False Altitude Results in Crash

[RocketBrew]

I was flying a mapping mission at a rock quarry using Map Pilot with the terrain awareness feature. I have used the same software with two different P4P aircraft on this site three times previously without any problems.

The first three batteries went fine, but after taking off on the fourth battery something went wrong. Although the aircraft was following the terrain, as it approached a steep slope it didn't gain enough altitude and hit a bush on the top. After an hour climb up along the Bighorn Sheep trails, I recovered the aircraft. (Came back down the blasting road I found at the top...)

Here is the odd part, the KML exported from Airdata in Google Earth shows the aircraft climbing as expected, and it is also shown above the top of the hill, where it was supposed to be, but obviously, it was not.

The DAT files also show something odd, there are multiple files with timestamps going forward and back in time.
01/31/2018 11:16 AM 165,138,432 FLY102.DAT
01/31/2018 11:41 AM 171,307,008 FLY103.DAT
01/31/2018 12:03 PM 156,315,648 FLY104.DAT
01/31/2018 12:27 PM 173,850,624 FLY105.DAT
01/31/2018 12:28 PM 61,276,160 FLY106.DAT
01/31/2018 12:33 PM 33,554,432 FLY107.DAT
01/31/2018 12:35 PM 45,109,248 FLY108.DAT
01/31/2018 12:28 PM 225,280 FLY109.DAT
01/31/2018 12:32 PM 25,780,224 FLY110.DAT
01/31/2018 12:28 PM 12,288 FLY111.DAT
01/31/2018 12:28 PM 118,784 FLY112.DAT

The battery was laying next to the aircraft, so it would appear to have popped out when it hit the ground.

The aircraft is operating fine on the workbench, the IMU was showing good but I recalibrated anyway. Will take it out this weekend for a cautious test flight, but not sure how to accurately test the altitude reading. Maybe fly both p4P's next to each other and compare readings? This was not a case of it being a few feet off as might be expected, it should have been 220 feet higher than the top of the hill.

 

[With The Birds]

How high above your home point was the point of impact?

 

[alokbhargava]

To OP: are you saying that you could follow. Same route with three batteries and hen you tried with 4th battery, it hit the bushes?

 

[Fly Dawg]

Looking at the data, the maximum altitude for this flight was only 140 feet above the takeoff point.
I would take the data from the 3 previous flights and compare the recorded home point altitude to this flight and see if they are significantly different. That would be a starting point.

 

[RocketBrew]

How high above your home point was the point of impact?

Airdata shows max altitude of 469.8 ft. Google Earth shows the top of the hill at 377 feet ATL.

 

[RocketBrew]

To OP: are you saying that you could follow. Same route with three batteries and hen you tried with 4th battery, it hit the bushes?

This is a large several hundred acre site, it takes four flights to cover completely. This was the fourth flight. I have flown this same site with this same aircraft twice previously, and once before that with another P4P.

 

[RocketBrew]

Looking at the data, the maximum altitude for this flight was only 140 feet above the takeoff point.
I would take the data from the 3 previous flights and compare the recorded home point altitude to this flight and see if they are significantly different. That would be a starting point.

I will have more time to dig into the files this weekend. The logs on the tablet appear to have different altitude data than the onboard DAT files at that rate...

Any idea why there are so many DAT files right at the end? I am thinking the falling/tumbling after hitting the cholla bush caused it to restart logging multiple times until the battery popped out when hitting the ground?

 

[sar104]

It looks like the problem here is probably in the DEM used for the flight plan. The aircraft DAT file shows that the GPS altitude and barometric altitude agree quite well. However, comparing the ground track to the flight path to estimate flight altitude AGL shows a problem:

At the end of the flight the calculated altitude AGL drops to just a few meters - that's not nearly enough clearance given the inherent errors in both the DEMs and the aircraft altitude readings.