Reports/Apollo 17/Saturn V flight evaluation/3 Launch Operations

This page is still under development.

[edit] 3.1 Summary

The ground systems supporting the AS-512/Apollo 17 countdown and launch performed satisfactorily with the exception of the Terminal Countdown Sequencer (TCS). The TCS malfunction, which is discussed in paragraph 3.3, resulted in a 2 hour and 40 minute launch delay. The space vehicle was launched at 00:33:00 Eastern Standard Time (EST) (05:33:00 UT) on December 7, 1972, from Pad 39A of the Kennedy Space Center, Saturn Complex. Damage to the pad, Launch Umbilical Tower (LUT) and support equipment was considered minimal.

[edit] 3.2 Prelaunch Milestones

A chronological summary of prelaunch milestones for the AS-512 launch is contained in Table 3-1.

Table 3-1. AS-512/Apollo 17 Prelaunch Milestones

DATE	ACTIVITY OR EVENT
October 27, 1970	S-11-12 Stage Arrival
December 21, 1970	S-1VB-512 Stage Arrival
June 16, 1971	Lunar Module (LM)-12 Ascent Stage Arrival
June 17, 1971	Module (LM)-12 Descent Stage Arrival
March 24, 1972	Spacecraft/Lunar Module Adapter (SLA)-21 Arrival
March 24, 1972	Command and Service Module (CSM)-114 Arrival
May 11, 1972	S-IC-12 Stage Arrival
May 15. 1972	S-IC Erection on Mobile Launcher (ML)-3
May 19. 1972	S-II Erection
June 2, 1972	Lunar Roving Vehicle (LRV)-3 Arrival
June 7, 1972	Instrument Unit (10-512 Arrival
June 20, 1972	TO Erection
June 23, 1972	S-IVB Erection
July 12. 1972	Launch Vehicle (LV) Electrical Systems Test Completed
August 1, 1972	LV Propellant Dispersion/Malfunction Overall Test (OAT) Complete
August 1972	LV Service Arm OAT Complete
August 13, 1972	LRV Installation
August 23, 1972	Spacecraft (SC) Erection
August 28, 1972	Space Vehicle (SV)/ML Transfer to Pad 39A
October 11, 1972	SV Electrical Mate
October 12, 1972	SV OAT No. 1 (Plugs In) Complete
October 20, 1972	SV Flight Readiness Test (FRT) Completed
November 10, 1972	RP-1 Loading
November 21, 1972	Countdown Demonstration Test (CDDT) Completed (Wet)
November 21, 1972	CDDT Completed (Dry)
December 5, 1972	SV Terminal Countdown Started (T-28 Hours)
December 7, 1972 (EST)	SV Launch

[edit] 3.2.1 S-IC Stage

S-IC stage and GSE systems performed satisfactorily during countdown with the exception of three failures which were subsequently corrected.

The failures were in the (1) Safe and Arm Devices (S&A). (2) Remote Digital Sub-Multiplexer, and (3) F-1 Engine No. 2 Gas Generator Igniter.

(1) The Safe and Arm Device failed to respond to a safe command. Possible causes for the failure were determined to be low voltage, improper installation, or a defective unit. The Safe and Arm Device and its mounting block were replaced and the replacement unit performed satisfactorily. Bench tests of the suspect unit failed to duplicate the problem and dimensional analysis of the unit and mounting block was satisfactory. Analysis did reveal, however, that output torque of the solenoid at the lower end of the voltage curve was marginal with respect to the torque requirements of the mechanical linkage of the S&A device. As a precautionary measure, the countdown procedure was changed to arm the device at T-33 minutes instead of T-5 minutes to eliminate the need for recycling to 1-22 minutes in the event of a hold. In addition, the provision was made to increase the stage bus voltage to 30V if the unit should fail to arm during the count.

(2) At the T-9 hour scheduled hold the Remote Digital Sub-Multiplexer (RDSM) failed and an 8 ampere current surge of one minute duration was recorded. The RDSM was replaced and satisfactorily retested. The cause was isolated to shorted ceramic capacitor (C7) in the power supply card. As a result of failure analysis it was concluded that the failure was random and no corrective action is anticipated.

(3) The F-1 Engine No. 2 Gas Generator (GG) igniter installed indication was lost at T-23 hours. Both GG igniters on Engine No. 2 were replaced and the problem was determined to be due to igniter failure. Failure analysis revealed an error in manufacture in that solder had been omitted from an electrical pin in the igniter, allowing intermittent contact. The lack of solder was seen in the X-ray picture which is made during receiving inspection. Corrective action taken was to review all remaining igniter X-ray pictures to assure no more omissions exist.

[edit] 3.2.2 S-II Stage

The S-II stage and GSE performed satisfactorily during the countdown. As a result of the unscheduled hold caused by the Terminal Countdown Sequencer (TCS) malfunction, some systems such as the J-2 engine start tank system were required to remain active.

During the first unscheduled hold at 02:52:30 UT (T-30 seconds), S-II stage systems were safed and recycled successfully during this 65.2 minute hold duration. At 03:57:41 UT (T-22 minutes), the countdown was resumed and continued to T-8 minutes when another hold occurred to resolve the TCS corrective action. This hold lasted 73.3 minutes and contingency hold Option 2 was utilized. S-II systems remaining active through this hold were LOX system helium injection, engine actuation hydraulic system temperature control, and engine helium and hydrogen start tanks pressurized. It was necessary to manually control engine helium tank venting as temperature changes dictated. The engine start tanks were chilled, pressurized, and then required one rechill cycle at 05:12:00 UT for proper temperature conditions. At 05:25:00 UT, the countdown resumed at T-8 minutes and proceeded without further problems to liftoff. Electrical batteries on the S-II stage were on internal power about 20 seconds longer than previous vehicles and were slightly more discharged at liftoff as a result of the repeated countdown.

[edit] 3.2.3 S-IVB Stage

Overall performance of the S-IVB stage and GSE was satisfactory during the countdown operations.

A hazardous gas detection sensor located at the LH₂ tank vent disconnect on Swing Arm No. 7, showed an intermittent indication of GH₂ for approximately 1-1/2 hours from T-3 hours 30 minutes. The leak was not large enough to cause a problem and was dispositioned acceptable for launch.

To keep the engine control helium sphere pressure below the redline limit of 3400 psia, the sphere was vented six times using the emergency vent during the hold period.

Prior to resuming the countdown at T-8 minutes, the start tank was rechilled to bring the temperature below the maximum limit acceptable for launch. After rechilling, the start tank emergency vent valve was cycled three times to keep the start tank pressure below the maximum limit.

A long term decay was noted on Forward Battery No. 2, open circuit voltage. The open circuit voltage at the time of installation was 34.74 V. The voltage decayed 1.50 V over a 24-hour period. During the hold at T-9 hours, a power transfer test was performed to verify battery performance under loaded conditions. Battery performance was normal. At T-8 hours 53 minutes, Battery Monitor Enable was turned on to provide a small load in order to stabilize the battery. The battery voltage stabilized at T-4 hours. The voltage decay was attributed to a greater than nominal silver-peroxide level in the battery cells. The battery met all specifications and criteria.

[edit] 3.2.4 IU Stage

The IU stage performed satisfactorily during the countdown.

[edit] 3.3 Terminal Countdown

The AS-512/Apollo 17 Terminal Countdown was picked up at T-38 hours on December 5, 1972. Scheduled holds were initiated at T-9 hours for a duration of 9 hours, and at T-3 hours 30 minutes for a duration of one hour.

At T-167 seconds the Terminal Countdown Sequencer (TCS) failed to issue the "S-IVB LOX Tank Pressurization" command. When it was visually observed that the S-IVB LOX Tank was not being pressurized, the console operator initiated action to manually control S-IVB LOX Tank Pressurization. The tank was pressurized, but because an interlock relay was not energized when the TCS failed to issue the T-167 second command, a countdown hold was experienced at T-30 seconds. This hold lasted for 2 hours and 40 minutes during which time the TCS failure was confirmed, a "Work-Around" was investigated, and the "Work-Around" was verified at the MSFC Saturn V System Development Facility (SDF). Also during this hold the countdown was recycled to T-22 minutes. After investigation of the failure and verification of the "Work-Around" it was concluded that the countdown could be successfully and safely accomplished by using a jumper to bypass the "S-IVB LOX Tank Pressurized" interlock relay and manually pressurizing the LOX tank from the LCC. The countdown sequence was restarted at T-22 minutes and completed successfully.

Figure 3-1. Electrical Support Equipment Partial Schematic

Figure 3-1 shows the electrical circuits associated with this anomaly and the following is a description of the functional operation of the circuits.

The T-167 second command from the TCS (Channel 3) is supplied to the Mobile Launcher (ML) Integration Patch Distributor to energize relay K3 which supplies a 28V signal to the ML S-IVB Patch distributor. This signal is used to initiate 1) S-IVB LOX tank vent closed, 2) S-IVB LOX tank pressurization valve open, and 3) energize relay K577 "Time for LOX Tank Pressurization." Without relay K577 energized the "S-IVB LOX Tank Pressurized" interlock relay K536 cannot be energized even if relay K492 "LOX Tank Minimum Low Pressure OK" is energized by manually pressurizing the LOX tank. When K536 is not energized the "S-IVB Ready for Launch" relay K607 will not provide a signal to the ML S-IC Patch Distributor "S-IVB Ready for Launch" relay K972 to complete the interlock chain to allow relay K465 "Swing Arm No. 1 Retract Preparation Complete" to be energized. If K465 is not energized when the T-30 second TCS command (Swing Arm No. 1 Carrier Retract) is received, a cutoff command will be initiated and a countdown hold will occur.

When the above condition occurred, the absence of the TCS T-167 second command was confirmed on the Digital Events Evaluator-6 (DEE-6) printout. Investigation of the DEE-6 printout disclosed that the T-176 second spare output from the TCS also did not occur. After investigation of various combinations of lost outputs and associated fixes, it was determined that the "LOX Tank Pressurized" relay K536 could be bypassed by moving the "LOX Tank Pressurized Bypass" jumper from "INHIBIT" to "ON" position. This jumper is located on S-IVB Patch Distributor in the LCC. The failure was simulated and the "Work-Around" was verified at the MSFC Saturn V SDF and a decision was made to proceed with the launch using the interlock bypass and manual pressurization. During the successful launch all TCS outputs were obtained except the T-176 second spare output. Therefore, the bypass and manual pressurization procedures were actually redundant to the normal circuitry.

Figure 3-2. Diode Chip Detail

Investigation of this failure at KSC subsequently centered on two diodes located in the logic circuitry of the TCS. One of these diodes inhibited the T-167 second S-IVB LOX Tank Pressurization command and the other inhibited the spare output. The two failures are functionally unrelated in the TCS circuitry. Excessive reverse current leakage through the partially shorted diodes caused intermittent operation of TCS outputs. The two failed diodes had been in service six years. Each TCS contains 1,827 of these diodes with approximately 1500 of these capable of causing a launch hold or scrub if they failed between CDDT and launch.

Testing of all similar diodes is being conducted where feasible. Of 2196 diodes tested, 7 additional diodes exhibited reverse current leakage in excess of the specification. The diodes that failed along with a number of non-failed diodes from the same printed circuit boards were subjected to extensive analysis. The following four causes of failure have been postulated: 1) inversion layer formation, 2) accumulation layer formation, 3) metallic precipitates in the depletion layer or 4) contamination in cracks partially or completely across the depletion layer.

Since deposition of contamination in microscopic cracks (Figure 3-2) was consistently observed in the failed diodes, this is considered to be the most probable failure mode. However, the investigation as to the cause of the cracks and subsequent contamination deposition is still underway and cannot be considered conclusive at this time.

The "Work-Around" with the TCS at KSC that resulted in a satisfactory terminal countdown would not be acceptable if a problem occurred with the TCS during the Skylab-2, -3, and -4 countdowns due to the short launch windows.

The following activities will be accomplished prior to the Skylab launches in order to eliminate the possibility of another failure.

a. The diodes will be tested and replaced as required in each of the existing TCS's to assure reliable performance.

b. Pad 39A and Pad 39B will be modified to provide three TCS's in each launch vehicle ESE rather than the present one.

c. Incorporate voting logic so that any two of the three TCS's will assure that the proper signals are provided.

d. All unused signals from each TCS will be unpatched and grounded so there will be no possibility of them causing problems.

The above activities will reduce the probability of a false command being initiated and also assure that no single electrical failure will result in loss of the proper terminal countdown command.

[edit] 3.4 Propellant Loading

[edit] 3.4.1 RP-1 Loading

The RP-1 system successfully supported countdown and launch without incident. Tail Service Mast (TSM) 1-2 fill and replenish was accomplished at T-13 hours and S-IC level adjust and fill line inert occurred at about T-60 minutes. Both operations were satisfactory, there were no failures or anomalies. Launch countdown support consumed 213,304 gallons of RP-1.

[edit] 3.4.2 LOX Loading

The LOX system supported countdown and launch satisfactorily. The fill sequence began with S-IVB fill command at 12:34 EST, December 6, 1972, and was completed 2 hours 40 minutes later with all stage replenish normal at 15:15 EST. Replenishment was automatic through the first Terminal Countdown Sequence but was switched to manual when S-IVB flight mass began cycling shortly before final countdown. This condition has been experienced during some previous loading operations and is a result of trapped LOX warming in the S-IVB inlet line. The LH₂/LOX Auto Load allows for manual replenishment when such cycling occurs.

When LOX loading was reinitiated shortly before recycling to T-22 minutes, LOX system logic did not reestablish replenish operations as expected. Instead, it sequenced into a dual mode configuring simultaneously for both "vehicle replenishment" and "S-IC chilldown." In this posture, the S-IC slow fill valve was opened allowing LOX to be pumped directly into the stage resulting in a slight overfill. The system was manually reverted to prevent further overfill. Subsequent investigation revealed that an S-IC discrete necessary few normal replenishment was missing when loading operations were resumed.

A real time procedure charge to LOX/LH₂ auto load, was prepared to initiate the discrete manually. Replenishment operations were reinitiated and continued normally through launch. This procedure change, which requires manual issue of Propellant Tanking Computer System (PTCS) discretes if tank level is at or above 98%, will prevent problem recurrence.

LOX consumption during launch countdown was 618,000 gallons.

[edit] 3.4.3 LH₂ Loading

The LH₂ system successfully supported countdown and launch. The fill sequence began with start of S-II loading at 15:27 EST, December 6, 1972, and was completed 85 minutes later when all stage replenish was established at 16:52 EST. S-II replenish was automatic until terminated at initiation of the Terminal Countdown Sequencer. Intermittent overfill indications were experienced after S-IVB auto replenish was achieved and had to be inhibited to avoid unnecessarily cycling the replenish valve. S-IVB replenish was switched to manual at T-1 hour and left in that mode through start of Terminal Countdown Sequencer at T-187 seconds.

During recycle operations at T-30 seconds the LH₂ system was reverted normally. Fill operations were reestablished when count was resumed and both stages replenished normally to flight mass.

Launch countdown support consumed about 520,000 gallons of LH₂.

[edit] 3.5 Ground Support Equipment

[edit] 3.5.1 Ground/Vehicle Interface

In general, performance of the ground service systems supporting all stages of the launch vehicle was satisfactory. Overall damage to the pad, LUT, and support equipment from blast and flame impingement was considered minimal.

The PTCS adequately supported all countdown operations and there was no damage or system failures.

The Environmental Control System (ECS) successfully supported the AS-512 countdown. All specifications for ECS flow rates, temperatures, and pressures were met and flow/pressure criteria were satisfactory during the air to O₂ changeover.

At 1-48 hours, ECS chiller No. 1 shut down due to a low refrigerant charge. The redundant chillers were placed in operation and Freon added to chiller No. 1. No impact resulted.

At T-2 minutes the S-IC forward lower compartment temperature indication became inoperative. Redundant measurement systems were utilized and no impact resulted.

The Holddown Arms and Service Arm Control Switches (SACS) satisfactorily supported countdown and launch. All Holddown Arms released pneumatically within a six (6) millisecond period. The retraction and explosive release lanyard pull was accomplished in advance of ordnance actuation with a 42 millisecond margin. Pneumatic release valves 1 and 2 opened within 21 milliseconds after SACS armed signal. The SACS primary switches closed simultaneously at 449 milliseconds after commit. SACS secondary switches closed 1.154 and 1.163 seconds after commit.

Overall performance of the Tail Service Masts was satisfactory. Mast retraction times were nominal; 2.760 seconds for TSM 1-2, 1.980 seconds for TSM 3-2 and 2.685 seconds for ISM 3-4, measured from umbilical plate separation to mast retracted.

The preflight and inflight Service Arms (S/A's 1 through 8) supported the countdown in a satisfactory manner. Performance was nominal during terminal count and liftoff.

The DEE-3 system adequately supported all countdown operations. A discrepant printed circuit board was replaced in the FR 1 subsystem and a failed vacuum motor was replaced in the Pad A DEE-3D magnetic tape station. The Pad A DEE-3F magnetic tape station became inoperative subsequent to the propellant loading operations. The remainder of the countdown was supported by backup tape and line printer recordings. There was no launch damage.

[edit] 3.5.2 MSFC Furnished Ground Support Equipment

Other than the TCS anomaly discussed in Section 3.3, the MSFC furnished electrical and mechanical ground support equipment successfully supported the Apollo 17 launch.