We destroyed silver and badly damaged blue. The root cause was not enough time and budget to do the necessary testing. Primary lesson only work on one vehicle at a time.
In Early September we flew multiple long flights, as long as 106 seconds. We then parked the vehicle. On Thursday I merged the navigation code I'd flown on the helicopter with the hover code on blue. I picked the wrong code branch and it did not have the wiggle correction in it, so it wasted too much thrust wiggling and did not have sufficent thrust to lift off. After sitting with what must have been an in inadequate purge the cat pack was badly corroded. On Saturday I changed the cat pack and added the gain scheduling to correct the wiggles. I was worried based on Friday that it might not get off the ground so I loaded what I calculated to be the minimum necessary propellant+ 2% , it flew for 85 Seconds then crashed on the 2nd pad. See: You Tube from on board and Video looking out. It looks like my code slicing and dicing also missed the roll or yaw control fix I put in the blue ball as yaw seemed largely uncontrolled. Based on the takeoff acceleration we could have put a lot more propellant into the vehicle and done the full 90 seconds.
I was really happy with the way silver turned out. It was very clean , light and capable. I just could not get io hover stably. It looked like a control loop integral wind up issue. From a sportsman ship stand point both Ben and Ian of Masten asked suggested that was the problem. I had turned off the Integral Gain on position hold as one of my first steps, alas the strange mix of code from three sources had a cut and paste error so there was still an integral term inplay even after I thought I'd turned it off. I'm 90 % sure that with 2 extra hours to slow down and review I'd have fixed it. We came very close: Silver ball hover.
The eventual fault that killed it was a bit subtle. On the blue ball when I hit the command abort it shuts the motor off cleanly it just closes the main valve. On the silver we purge on shutdown, and we purge hard, this causes a shut down thrust transient as residual propellants are forced through the motor. Couple that with the choice to center the motor and kill active attitude control when aborting and one gets wild gyrations on abort. Yes can see Ben's video of its final flight.
On a personal level I'm really glad I got the one 94% successful flight out of the blue ball. It at least shows we were close. It also shows what a small team can accomplish. On the flip side we walk away from the contest with nothing tangible to show for the effort. This project has been a really big part of my life for the last year and it leaves a big hole. At the end it became an unhealthy all consuming obsession as the clock wound down and we ran out of time and budget. From a financial stand point it also means no playing with rockets for at least 6 months to a year as I try to rebuild the rocket fund.
Only time will tell if I eventually see the overall project as a positive thing. It does not seem that way this morning.