I attempt to use automatic landing, but there was a obviously hard touch down, after check the log, I found the aircraft didn’t establish the pitch angle during flare phase, causing touch down with a very high descent rate.
BTW, I used rangefinder to achieve more accurate altitude control
But only just! The actual sink rate didn’t have time to catch up. This is in part related to the inherent delay of the actual pitch angle to track the pitch angle demand.
I recommend that you pull your flare start a little earlier, by a second or two, and adjust to whatever that means for the flare altitude.
Additionally, if your stall speed is well below 16m/s (your landing speed), you could also add in a preflare segment.
and then tecs try to establish positive pitch angle, but had waste too much time because vehicle need to pitch down first and then pitch up starting from a large negative pitch angle
if we try to flare a bit earlier, that may cause stall due to throttle had cut off, anyway, I will try to add 1 or 2 meter above the current flare altitude
so could you help to figure out why the desPitch show a rapid drop when flare start
we also verify in simulator (Realflight) and have same behaviour
Notice that TECS had already started the pitch-down before the flare (blue cursor). And immediately starts increasing the pitch demand after the flare starts (red cursor).
But during that time, the attitude control loop has ran (or at least logged) twice, propagating that low pitch demand.
All of this happens in less than 100ms. And indeed the update rate of the TECS controller is 10Hz. The attitude control loop rate is 50Hz. So you couldn’t get a faster response in any circumstance.
Raising the flare height is the correct option here, you are operating too close to the limit.
I try to retest the flare today with higer flare altitude (LAND_FLARE_ALT = 4) and it work, noticeable flare maneuver was observed with smoothly touch down.
we also use the flap to achieve a lower touch down speed (TECS_LAND_ARSPD=12)
