays to discuss all other (less urgent) program issues; in addition, other special meetings are called by the PRCB secretary when deemed necessary.

¶Each of the supporting organizations also has regularly sched- uled meetings, often by teleconference when they involve more than one location. Regularly scheduled (often daily) teleconferences are also held between various directors and managers.

¶Level I at headquarters conducts daily status meetings and also participates in the noon teleconference. These meetings plus all the Flight Readiness Reviews provide ample opportunity to surface problems. Issue 2

¶Is too much information being disseminated so that important in- formation is lost? Finding

¶Large amounts of information are disseminated on a routine basis, often with little or no indication of its importance to all of the recipients. Recommendation

¶NASA management should review the process of providing infor- mation on significant actions so that awareness by concerned man- agers is assured. Discussion

¶In a NASA briefing to staff on Mission Operations (May 21, 19861, NASA managers revealed that they routinely received infor- mation copies of all sorts of memoranda, such as directives, re- quests, approvals for changes, etc. Often the individual receiving these copies had no direct involvement with the specific subject of the memoranda, and they acknowledged that it was entirely likely that an important piece of information could cross their desk with- out their awareness. Issue 3

¶Are communications filtered so that important information is prevented from reaching the decision makers?

172

¶Finding

¶NASA managers delegated the responsibility for making techni- cal judgments to lower level managers or assistants. Therefore, the information that reached the top decision makers was "filtered" in that it was interpreted by others that were presumed to have more specialized experience or expertise in a given area. There is no evi- dence that middle level managers suppressed information that they themselves deemed to be significant. In fact, as discussed in the Section on Technical Expertise, the failure was not the problem of technical communications, but rather a failure of technical deci- sionmaking. Discussion

¶It is typical in any large, complex organization that as managers rise higher in the organization the scope of their responsibilities broadens to encompass technical areas beyond their own special- ized expertise. Therefore they must rely increasingly on the techni- cal judgments of lower level managers or assistants. There is the additional risk of subordinates' reluctance to transmit unpleasant information upwards; however, it is not evident that NASA manag- ers suppressed information about problems they themselves under- stood.

¶Throughout the hearings, witnesses said that had they known about the seriousness of the problem with the SRM joint, they would have stopped the flights; or (in their opinion), had the deci- sion makers known about it the flights would have stopped. The witnesses acknowledged that the problems with the SRM joint had been briefed at all levels, but always in a way that didn't communi- cate the seriousness of the problem; it was not viewed as life- threatening. Yet the witnesses appeared reluctant to attribute this to poor technical judgments on the part of the managers or techni- cal staff with expertise in propulsion, preferring instead to bl,ame it on poor communications or a poor "decision-making process.

¶Mr. Scheuer questioned Jesse Moore specifically on this point when he asked, referring to Mr. Weeks' summary of the August 19 meeting, "Was it a failure of decision-making on his part or com- munications on his part?" Mr. Moore responded:

Sir, I think that in a position like Mr. Weeks is in, we have to work as a team, for example, and people have to make assessments on situations and I think Mr. Weeks looked at the data and his assessment was that he thought we had a program adequate to cover the activities in the SRB and he believed that after he had talked to the people at Thiokol and he also believed that, I think, after talking to the people at Marshall and I believe his position was that in fact was an acceptable posture for him to take. Part of his responsibility is to make technical judg- ments2O 0

¶Mr. Moore went on to explain that he believed the lack of under- standing of the SRB joint extended throughout the agency:

173

I would say, sir, in looking up and down the system and what has been determined about the SRB from the many analyses and work that has been done in the past, I don t think the system all the way from day one of the program really understood all the implications of how the SRB joints worked and I think that we have learned, all of us have learned, an awful lot about the SRB . . .201

¶And again referring to the August 19 meeting:

That was a report from my deputy (Mr. Weeks), that he believed the situation was acceptable as far as assessment of the data presented to him, and I trust the people in the organization to make those kinds of judgments. We have to make those judgments on a day-to-day kind of basis, but I did hear at Flight Readiness Reviews, as ev- erybody as a member of the overall Shuttle team heard about issues associated with the O-ring problem. I believe the first time this was experienced on the Shuttle Program was all the way back to flight 2 . . . I did not, as the head of the Level I office, believe the problem with the SRB 0- rings was serious enough to consider stopping the launches. If I did, I would have stopped the launches, sir.2o2

¶Mr. Scheuer again asked Mr. Moore to identify where the failure was, "Was it in your being communicated with by Mr. Weeks? Was it a failure of judgment on Mr. Weeks' part that all systems were go.? Where was the failure?" At that point, Mr. Moore blamed the failure on communications:

I think in looking at the whole situation, I think there was a failure to communicate the technical seriousness from the contractors involved in this program through . . . .203

¶But, Mr. Scheuer suggested that the contractors had communi- cated the problem at the August 19th meeting. Mr. Moore then suggested that perhaps someone should have made a stronger statement; however, in their collective judgment it was not a seri- ous problem:

On the basis of the specific August 19 briefing that was presented, I believe there should have been a stronger

¶statement made to me that we have a much more serious

problem by Mr. Weeks or any of the people who attended that briefing. Mr. Weeks was not the only one at the brief- ing. There were others at the briefing who had some knowledge about the SRB . . . . I don't recall the specific list of attendees at that par- ticular meeting, but people that were in the overall pro-

¶pulsion area of the office of space flight-and the office of space flight is level one-that is the level one-people who had experience in this thing.

¶20' bid., p. 95.

174

I believe if they felt after that August 19th briefing that we had a problem, that the system should be grounded, that somebody would have come and said, "We have got a problem serious enough to ground the Shuttle flight." That did not occur, and I believe it was based on a col- lective set of judgments that we did not believe the prob- lem was as serious.204

¶The Committee finds no reason to doubt Mr. Moore's observa- tions that no one within NASA understood the problem with the 0- ring and accepts his conclusions:

In hindsight, I think we should have taken much strong- er action after the August 19th briefing . . . if I had the knowledge then that I have today, we would have ground- ed the fleet. I did not have it at the time.205

¶In hindsight, the August 19th briefing, as well as the January 27th telephone conversation clearly identified a serious problem. Perhaps the Thiokol engineers understood the seriousness of the problem; however, Thiokol's own summary and recommendation at the conclusion of the August 19th briefing stated:

Analysis of existing data indicates that it is safe to con- tinue flying existing design as long as all joints are leak checked with a 200 psig stabilization pressure, are free of contamination in the seal areas and meet O-ring squeeze requirements.20

¶This conclusion was accepted by all who heard the briefing, and this was the information that was transmitted throughout NASA. The evidence does not support a conclusion that the top decision makers would have arrived at a different conclusion from the man- agers at Marshall and the Level I managers with propulsion back- grounds. (For additional discussion on this issue, see Section VI.B.l.c.1 c. Safety, Reliability and Quality Assurance Issue 1

¶Is NASA's decision to establish a new Office of Safety, Reliabil- ity, and Quality Assurance appropriate and, if so, what should its role be? Finding

¶The Committee finds that the Rogers Commission recommenda- tion that NASA should establish an Office of Safety, Reliability and Quality Assurance that reports directly to the Administrator is indeed appropriate. However it is not clear what the activities of this office will encompass.

bid, p. 98.

175

¶Recomrnendations

The Associate Administrator for Safety, Reliability and Qual- ity Assurance (SR&QA) should provide to the Committee the agen- cy's draft plan delineating the organization, goals, implementation strategies and resource requirements of the office of SR&QA.
After the Office of SR&QA is fully operational, the Committee will wish to continue oversight over its activities. Discussion

¶Chapter 7 of the Rogers Commission report deals with the sub- ject entitled "The Silent Safety Program." The Commission identi- fied shortcomings in NASA's overall Safety, Reliability and Quality Assurance Programs, and recommended the formation of a sepa- rate Office of Safety, Reliability and Quality Assurance that would report directly to the Administrator. The role of safety and quality assurance in the decisionmaking processes associated with Shuttle flight production requirements has been relatively undefined and ambiguous. The formation of a centralized coordination and control organization should serve to remedy the situation. As the Rogers Commission report notes, ". . . No one thought to invite a safety representative or a reliability and quality assurance engineer to the January 27, 1986 teleconference between Marshall and Thio- kol."²² Rogers Commission Report, Volume I, p. 199. o bid., Chart SRB-4. 2 1 Larry Mulloy, NASA, Marshall Space Fli ht Center, "STS-51L Level I1 Flight Readiness Review,'' January 14,1986. See Appendix VIII-g. 2 2 Discussion with Allan McDonald, September 4, 1986. Z3 Rogers Commission Report, Volume 11. See Chart 15 (p. H-10) and Chart 19 (p. H-12). s Ibid., Chart 30 (p. H-18). 0 7

¶On July 8, 1986, the Administrator established the position of As- sociate Administrator for Safety, Reliability and Quality Assur- ance, and briefly delineated the responsibilities of this office in NASA's responsive document to the Rogers Commission report.208

¶According to NASA, the purpose of this office is to strengthen the role of the SR&QA functions across all the the NASA pro- grams. This will be accomplished by establishing centralized coordi- nation under the Associate Administrator €or SR&QA who reports direct1 to the Administrator on all pertinent matters related to the NlTS. The Associate Administrator is chartered to examine the adequacy of the agencies resources in these areas and to make recommendations for improvements as appropriate. Functional or- ganizations that were previously under the purview of the Chief Engineer's office will now report directly to the Associate Adminis- trator for SR&QA.

¶The major contractors to the NSTS agree with the Commission's recommendation to form a separate NASA SR&QA organization re- porting directly to the Administrator. They are, however, of the opinion that responsibility for the work required to recommend or implement changes or modifications in the quality assurance area must remain with Level I11 and the contractors themselves.

¶The Committee does not argue with the contention that strong SR&QA capabilities must reside at the contractors' plants. Further, the Committee suports NASA's efforts to enhance its in-house ca- pabilities in order to improve the agency's monitoring and over- sight capabilities in the areas of SR&QA. Strengthening Headquar- ter's ability to provide guidance and centralized coordination in the

¶*O' bid., Volume I, p. 152. *08 NASA Response to Rogers Commission, July 14, 1986, p. 20.

176

¶areas of configuration management, product reliability and quality assurance and risk management, are essential to returnng the Shuttle to flight readiness condition. Issue 2

¶Ha s NASA applied sufficient resources to support adequate S M Q A efforts within the NSTS program? Findings

The Committee finds that reductions in NASA civil service personnel that have occurred over the past decade have adversely impacted the agency's ability to maintain the appropriate level of oversight control of the Safety, Reliability and Quality Assurance activities within the NSTS.
NASA has become increasingly dependent upon outside SR&QA support from the Department of Defense (Defense Contract Administration Services [DCAS] and Air Force Plant Representa- tive Office [AFPRO]) and contractors.
NASA has reduced or reassigned to other program areas in- house safety, reliability and quality assurance tasks such as test- ing, analyses and instrumentation and has reduced or shut down in-house facilities for performing SR&QA research and technology development. The degree to which these factors have adversely im- pacted the safety, reliability and quality assurance activities within the NSTS program has not been adequately assessed. Reconmenda t ions
NASA should establish and maintain a strong and effective SR&QA Program. Continuing support for such a program must come directly from the Administrator.
Although it is appropriate to establish strong contractor capa- bilitie in the areas of SR&QA the internal oversight responsibilities and coordination of SR&QA tasks must be the responsibility of NASA itself. In order to assure that the appropriate interfaces among the various subsystem elements that comprise the NSTS, are maintained, a sufficient complement of NASA SR&QA manage- ment and support staff must be available to perform the necessary oversight and coordination tasks. Discussion

¶Reductions in force over the past several years have reduced per- sonnel across the agency from a complement of some thirty-six thousand people down to twenty-two thousand people. A dispropor- tionate decline in Reliability and Quality Assurance (R&QA) staff- ing occured as a result of these reductions. In the Shuttle program, many of the quality control functions and government inspection activities have been performed by contractors in conjunction with the Department of Defense support personnel (DCAS and AFPRO). NASA has expressed some concern about their ability to maintain adequate in-house staffing in these areas. The total number of civil servant employees within NASA dedicated to the SR&QA program is presently about 500 professionals.²² Rogers Commission Report, Volume I, p. 199. o bid., Chart SRB-4. 2 1 Larry Mulloy, NASA, Marshall Space Fli ht Center, "STS-51L Level I1 Flight Readiness Review,'' January 14,1986. See Appendix VIII-g. 2 2 Discussion with Allan McDonald, September 4, 1986. Z3 Rogers Commission Report, Volume 11. See Chart 15 (p. H-10) and Chart 19 (p. H-12). s Ibid., Chart 30 (p. H-18). 0 9 This represents a reduction of 71% from the 1970 complement.

177

¶NASA attributes this reduction to the termination of "in-house flight programs, along with the transfer of certain functions . . . to other organizations within the NASA centers," In their response to Mr. Roe's inquiryY2 l o NASA makes the following statement:

"Even though we had a reduction in R&QA personnel, our detailed review of the quality operation did not reveal that we missed any of the quality control check points which may be contributed to the accident.''

¶The Committee cannot support NASA's assessment on this matter. Although NASA may argue that the quality control check points for the certification tests required on the ambient and in- duced temperature effects on the O-ring seals were checked off by the &A representative at Thiokol as having been satisfactorily com- pleted, in actuality these tests were never performed. To what extent this failure of the QA function to do its job conbributed to the accident may be questioned, but the fact that the control didn't work in this case cannot be denied.

¶It should be noted, however, that according to some of the prime contractors, SR&QA staffing has actually improved over the years. For example, at the Rocketdyne Corporation, there has been an increase in QA staffing to a level that represents nearly 40% of the corporation's manufacturing staff. Issue 3

¶Are the responsibilities of safety engineers and design engineers adequately specified within NASA's "risk management" program? Finding

¶The roles of safety, design as well as reliability engineers are not adequately and uniformly defined throughout the NSTS program. In some cases, the Committee learned that safety engineers were not participating in major decisions related to flights of the Shut- tle. Recommendations

¶It should be the responsibility of the new Associate Administra- tor for SR&QA to fully specify the roles of safety and reliability en- gineering as well as quality assurance personnel within the NSTS program so that all critical aspects of the program and decisions related to the adequacy of hardware and subsystem performance are fully reviewed by these disciplines. Discussion

¶The function of the safety engineers within the NSTS program has been to determine whether or not certain prescribed tests, analyses, and design descriptions have been followed appropriately

¶*OQDiscussionwith the NASA Chief Engineer's Ofice, May 13, 1986. Also, Cmte Hgs, re- sponse to question by Mr. Roe,Transcript, June 11, 1986, pp. 59-60.²² Rogers Commission Report, Volume I, p. 199. o bid., Chart SRB-4. 2 1 Larry Mulloy, NASA, Marshall Space Fli ht Center, "STS-51L Level I1 Flight Readiness Review,'' January 14,1986. See Appendix VIII-g. 2 2 Discussion with Allan McDonald, September 4, 1986. Z3 Rogers Commission Report, Volume 11. See Chart 15 (p. H-10) and Chart 19 (p. H-12). s Ibid., Chart 30 (p. H-18). 1 0 Cmte Hgs, response to question by Mr. Roe, Transcript, June 11, 1986, pp. 59-60.²² Rogers Commission Report, Volume I, p. 199. o bid., Chart SRB-4. 2 1 Larry Mulloy, NASA, Marshall Space Fli ht Center, "STS-51L Level I1 Flight Readiness Review,'' January 14,1986. See Appendix VIII-g. 2 2 Discussion with Allan McDonald, September 4, 1986. Z3 Rogers Commission Report, Volume 11. See Chart 15 (p. H-10) and Chart 19 (p. H-12). s Ibid., Chart 30 (p. H-18). 1 1 Telephone Conversation, August 13, 1986.

178

¶as they relate to safety concerns, using the techniques of HA. The safety office has not been significantly involved in the engineering design efforts. If an engineering problem arises that could effect the safety of the overall system, it is the responsibility of design engineering teams to perform technical evaluations rather than having these analyses performed by the safety engineers. Prior to the Challenger accident, the safety program did not have the per- sonnel, facilities or expertise to review decisions by design engi- neers that the O-ring erosion problem was a manageable risk. Even though this erosion was a continuing problem, there was, according to testimony provided to the Rogers Commission, no second set of "eyes" available to question waiver applied to this problem.212 Issue .I

¶Does the SR&QA program require improved coordination be- tween centers, contractors and NASA Headquarters? Findings

Although guidelines have been published that describe the re- sponsibility of contractors' in the areas of SR&QA,213 NASA's guidelines do not adequately distinguish these various activities as distinct disciplines requiring specialized skills and centralized co- ordination.
In its review of the agency's reliability and quality assurance programs as they relate to the Space Shuttle, the Committee found there was little commonality among the cognizant officials at MSFC, JSC, KSC, and Headquarters in the perception of the vari- ous responsibilities associated with these separate and distinct dis- ciplines. Recommendations
It is important that a clear delineation of responsibilities for the separate SR&QA disciplines be appropriately documented. It is also essential that the relative importance of each of the three sep- arate disciplines be established as an integral part of the NSTS program. These functions are the responsibility of NASA Head- quarters.
NASA must carefully review the staff and resources devoted to the SR&QA function within NASA and contractor organizations for adequacy. The Administrator shall report to the Committee with his findings and recommendations. Discussion

¶Although the controlling document describing the SR&QA func- tions for the Shuttle contractors was provided to the Committee, no corresponding document was identified that describes the imple- mentation of these functions for the SR&QA engineers that are direct employees of NASA. NASA contends that the same control- ling document applies to agency employees. The specific oversight

¶Testimony before the Rogers Commission, Mr. Jack Walker, Deputy Director, MSFC Safety Office, ,$ ril, 1986.²² Rogers Commission Report, Volume I, p. 199. o bid., Chart SRB-4. 2 1 Larry Mulloy, NASA, Marshall Space Fli ht Center, "STS-51L Level I1 Flight Readiness Review,'' January 14,1986. See Appendix VIII-g. 2 2 Discussion with Allan McDonald, September 4, 1986. Z3 Rogers Commission Report, Volume 11. See Chart 15 (p. H-10) and Chart 19 (p. H-12). s Ibid., Chart 30 (p. H-18). 1 5 NASA, &fety, Reliabiliz Maintainability and Quality Provisions for the Space Shuttle Program": NHB 5300.4 (1D-2), tober, 1979 responsibilities of these employees and their independent reviews and analyses requires a more complete delineation in the Commit- tee's view. The Rogers Commission report provides definitions for the SR&QA disciplines. An expansion upon these definitions is required in order to establish a commonality of understanding of the various functions as they apply to the Shuttle program.

179

¶The management structure within NASA that coordinates and performs the activities associated with the SR&QA tasks for the NSTS has become decentralized over the past decade. Until recent- ly many of the oversight duties that at one time were handled through Level I were moved to the field centers. Responsibilities for various systems that comprise the Space Shuttle are delegated to the Level I11 field centers. These centers establish and coordi- nate SR&QA activities at the contractor facilities. They are also re- sponsible for reporting any anomalies, inconsistencies, or problems to Level I1 program management.

¶Until recently, the Office of the Chief Engineer had responsibil- ity for SR&QA activities. For various reasons, the operations of this office in the areas of SR&QA appear to have lost effectiveness, either through reductions of personnel and support of these pro- grams at the Headquarters level or through the diffusion of these functions into various organizations within the operating divisions at the field centers. These changes reduced Headquarter's abili- ty to participate in field center status reviews with the prime con- tractors, limited the Level I manager's ability to survey the effec- tiveness of the SR&QA programs agency-wide and reduced the co- location of SR&QA personnel within Headquarter's program of- fices. The Committee expects that the new Office of SR&QA will be chartered to make appropriate corrections to augment the safety, reliability and quality assurance functions within the NSTS Pro- gram.

¶d. Contractor Incentives Issue

¶Key Shuttle contracts (e.g., the Solid Rocket Booster Production Contract and the Shuttle Processing Contract (SPC))provide incen- tives both for reliability, integrity, and safety of products and serv- ices on the one hand, and for cost and schedule on the other. Do these contracts provide an appropriate balance between the two types of incentives? That is, does NASA utilize contracts to reward and promote operational safety? Findings

The SPC provides far greater incentives to the contractor for minimizing costs and meeting schedules than for features related to safety and performance. SPC is a cost-plus, incentive/award fee contract. The amount of the incentive fee is based on contract costs (lower costs yields a larger incentive fee) and on safe and successful launch and recovery of the Orbiter. The award fee is designed to permit NASA to focus on those areas of concern which are not sen-Rogers Commission Re rt, Volume I, p. 152. 21bDiacussion with the N E A Chief Engineer's Oftice, May 13, 1986.

180

¶sitive to the incentive fee provisions, including the safety record of the contractor. However, the incentive fee dwarfs the award fee- while the maximum value of the award fee is only one percent of the value of the SPC, the incentive fee could total as much as 14 percent of the SPC.

During the developmental phases of the Thiokol contract for Solid Rocket Booster production (1980-1983), the contractor re- ceived consistent ratings of "Excellent-Plus" or "Superior" under the cost-plus, award-fee contract. NASA contracted with Thiokol on a cost-plus, incentive-fee (CPIF) basis beginning in July, 1983. The CPIF contract pays strictly on the basis of costs, although penalties may be invoked for delays in delivery or for Shuttle accidents due to SRB failure. At the time of the Challenger accident, Thiokol was eligible to receive a very large incentive fee, probably on the order of $75 million. Recommendations
NASA should reexamine all Shuttle contracts and report to the Committee with its findings and recommendations on whether more incentives for safety and quality can be built into these con- tracts. This report should address, inter alia, the SRB Production Contract and the SPC.
NASA's new Office of SWQA should be involved in the pro- curement and award fee processes, both to establish reasonable guidelines and rewards in new contract and to judge performance of ongoing contracts. Discussion

¶Mr. Robert Thompson, Vice President of McDonnell Douglas, summarized the position of several Committee witnesses when he stated:

I have never detected that a contractor would deliberate- ly infringe on safety for a profit motive.2

¶On the other hand, Thompson also admitted that contracts do vary in the extent of their safety incentives and that, to a certain degree, such incentives can make a difference in operational safety:

. . . the type of safety that we are looking for, for a system like the Shuttle, I think they can be enhanced with these kind of stipulations in a contract. They can't truly be bought that way. Certainly you [could] hang a larger incentive toward safety. You may enhance a strong focus on safety and I would not say that it wouldn't do some good to enlarge those enhancements. '

¶The more difficult question is whether existing NASA contracts, such as the SPC and the SRB Production Contact, strike an appro- priate balance between safety incentives and cost/schedule incen- tives. This question is particularly critical in light of reductions in

¶01aCrnt.e Hgs, Transcript,July 24, 1986, p. 135. 117 mid.

181

¶NASA's SR & &A programs detailed in Section VI.B.2.c. of this report.

¶Both Thiokol and NASA witnesses on June 17, 1986, argued that the penalties inherent in the Thiokol contract with Marshall Space Flight Center provided more than adequate incentives for Thiokol to deliver safe, reliable products. These penalties are of two types. Late-delivery penalties amount to $100-200 thousand per unit. Pen- alties for mission failures are much larger:

If findings of this Board of Investigation determines (sic) that the cause of the failure is attributed to the Solid Rocket Motor/Motors not performing in compliance with the specification requirements of the contract, a fee reduc- tion of $10,000,000 for each category I failure and $5,000,000 for each category I1 failure shall be deducted from any fee otherwise earned under this contract.218

¶Similarly, in briefings for Committee staff, NASA contract man- agers have stated that the award fee portion of the SPC, though small, is highly visible and that contractors take the award fee and the semiannual contract ratings very seriously. In NASA's view, high ratings enhance a company's reputation and, therefore, its likelihood of competing effectively for additional contracts.

¶Nevertheless, there are several reasons to believe that NASA could utilize contractual terms more effectively to enhance pro- gram safety. First, there can be no argument, for both the SPC and the SRB contract, that absent a major mission failure, virtually all the financial incentives are tilted toward cost-savings and timely delivery.

¶Secondly, because of the complex and overlapping division of re- sponsibilities between NASA and its contractors, it is not clear that contractors will be fully penalized even in cases where their actions or their hardware appear to be directly responsible for a mission failure. Mr. Scheuer's questioning of Mr. Charles Locke, Chairman of the Board of Morton Thiokol, showed that Thiokol i s not pre- pared to accept the full contractual penalties for the Challenger ac- cident.219

¶Finally, it is revealing that, under its NASA contract, Thiokol was never penalized for any of the numerous SRB flight anoma- lies.22oThe booster joint had never worked as intended, nor was its behavior at ignition ever clearly understood. Occurrences of O-ring erosion and/or blow-by exceeded twenty-five at the time of the Challenger accident. In fact, the rute of erosion/blow-by had in- creased steadily since the beginning of the SRB contract in 1983. The seal problem was serious enough to lead both to briefings at Headquarters and to establishment of a redesign task force. Yet, in spite of all these problems, Thiokol was eligible to receive a near-maximum incentive fee of approximately $75 million. But, in the final analysis, it was NASA that both approved the SRB design z:: NASA,

C& h F eI Schedule D! NAS8-30490, A & Transcript, June 1 9 d , 55-56. 17, 1983, ril 12, pp. p. 71. discussion focuses on possible contract penalties related to flight anomalies, it

¶is also interesting to note that Thiokol has never been penalbed for numerous safety and proc- ess violations at its Utah facilities. Several of these violations have resulted in serious fim and/ or explosions.

182

¶and drew up an SRB contract which contained no provisions for performance penalties or flight-anomaly penalties. One must not fault Thiokol for collecting the bonus; one must fault NASA for al- lowing the bonus to be collected at all.

¶The problem with the kinds of penalties that were contained in the SRB contract is that, so long as management is convinced that a festering problem like the seal problem is not likely to cause mis- sion failure, there is little incentive for the company to spend re- sources to fix the problem. In fact, if the solution involves signifi- cant delays in delivery, there may be a strong financial dis-incen- tive for the company to pursue a short-term solution aggressively. For example, Thiokol engineer R.M. Boisjoly provided a clear warn- ing of the seriousness of the O-ring problem in July, 1985, and Thiokol engineer A. R. Thompson laid out a plan for a possible short-term solution to the problem.221 Whatever its efficacy, why was Thompson's plan apparently dismissed so summarily? Part of the answer may be found in the June 18, 1986, exchange between Mr. Scheuer and Mr. Thompson:

Mr. SCHEUER. Would the research and development of your fixes have delayed the delivery of ths SRMs to

NASA?

Mr. THOMPSON. . . . It probably would have delayed it a month or two, at least for the hardware and some of the research work. . . .222

¶The Committee is certainly not suggesting that anyone in NASA or Thiokol would recommend launch or would refuse to spend re- sources fixing a problem if it was known that the problem consti- tuted a real threat to mission safety. However, in the case of the SRB joint, both NASA and Thiokol managers clearly misjudged the threat to mission safety. In situations of this sort, contractual pro- visions rewarding performance rather than cost and schedule would have provided a far stronger incentive to f m a long-festering problem. Ultimately, the balance between safety incentives and cost/schedule incentives in the SRB contract may illuminate a number of issues raised by the Challenger accident.

¶as1 Rugem Commission Report, Volume I, pp. 249-51.

¶Cmte Hgs, Transcript, June 18, 1986, p. 18.

CASING JOINT DESIGN

¶Discussion

¶(a)Introduction

¶The fact that the aft field joint of the right-hand Solid Rocket Booster failed at the 300 degree location is overwhelmingly sup- ported by the evidence. Retrieval of two large pieces of the joint clearly show that they were destroyed by the heat and velocity of the gas flame emanating from the right-hand booster. Additional supporting evidence was found by reviewing the telemetry data and the photographs taken during launch and flight.'

¶For the purpose of redesigning the joint it is important that the way in which the joint failed be determined as closely as possible. This determination, however, is difficult, if not impossible, to make with one hundred percent certainty. The evidence to support progress of the failure through the joint is incomplete. However, based on the recorded history of the joint problems encountered in flight and in test, based on the laws of physics, and based on behav- ior of the materials used in the joint, the following PROBABLE CAUSE is offered.

¶(b) Probable Cause of Failure

Both the primary O-ring and the secondary O-ring were seated when the steel casings were mated. The pressure check verified this fact. However, from experience, the primary O-ring was seated in the upstream position as had been previously recognized by NASA and Thiokol engineers. (See Figure VII-1.)

¶Rogers Cornmiasion Report, Volume I, pp. 22-23 and 78-79.

¶(183)

184

. PRIMARY EEATED

/ -RING IN BUT

UPSTREAM WRONG) POSITION

DOWNSTREAM SECONDARY

0-R ING h E A T E D

DOWNSTREAM (PROPER) POSIT O

¶IN

)I; FIGURE

¶VII-1

185

Upon ignition, the primary O-ring could not reseat at the 300 degree location in the downstream position, where it needed to be in time to prevent blowby. At this point there were too many defi- ciencies acting in unison which prevented the O-ring from reseat- ing in that location. First, proper spacing between the inner face of the tang and the opposing face of the inner leg of the clevis ap- proximately 0.020 inches, is critical. That spacing for Flight 51-L was too small, at the 300 degree location where the smoke was ob- served, to facilitate prompt reseating of the primary O-ring. Calcu- lations of segment diameters indicate the gap spacing was only 0.004 inches, near metal-to-metal contact. The ignition gases passed the O-ring at this location (See Figure VII-2). This condition did not exist elsewhere in the joint around the casings since the pri- mary O-ring was able to seat around the joint in other locations.

¶Second, the low temperature throughout the night prior to launch left the fluorocarbon elastomer primary O-ring stiff and lacking in ability to spring into the downstream (seated) position at the 300 degree location in time, relative to the buildup of motor pressure, to provide a tight seal. The temperature of the aft field joint at time of launch was calculated by Thiokol after the accident to be 16 degree F. Part of the reason for this low temperature was the heat transfer away from the joint, by conduction through the aft attachment strut. The conduction was driven by liquid hydro- gen, which remained in the external tank overnight. The supercold fuel created a 430 degree temperature differential across the ship, drawing heat out of the joint and O-rings.

¶A t ignition, blowby occurred, either with erosion of the primary O-ring or without erosion.

¶64-420 0 - 86 - 7

186

-SPAC I NG

¶"Too TIGHT"