FW raise min airspeed based on commanded bank

[dagar]

Needs review and testing.

• closes FW Minimum Airspeed scaling for banked stall avoidance #7572

@JonReacher @Antiheavy FYI
I'll share a link to the binary once it's available.

[see3peoh]

@dagar Question about this and the math. Maybe the commit says the following but I'm couldn't get it to do so (granted its the middle of the night here).

Load factor at bank should be 1/cos(bank angle). Let's assume 45 degrees so thats a new 1.41 load factor

Isn't stall speed at bank just = level Stall speed * load factor at bank? Let's assume 10m/s stall speed so 10 *1.41 is 14.1 m/s (new stall speed at 45 deg bank)

Am I off here or does the above in the commit say this as I kept getting a different stall speed and load when following the math in the commit.

[JonReacher]

@see3peoh You're correct about formulating the load factor, but remember that stall speed increases by the square root of the load factor, not the load factor itself. So at 45 deg bank angle, you calculate a 1.41 load factor. You then find the square root of 1.41. You get ~1.19. So the new scaled stall speed is Stall speed*1.19 @ 45 degrees.

[see3peoh]

@JonReacher that makes soo much more sense. Thanks for the clarification. @dagar sorry for doubting you!

[dagar]

@see3peoh don't apologize, thanks for looking.

@JonReacher do you want a binary to test? Which board do you have?

[JonReacher]

@dagar Didn't see your question until now. I have a standard pixhawk. I'm down at the moment, need to build a new airframe for testing. It shouldn't be too long though. I'll be happy to have a binary to test.

[Antiheavy]

I'm also willing to test this.

[LorenzMeier]

@JonReacher @Antiheavy we're blocked on your tests with this PR.

[Antiheavy]

Got a couple flight tests done today with a build including this PR.

Survey and loiter patterns with increasing FW_AIRSPD_MIN:
https://logs.px4.io/plot_app?log=018edc80-b2ea-4707-90dd-ba899bdc0cf8

Survey and loiter patterns with increasing FW_R_LIM:
https://logs.px4.io/plot_app?log=20f8405a-ca05-441c-914b-60f9a460af8a

@dagar should take a closer look, but it seems be be behaving as intended.

[Antiheavy]

More data:

Here are the airspeed and roll angle set points from a recent flight - airspeed setpoint definitely seems to increase whenever max roll angle is commanded. It seems to match the math in the PR. Airspeed setpoint is dark blue line, scaled 2x (m/s):

from this log:
https://review.px4.io/plot_app?log=f3da31a6-c015-4bab-ad23-8de3ccdd321b

[bresch]

It's theoretically correct and can only help. It's fine for me to merge those modifications.

[dagar]

Thanks for the review.