(d) There is no indication from the heater voltage recording that the thermostatic switches functioned and cycled the heaters off and on during these special detanking procedures. (e) At the completion of detanking following CDDT, the switches are only checked to see that they remain closed at -75° F as the tank is warmed up. They are not checked to verify that they will open at +80° F. (f) Tests subsequent to the flight showed that the current associated with the KSC 65 V de ground powering of the heaters would cause the thermostatic switch contacts to weld closed if they attempted to interrupt this current. (g) A second test showed that without functioning thermostatic switches, temperatures in the 800° to 1000° F range would exist at locations on the heater tube assembly that were in close proximity with the motor wires. These temperatures are high enough to damage Teflon and melt solder. Determinations (1) Oxygen tank no. 2 (XTA 0008) did not detank after CDDT in a manner comparable to its performance the last time it had contained liquid oxygen, i.e., in acceptance test at Beech. (2) Such evidence indicates that the tank had undergone some change of internal configuration during the intervening events of the previous 3 years. (3) The tank conditions during the special detanking procedures were outside all prior testing of Apollo CSM cryogenic oxygen storage tanks. Heater assembly temperatures measured in subsequent tests exceeded 1000° F. (4) Severe damage to the insulation of electrical wiring internal to the tank, as determined from subsequent tests, resulted from the special procedure. (5) Damage to the insulation, particularly on the long unsupported lengths of wiring, may also have occurred due to boiling associated with this procedure. (6) MSC, KSC, and NR personnel did not know that the thermostatic switches were not rated to open with 65 V de GSE power applied. 15. Findings (a) The change in detanking procedures on the cryogenic oxygen tank was made in accordance with the existing change control system during final launch preparations for Apollo 13. (b) Launch operations personnel who made the change did not have a detailed understanding of the tank internal components, or the tank history. They made appropriate contacts before making the change. (c) Communications, primarily by telephone, among MSC, KSC, NR, and Beech personnel during final launch preparations regarding the cryogenic oxygen system included incomplete and inaccurate information. (d) The MSC Test Specification Criteria Document (TSCD) which was used by KSC in preparing detailed tank test procedures states the tank allowable heater voltage and current as 65 to 85 V de and 9 to 17 amperes with no restrictions on time. Determinations (1) NR and MSC personnel who prepared the TSCD did not know that the tank heater themostatic switches would not protect the tank. (2) Launch operations personnel assumed the tank was protected from overheating by the switches. (3) Launch operations personnel at KSC stayed within the specified tank heater voltage and current limits during the detanking at KSC. 16. Findings (a) After receipt of the Block II oxygen tank specifications from NR. which required the tank heater assembly to operate with 65 V de GSE power only during tank pressurization, Beech Aircraft did not require their Block I thermostatic switch supplier to make a change in the switch to operate at the higher voltage. (b) NR did not review the tank or heater to assure compatibility between the switch and the GSE. (c) MSC did not review the tank or heater to assure compatibility between the switch and the GSE. (d) No tests were specified by MSC, NR, or Beech to check this switch under load. Determinations (1) NR and Beech specifications governing the powering and the thermostatic switch protection of the heater assemblies were inadequate. (2) The specifications governing the testing of the heater assemblies were inadequate. 17. Finding The hazard associated with the long heater cycle during detanking was not given consideration in the decision to fly oxygen tank no. 2. Determinations (1) MSC, KSC, and NR personnel did not know that the tank heater thermostatic switches did not protect the tank from overheating. (2) If the long period of continuous heater operation with failed thermostatic switches had been known, the tank would have been replaced. 18. Findings (a) Management controls requiring detailed reviews and approvals of design, manufacturing processes, assembly procedures, test procedures, hardware acceptance, safety, reliability, and flight readiness are in effect for all Apollo hardware and operations. (b) When the Apollo 13 cryogenic oxygen system was originally designed, the management controls were not defined in as great detail as they are now. Determination From review of documents and interviews, it appears that the management controls existing at that time were adhered to in the case of the cryogenic oxygen system incorporated in Apollo 13. 19. Finding The only oxygen tank no. 2 anomaly during the final countdown was a small leak through the vent quick disconnect, which was corrected. Determination No indications of a potential inflight malfunction of the oxygen tank no. 2 were present during the launch countdown. 20. Findings MISSION EVENTS THROUGH ACCIDENT (a) The center engine of the S-II stage of the Saturn V launch vehicle prematurely shut down at 132 seconds due to large 16 hertz oscillations in thrust chamber pressure. (b) Data indicated less than 0.1g vibration in the CM. Determinations (1) Investigation of this S-II anomaly was not within the purview of the Board except insofar as it relates to the Apollo 13 accident. (2) The resulting oscillations or vibration of the space vehicle probably did not affect the oxygen tank. 21. Findings (a) Fuel cell current increased between 46:40:05 and 46:40:08 indicating that the oxygen tank no. 1 and tank no. 2 fans were turned on during this interval. (b) The oxygen tank no. 2 quantity indicated off-scale high at 46:40:08. Determinations (1) The oxygen tank no. 2 quantity probe short circuited at 46:40:08. (2) The short circuit could have been caused by either a completely loose fill tube part or a solder splash being carried by the moving fluid into contact with both elements of the probe capacitor. 22. Findings (a) The crew acknowledged Mission Control's request to turn on the tank fans at 55:53:06. (b) Spacecraft current increased by 1 ampere at 55:53:19. (c) The oxygen tank no. 1 pressure decreased 8 psi at 55:53:19 due to normal destratification. Determination The fans in oxygen tank no. 1 were turned on and began rotating at 55:53:19. 23. Findings (a) Spacecraft current increased by 12 amperes and ac bus 2 voltage decreased 0.6 volt at 55:53:20. (b) Stabilization and Control System (SCS) gimbal command telemetry channels, which are sensitive indicators of electrical transients associated with switching on or off of certain spacecraft electrical loads, showed a negative initial transient during oxygen tank no. 2 fan turnon cycles and a positive initial transient during oxygen tank no. 2 fan turnoff cycles during the Apollo 13 mission. A negative initial transient was measured in the SCS at 55:53:20. (c) The oxygen tank no. 2 pressure decreased about 4 psi when the fans were turned on at 55 :53:21. Determinations (1) The fans in oxygen tank no. 2 were turned on at 55:53:20. (2) It cannot be determined whether or not they were rotating because the pressure decrease was too small to conclusively show destratification. It is likely that they were. 24. Finding An 11.1-amp spike in fuel cell 3 current and a momentary 1.2-volt decrease were measured in ac bus 2 at 55:53:23. Determinations (1) A short circuit occurred in the circuits of the fans in oxygen tank no. 2 which resulted in either blown fuses or opened wiring, and one fan ceased to function. (2) The short circuit probably dissipated an energy in excess of 10 joules which, as shown in subsequent tests, is more than sufficient to ignite Teflon wire insulation by means of an electric arc. 25. Findings (a) A momentary 11-volt decrease in ac bus 2 voltage was measured at 55:53:38. (b) A 22.9-amp spike in fuel cell 3 current was measured at 55:53:41. (c) After the electrical transients, CM current and ac bus 2 voltage returned to the values indicated prior to the turnon of the fans in oxygen tank no. 2. Determination Two short circuits occurred in the oxygen tank no. 2 fan circuits between 55:53:38 and 55:53:41 which resulted in either blown fuses or opened wiring, and the second fan ceased to function. 26. Finding Oxygen tank no. 2 telemetry showed a pressure rise from 887 to 954 psia between 55:53:36 and 55:54:00. It then remained nearly constant for about 15 seconds and then rose again from 954 to 1008 psia, beginning at 55:54:15 and ending at 55:54:45. Determinations (1) An abnormal pressure rise occurred in oxygen tank no. 2. (2) Since no other known energy source in the tank could produce this pressure buildup, it is concluded to have resulted from combustion initiated by the first short circuit which started a wire insulation fire in the tank. 27. Findings (a) The pressure relief valve was designed to be fully open at about 1000 psi. (b) Oxygen tank no. 2 telemetry showed a pressure drop from 1008 psia at 55:54: 45 to 996 psia at 55:54:53, at which time telemetry data were lost. Determination This drop resulted from the normal operation of the pressure relief valve as verified in subsequent tests. 28. Findings (a) At 55: 54: 29, when the pressure in oxygen tank no. 2 exceeded the master caution and warning trip level of 975 psia, the CM master alarm was inhibited by the fact that a warning of low hydrogen pressure was already in effect, and neither the crew nor Mission Control was alerted to the pressure rise. (b) The master caution and warning system logic for the cryogenic system is such that an out-of-tolerance condition of one measurement which triggers a master alarm prevents another master alarm from being generated when any other parameter in the same system becomes out-of-tolerance. (c) The low-pressure trip level of the master caution and warning system for the cryogenic storage system is only 1 psi below the specified lower limit of the pressure switch which controls the tank heaters. A small imbalance in hydrogen tank pressures or a shift in transducer or switch calibration can cause the master caution and warning to be triggered preceding each heater cycle. This occurred several times on Apollo 13. (d) A limit sense light indicating abnormal oxygen tank no. 2 pressure should have come on in Mission Control about 30 seconds before oxygen tank no. 2 failed. There is no way to ascertain that the light did, in fact, come on. If it did come on, Mission Control did not observe it. Determinations (1) If the pressure switch setting and master caution and warning trip levels were separated by a greater pressure differential, there would be less likelihood of unnecessary master alarms. (2) With the present master caution and warning system, a spacecraft problem can go unnoticed because of the presence of a previous out-of-tolerance condition in the same subsystem. (3) Although a master alarm at 55: 54: 29 or observance of a limit sense light in Mission Control could have alerted the crew or Mission Control in sufficient time to detect the pressure rise in oxygen tank no. 2, no action could have been taken at that time to prevent the tank failure. However, the information could have been helpful to Mission Control and the crew in diagnosis of spacecraft malfunctions. (4) The limit sense system in Mission Control can be modified to constitute a more positive backup warning system. 29. Finding Oxygen tank no. 2 telemetry showed a temperature rise of 38° F beginning at 55:54:31 sensed by a single sensor which measured local temperature. This sensor indicated off-scale low at 55:54:53. Determinations (1) An abnormal and sudden temperature rise occurred in oxygen tank no. 2 at approximately 55: 54: 31. (2) The temperature was a local value which rose when combustion had progressed to the vicinity of the sensor. (3) The temperature sensor failed at 55:54:53. 30. Finding Oxygen tank no. 2 telemetry indicated the following changes: (1) quantity decreased from off-scale high to off-scale low in 2 seconds at 55: 54:30, (2) quantity increased to 75.3 percent at 55: 54: 32, and (3) quantity was off-scale high at 55:54:51 and later became erratic. Determinations (1) Oxygen tank no. 2 quantity data between 55:54:32 and 55:54:50 may represent valid measurements. (2) Immediately preceding and following this time period, the indications were caused by electrical faults. 31. Findings (a) At about 55: 54: 53, or about half a second before telemetry loss, the bodymounted linear accelerometers in the command module. which are sampled at 100 times per second, began indicating spacecraft motions. These disturbances were erratic, but reached peak values of 1.17g, 0.65g, and 0.65g in the X, Y, and Z directions, respectively, about 13 milliseconds before data loss. (b) The body-mounted roll, pitch, and yaw rate gyros showed low-level activity for 4 second beginning at 55: 54: 53.220. (c) The integrating accelerometers indicated that a velocity increment of approximately 0.5 fps was imparted to the spacecraft between 55:54:53 and 55:54:55. (d) Doppler tracking data measured an incremental velocity component of 0.26 fps along a line from the Earth to the spacecraft at approximately 55:54: 55. (e) The crew heard a loud "bang" at about this time. (f) Telemetry data were lost between approximately 55: 54: 53 and 55: 54: 55 and the spacecraft switched from the narrow-beam antenna to the wide-beam antenna. (g) Crew observations and photographs showed the bay 4 panel to be missing and the high-gain antenna to be damaged. Determinations (1) The spacecraft was subjected to abnormal forces at approximately 55:54:53. These disturbances were reactions resulting from failure and venting of the oxygen tank no. 2 system and subsequent separation and ejection of the bay 4 panel. (2) The high-gain antenna was damaged either by the panel or a section thereof from bay 4 at the time of panel separation. 32. Finding Temperature sensors in bay 3, bay 4, and the central column of the SM indicated abnormal increases following reacquisition of data at 55: 54:55. Determination Heating took place in the SM at approximately the time of panel separation. 33. Findings (a) The telemetered nitrogen pressure in fuel cell 1 was off-scale low at reacquisition of data at 55: 54: 55. (b) Fuel cell 1 continued to operate for about 3 minutes past this time. (c) The wiring to the nitrogen sensor passes along the top of the shelf which supports the fuel cells immediately above the oxygen tanks. Determinations (1) The nitrogen pressure sensor in fuel cell 1 or its wiring failed at the time of the accident. (2) The failure was probably caused by physical damage to the sensor wiring or shock. (3) This is the only known instrumentation failure outside the oxygen system at that time. 34. Finding Oxygen tank no. 1 pressure decreased rapidly from 879 psia to 782 psia at approximately 55:54:54 and then began to decrease more slowly at 55:54:56. Determination A leak caused loss of oxygen from tank no. 1 beginning at approximately 55:54:54. 35. Findings (a) Oxygen flow rates to fuel cells 1 and 3 decreased in a 5-second period be ginning at 55: 54: 55, but sufficient volume existed in lines feeding the fuel cells to allow them to operate about 3 minutes after the oxygen valves were cut off. (b) The crew reported at 55:57:44 that five valves in the reaction control system (RCS) were closed. The shock required to close the oxygen supply valves is of the same order of magnitude as the shock required to close the RCS valves. (c) Fuel cells 1 and 3 failed at about 55:58. Determination The oxygen supply valves to fuel cells 1 and 3, and the five RCS valves, were probably closed by the shock of tank failure or panel ejection or both. 36. Findings MISSION EVENTS AFTER ACCIDENT (a) Since data presented to flight controllers in Mission Control are updated only once per second, the 1.8-second loss of data which occurred in Mission Control was not directly noticed. However, the Guidance Officer did note and report a "hardware restart" of the spacecraft computer. This was quickly followed by the crew's report of a problem. 47-476 0-70-2 |