Ryerson flying- Sat. Dec 3 12-2:30

Started by Frank v B, November 30, 2016, 10:05:55 PM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

Frank v B

We are planning to meet at the field for a Ryerson test flight this Saturday from 12 to about 2:30.  Fingers crossed about the weather.  We have to keep the white stuff (Andy's dandruff) away.   ;D

Frank

"Never trade luck for skill"

sihinch

If the weather''s good I'll join you!!!

sihinch

Man it was cold today - the wind helped! We even had a bit of snow.

But it was great fun! The Ryerson plane flew the best I'd ever seen it. I had the pleasure of meeting "the Prof" and we talked about some exciting potential projects with Quads. And I flew my new Theory Type W alternative, the Wing Wing Z-84.

Frank v B

Alton achieved autonomous flight today!!! Congratulations.

We had two flights. 

Flight one:
The first was about 6 minutes at full throttle to complete the auto tuning so the plane could settle into the proper roll and pitch angles.  Spent about two circuits to trim the plane for today's conditions, then two circuits of roll axis tuning (alternating max left and max right while flying in a straight line).  Then two circuits of pitch tuning (max up and down) while flying in a straight line.

Flight two:
Alton had loaded a flight plan that would see the plane loiter (lazy circle) while maintaining a constant altitude programmed for 150 meters.  The loiter flight plan was over the large area in the south east corner of the field.  The flight plan clearly showed on his laptop computer screen superimposed over a Google earth photo of the field.

The take-off was to the north (road) and then a right tun and cross over the field coming towards us.  This was done so that when Alton flicked the switch to autonomous flight, the plane would immediately turn left and go to its corner.  This is so we would know immediately that it had taken command.  As soon as he switched to autonomous flight on the transmitter, the plane veered left and headed to its GPS corner.  It did absolutely perfect lazy circles at full throttle and 150 meters.  Rock steady, not a single twitch or glitch.  Guaranteed no pilot who has ever flown at our field could duplicate its accuracy and steadiness.

The plane flew on its own for 6 minutes without a single input from me.  Alton then enlarged the circles on the screen and sent the command to the plane and it made visibly wider circles.  Again, rock steady.
He then changed the altitude from 150 meters to 100 meters.  When he sent the command, the motor speed backed off automatically until it reached 100 meters AGL and then automatically returned to full throttle.

BJ was there to witness the flight.  Here's to hoping Alton can download the video soon and BJ can comment on it.

A great day for Ryerson and a great day for TEMAC.  Congrats Alton.

Oh, and yes Simon... it was cold and blustery.  Agreed.

Frank
"Never trade luck for skill"

altonyeung

#4
What a wonderful day. Thanks again for helping us out Frank.
Yes our plane can finally fly on its own.

I absolutly love the new camera angle, the rear camera can clearly capture the aileron deflection and without too much of the wing blocking the view.  ;D

Flight 1 (Autotunning)
https://www.youtube.com/watch?v=xm4YFQjge0k

Flight 2 (Loitering)
https://www.youtube.com/watch?v=5MnBG7ipoRw

sihinch

Congrats Alton! That's mighty impressive!

BJROB

Yes it was very cold.
The wind was gusty. This plane flew so steady.
The flight path pattern was change. It did not matter.the plane flew stead.
What an awesome job Our students did....
I'm so glad I froze a little to see such a heavy plane fly. It was .... Awesome
My Work??? is so secret....
I Don't even Know what I'm Doing!!!

Frank v B

Wow! very revealing.  Run both video 1 and video 2 at the same time and notice the difference in aileron deflection. 

Video 1- manual control for the entire flight.
Video 2- autonomous flight starting at 2:58 video run time.

1) In video 1, once the turn is initiated with left aileron down, the roll correction (left aileron up) starts immediately and lasts for the entire rest of the turn.  This airplane is very unstable along the roll axis once it is initiated.  The aileron deflection is fighting to correct it.
2) In video 1, the aileron deflection is about twice as much as video 2 when on autopilot... proving there is a drunk (me) at the sticks in video 1.  8)
3) The aileron deflection in video 2 is very smooth and gentle with the only small changes when it crosses the wind from the northwest.  The aileron deflection changes from fighting the wind when turning into it to easing out of it when it pushed further into the turn.  Remember this plane is holding a GPS position and perfectly concentric circles, not just random lazy circles while it is being carried downwind.  It automatically compensates for the effect of the wind and gusts.
4) In video 2, the landing started as a cross runway approach but ended up straight down the runway.  I was aiming for a grass landing but suspect that once the plane got safely below the tree line the plane automatically overcorrected to the runway because I was crabbing* to get onto the grass.

Note: this airplane is a one landing attempt plane.  Attempting a go-around would have a nasty end with very high probability.  It has gone on a diet but still weighs 16 pounds 12 oz.!!

Video 2 timeframes
2:58- autonomous flight starts with clockwise circles.
8:52- loitering in larger circles.  Note that it actually used left roll to immediately assume the larger circle.
11:03- altitude reduced from 150 meters to 100 meters.
13:58- the drunk takes back control for the landing.

Phew.

Thanks to Temac members for letting us share our sandbox with Ryerson. ;D


Frank

* Andy, a term used in flying... not describing my personality. ;)


"Never trade luck for skill"

altonyeung

I have prepared the KMZ file of the autonomous flight for those playing along at home. You can superimpose the 3D flight path using Google Earth.  ;D ;D


sihinch

I hope they don't build that big green & yellow cylinder thing on top of TEMAC!

altonyeung

Hi TEMAC,
We got more results from Dec 3 flights. Using the onboard flight log and camera, I have combined the two to display flight info beside the footage. Having the data synced with the video is a bit tricky but it's worth it! Now we know what Frank has done wrong flying our plane.  :P
This is still in development and I am planning to put more data on it and a map on the side with GPS info as well. Let me know what you think!

Alt: Altitude Above Ground Level (AGL) from GPS, [m]
VZ: Vertical Speed from GPS, [m/s], positive UP
GS: Ground Speed from GPS, [m/s]
ARSP: Airspeed from pitot tube, [m/s]
L ARN: Left Aileron Deflection, [deg], positive UP
ELV: Elevator Deflection, [deg], positive UP
THR: Throttle Output [%]


Dec 3, Flight 1
https://www.youtube.com/watch?v=T3c_cVdD9eo

Dec 3, Flight 2
https://www.youtube.com/watch?v=uStFfQezSzc

Frank v B

Alton,

re:"Now we know what Frank has done wrong flying our plane.  :P"

You don't have to be that obvious to get Andy going on my flight evaluation.  Given a chance, Andy has an auto-start button. ;D


Frank
"Never trade luck for skill"

sihinch

Hey Alton, great data. What's the units of g/s and a/s, is it meters per second?

altonyeung

Thanks Simon, and yes those speed readings are in meters per second

Candu

Congratulations Alton for achieving autonomous flight! Very informative videos.

Aside from the landing approach, I notice in the video its flying at 100% throttle all the time either in manual, FBWA, or Guided flight modes. In Manual or FBWA modes, throttle is controlled by Frank and we all know Frank likes to fly fast! But in Guided mode, autopilot takes the throttle stick away from Frank but the autopilot is still keeping the throttle at 100%. I'm curious if the heavy plane is under power by any chance.  I'm running the same control software on my much lighter twin boom aircraft and it usually cruse at around 50-60% throttle, manual or autopilot controlled.