endotherm
 Captain
Flight distance : 503241 ft
Australia
Offline
|
I've been going over the flight data for the past few hours and this one is a doozy.
As an observation, your flight control movements during this flight are very abrupt, quickly going from zero to full deflection. This results in violent movements, including full acceleration to maximum braking in short order. This has been known to crack airframes and throw props on various aircraft. We regularly see your pitch and roll angles at 40° or more quite frequently, which is the limit of controlled flight for P4.
I examined your comments and tried to match them to the flight record.
While Flying I Received Low Battery Warning,
There was a low power Go Home initiated at about 36%. This would be due to the rate of power consumption, and a calculated safe return home with sufficient reserve from that point.
Returning GO Home initiated,
I Canceled RTH and Manually Returned (sport Mode),
While there is nothing wrong with flying it back home normally, resuming a maximum power Sport mode flight home on a low battery is questionable. Fortunately you were close enough and the battery didn't go anywhere near critical, but it is something to keep in mind in future. It would be more sensible to fly it back gently once a low power warning is initiated.
High Wind Velocity warning, (not uncommon) it was not a windy day
We hear this one all the time. You were standing on the ground at the home point. You can comment on the wind strength at your location. You are likely to have no idea what the wind is doing at the aircraft's position and altitude. It is very common to have high winds a few feet above the ground, where it is calm. I did not see any wind error in the flight record.
Received Warning: Weak GPS Signal,
Received Warning: Ultrasonic System Error: Land as soon as possible,
Received Warning: Motor Obstructed,
Received Warning: Gimble Roll Reached Movement Limit,
The Controls Auto Changed to Atti Mode,
Received Warning: Compass Error: Exit P-GPS Mode:
Received Warning: Weak GPS Signal. Positional accuracy may be compromised.
These warnings are typical of an aircraft that is in the process of tumbling and crashing.
Phantom 4 Pro goes into a Sudden uncontrollable clockwise spiral inverting several times and Registered horizontal speed of 108.2 mph, 128 Ft. off the ground.
I notice at around 18 minutes into the flight, your incident occurs. There appears to be a clockwise spiral, however there is a distinct manual left stick right and right stick down command issued and held for seven seconds and consistent with the flight path recorded.
Spinning and crashing, all control over the Phantom 4 Pro was lost, The Phantom 4 pro crashed down into cedar trees in a vacant lot next to us, sustaining damage to housing, props, Camera, gimbal, ex. I was in complete Shock that this could ever happen. There were no obstacles or collisions, no bird strikes, weather was clear blue skies, no wind ex. Just sudden uncontrollable clockwise spinning. All of the Errors and crash happened within seconds of each other and the video and control inputs show hover, descending and then sudden loss of control.
The issue is what initiated this crash sequence. Up until 18 min the flight was pretty normal and predictable. The flight log shows you held maximum down, rotate right, and move backwards on the sticks at all times during the crash. That is 3 stick positions out of 4 on the way to a CSC by the way.
Referring to the flight record alone, I'd question the accuracy of the altitudes recorded. It started in sync with the VPS altitude, but drifted further and further apart during the flight. The last GPS location recorded does not line up with the last point of the kml data. It seems to continue on further into the trees and comes to rest against a fence. Can you confirm on the map/picture where the aircraft was recovered?
The post-crash altitude appears to go to around -92ft, which is impossible. The land changes only a couple of feet between home and the crash site, and is typically flat. This is in contrast to your claim of "horizontal speed of 108.2 mph, 128 Ft. off the ground". Typically the crash speed was no more than 40mph for short instances whilst tumbling. We can see the tree line is no more than 10-20ft.
I'm going to suggest that the aircraft was far closer to the ground than indicated, maybe 20 feet? If we reduce the indicated altitude by 100feet, to account for the "negative altitude" recorded in the crash, things start to look more natural and explainable. We can see around 17 min the VPS sensor records an altitude of about 10-12 ft while zooming along at 40mph, while the barometer records 438feet! As expected the VPS sensor readings cut out when the altitude increases beyond its (40ft?) limit. To me, the VPS sensor is providing more reliable readings than the barometer.
If we extend this to 18 minutes and the crash starts, and looking at the map, I suggest the aircraft clipped and "bounces off" the side of the house at the back corner at 10-20ft altitude. This resulted in a sudden violent nose pitch down at -40° briefly upon impact, followed by a reactionary +40° nose pitch up to counter. At the same time we see the airframe roll +25° through -35° over the next 3 seconds. (See pitch/roll columns from 18:00 onwards, this is recorded exactly as the described "bump" would react). The rest of the path and various rolls and tumbles would coincide with clipping and crashing into the tree tops in the image. At the reported altitudes, there is nothing to hit, clip or fall through that would cause these rolls and tumbles. It is more reasonable and rational to attribute them to interaction with the trees themselves, as the movements correspond with locations of treelines. It is possible a prop blade or tip could have snapped off at the initial location, initiating a spiral tumble -- try a thorough search at the corner of the house, and inspect the wall for impact damage.
From the pilot's perspective, it would appear the flight was normal followed by an inexplicable spin and crash, especially if he believed he was flying at hundreds of feet and not close to the ground. In reality, the aircraft clipped something and tumbled out of control.
I'd suggest the reason for the incorrect altitude needs to be investigated. It started in sync at takeoff, but was wildly incorrect at the point of the crash. Had the aircraft been taken out of an airconditioned/pressurised environment immediately before flight? Was the temperature wildly different? This could explain a drifting barometric pressure and a change in reported altitude.
Normally I'd consider this to be a case of pilot error, having flown it into the side of his house on the way home. However his instrumentation seems to be faulty and misrepresenting the true altitude. Perhaps if it were left to its own devices to perform it's Go Home or RTH by itself, this wouldn't have happened (or if it did, it would be 100% a DJI responsibility). I would suggest DJI should provide warranty if the crash is due to faulty instruments. You are fortunate that you can fall back to your own insurance if you are not satisfied. If DJI repair it or replace it with a refurbished item, you will still have 100% warranty and it should be no different than a brand new one. It may be replaced with a brand new one due to it's rarity and age. If you are particularly concerned with this, make an insurance claim and get paid out or get it replaced by a brand new one.
I still have the DAT file to examine. It has a greater number of parameters and data points to examine so it may take a while... I'll see if it reveals any clues.
EDIT: I just got the video to download and play, this does NOT support my altitude theory at all, although it does look a lot closer to the ground than 200 feet, and it doesn't explain the negative altitude during the crash.
I've examined the DAT file recovered from inside the aircraft. It shows a failure of the right front motor immediately prior to the spiral commencing. I haven't been able to see what caused the stoppage, although it does not look like a prop failure. The associated ESC temperature dropped and was cooler than the others, so we can assume it was not feeding the motor like the remainder were. This suggests the motor was not rotating; with a damaged prop it would still be trying to rotate.
|
|
lvl.3
lvl.4
lvl.1
