Date: 10/21/10 [[BR]]
Location: Tri Valley [[BR]]
Aircraft: Loki [[BR]]
Pilot: Arion Mangio [[BR]]
Flights: 5 Loki [[BR]]

[source:/trunk/FlightData/2010_10_21 Link to Flight Data]

Arion and myself went to Tri Valley with Loki, the new avionics build. Weather forecast was for northwest winds at 10MPH (Friday's forecast was for 14-15MPH winds). Actual winds seemed closer to 15MPH out of the north. Temperature was in the 40s. Wind chills were difficult to deal with. 

Deconstruction of Tri Valley and Jensen Field continues. Some hangars at Jensen are gone, others haven't been touched.

The test objectives were:
   1. Test FDI filter with baseline, LQR, and LQR+MRAC controllers, with 10 deg aileron fault injection and revised doublet sequence.
   2. Test FDI filter with LQR+MRAC, no fault injection (get nominal performance)
   3. Single axis PID inputs (time permitting)

We completed 5 flights and partially met objectives 1 and 3; notes as follows for each flight number:

Loki Flight 13: FDI filter, LQR+MRAC controller, fault injection. Upon controller engagement aircraft nosed down significantly. Inspection of the ground station revealed elevator command pegged at zero. Flight terminated. Ground tested the other MRAC program (no fault injection) and found similar behavior. 

Loki Flight 14: FDI filter, LQR controller. 5 runs of approximately 25 seconds each.

Loki Flight 15: FDI filter, baseline controller. 5 runs of approximately 25 seconds each. Large heading changes due to wind and controller being unable to maintain a good wings level flight condition.

Loki Flight 16: LQR controller, multi-sine PID input on elevator only. Multi-sine sweep (.2 to 1.9Hz) on elevator. 5 runs. Aileron/rudder under LQR control for duration of PID maneuver.

Loki Flight 17: LQR controller, multi-sine PID input on aileron only. Multi-sine sweep (.2 to 1.9Hz) on aileron. 2 runs. Elevator under LAR control for duration of PID maneuver. Rudder held fixed at last LQR command. When PID input was engaged (t0+4sec) the aircraft entered a right hand spiral. The maneuver was abandoned and attempted again, with similar results. Flight terminated. Landing was difficult to increasing winds and gusts. Decided to terminate deployment.


Actions & notes:
1. Effective immediately, no program will be flown that does not compile from the SVN repository. PIL simulation plots must be presented prior to flying.
2. Winter flying will be restricted to low-wind (<10MPH) days. Finger tip gloves and hand warmers are essential. A wind break would be nice as well.
3. On flights 16 & 17, the static pressure reading was around 874KPa, which is about 8X too high. This scenario also happened on [../2010_09_16], Loki Flight 01. It may be related to the sensor, but its worth noting these pressure values are outside the viable range of the sensor.
4. Prior to these flights, the motor was vibration isolated with rubber washers. Ground tests showed mixed results- some tests showed improvement in the noise levels, others showed no change.  The following two figures compare the auto spectral density of the Y acceleration Loki Flight 12 from [../2010_10_15] and Loki Flights 16 and 17 from this test. The first figure shows improvement in the noise content, the second figure shows little difference.

[[Image(flight4_vs_flight4.jpg, 75%)]]
[[Image(flight4_vs_flight5.jpg, 75%)]]



Back to FlightReports