Downthread: Spacecraft computer program status (Aug 22, 1967)
See attached listAUG 1 196767-FM-T-52FM/Deputy ChiefSpacecraft computer program status – particularly Sundisk
1. I have been putting off writing this note because it is rather pain- ful. To put it bluntly, the release of Sundisk–the interim earth orbit command module spacecraft computer program–has slipped rather badly. What makes it painful is that it's not all together clear why it is slip- ping and we still haven't found a good way to monitor and predict these slippages further in advance of their occurring.
2. I am sure a substantial contributor was the digital computer facility saturation we experienced in January, February and March, when we were getting out Sunburst–the unmanned LM-1 program. The lack of computer time for Sundisk development during those critical months must have im- pacted us severely. Another cause was the delay in getting the hybrid facility converted to Block II as quickly as we hoped. It did not become operational until May. And it is my understanding that a backlog of hybrid testing in large measure is pacing program delivery even now. third contributor was the substantial number of program changes. The last of these changes slipped the configuration control date from August 18 to August 28 where it stands today. Configuration control comes about a month prior to program release for rope manufacture which will be around October 1, I suppose. It takes about two months to make the ropes so they should be available well in advance of spacecraft delivery.
3. Of course, another consideration is to make sure we are not delaying flight crew training, but as far as I can determine, we are not! Sub- stantial parts of the program are working and are now available for crew training. As I said, our problem is determining what is going on and what we can do about it.
4. Most troubles in the program now are in the Boost and Entry processing. Actually, Boost does work in the bit-by-bit digital facility. The problem is in the computer program simulating the Saturn in the hybrid facility. Entry apparently does still have some work required on it. The only major problem area that I know of is associated with the Thrust Vector Control (TVC) digital auto pilot. Apparently, even at this late date, configuration has not been entirely frozen – people are still diddling around with the design. I might also point out that we have decided to move some of the entry constraints defining L/D of the spacecraft from fixed to erasable memory so that we may respond as people play around with the spacecraft ballast, etc. All testing, however, will be carried out assuming a L/D of .30. This verification testing, or course, is to prove that the program is bug free and runs. When the actual L/D is established, it may be desirable to run some more tests to determine the performance of the program under those conditions but not for verification purposes.
5. We have agreed with MIT on a delivery schedule for the final version of the GSOP. The blue cover of versions of Chapters 3, 4 and 5 will be avail- able by September 30. It is our understanding that Chapter 5 (program formulation) is in the final stages of typing including all MDRB's; Chapter 4 (program sequencing) is kept up to date at all times and will be ready for release; and Chapter 3 which defines the digital auto pilot paces the whole works, but is supposed to be finished by that date. We are told that the entry DAP is in good shape, the RCS DAP requires some modification and the TVC DAP is being redesigned. No documentation of it exists anywhere. Deplorable situation.
6. It is pretty obvious to everyone that the delay we are experiencing in getting out Sundisk will reflect on our delivery of the subsequent programs – Colossus (CSM) and Sundance (LM). This, of course, results from not getting the full strength effort on these programs as soon as we had planned, but is is also evident from our experience that program development takes substantially longer than we had previously estimated. As if now Colossus and Sundance are scheduled for a release of February 1968, and that's a good target date to shoot for. I wouldn't be too surprised thought if the actual release date of the first flight-worthy version of the entire lunar landing Sundance and Colossus programs was as late as June 1968. Rest assured both MIT management and we here will make every effort to minimize these slips and to at least get a system working so we'll know what's going on.