STACKING OPERATIONS

¶Issue

¶Was there any damage to the casing joints or contamination that occurred during the stacking operations when the Shuttle was as- sembled in the Vehicle Assembly Building (VAB) that could have contributed to the failure? Finding

¶There was no evidence of joint contamination, fracture, or other damage from foreign objects or due to casing ovality that contribut- ed to the joint failure. Although certain problems occurred during stacking and the procedures were violated once, there was no evi- dence that these events contributed to the Flight 51-L accident. Discussion

¶The discussion of the assembly of the aft field joint on the right hand Solid Rocket Booster is drawn from the "STS 51-L SRB Joint Mate Review Team" report 33a The report was provided to Commit- tee staff during the Committee's trip to KSC on June 6, 1986.

¶There were 24 Solid Rocket Booster sets (48 SRBs) stacked prior to STS 51-L. The stacking experience of the technicians involved in STS 51-L ranged from 5 to 20 stacking operations. Sixty percent of the technicans and all of the supervisory personnel, including lead technicians, had participated in the 14 stacking operations per- formed since the Shuttle processing contract was awarded to Lock- heed. Thiokol managed the stacking operations for Lockheed under a subcontract. The NASA Accident Review Team found that all personnel assigned to the stacking of STS 51-L were experienced and qualified to perform their assigned tasks.

¶Aft segment receiving inspection and processing in the Rotation, Processing and Surge Facility (RPSF) was normal. No problems were reported relative to the aft segment clevis during offload from the railcar, mate to the aft skirt, aft booster assembly, or in prepa- ration for transfer to the Vehicle Assembly Building (VAB) for stacking. Some surface defects were identified on non-sealing sur- faces of the aft segment clevis, but were found not to exceed the specification in the Operations and Maintenance Requirements Specification (OMRS) document. There were no defects identified in the clevis O-ring grooves. The aft segment was processed normally in the RPSF to prepare for stack.

NASA, J. Robert Lang, KSC, "STS 51-L SRB Joint Mate Review Team Report," March 18, 1986.

¶A problem was reported at the 165-168 degree location where a segment case-to-insulation bondline separation 0.109 inch in depth (longitudinally) was found. The OMRS document specifies no sepa- rations in excess of 0.050 inch. A Material Review Board (MRB) repair was approved and the separation was filled with an asbestos float-filled, liquid epoxy resin sealant. This repair is standard for this type of separation, and has been performed on numerous seg- ments. Some surface defects were identified on the tang, but none were found to exceed specification.

¶The right aft booster assembly was transferred from the RPSF directly to the VAB transfer aisle. Positioning of the aft booster as- sembly on the Mobile Launch Platform (MLP) holddown posts was normal. One iteration of shimming was performed and the subse- quent holddown post strain gauge output indicated proper distribu- tion of aft booster assembly loads.

¶Holddown hardware was installed and stud tensioning began with ultrasonic measurement of stud initial lengths. A problem was reported a t holddown post # 1 when ultrasonic measurements indicated a stud length twice the actual. The stud and associated hardware a t post #1 were removed for offline bench testing. The problem was isolated to a faulty ultrasonic transducer. While awaiting replacement hardware, studs at holddown posts # 3 and # 4 were tensioned satisfactorily. The stud at holddown post #2 was tensioned but adequate margin was not attained.

¶This problem in tensioning studs at holddown posts # 1 and # 2 led to a revision in the schedule. All left hand Solid Rocket Motor segments were stacked while problems on the right hand side were resolved. This procedure had been employed in one-third of the pre- vious stacking operations and was not an uncommon method of stacking.

¶After installation of the replacement hardware at holddown post #1, studs a t posts #1 and # Z were tensioned. The replacement stud at post #1 was brought up to satisfactory tension, but con- cerns over stud tension at post #2 prompted engineering to re- quest that a problem report be generated.

¶Engineering determined the tension (approximately 690,000 lbs.) was adequate for SRB stacking, but marginal for launch loads. Therefore, stud removal and replacement was planned after SRB stack but prior to Orbiter mate.

¶While holddown post stud tensioning proceeded, preparation and inspection of the aft segment clevis was put in work. No problems were identified on the aft segment clevis during this inspection. Since a stacking delay was evident, the clevis was secured and sealed to maintain inspection integrity until stacking could resume.

¶A Solid Rocket Motor configuration change was released as a result of a handling incident. The SRM-25 left forward center seg- ment was damaged during processing in the RPSF.34 Deviation A p proval Request (DAR) Number RWW-376R1 was approved to re- place the damaged segment with a left forward center segment for SRM-26 motor set. In order to prevent flight performance imbal- ance, the right aft center segment was also reassigned from SRM26 to the SRM-25 flight motor set. The SRM-26 right aft center segment was transported to the VAB for stacking.

¶While this segment was outside the VAB, a storm occurred. A Problem Report was generated as rain water was reported leaking from under the segment protective covers. The segment was brought into the VAB transfer aisle and hoisted for pre-stack in- spection. At this time, all visible moisture was removed from the aft surfaces of the segment. Inspections were performed and no problems were identified as a result of rain water intrusion. The MRB repair of the separation of insulation from the segment case located at 165/ 168 degrees was reinspected and final acceptance was verified. A complete inspection of tang surfaces and aft insula- tion surfaces was performed and no problems were identified.

¶The use of this segment violated assembly procedure which re- quires that the segments be protected from direct water entry and it should not have been employed. While there is no evidence of a direct connection to the joint failure, the decision to use this aft center segment was a compromise that need not have been made.

¶The aft segment clevis diameter was measured at six locations and corresponding measurements were taken of the aft center seg- ment tang. Measurements indicated that a potential for interfer- ence existed along the 0/180 and 30/210 degree axes where the tang diameter was larger than that of the clevis. The normal proce- dure for changing the shape (ovality) of the tang was initiated. The procedure calls for reconfiguration of the segment lifting beam from a four-point to a two-point lift configuration to decrease the tang outside diameter along the axis of interference. The procedure was followed and after stabilization, a decrease in tang diameter of 0.178 inch was measured along the axis of potential interference. Shuttle Processing Contractor (SPC) engineering was called on to evaluate the latest overall characteristics of the joint. At that time the aft center tang was larger in diameter than the aft segment clevis by more than 0.31 inches along the two axes, 0/180 and 30/ 210.

¶These measurements still indicated a potential for interference based upon normal KSC experience. SPC engineering determined that additional deflection of the aft center segment case was neces- sary and prescribed installation of the SRM Circumferential Align- ment Tool along the 16/196 degree axis of the tang. The Circumfer- ential Alignment Tool was installed and maximum hydraulic pres- sure was applied (1200 psig), producing a deflection of 0.196 inch. Later an unspecified torque on the Circumferential Alignment Tool tension rod nut produced an additional deflection of 0.040 inch. This additional torque caused an additional load and exceeded the safe working limit of the tool. Technicians noticed an increase in hydraulic pressure on the pumping unit gauge to 1300 psig at the time torque was applied. This pressure indicates a force of up to 3250 pounds may have been applied to the segment case. Currently, a force of 5000 pounds may be applied to the segment case. The safety limits of the Circumferential Alignment Tool were exceeded (safety factor reduced to 1.2), but the force applied to the segment case was still well below the established maximum. However, the procedure was determined to be inappropriate by the post-accident investigation.

¶The alignment tool used on the aft center segment of STS 51-L's right hand Solid Rocket Motor is now considered inappropriate by NASA due to the concentrated loads applied at two points. A new alignment tool is now being designed. However, the use of this tool did not appear to have contributed to the STS 51-L's accident.

¶With the Circumferential Alignment Tool installed, the right aft center segment was hoisted from the transfer aisle and positioned above the aft segment in the VAB High Bay. Installation of pri- mary and secondary O-rings was performed, and no problems were identified. Closeout photographs were then taken showing the 0- ring and zinc chromate putty installation.

¶The joint mating operation proceeded with final inspection of the greased joint surfaces. No problems were identified during engage- ment of the tang into the clevis, aided by the nearly co-planar rela- tionship of the mating surfaces (within 0.15 inch). The joint mate was completed with installation of all clevis pins and pin retainer clips per the normal procedure. No difficulties were encountered.

¶After disconnection of the segment lifting beam, the SRM field joint leak test was performed. Following the 200 psig pressuriza- tion, the 50 psig decay test was performed and zero pressure decay was recorded, indicating successful assembly of the joint (maximum allowable decay is 1.0 psig over a 10 minute period).

¶Field joint closeouts were performed in the normal fashion. No problems were reported during pin retainer band and cork insula- tor installation. Data also indicated normal application of the bead of grease around the seam of the joint. Installation of the systems tunnel floor splice plate across the field joints at the 90 degree loca- tion completed the closeout.

¶Because of its unique design, the clevis of the aft case must always be used as the field joint at the forward end of the aft seg- ment. It was previously flown on the left booster segment on STS 51-C. It was also utilized in qualification test motor QM-4 which was static test fired a t Thiokol's Utah plant.

¶The field joint tang of the STS 51-L aft center segment (serial number L60 had flown previously as forward center segment to aft center segment field joint tang on the left booster on STS 41-D. In a memo to J. Harrington of NASA's Data and Design Analysis Task Force on February 24, 1986, the Chairman of the SRB Joint Mate Review Team noted the conclusion that the 200 psig O-ring seating operation could produce a blowhole in the putty. Such a blowhole would not be known prior to launch. Since the putty is intended to provide a heat shield to protect the O-rings, the O-rings would be unprotected in cases where blowholes occurred.

SUMMARY OF LAUNCH OPERATIONS

¶Issue 1

¶How was the decision to launch STS 51-L arrived at and why was it wrong? Findings

The Flight Readiness Review for STS 51-L was conducted in accordance with established procedure.

The decision to launch STS 51-L was based on a faulty engi- neering analysis of the SRM field joint seal behavior.
Compounding this erroneous analysis were serious ongoing weaknesses in the Shuttle Safety, Reliability, and Quality Assur- ance program which had failed to exercise control over the problem tracking systems, had not critiqued the engineering analysis ad- vanced as an explanation of the SRM seal problem, and did not provide the independent perspective required by senior NASA managers at Flight Readiness Reviews.
The initial response of Marshall managers to the attempts of Thiokol engineers to raise the issue of temperature effects on the SRM seals caused Thiokol management to discount proper techni- cal concerns and engineering judgement in their recommendation to launch.