Re: [Paparazzi-devel] Autopilot crashing mid-flight

[Loic Drumettaz]

Hi,

I had a quick look at your airframe file. You may have too much elevator_of_roll. This will reduce speed while turning.

Some tips about the previous messages you sent:

- If the cruise speed is significantly faster than stall speed, you probably don't need rudder_of_aileron mixing. 0.75 sounds a lot. Too much would be as bad as nothing. Have a look at adverse yaw on google to understand this setting.

- If your cruise speed is 23 m/s and maximum bank angle is set to 29 degrees, then your minimum turn radius must be around 97 m (turn_radius = ground_speed^2/(9.81*tan(bank_angle)). If your flight plan or photogrammetry function ask for a turn radius less than 97m, then your aircraft won't be able to follow the setpoints.

Good luck

Loïc

2014-04-28 7:55 GMT+02:00 Loic Drumettaz <address@hidden>:

Hi Matthew,

Aileron_of_throttle is supposed to compensate roll disturbance due to throttle input. (This can also be done mechanically by motor mounting adjustment.)
So this is not what you wanted to do: add throttle with ailerons.
I don't think such a setting exist: anyway you don't want that, it wouldn't help.
What you need to keep a constant speed is the appropriate elevator_of_roll setting.
I'll have a look to your airframe config file to see if i can be any help.
Regards
Loic

Le 28 avr. 2014 06:53, "paparazzi_guy" <address@hidden> a écrit :

Hello everybody,

The weirdest thing happened today during our test flight. The plane was
flying perfectly in manual, fly by wire, and full auto. We wanted to
increase the throttle during the rolls since it was always losing speed when
turning. So we changed the gain AILERON_OF_THROTTLE from 0 to about 500. I
thought this was a pretty reasonable amount since it gave me a range of 0 to
5000 with this value. However, once we committed this change the plane
started spiralling and dove straight towards the ground. The pilot switched
back to manual but had to land the plane because it had taken some damage. I
took a picture of some of the logs and will provide the logs themselves in
this message as well. The time that I committed the change was around 2125s
and I reverted the change around the same time the pilot switched to manual
at 2200s. It appears that the IMU and some other sensors died during this
time for some reason, which is probably why it was starting to crash.

Do you guys know why the plane would want to do this based on what happened
and from these logs? Any input would be very helpful. We're going to try and
repair the plane, but we leave for competition tomorrow. Luckily we have a
backup, but it's not as nice.

I will also include our airframe configuration if it would help.

Thanks,
Matthew

14_04_27__18_09_14.log
<http://lists.paparazziuav.org/file/n15211/14_04_27__18_09_14.log>
14_04_27__18_09_14.data
<http://lists.paparazziuav.org/file/n15211/14_04_27__18_09_14.data>
Snitch2.xml <http://lists.paparazziuav.org/file/n15211/Snitch2.xml>
Death_of_the_autopilot.png
<http://lists.paparazziuav.org/file/n15211/Death_of_the_autopilot.png>



--
View this message in context: http://lists.paparazziuav.org/Autopilot-crashing-mid-flight-tp15211.html
Sent from the paparazzi-devel mailing list archive at Nabble.com.

_______________________________________________
Paparazzi-devel mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel