First up is a plot straight from the RDAS GUI.
This shows that the RDAS commanded the apogee event at 30 seconds which means that the backup timer triggered the event. The RDAS never had a chance to trigger the event via the acceleration algorithm because of the early deployment.
The first hint of trouble occurs at about 15.5 seconds. There is a large spike in the acceleration data which is usually an indication that one of the ejection charges has fired. There is then a period of chaotic acceleration as the parachutes and body sections sort themselves out.
Here is the Kalman filtered version of the data.
This shows that the velocity at the beginning of deployment was around 200 ft/sec. A bit higher than normal. The only damage to the rocket was an internal zipper of the motor mount tube. I had realized that this was a weak point in my recover design. The simple way to work around it is to drill holes in the forward centering ring and bring the tubular Kevlar through those holes. The thick plywood centering ring can handle sideways forces much better than than the thin phenolic motor mount tube.
Repair will be fairly easy. Just poke the Kevlar back in place and then epoxy a tube coupler over the damaged area. This doesn't extend too far into the tube. The plus side of this is that I will now have an internal thrust ring that I had procrastinated over putting in. Now my 54mm motor adapter will work and I can fly using cheap (relative to the 75mm motors) K550's.
This next plot is of just the raw pressure altitude data.
You can clearly see the periodic glitches in the data. I matched several of these with the APRS transmission times. (See the GPS data for further discussion.)
If you want to look at the raw data, here is the RDAS data file. It has been run through gzip because I have had problems downloading .rd files in the past. Winzip understands gzip format.