================================================================================

Flight Test Log 23

================================================================================

Summary
-------

This was a sucessful test of GPS waypoint tracking, where the waypoints were
in the form of a square pattern 200 meters on a side. This was a repeat of
test 52, except that aileron gain was reduced to its minimum value.

After some initial transients in autonomous mode, the airplane flew very well.
Waypoint tracking was excellent and much improved over test 52. One complete
circuit of the square was made and the vehicle flew close to all four corners
of the square. Altitude hold was good. Pitch trim was acceptable.

At the start of the second cycle, some sort of glitch caused a sudden rapid
turn. The autopilot was switched to manual mode as a precaution, but the
pilot had difficulty regaining control due to the long visual range, and
the airplane crashed. Damage was relatively minor.

================================================================================

   Test_ID:  53
      Date:  2006/08/18 (Fri)
      Time:  18:50 (approx)
      Wind:  Calm
   Vehicle:  SVC
     Scale:  55 %
      Mass:  --
     Pilot:  Frank
    Camera:  none
  Location:  PCC West Campus
 Elevation:  730 m                             (2400 ft)

Changes to airplane:

   1. Cracked landing skid repaired (crack was from crash of 2006/7/6).

   2. Autopilot aileron gain was minimized.

................................................................................

Crash damage is shown below. The starboard wingtip and vertical tail broke
off as a unit. The starboard elevon and part of the inboard wing also
separated.

The propeller and motor mount were damaged. The motor mount uses three
nylon tie wraps to clamp the mount to a pair of carbon arrowshafts. The
crash shoved the motor backwards, with the motor assembly absorbing some
of the crash energy as it slid along the arrowshafts.



Wing and elevon damage



Nose damage

................................................................................

Regarding the sudden turn problem -- we're currently investigating whether an
electronic v-tail mixer is causing the problem. The mixer is used for the
elevon servos. The autopilot does not have built-in elevon capability, so
we're using the mixer to combine elevator and aileron channels.

The sudden turn problem also showed up at least once during the flight test
on 16 August. At that time the autopilot recovered by itself, and we have no
reason to believe the same thing wouldn't have happened during test 53 if the
airplane had remained in autonomous mode.

................................................................................

[Added 2006/08/27]

The autopilot manufacturer confirms that most electronic v-tail mixers glitch
with the autopilot because of pulse synchronization issues. They recently
found a mixer that is tolerant of unsynchronized pulses. Our plan is to try
this mixer.

More detail on crash energy-absorbing motor displacement, with before-and-after
photo comparisons:




================================================================================
Last updated 2006/08/27 FLM