LM Descent to the Moon - Part 4 - Descent Monitoring

REAL-TIME DESCENT MONITORING AND ANALYSIS


During the real-time situation, monitoring of the spacecraft systems and of the trajectory was performed continually both on board by the crew and on the ground by the flight controllers. The real-time monitoring determined whether the mission was to be continued or aborted, as established by mission techniques prior to flight. The real-time situation for the Apollo 11 descent is described here.

LM (Apollo 11) DOI (Descent Orbit Insertion) is done 180 degrees before PDI (Powered Descent Initiation)

Descent Orbit Insertion (DOI)

The DOl maneuver is performed on the farside of the Moon at a position in the orbit 180 degrees prior to the PDI (Powered Descent Initiation) and is, therefore, executed and monitored solely by the crew. Of major concern during the burn is the performance of the PGNCS (also called PGNS, Primary Guidance and Navigation [Sub]System) and the DPS (Descent Propulsion [Sub]System). The DOl maneuver is essentially a retrograde burn to reduce orbit altitude from approximately 60 nautical miles to 50 000 feet for the PDI and requires a velocity reduction of 75 fps.

This speed reduction is accomplished by throttling the DPS to 10-percent thrust for 15 seconds (for the center-of-gravity trimming) and to 40-percent thrust for 13 seconds. An over burn of 12 fps (or 3 seconds) would cause the LM to be on an impacting trajectory prior to PDI. Thus, the DOl is monitored by the crew with the AGS (Abort Guidance [Sub]System) during the burn and by range-rate tracking with the rendezvous radar (RR) immediately after the burn. If the maneuver is unsatisfactory, an immediate rendezvous with the CSM (Command Service Module) is performed with the AGS. For Apollo 11, this maneuver was nominal. Down-range residuals after the burn were 0.4 fps.



Powered Descent (PDI)

Trajectory limits. - During real time, trajectory limits are monitored for flight safety (green area in the following figure). The prime criterion for flight safety is the ability to abort the descent at any time until the final decision to commit to touchdown. Thus, flight dynamics limits are placed on altitude and altitude rate, as shown below.

Altitude as a function of altitude rate during powered descent

Notice that the nominal trajectory design (blue curve) does not approach the limits until late in the descent, after the crew has had ample time for visual assessment of the situation. The limits shown are based on APS (Abort Propulsion [Sub]System) abort with a 4-second free fall for crew action delay or on DPS (Descent Propulsion [Sub]System) abort with a 20-second communications delay for ground notification [The abort limit line above is calculated using 20 s at any given sink rate]. The flight controllers and the crew monitor altitude and altitude rate, but because of communication delays with the ground, the flight controllers only advise, based on projected trends. The profile shown in the above figure was near nominal (red curve).

[A more accurate (?) descent trajectory can be found in the Apollo 11 Mission Report. The descent was measured by three computers: PGNS (onboard LM primary computer), AGS (on board secondary computer), and MSFN (Manned Space Flight Network on ground [Earth]). These three trajectories are shown in the following figure. It is worth to mention that there are several reasons these curves might not be accurate since for example the altitude was corrected 2200 ft (too low) manually during the descent (Apollo 11 Experience Report) and that correction was slowly applied by the computer(?s?). Most likely the PGNS curve might be the most close to the actual.]

A (maybe more) accurate altitude vs altitude rate figure from the Apollo 11 Mission Report 1969

[The 3-sigma curves are calculated using standard 3-sigma dispersions. The following figure shows that almost all cases are inside the 3-sigma curves for a nominal line.]

99.7 % of the standard distributed cases are inside the +/-3-sigma limit lines.

The DPS and PGNCS interface. - To determine in real time if the DPS is providing sufficient thrust to achieve the guidance targets, the flight controllers monitor a plot of guidance thrust command (GTC) as a function of horizontal velocity, as shown below.

Guidance thrust command (GTC) as a function of horizontal velocity.

Nominally, the GTC decreases almost parabolically from an initial value near 160 percent of design thrust to the throttleable level of 57 percent, approximately 2 minutes (horizontal velocity being 1400fps) before high gate (horizontal velocity being 500 fps). If the DPS produces off-nominal high thrust, horizontal velocity is being reduced more rapidly than desired to reach high-gate conditions. Therefore, the GTC drops to 57 percent earlier with a higher-than-nominal velocity to guide to the desired position and velocity targets. This early throttledown (also called throttle recovery) results in propellant inefficiency.

If the DPS produces off-nominal low thrust, horizontal velocity is not being reduced rapidly enough. Therefore, the GTC drops to 57 percent later at a lower velocity to guide to the desired position and velocity. This later throttledown results in increased propellant efficiency (I. e., longer operation at maximum thrust). However, if no throttle down occurs prior to high gate (program switch from P63 to P64), the targets will not be satisfied, and the resulting trajectory may not be satisfactory from the standpoint of visibility.

CAPCOM Charles Duke, with backup pilots James Lovell and Fred Haise listening in during Apollo 11's descent

 
In fact, for extremely low thrust, the guidance solution for the GTC can diverge (previous figure curves 9400 and 9350); as TGO (Time to Go) becomes small, the guidance calls for more and more thrust in order to achieve its targets. This divergence can result in an unsafe trajectory, one from which an abort cannot be satisfactorily performed. The 2-minute bias for throttle recovery before high gate provides sufficient margin for 3-sigma low thrust even with propellant valve malfunction. However, the flight controllers monitor the GTC to assure satisfactory interface between DPS and PGNCS operation. A mission rule was established that called for an abort based on the GTC divergence. During the Apollo 11 landing, the DPS thrust was nearly nominal; thus, no DPS and PGNCS interface problems were encountered.


The LR and PGNCS interface. - Normally, the LR (Landing Radar) update of the PGNCS altitude estimate is expected to occur by crew input (always manually updated by te crew) at an altitude of 39 000 ± 5000 feet (3-sigma dispersion). Without LR altitude updating, system and navigation errors are such that the descent cannot be safely completed. In fact, it is unsafe to try to achieve high gate where the crew can visually assess the approach without altitude updating. Thus, a mission rule for real-time operation was established that called for aborting the descent at a PGNCS-estimated altitude of 10,000 feet, if LR altitude updating had not been established.

The Landing Radar antenna was located under the LM.

In addition to the concern for the time that initial altitude updating occurs is the concern for the amount of altitude updating (I. e., the difference between PGNCS and LR altitude determinations dh). If the LM is actually higher than the PGNCS estimate, the LR will determine the discrepancy and update the PGNCS. The guidance then tries to steer down rapidly to achieve the targets. As a result of the rapid changes, altitude rates may increase to an unsafe level for aborting the descent. That is, should an abort be required, the altitude rates could not be nulled by the ascent engine in time to prevent surface collision.

The dh limits necessary to avoid these rates are shown below.

Landing radar altitude updates.

Notice that over the estimated 3-sigma region of LR initial updating (which at the time of that analysis was centered at an altitude of only 35 600 feet instead of 39 000 feet), the dh limits are much greater than the 3-sigma navigation estimates of dh.

However, the flight controllers, as well as the crew, monitor dh to assure that the boundary is not exceeded before incorporation of the LR altitude updating. If the boundary is exceeded, then the data are not incorporated, and an abort is called. When the LM is actually lower than estimated, no excessive rates are encountered upon LR updating. It is necessary only that the LM altitude and altitude rate be above the abort limits, defined in the above section entitled "Trajectory Limits."


During the Apollo 11 mission, the LR acquired lock-on to the lunar surface during the rotation to face-up attitude at an altitude of 37,000 feet. The dh was -2200 feet, indicating that the LM was actually low.

[Some sources say that the Apollo 11 LM PDI was at 52,000 ft instead of the nominal 50,000 ft, indicating that LM was 2,000 ft high. - If this information here is correct then the PDI altitude was most likely 48,000 ft instead of the nominal 50,000 ft assumetd at the PDI.]

This small amount of dh can readily be attributed to terrain variations. Because no limits were violated, the data were incorporated after a short period of monitoring at an altitude of 31,600 feet. The dh readily converged to a small value of 100 feet within 30 seconds. The LR velocity updates were incorporated nominally, beginning at a 29,000 foot altitude. As expected, LR signal dropouts were encountered at low altitudes (below 500 feet) but presented no problem. When the velocity becomes small along the LR beams, depending on the attitude and approach velocity, zero Doppler shift is encountered; hence, no signal occurs.


Crew visual assessment. - As stated previously, the approach and landing phases have been designed to provide crew visibility of the landing area. This provision allows the crew to assess the acceptability of the landing area, to decide to continue toward the landing area, or to redesignate a landing away from it with LPD or manual control. During the Apollo 11 mission, because of the initial navigation errors, the descent was guided into the generally rough area surrounding West Crater (se below).

Apollo 11 LM final approach, West Crater in the center.

West Crater is inside the premission mapped area, approximately 3 nautical miles west of center. Unfortunately, because of the guidance program alarms, the commander was unable to concentrate on the window view until late in the descent (near low gate). Thus, crew visual assessment during the approach phase was minimal, which resulted in continued approach into the West Crater area.


The PGNCS monitoring. - To determine degraded performance of the PGNCS, the ground flight controllers continually compare the LM velocity components computed by the PGNCS with those computed by the AGS and with those determined on the ground through Manned Space Flight Network (MSFN) tracking. That is, a two-out-of-three voting comparison logic is used to determine whether the PGNCS or the AGS is degrading. Limit or red lines for velocity residuals between the PGNCS and the MSFN computations and between the PGNCS and the AGS computations are established before the mission, based on the ability to abort on the PGNCS to a safe (30,000-footperilune) orbit.

In real time, the Apollo 11 PGNCS and AGS performance was close to nominal; however, a large velocity difference in the radial direction of 18 fps (limit line at 35 fps) was detected at PDI and remained constant well into the burn. This error did not indicate a systems performance problem, but rather an initialization error in down-range position. This effect is illustrated geometrically in the next figure.

Effect of position error on velocity comparison.

The PGNCS position vector RE and velocity vector VE estimates are used to initiate the MSFN powered-flight processor. The MSFN directly senses the actual velocity VA at the actual position RA' but, having been initialized by the PGNCS state, the MSFN applies VA at RE. Thus, a flight-path-angle error dy is introduced by a down-range position error and shows up as a radial velocity difference dVDIFF.

The magnitude of the velocity difference indicates that the Apollo 11 LM down-range position was in error by approximately 3 nautical miles at PDI and throughout the powered descent to landing. The reason for the down-range navigation error was attributed to several small dV inputs to the spacecraft state in coasting flight. These inputs were from uncoupled RCS attitude maneuvers and cooling system venting not accounted for in the prediction of the navigated state at PDI.

LGC was located at the back of the LM.

The LM guidance computer (LGC) also monitors the speed at which it is performing computation tasks: navigation, guidance, displays, radar data processing, and auxiliary tasks. If the computer becomes overloaded or falls behind in accomplishing these tasks, an alarm is issued to inform the crew and the flight controllers, and priorities are established so that the more important tasks are accomplished first. This alarm system is termed "computer restart protection."
 

Apollo Guidance Computer (AGC) or LM Guidance Computer (LMC) and DSKY.

During real time, because of an improperly defined interface, a continuous signal was issued to the LGC from the RR (Rendezvous Radar) through coupling data units (CDU's). These signals caused the LGC to count pulses continually in an attempt to slew the RR until a computation time interval was exceeded. As a result, the alarm was displayed and computation priorities were executed by the computer. The alarm was quickly interpreted, and flight-control monitoring indicated that guidance and navigation functions were being performed properly; thus, the descent was continued. In spite of the initial position error and the RR inputs, the PGNCS performed excellently during the Apollo 11 powered descent.


RESOURCES

/1/ Bennett, Floyd V., Manned Spacecraft Center (MSC) - Apollo Experience Report - Mission Planning for Lunar Module Descent and Ascent

/2/ Wikipedia

/3/ Apollo 11 Mission Report 1969, NASA

/4/ Apollo Experience Report(s)

* * *