Andrew was absent due to an emergency. The meeting proceeded in a less-than-usually directed fashion coordinated by Brian, and unexpectedly, Tim.
A team round-up confirmed that we all believe the systems are (mostly) in good shape. Airframe is unchanged and ready. Recovery is ready with the new X-Form parachute, and the line cutters aren't yet past their shelf-life. Propulsion is ready with the same 4.3 second burn, 13307 Ns motor we've had in the can for a year+ now. The avionics are done in the sense that we have no known serious problems effecting a launch.
Avionics team wants to fix-up the safety critical stuff for launch tower operations. This shouldn't be a lot of work, and ought to improve an already fairly safe system considerably. The plan is to have a TWiki page up on this soon.
Brian and Jimmy will work on the launch tower some day soon to do some outstanding maintenance.
Software is thought to be behaving reasonably, however, after the somewhat mysterious performance at last Sunday's integration test, software behavior will be a prime focus of next Saturday's test.
The new cable-ends for the three launch control RF-communication cables have been ordered. It's not clear if they can be installed by Saturday, but we think we can finagle the cable ends we have for the test.
The conclusions drawn from the last integration test continue to evolve. It appears that we did have some unexpected and as yet un-explained freezes in software systems. So these will be getting a close watch during the next test.
The next integration test is scheduled for Saturday July 9th. We'll have at least Brian, Jamey, Glenn, and Tim representing team-leads, plus enough other PSAS'ers to get things done reasonably fast and painlessly. Possibly not Andrew but we think we can work around. We decided, and Dave and Holly seemed willing, to include the launch tower in the testing to validate the communications while the rocket is on the tower.
A specific goal of this meeting was to pre-define what we expect for a successful integration test, which is hopefully what we'll get on Saturday. No long list was elucidated but we did come up with a solid short one.
Integration Test Goals
- Full-Rate Communication Links (2500 IMU messages per second, plus other normal packets)
We expect this includes no unexplained losses in logging, network, etc. - Smooth and convincing demonstration of the launch procedure, from setup to rocket-ready relay click.
This means no debug mode overrides or other abnormal measures. - Clean log files. We want to do a post-mortem on our tests to see if any abnormalities exist.
- Record the ATV video. This will allow review similar to the previous goal.
- Repetitive success. We would like to have a good test, tear down and setup again and have another successful test back to back.
The back to back testing requirement contains a time-allowing clause. If we can't do it, the test is still successful, but we then strongly hope to do another test on another day. (We want to do that anyway.)
There was also some discussion of testing post-launch code. Jamey expressed some skepticism about the quality of validation added by such testing, but seems willing to add some use-the-force override for launch detect particularly since we already have such overrides for apogee and main detect. I think the general feeling was that a little post-launch sequence testing was cheap insurance even if it's worth about exactly the price.
We concluded the coherent portion of the meeting with a logistics power circle lead of course by Holly, and a naive hope for a confidence building integration test on Saturday.
After that we divided into group discussion while following the laws of diffusion. Soon the whole thing degenerated into a software meeting.