Ascent newsletter

See list attachedMay 6, 196969-PA-T-71APA/Chief, Apollo Data Priority CoordinationAscent newsletter

This memo is to report several interesting things regarding lunar ascent, both nominal and after a descent abort.

1. It turns out we demand better performance of the PGNCS to support ascent to orbit than we do descent. Accordingly, if it is necessary to abort during descent due to degradation of the PGNCS, it is automatically necessary to switchover from the PGNCS to the AGS. Of course, this assumes that the AGS is performing better than the PGNCS.

2. We have recently had a running philosophical argument regard- ing ascent switchover. Of course, switchover in itself is not cata- strophic as is an abort; if the system you switch to is working okay, the mission continues just as planned. This led me to push for establish- ing fairly tight switchover limits since I felt that it was highly desirable to assure as near nominal rendezvous characteristics as possible. That is, why stick with a degraded PGNCS if the AGS is working better? The only disadvantage seems to be the hazard involved in the act of switchover itself; all the switches, relays, and so forth have to work. In other words, it comes down to a tradeoff between the hazards involved in switch- ing over versus the dispersions in the rendezvous situation which could be avoided by switching over.

More recently we have adopted a procedure for eliminating dis- persions at insertion following descent aborts by making an adjustment maneuver immediately after insertion. This so-called tweak burn is used specifically to assure satisfactory rendezvous conditions. This procedure may also be used to compensate for degradation of the PGNCS during ascent and makes it possible to leave the PGNCS in control as long as it is still capable of providing a safe orbit. However, if the PGNCS degradation is sufficient to justify it (say, worse than 3 sigma) the crew should be advised of the situation during powered flight such that they will stand by for a tweak burn to be executed immediately after insertion using the same procedures as for the descent abort.

Having adopted this technique, it seemed reasonable to set the PGNCS switchover limits fairly wide. The value chosen was 6 sigma. The compromise here, of course, is the operational messiness of a tweak burn traded off against the switchover to AGS “hazard.”

3. One thing. which could give us bad trouble is a misaligned PGNCS prior to ascent, particularly if we align the AGS to it as was planned. The problem, of course, is that small misalignments can result in unaccept- able insertion conditions and, even though ground monitoring would probably detect the situation during ascent, switchover would do no good since the AGS would be equally misaligned. To avoid this situation entirely, we have concluded that the best course of action is to independently align the AGS while on the lunar surface rather than to align it to the PGNCS. This makes the two systems truly independent, which not only gives us a cross-check on the accuracy of the alignment of each but also permits a useful switchover if somehow a PGNCS misalign escapes our detection techniques. Incidentally, this also eliminates the problem of CDU tran- sients in the AGS lunar surface alignments. Accordingly, we are proposing that the procedures be changed to always utilize the AGS gravity lunar surface alignment technique rather than alignments to the PGNCS. I expect this will—be-done once some details have been worked out.

4. It is interesting to note that the problem just discussed is not quite as severe in the event of a descent abort. In that case, of course, the AGS must have been aligned to the PGNCS and so they both will suffer the same misalignment at PDI. What happens then if we have a descent abort and try to achieve orbit with both systems misaligned? It turns out that this particular error is partially compensating – that is, the trajectory dispersion during descent is partially eliminated by the trajectory dispersion during ascent back into orbit. In addition, the descent abort limits will be tight enough that unacceptable dispersions should not occur prior to descent. In other words, we feel we have a safe situation here.