Over the summer, we flew a test rocket to validate our parachute deployment avionics. Having had great experiences with the Friends of Amateur Rocketry site for past rockets, and with most team members in LA for the summer, launching at FAR again was an obvious choice.
Three members of the team spearheaded preparations needed during the two weeks leading up to launch. They sized a kit rocket and motor that would be suitable for testing, put the kit rocket together, and pulled together all the electrical components needed for the test.
On launch day, our team of six drove out into the Mojave Desert. Once at the FAR site, we quickly unpacked and started to integrate the rocket for launch. As is often the case in rocketry, that’s a lot more work than it sounds. Our avionics team started laying out the avionics sled and wiring it up to the rocket, and the rest of the team worked on friction fits, parachute packing, and black powder ejection tests.
Winds started picking up at around 3:30pm, as did our working pace. Our first ejection test revealed that we had estimated too little black power for the deployment charges, so we recalibrated and had a successful second test. Confident in our deployment system, we prepared our rocket for launch and brought it out to the pad.
One of the reasons we test is so that we can find bugs in the software, and we certainly found one. As soon as we turned on the flight computer on the pad, it immediately commanded the drogue parachute to deploy, which is off-nominal behavior. This is is how our rocket came to be known as “Trigger Happy.” After assessing the situation, we decided to launch using the backup commercial off-the-shelf computer for parachute deployment.
With the all-clear from range personnel, we launched the rocket and watch it lift off. Unfortunately, the rocket was overstable and weathercocked into the wind, and we lost visual contact when it flew directly into the sun. Only one team member even caught a glimpse of the parachute before we lost it completely.
After sufficient time squinting into the sun, we packed up and started combing the area with the help of Rick, one of the FAR site personnel. We searched for hours and did not find our rocket, though we did find the remains of other rockets stuck deep into the ground.
As the sun disappeared behind the mountains, we called it quits and headed home. However, at around 4am, Rick wrote us to say that he had found our rocket, with both parachutes successfully deployed! Mission accomplished. It turned out, just as the rocket had turned into the wind under thrust, the same wind had carried it under parachute back past us; we were looking in the complete wrong direction! For future flights we will explore multiple layers of solutions so we do not repeat this experience, including live telemetry and possibly drone search and rescue.
Post analysis revealed that the experimental avionics did correctly deploy main parachute at 700ft, but that the launch-detect logic was not calibrated correctly. This caused the rocket to think it had already launched - and was well below 700 ft, within the deployment zone - when it was still on the pad! We have since fixed the code to make it more robust and prevent this failure in the future.