Reports/Apollo 17/Saturn V flight evaluation/9 Guidance and Navigation

This page is still under development. Numbers need to be checked.

[edit] 9.1 Summary

The Stabilized Platform and the Guidance Computer successfully supported the accomplishment of all guidance and navigation mission objectives with no discrepancies in performance of the hardware. The end conditions at Parking Orbit Insertion and Translunar Injection were attained with insignificant navigation error.

Two anomalies related to the flight program did occur. At approximately 5421 seconds range time (T5 +4718.8) minor loop error telemetry indicated an unreasonable change in the yaw gimbal angle during one minor loop. At the re-initialization of boost navigation for S-IVB second burn the extra accelerometer readings normally telemetered from Guidance Reference Release (GRR) to liftoff plus 10 seconds were restarted and continued throughout second burn boost navigation. Neither of these anomalies significantly impacted navigation, guidance and control. A detailed discussion is included in Section 9.3.3 and 9.3.4.

A minor discrepancy occurred during S-II burn, when the yaw gimbal angle failed the zero reasonableness test twice, resulting in minor loop error telemetry at 478.3 seconds (T3 +317.2) and 559.4 seconds (T3 +398.2). Detailed discussion of this occurrence is included in Section 9.3.2.

[edit] 9.2 Guidance Comparisons

The postflight guidance error analysis was based on comparisons of telemetered position and velocity data with corresponding values from the final postflight trajectory (21 day observed mass point trajectory) as established from telemetry and external tracking (see paragraph 4.2). Comparisons of the inertial platform measured velocities (PACSS 12) with corresponding postflight trajectory values from launch to earth parking orbit (EPO) are shown in Figure 9-1. At EPO insertion these differences were 0.47 m/s (1.54 ft/s), 3.D7 m/s (10.07 ft/s), and D.18 m/s (0.59 ft/s) for vertical, crossrange and downrange velocities, respectively. The inplane differences are very small. The crossrange velocity difference is somewhat larger than expected from laboratory measured hardware errors. However, this difference includes trajectory errors as well as platform measurement errors and is well within the combined accuracies. There was no indication of either inplane or crossrange velocity error caused by an accelerometer hitting its mechanical stop during thrust buildup on AS-512.

Platform velocity differences for the translunar injection burn are shown in Figure 9-2. At Time Base 6 (T6) minus 7.21 seconds, the platform velocity measurements were properly set to zero in the LVDC and the corresponding trajectory data were adjusted accordingly for comparison with the LVDC outputs. The differences shown in Figure 9-2 reflect adjustments made to the telemetered platform velocities during construction of the trajectory initialized to a parking orbit state vector and constrained to a state vector near TLI which was determined from post TLI tracking. The inplane (vertical and downrange) velocity difference profiles are not characteristic of hardware errors. However, the deviations are small and reflect an inconsistency between the initial and terminal trajectory state vectors. The crossrange velocity difference is greater than expected but well within the accuracy of the trajectory and 3 sigma hardware errors and the error profile is characteristic of platform misalignment due to drift over the long coast before second burn.

Telemetered platform system velocity measurements at significant event times are shown in Table 9-1 along with corresponding data from both the postflight and Operational (predicted) Trajectories (0T). The differences between the telemetered and postflight trajectory data reflect some combination of small guidance hardware errors and tracking errors. The differences between the LVDC and OT values reflect differences between actual and nominal performance and environmental conditions. The values shown for the second burn are velocity changes from T6. The characteristic velocity accumulated during second burn was 0.44 m/s (1.44 ft/s) greater than the OT which indicates slightly more stage performance was required to meet the targeted end conditions. The telemetered data indicated 0.32 m/s (1.05 ft/s) less than the postflight trajectory. The difference in indicated performance between the telemetered and postflight trajectory data reflects small errors in the state vectors to which the guidance velocities were constrained to generate the boost-to-TLI trajectory. The velocity increase due to thrust decay was 0.01 m/s (0.033 ft/s) less than the OT after first ECO and 0.05 m/s (0.16 ft/s) greater than the OT after second ECO, indicating very good prediction in both cases.

Comparisons of navigation (PACSS 13) positions, velocities and flight path angle at significant event times are presented in Table 9-2. Differences between the LVDC and PT values reflect off-nominal flight environment and vehicle performance. At first S-IVB ECO total velocity was 0.20 m/s (0.66 ft/s) less than the OT and the radius vector was 30.8 m (101.0 ft) greater than the OT. At S-IVB second ECO orbital energy (C3) was 1849 m2/s2 greater than the OT value of -1,769,443 m2/s2. The LVDC and postflight trajectory were in excellent agreement, except for crossrange, for the boost-to-EPO portion of flight. The crossrange component differences are within the accuracy of the data compared. The state vector differences during parking orbit were very small as compared to prior Saturn V flights. These small differences during parking orbit indicate that the vent thrust was effectively the same as programmed in the LVDC. The postflight trajectory and LVDC state vectors at TLI were in relatively good agreement. The difference in C3 at TLI was -1887 m2/s2 (trajectory minus LVDC). Figure 9-3 presents the state vector comparisons during EPO. The LVDC data not received because of non-continuous station coverage were simulated by initializing to a telemetered state vector and integrating a trajectory using flight program navigation equations and programmed vent accelerations. At T6, the differences in total position and velocity were 872 meters in radius and 1 m/s in velocity and are not significant.

The AS-512 vehicle was guided to the targeted end conditions with a high degree of accuracy. Vent thrust was effectively nominal during EPO. Figure 9-4 presents the continuous vent thrust reconstruction along with OT predictions and three-sigma envelope. The upper portion of Figure 9-4 shows the orbital acceleration derived from the platform measurements adjusted for accelerometer bias. The LVDC programmed acceleration is also shown. The oscillations in acceleration from orbital navigation (804.2 seconds) to about 2500 seconds may not be real. During this period only compressed data were available for a curve fit of the telemetered velocity outputs. However, the area under the curve which represents the accumulated velocity over this time span is essentially nominal.

The LVDC state vector at TLI was compared with the OT and postflight trajectories and the differences are presented in Table 9-3. The LVDC radius vector was 5093.1 meters (16,709.6 ft) higher than the OT and 686.7 meters (2253.0 ft) lower than the postflight trajectory value. Telemetered total velocity was 4.24 m/s (13.91 ft/s) less than the OT and 0.83 m/s (2.72 ft/s) higher than the postflight trajectory. The guidance system was highly successful in measuring the vehicle performance and generating proper commands to guide the vehicle to desired conditions as shown in Table 9-4.

[edit] 9.3 Navigation and Guidance Scheme Evaluation

The LVDC flight program performed all required functions properly. Two anomalies are reported in paragraphs 9.3.3 and 9.3.4. Neither significantly affected flight program performance.

[edit] 9.3.1 Variable Launch Azimuth

Due to the unscheduled hold in the countdown at approximately T-30 seconds, the variable launch azimuth function of the flight program was required to perform over a time variation greater than for any previous Saturn V vehicle. The two hour 40 minute launch delay resulted in a change of the flight azimuth from 72.141 degrees to 91.504 degrees East of North. The performance of flight program in achieving the targeted parameters was satisfactory.

[edit] 9.3.2 First Boost Period

All first stage maneuvers were performed within predicted tolerances and Iterative Guidance Mode (IGM) performance for first boost was nominal. The steering commands telemetered during first boost are illustrated in Figure 9-5. Table 9-4 shows the terminal end conditions for first burn. Terminal conditions were obtained by linear forward extrapolation using the velocity bias at, = 1.514 meters/second to establish the extrapolation interval beyond velocity cutoff.

Minor loop error telemetry indicated an unreasonable zero reading of the yaw (Z) gimbal at 478.4 seconds (T3 +317.2) and again at 559.4 seconds (T3 +398.2). The test for an unreasonable zero reading was designed to detect a failure of the gimbal resolver power source. If two successive readings of the gimbal are found to be zero while the past attitude error magnitude exceeds the test constant (0.D6 degrees) the zero reasonableness test is failed and minor loop error telemetry is generated. If the fine resolver fails the zero test three times in 0.8 seconds during boost, a failure of the fine resolver is assumed and the corresponding backup resolver is selected for attitude information for the remainder of the mission. Since gimbal and ladder data at the times of the error telemetry indicate zero yaw with yaw ladders (indicative of yaw attitude error) greater than the test constant, the flight program apparently responded correctly. Only one unreasonable zero reading was found in each case and no change to backup readings was initiated. Although the improper selection of a backup resolver would not significantly degrade system accuracy, the current zero test is being studied for possible changes to either the test method &r the magnitude of the test constant for future missions.

[edit] 9.3.3 Earth Parking Orbit

Parking orbit guidance proceeded as expected. Table 9-5 presents the commanded steering angles for major events.

Orbital navigation was within the required tolerances for parking orbit. Termination of orbital navigation occurred at 10,971.4 seconds (T6 -7.2). Minor loop error telemetry issued at approximately 5421 seconds (T5 +4718.8) indicated an unreasonable change in successive readings of the yaw gimbal angle. The test for a reasonable change is made by comparing the difference in past and current gimbal readings with a preset test constant. If the change between past and current gimbal readings exceeds the respective test constant for pitch, yaw, or roll the change is considered unreasonable. The magnitude of the yaw test constant at the time of the failure was 0.2 degree/minor loop. If a fine resolver fails the reasonableness test three times in one second during orbit the corresponding backup (coarse) resolver reading is selected for attitude information for the remainder of the mission. Since only one unreasonable change was found, the backup yaw gimbal was not selected. Evaluation of the gimbal angle data from the time of the error telemetry indicated that the yaw (Z) backup gimbal reading was erroneously compared with a fine resolver reading instead of the proper comparison of two successive fine resolver readings. Further investigation revealed the initiation of the once per 100 second data compression module at the time of the minor loop interrupt. The occurrence of the minor loop interrupt during a particular six instruction interval at the start of the data compression resulted in the replacement of the fine yaw gimbal reading by the backup yaw gimbal. Since the backup reading was rejected as unreasonable, the next fine gimbal reading was properly compared with the last reasonable fine gimbal reading and all subsequent reasonableness tests were passed. The possibility of a similar occurrence on subsequent missions has been eliminated by starting a read of the currently selected Z gimbal ..solver (fine or backup) at the end of data compression.

[edit] 9.3.4 Second Boost Period

The December 6 target objectives resulted in nearly constant-time-of-arrival trajectories across the launch window. Therefore the targeting parameters calculated in preparation for second burn defined a higher energy transfer orbit which compensated for the 2 hour 40 minute launch delay and enabled completion of the lunar landing and exploration on the originally planned timeline.

Sequencing of restart preparations occurred as scheduled. T6 was initiated at 10,978.6 seconds. Extra accelerometer telemetry was noted throughout the second boost navigation periods. This is discussed in the following paragraphs.

Upon reinitiation of boost navigation at 10,971.4 seconds the extra accelerometer readings, that should have been telemetered only from GRR to T +10, were reinitiated and continued throughout second boost navigation. This resulted from the extra accelerometer read module being queued in with the periodic processor at GRR and again at second boost initialize. The readings were not stopped as in first boost, because there was no counterpart to the T +1D second cue during second boost. In previous flight programs the extra accelerometer readings were Queued in separately after GRR and were not queued in again at second boost. A class II change effective with AS-512 reduced the priority of these accelerometer readings and placed their start time at GRR. The only effect of this problem was a slight lengthening of the computation cycle during second boost but this was accounted for by the flight program without adverse results. Since no further missions with a S-IVB second burn are planned no program changes are recommended but documentation of the occurrence has been accomplished for future reference.

IGM for the S-IVB second burn was implemented at 11,562.7 seconds (T6 +584.1). Pitch, yaw and roll attitude angles for second burr are shown in Figure 9-6.

Table 9-4 shows the terminal end conditions for the S-IVB second burn. Desired values are the telemetered target values and actual terminal values were obtained by linear forward extrapolation using a velocity bias of 1Vbra = 3.660 meters/second.

[edit] 9.3.5 Post-TLI Period

Post TLI guidance proceeded as expected. Table 9-5 presents the commanded steering angles for some major events.

Two lunar impact APS burns were commanded from Mission Control Center-Houston (MCC-H) at 21,735 seconds (6:02:15) and 39,754 (11:02:34), respectively. The first burn of 98 seconds duration was started at the commanded time of 22,200 seconds (6:10:00). The second burn was commanded to start at 40,500 seconds (1:15:00) with a duration of 102 seconds. Both burns were properly implemented by the flight program with the desired attitude changes occurring upon acceptance of the Digital Command System (DCS) commands, ignition times and burn durations occurring as commanded.

The three-axis tumble was started by a zero burn set of lunar impact commands beginning at 41,502 seconds. Changes of +31 degrees to pitch, yaw and roll were commanded establishing tumble rates, followed by Flight Control Computer power off "A" and "B" commands at 41,519 seconds and 41,530 seconds, respectively. (Power off "A" and "B" switch selectors were issued at 41,521 and 41,532 seconds, respectively.)

The telemetry subcarrier oscilator was commanded off by the flight program at 49,620 seconds after which no further telemetry data was available.

[edit] 9.4 Navigation and Guidance System Components

The navigation and guidance hardware satisfactorily supported the accomplishment of mission objectives. No anomalies were observed during the AS-512 flight.

[edit] 9.4.1 ST-124M Stabilized Platform System

The three gyro servo loops responded properly to all vehicle perturbations. Maximum deflection during the liftoff period was 0.3 degree on the Z gyro pickoff. As on previous vehicles the 5 Hz oscillation (0.2° peakto-peak) occurred from S-IC CECO to S-IC OECO.

The largest disturbance occurred at Spacecraft/IU separation when the X gyro pickoff deflected 0.8 degree, well within limits for proper control. The three accelerometer servo loops operated within previously experienced limits. Peak deflections of the accelerometer gyro pickoffs occurred during the heavy vehicle vibration period at liftoff. Maximum excursions were as follows:

[edit] 9.4.2 Guidance Computer

The LVDC and LVOA performed satisfactorily, and no hardware anomalies were observed during any phase of the A5-512 mission.