Need Help Determining Cause of Crash

[grantcoll]

Hi,

Two days ago, I crashed my Mavic Pro 2 while hovering and taking pano photos at the edge of a cliff face. I suddenly got a warning that motor was at max speed and that gimbal roll reached movement limit. I tried to bring it home but saw the scenery on screen go to a 45 degree angle, then lost signal as it veered away from me atop the cliff, never to return. (As an aside, if there's anyone near Moab, Utah who can drive up a very rough jeep road to look for drone, let me know) Here's the fight log:

DJI Flight Log Viewer | Phantom Help

Upload and view flight logs from your DJI drones

www.phantomhelp.com

Note that, although it says I was at 1400', I took off near bottom of a 1200' canyon wall, so I was shooting images just 200 feet AGL.

I did get high wind warnings on my screen, but winds that day were 11 mph with gusts up to around 20 mph. Drone was flying fine and hadn't had issues with that wind speed before, so I kept flying. Before taking off, I got a warning that vision sensor needed calibrating. I ignored this since I had the obstacle avoidance sensors disabled. Or should I have heeded this warning anyway?

I had done 250 flights in the last 3 years without anything close to a crash, so I was probably overconfident and should have heeded warnings. But I don't think the warnings were the cause of crash. I'm hoping to find out for sure, so I know how to prevent this in the future.

The flight log shows that after I got the motor speed and gimbal warnings at around 10 min 28 sec, the drone suddenly spun around counter-clockwise and lost altitude despite the remote control not being used. The log shows I tried to turn it back clockwise and fly up. I got it positioned towards home and tried to go full throttle forward. But it careened backwards to the NE at close 40mph before losing signal. My best guess is that a propeller fell off. I check the propellers for damage before every flight but didn't check to make sure they were secure.

Thanks,

Grant

 

[Meta4]

Hi,

Two days ago, I crashed my Mavic Pro 2 while hovering and taking pano photos at the edge of a cliff face. I suddenly got a warning that motor was at max speed and that gimbal roll reached movement limit. I tried to bring it home but saw the scenery on screen go to a 45 degree angle, then lost signal as it veered away from me atop the cliff,

The problem first shows in your flight data at 10:25.2 when the pitch and roll data show a sudden change.
The roll had been steady at 15° (roll to the left) and it suddenly rolls to the left 42° past level.
There is also a smaller change in the pitch angle.

1.7 seconds later the drone starts to slowly spin in an anti-clockwise direction and lose height.
About a second later, the spin and height loss are faster.

Spinning and losing height usually indicated loss of a prop or motor.

You managed to regain some height and slow the spinning, so it may have just been a lost blade rather than the whole prop.
The drone was well above the terrain height, so bumping into a rock or tree can be discounted.

Signal was lost when signal was blocked by the approx 200 ft ridge near the top of the canyon.
The drone would have come down somewhere near the last GPS location, but not necessarily right at that spot.

Before taking off, I got a warning that vision sensor needed calibrating. I ignored this since I had the obstacle avoidance sensors disabled. Or should I have heeded this warning anyway?

That wouldn't have contributed to this incident.

 

[eEridani]

As far as the wind goes, what is down in the valley can be quite a lot different than higher up on top of plateaus. On top it was blowing 30 mph with gusts over 35 mph.

 

[Meta4]

As far as the wind goes, what is down in the valley can be quite a lot different than higher up on top of plateaus. On top it was blowing 30 mph with gusts over 35 mph.

Strong winds might cause a reduced speed over the ground, or in extreme cases could lead to a drone being blown backwards.
But it won't blow the props off.

 

[slup]

...I got the motor speed and gimbal warnings...the drone suddenly spun around counter-clockwise and lost altitude...My best guess is that a propeller fell off.

The incident starts off by a quick deflection in both pitch (green graph), yaw (purple graph) & especially in roll (blue graph ) at 625sec into the flight (where I have placed the chart marker below). Slightly later "NotEnoughForce" errors are shown (yellow background) & remains active until the log ends ... this indicate that some of the motors have reached max rpm & can't rotate faster.

At 628sec pitch & roll stabilize somewhat but the craft starts to slowly yaw CCW (it take over 10sec for it to rotate 1 full turn).

So this behavior isn't at all typical for a complete loss of thrust from one of the 4 corners ... instead it points to a partial damage influencing the thrust generation & making some motors rev up to make up for the lower amount of generated thrust during normal revs.

If you share the mobile device .DAT log from this flight we possibly will be able to say more ... the .DAT log for this flight ends with FLY061.DAT & is stored where you found the .TXT log you uploaded to PhantomHelp ... but in a sub-folder MCDatFlightRecords.

(Here the relevant telemetry ... click on the chart to make it larger)

...I did get high wind warnings on my screen

Wind will not cause a craft to yaw uncommanded or lose height ... if the wind is to strong the craft will instead drift away in the wind direction tilted to the max allowed angle to fight the wind. So winds was not the problem here ...

 

[Capt KO]

I remember an earlier post where one of the tabs broke off the bottom of the prop and went into the motor. With the blade still attached could this have caused the additional power draw to that motor while still remaining airborne? Or would that have caused a “blocked motor” warning?
Looks like it will require another drone for retrieval. Sorry for your loss.

 

[diablo2112]

The broken prop mounting tab causing a stalled motor was mentionned in the Mavic 3 Lands-on-Karen's-Roof thread. That was my thought here, too. This particular design element of the DJI prop mounting system appears like a possible issue. Those are tiny plastic tabs, and they are stressed during rotor acceleration/deceleration, something that happens constantly on a quadcopter. Those tiny pieces can potentially end up inside the motor, where they might interfere with the free-rotation of the motor housing.

 

[Capt KO]

The broken prop mounting tab causing a stalled motor was mentionned in the Mavic 3 Lands-on-Karen's-Roof thread.

That’s the one, from PhantomRain. Think op will need a longer tape measure retrieval hook.
Hope he gets it back…eventually

Learned important lesson from the Karen post

 

[slup]

...tabs broke off the bottom of the prop and went into the motor...Or would that have caused a “blocked motor” warning?

The broken prop mounting tab causing a stalled motor... was my thought here, too.

We will know for sure when/if the OP shares the DAT log ... but NotEnoughForce messages are usually due to motors being on max RPM ... if they are prevented to rotate that usually generates a Blocked Motor message.

 

[diablo2112]

Good point. This is likely a broken/failed prop then. The rest of the behavior is consistent with this.

 

[fguthrie]

Sorry you lost your drone ... looks like a beautiful place to fly.

 

[Defens]

Looks like there is access up there. If you don't own (or have a friend who owns) a Jeep or other off-road capable vehicle, might be worth renting a Raz'r or other UTV in Moab and going for an outing for retrieval. Take another drone with you to recon the general area where it went down once you're up on the plateau.

 

[slup]

...might be worth... going for an outing for retrieval.

Will most probably turn out to be like searching for a needle in the haystack unfortunately ...

The craft was approx 115m above ground when the RC-AC connection was lost ...in the end the craft was slowly rotating uncommanded & it couldn't quite hold position instead drifting away to a approx north easterly direction (looks to be straight with the wind direction) & slowly losing height.

After the connection was lost the failsafe RTH would have been activated, which would have started with a ascent of 44m up to the RTH height ... so where could it have ended up, would it actually be able to ascend, and then... continued further down wind or would it have been able to make any head way towards the HP..?

When the log ended nothing were indicating that the craft had started to fall straight down ...

 

[grantcoll]

Thanks for all the feedback. It is helpful. I've attached 3 DAT files from date of crash. These were in the "flight logs" folder, not in the "MCDatFlightRecords" where someone told me to look. In that folder, there were only 10 files with dates from 2021. So I'm not certain if the attached files are the right ones.

 

[grantcoll]

Will most probably turn out to be like searching for a needle in the haystack unfortunately ...

Yeah, I figure it will be hard to find. Good thing is that it was near a jeep road and moving in same direction as jeep road when signal lost. But I can't drive up there, as I've two spinal fusions and the road is too rough for me. Been looking for someone in Moab who might be up that way who could look. I have a bright orange skin on it for the very purpose of making it easier to find. Hopefully skin didn't contribute to crash by peeling off, but highly doubt that since it had stayed almost fully intact for 3 years.

 

[grantcoll]

Looks like there is access up there. If you don't own (or have a friend who owns) a Jeep or other off-road capable vehicle, might be worth renting a Raz'r or other UTV in Moab and going for an outing for retrieval. Take another drone with you to recon the general area where it went down once you're up on the plateau.

I have a Land Rover that could probably make it up there, but my bad back would prevent me from making it up one stretch that is very rough. Might see if I can get close enough to fly my older Mavic up there to recon, should I find anyone who can go look for it for me.

 

[grantcoll]

As far as the wind goes, what is down in the valley can be quite a lot different than higher up on top of plateaus. On top it was blowing 30 mph with gusts over 35 mph.

Yeah, that was my first thought when it crashed. I was relying on forecast from airport, since that is up above the canyon as well, so I figured it would be 10-20 mph. Is there a way to see actual wind speeds while flying in GO4? I don't feel quite so stupid now that it appears the wind didn't directly cause the crash. But added forces from wind could have caused something to break, so I think I've learned the hard way to heed the high wind warnings.

 

[eEridani]

Try something like these meets: there are Jeep specific, Toy specific, and general meets. They happen when the trails aren't muddy (normally). But there are lots of them. Maybe one of the groups know locals who'd be happy for an excuse to go wheeling.

https://moab4x4expo.com/

Trail of the Month - Red Rock 4-Wheelers, Inc.

Red Rock 4-Wheelers Inc. - Moab, Utah.

www.rr4w.com

 

[slup]

...I'm not certain if the attached files are the right ones.

Unfortunately those ones aren't the type we need ... meaning that we won't be able to get data support for further conclusions.

The Mavic 2 series is randomly? paired with 2 different types of controllers ... one labeled with RC1A & another with RC1B. The B variant doesn't write .DAT logs ... the A type does. So this could be an explanation that the DAT wasn't there ... just odd that you had DAT's from earlier dates, have you 2 different controllers or have you replaced it earlier?

...it was near a jeep road and moving in same direction as jeep road when signal lost.

Don't forget that once the connection is lost the failsafe action will kick in ... you had that set to RTH. So one other possible direction could also have been a path directed towards your HP.

or ...

When the RTH started, your M2 had as a first step to ascend up to the RTH height ... so it should go up further approx 44m, that might have "killed" it & made it spiral straight down ... so another possibility is that it actually is very near the last known position in the log ... Lat 38.6898 Lon -109.4676

...Hopefully skin didn't contribute to crash by peeling off.

...added forces from wind could have caused something to break.

I doubt either of those is responsible for this incident ... instead it looks like a possible bird strike with that big deflection mostly concentrated to the roll axis, from the telemetry something could have hit & damaged the front left prop.

...Is there a way to see actual wind speeds while flying in GO4?

To get a forecast on different heights many use the UAV Forecast app, but that is just ... forecasts, but better than nothing. Good for some pre-flight initial indications at least.

Once air borne you first direct your craft, nose straight into the wind ... use the GO4's attitude indicator for knowing that you're directed with a head wind ... the lower blueish area there should be horizontal as it is in the pic below. Once directed correctly you fly with full speed into the wind & take note of the ground speed the app says your craft can achieve ... max air speed in Sport mode is according to the spec. 72km/h (or 44,7mph or 20m/s).

So if your craft only achieves a ground speed of ... let's say 30mph, you know that the head wind speed is:
44,7(max air speed) - 30(achieved ground speed) = 14,7mph.

 

[Yorkshire_Pud]

Sorry about the drone.
This interested me, I have two damaged mavic 2 propelors so I cut the damage off one blade on one propelor and cut down both blades on the other propelor.
These were flights inside the house.
With out of balance propelor on rear motor the drone flew well. Kept height, staedy, level.
With the out of balance propelor on the front motor the drone required much throttle to be controlable, with not much throttle it sank and was difficult to keep steady but was level. I did not see yaw.
I did only short flights. I did not want damage my drone,
The drone flew well with balanced propelor on ether the front or rear motor.