Ryerson Training/experimental flights Sunday April 23 '17

[Frank v B]

Both Zach and Alton are hoping to be at the field on Sunday, April 23 from about noon until 2 pm.

Zach flies the basket of wiring and electronics with 4 motors flying in formation (quad).    Last week he photo-mapped the field totally autonomously from take-off to landing.  The GPS driven flight path was pre-selected on the computer screen at the set-up tables.  It even used GPS triggers for the shutter.

Alton is back with the acre of wing with a few tail feathers at the back to continue his quest for totally autonomous flight via GPS.  This plane has more electronics that Frank Candu has on all his planes combined.  When we left the project last fall it successfully flew totally autonomously after take-off at a prescribed height (50 meters).  It needs human intervention for landing.... just someone like me can be blamed for screwing it up.

Frank

[Frank v B]

Alton showed up this morning for the first time this year with the 7' span electric plane.

As a reminder- this is for his Masters Thesis and the purpose is to achieve and maintain autonomous flight. 

We had 3 flights.  The first one was to auto-tune the electronics to have it learn the sensitivity of the plane.  The last flight was at about 80% throttle and a full 13 minutes was totally autonomous with the flight path/plan determined by the person operating the computer beside me.  All of the inputs were GPS determined and the altitude was pre-set.  It kept altitude to within one foot AGL for the entire time, no matter what the throttle setting.  For 13 minutes I did not touch the sticks at all.


We had full and immediate read-out of altitude, speed, wind speed, elevator setting.

The plane flew a lot better this year after they incorporated the jury-rig adjustments that Glenn and I suggested last season (lager rudder, larger fin). Last year we taped plastic card to the rudder and fin to increase the area.  The plane had all new tail feathers today.

Congrats to Alton and the entire Ryerson pit crew.

Frank

[altonyeung]

Thank you Frank for flying the plane for us.   
As he said we performed  totally autonomous flight. On the last one we did over 18 minutes of flight time. I think we can do every better. I designed the plane to do 20 minutes and I will prove it one day 
The plane has a brand new tail to adjust the slow speed flight performance as we worked it out from flight data from last year. It provides greater lateral stability and more rudder and elevator surfaces. The brand new tail design is why more autopilot parameters tuning had been done. The tuning involved the plane rolling and pitching back and forth as you can see from the video below.

Flight 1: There was a glitch in telemetry system and realtime downlink isn't showing on Bill's laptop. We had to end the test early.
https://www.youtube.com/watch?v=xzjm0fR9TlM

Flight 2: Pitch and Roll tuning
https://www.youtube.com/watch?v=pAE9TAsohWk

Flight 3: A bit more tuning and full autonomous flight heading back and forth the waypoints while maintaining 150 m AGL.
https://www.youtube.com/watch?v=feZJLbBJUuU


Once again, Thank you TEMAC for letting us fly here.

- Alton

[Frank v B]

Just in case you guys think I was drunk while flying.  In the first two flights when Bill said he wanted to autotune the electronics, I had to fly several circuits to autotune the pitch axis.... that's why all the up and down stuff.  Full deflection elevator up and down.

When auto-tuning the roll axis, the aileron inputs were full left then full right for several circuits.

Notice in flight 3 at 1:21, the plane suddenly dove to the ground in the left turn.  We turned auto-pilot off and I took over manually to level it out.  After that, 17 minutes without touching the sticks.

Great achievement Alton.

Frank


Frank

[altonyeung]

Just in case you guys think I was drunk while flying. ...

The plane is doing the roll especially drunk on these few flights and Bill and I are not sure why. The data log was showing the rudder wasn't deflecting during the rolling which it should have. This only happened during the autotune segment where we roll and pitch to the extreme. The rest of the flight the rudder did deflect according to the roll input to maintain a coordinated turn by counteracting the adverse yaw effect. We will investigate.