PSAS/ news/ 2003-09-21 - Launch of LV2.2 in the Black Rock Desert of Nevada

Summary: The PSAS team drove out to the Black Rock Desert of Nevada to attempt two launches of Launch Vehicle No. 2 (LV2): LV2.1 which was flown in September of 2002 and LV2.2 which was a new airframe. After all day work and integration sessions in the middle of the desert on Friday and Saturday, LV2.2 was launched on an "O" motor Sunday morning as an airframe only (i.e., without the LV2 avionics package). Unfortunately, LV2.2 was destroyed about 8 seconds into flight near 5,000ft and between Mach 1 and 2. We think the fin canister tore off because of the modifications we did in order to make it fit on the the "O" motor case. The nosecone, commercial flight computer, and 2m uplink board were recovered, but the rest of the rocket was destroyed. The failure of LV2.2 forced us to scrub the launch of LV2.1 with the avionics package because of the loss of the parachutes and unknown condition of the commercial flight computer and 2m uplink board.

Thursday September 18th, 2003

After a week of hectic work trying to finish the airframe and avionics system, the team packed up our stuff and headed off to the Black Rock Desert of Nevada. We left Portland around 9:00am and arrived around 9:00pm, making for a long day's drive.

We unpacked one of the tents and started setting up camp, but quickly decided to leave most of the work until the next morning. Thursday night was the first time many of us had gotten a chance to rest in the last week; we took the night off and watched the amazing night sky through a telescope, as well as kibitzed with the other amateur rocketry folks who had already shown up.

Friday September 19th, 2003

Friday was a serious work day. We set up shop under the awning and got to work integrating and prepping the rocket for flight. The flight range opened up at 9:00am for the smaller, less experimental rockets, so the team was given a pretty good show while working.

The airframe team focused on getting the LV2.2 airframe ready: last minute rivnut insertions, drilling and tapping holes, cutting the fins out of the rough canister, and of course... Brian sanding.

The avionics and software teams set up shop under one of the tents and deployed most of Andrew's basement in order to get down to work. We had a pretty impressive setup; people wandering by had lots of great things to say about our elaborate setup. In the future, we hope to be done with the avionics system before we leave for a launch, and thus be able to leave Andrew's basement where it is.

That day we built three cylindrical patch antennas (CPAs): one for the 2.4GHz telemetry system, one for the 1.27GHz amateur TV broadcast system, and one at 1.57GHz for the onboard GPS receiver. It took, of course, way longer to make the antennas than we wished - about an hour plus each. Steve made an amazing should-mounted tracking system using Glenn's helix antennas... and thus was born the TrackMaster 2000.

Then we worked on debugging firmware and software for the avionics system... something that should have been done weeks ago. We made great progress on Friday, that's for sure, but ran into the integration problems you can always expect if you try building a system at the very last minute.

We also set up the launch tower computer and did the final integration on that system. Although overkill, it will allow us a lot of flexibility on future launches. Simply cabling up the various boards, getting the cables routed, and getting the RF connection up and running turned out to take most of the day. We also finally found out that our CAN driver had a serious bug which didn't allow the CAN messages with the RTR bit set to ever leave the CAN chip. Whoops. We were still able to use the system, but we couldn't get any status messages back.

We set up the second work tent Friday night, getting ready for the big day coming up... and except for some of the obsessive compulsive avionics team, relaxed a bit on the playa.

Saturday September 20th, 2003

Saturday morning was a flurry of activity, trying to get LV2.2 ready for flight on the O motor. By noon, it was apparent that the avionics system was not going to be ready for a Saturday launch. So we decided to launch LV2.2 as an airframe test only: Matt's commercial flight computer and 2m uplink board being the only avionics onboard.

There was plenty to do that day just to get the airframe ready: it took us a surprisingly long time just to prep the commercial flight computer board, the 2m uplink system, and the recovery system.

That morning we also discovered that the fin canister didn't fit over the "O" motor casing, so we had to slot the casing to make it fit.

Despite our concerted efforts, at 4:30pm we were still too far away from launch to make the waiver's window that day, so we scrubbed and decided to leave the airframe exactly as it was and launch first thing in the morning.

That evening, the avionics and software teams spent a hellishly long evening trying to get the avionics system ready for a flight in the LV2.1 airframe scheduled for Sunday afternoon. The cylindrical patch antennas were mounted on the avionics module, the avionics system was buttoned up for flight, and we desperately tried to fix the broken Linux CAN driver that was dropping a certain kind of packets on the launch tower computer and flight computer.

Sunday September 21st, 2003

Sunday morning we woke early and prepped LV2.2 and the recovery teams for flight. Around 10:00am we drove LV2.2 out to the tower and loaded it; we were ready for launch!

The ignition, liftoff and first 8 or so seconds of flight were absolutely perfect.

Then around 8 seconds or so, we saw debris fall off the rocket. We believe that was the fin canister tearing off. Without fins, the rocket body was unstable and started to tumble. The tumbling caused the nose cone and "O" motor casing to be ripped off of the three airframe modules. The nosecone and motor canister fell fairly close to the launch tower, and the three airframe modules fell on the shreded main parachute pretty much in the middle of the flight line. Later we were able to recover the drogue and various pieces of the airframe and fin canister.

Baker - the team's intrepid sockmonkeynaut - was once again hidden in the nose cone. Luckily, he rode out the harrowing ride and was unscathed. According to James, however, he's demanding a pay raise. Or at least sockmonkeynaut life insurance.

After discussing the failure for a while, we decided that the failure was due to the slot we had to cut in the fin canister in order for it to fit over the 136mm "O" motor casing. Our evidence was that the recovered fin canister was almost perfect flat and the remaining unslotted Aluminum was torn through, suggesting that the fin canister was peeled off along the slot by aerodynamic forces.

Because of the shared systems between LV2.1 and LV2.2 - which included parachutes and launch lugs - we decided to scrub the launch of LV2.1 planned for later that day.

Many of the team left as planned that afternoon to head back to Portland. A smaller contingent stayed around and enjoyed the BRD for another night before heading home the next day.

Overall, we accomplished an amazing amount of integration work, got our launch infrastructure squared away, and spent another amazing weekend in the Black Rock Desert. Although we're saddened by the loss of LV2.2, we've chalked it up to yet another "learning experience". We're looking forward to launching LV2.1 with the avionics system in place later this fall or in early spring.

A giant thank you to the PSAS team members and all of the people and groups who helped get us there.

09/28/2003 COMING SOON: Video clips

Misc Links: