Upthread: D Rendezvous Open Items, Action Items or whatever you call them (Oct 17, 1968)
Downthread: D Rendezvous Mission Techniques (Dec 13, 1968)
See list attachedNovember 29, 196868-PA-T-262APA/Chief, Apollo Data Priority CoordinationD Rendezvous Mission Techniques
This memo is to tell you about the results of the November 25 D Rendezvous Mission Techniques meeting. Except for a number of small clean up items, we spent most of our time talking about how to handle slippage of TPI time and incomplete Insertion and Phasing maneuvers. After an exhausting discussion, I think we have those items under pretty good control now.
1. There was a discussion of various techniques for aborting from the mini-football. The only procedure we will pursue is for the CSM to make a “tweak” maneuver at the horizontal crossing if necessary to return the two spacecraft to a nominal relative motion mini-football. This maneuver will be made only if it is known that an abort is required. It shall be based on a chart the command module pilot carrys.
2. It had previously been decided to stage the DPS if the LM must make the TPI₀ maneuver – abort from the football. Of the several techniques proposed, the one most favored now to preclude DPS recontact is to impart an out-of-plane ΔV to it as part of the TPI₀ maneuver. The crew is going to try out the following procedures in the simulator and if acceptable we will stick with them for flight.
a. Just prior to TPI TIG but after Average g comes on, the LM will thrust laterally using the y-axis RCS jets to build up approximately 5 fps out-of-plane.
b. At TIG they will start thrusting with the plus x-axis RCS jets and stage the DPS as soon as acceleration exists. The out-of-plane ΔV will be removed with the TPI thrusting with the x-axis jets by yawing the spacecraft (i.e., spacecraft roll). (We are told there is no problem in reinitializing the attitude control DAP for the staged configuration in SUNDANCE.) If the CSM is active for TPI₀, the LM shall not stage the DPS.
3. It had been recognized that when computing TPI in the football trajectory it is possible to get two different solutions since there are two times the relative angle between the spacecraft passes through 27.5 degrees. Both MPAD and MIT have run analysis to determine what happens and how to handle the situation. The following table summaries the results: Time From Phasing PGNCS Operation 0 to 45 minutes Alarm code (no solution) 45 to 85 minutes Desired solution obtained (TPI = 70 minutes) 85 to 87 minutes Wrong solution obtained (TPI = 87 minutes) Greater than 87 minutes Fails to converge
The nominal TPI₀ time we want to use is about 70 minutes after the Phasing burn and if the crew uses that value as an input to P34, there should be no trouble since it's well inside the boundaries which yield the desired solution.
4. Experience has shown that the crew simulators – LMS and CMS – do not accurately duplicate the true spacecraft guidance system with respect to the time the computers take to perform their operations. Specifically, the crew trainers run considerably faster than the actual flight computer and if not taken into account, this characteristic can badly mislead those responsible for setting up crew procedures. As a result, we levied an action on MIT to determine the actual, real-life computer time required to perform a list of specific operations. This list is included as an attachment to this memo. Based on this, I'm told the simulators can be fixed to be more realistic.
5. At this meeting we finally defined the acceptable TPI window and the procedures to be followed in the event TPI falls outside the window. MPAD reports that the current three sigma estimate of TPI time dispersion is ± 4 minutes. What I mean by this is that by using the LM radar navigation to perform the CSI and CDH targeting, errors can result causing the time at which the nominal TPI elevation angle actually occurs to be as much as four minutes from the time the tar- geting was aiming for. FCSD reports that the acceptable DPI window is 3.5 minutes which you recall, is centered about the nominal TPI time – 25.5 minutes before the CSM breaks into the sunlight. You can see from this that we have a very good chance of being within the acceptable window. However, obviously techniques must be developed to handle the case when we miss.
a. Our discussion revealed that it is unacceptable for TPI to slip earlier than the 3.5 minutes before nominal, since that would cause braking to occur in darkness. Accordingly, if that occurs the crew will recycle into the TPI targeting program (P34) using the Time Option with an input of the nominal TPI time.
b. Discussion also showed that, although undesirable, late TPI is not unacceptable and, in fact, it is preferable to continue to use the elevation angle option with a nominal 27.5 degree value regardless of how late TIG occurs. And, so this is what we shall do.
As you see then, we have a fairly simple logic to guide the crew in choosing their procedure. That is, the crew procedure is based on whether the TPI time as determined onboard the spacecraft occurs earlier than 3.5 minutes before nominal TPI. Since they only have to recycle the TPI computation switching to the Time Option if the TPI is too early by more than 3.5 minutes, they always have at least an additional 3.5 minutes to take action. This makes it possible for the crew to wait for the final computation of TPI after the last rendezvous navigation to make the decision of which way to go.
6. There is a problem brought about by this procedure with regard to what the MCC-H must do for the TPI PAD message. This data – relayed by voice to the crew – is normally used for two things. First to verify that the onboard guidance system is working acceptably and the second is to provide a backup maneuver to be executed in the event it is not. The procedure noted above presents an obvious problem if the crew has to go into the Time Option since there is no way for the ground to com- pute a compatible solution for comparison. Accordingly, the following procedures were developed, which are only used if the onboard solution of TPI time is more than 3.5 minutes early:
a. The MCC-H computes and relays only one maneuver PAD message – namely, a maneuver based on executing TPI with an elevation angle of 27.5 degrees, regardless of when TIG occurs.
b. Even though the LM crew determines that TPI time is too early, they will call for the 27.5 degree ΔV solution and compare it with the ground data to determine if their PGNCS is working. If it is acceptable, they will use the procedure noted in 5a above, calling for the Time Option with nominal TPI and continuing on without a ground backup maneuver.
c. If the LM comparison with the ground solution is not favorable, the CSM also compares its 27.5 degree TPI solution with the ground and if acceptable, will recycle into the Time Option of P34 using the nominal TPI time and will execute the resultant maneuver. In other words, if the LM PGNCS is broken and the CSM GNCS is working, the CSM should become active for TPI.
d. If the CSM solution is also found to be unacceptable, the LM crew should compare their chart solution with the ground and execute it if acceptable.
e. If all of these fail, we have a situation in which TPI has slipped too early, both spacecraft guidance systems have failed, as has the LM backup chart solution and there seems nothing to do but to perform the MCC-H solution. Boy!
7. A lengthy detailed discussion of what to do in the event of incomplete Phasing and Insertion maneuvers led to the following Mission Techniques:
If the DPS does not light or if the DPS lights but shuts down prematurely, do not stage, null horizontal ΔV's and if possible, trim radial (x-body) ΔV to within 2 fps of nominal. This places the LM in a football, its size dependent on the extent of the ΔV gained. Then it is necessary to choose one of the following courses of action in Real Time, dependent on what caused the premature shutdown.
(1) Execute TPI₀ from the present trajectory this rev or next.
(2) Complete the phasing one rev later (CSM shall be mirror image targeted for this maneuver) using DPS under PGNCS control, RCS (Staged), APS, or CSM (RCS or SPS) followed by TPI₀ at the next opportunity or insertion a quarter rev after that.
This is an appalling number of choices which must be substantially reduced before the flight based on systems considerations, mission objectives and extent of flexibility affected by the crew procedures. The latter is extremely important since the procedures are complex and completely time dependent; they are not easy to recycle into.
(1) If DPS does not start, stay in football by nulling out ullage.
(2) If DPS does start, the primary goal is to complete the burn using RCS with APS interconnect. If the ΔV required is greater than about 8 fps, staging is required.
(3) In order to be prepared for some mysterious time critical problem discovered within one minute after TIG, the CSM will be targeted with the same burn as the LM to be executed with a one minute delay. This is not a mirror image burn. It nulls the LM burn.
8. MIT reported on an old action item that the CSM PIPA bias check cannot be conveniently reduced below 256 seconds duration.
9. In case everyone has not heard, the SUNDANCE program has been fixed so that the crew can use the rendezvous radar self-test program (R04) during terminal breaking with the Average g program (P47) running simultaneously. That is great!
10. Although not part of the D mission rendezvous, our final discussion of the day involved what the CSM should do during the docked DPS maneuver. Options for the CSM are to use the SPS thrust program (P40), the RCS thrust program (P41), or the Average g program (P47). Due to a limitation in the displays available in P47, which we know would work, the crew would prefer to use P40 or P41. We're not too sure how they will do so we asked MIT to look into how each of these programs would operate during the docked DPS burns such that we may make a final choice.
I don't expect to have any more full blown D rendezvous meetings until the final review of the Mission Techniques Document now scheduled for distribution about December 16. This review will probably be about January 10, 1969.