Emergency Unpowered Descent

[OmegaOmega]

Hey All - question for anyone who either has experience with this - I can't seem to find an answer or video.

Barring any violation or concern of regulations/laws, if an individual were to take their TH up to 1500ft+ altitude but needed to rapidly descend, it would seem the fastest way would be to cut power to the motors; allowing the device to begin a free-fall.

In the simulation software, when you fly the Q, you can restart the motors and the craft will correct itself and respond to throttle commands, however, I am wondering if this is a situation that would happen in reality?

I've seen posts on the DJI boards that "people have done this" but when pressed for videos or proof, they simply reply with "just google for a video on YouTube" - I have, and I can't find anything. I've tried 14 different keywords but the closest is a CSC video from about 20' where the individual was testing whether the system would actually shut off if it knew it was mid-air. It did. The next closest are simple wreck videos of drones failing mid-flight... etc. None that pose an intentional power cut mid-flight with attempted power restore and in-descent stabilization.

On my DJI, I would suspect this wouldn't work because the IMU doesn't show interactive function once you land and power down. However, when I've hooked directly to my TH, I notice that nearly all of the sensors remain live/active and appear to be functional.

If I were rich and carefree I'd be tempted to try it out myself. This isn't a question to challenge legality or rationale of such a maneuver, more just curiosity around the general physics and capability of the craft itself.

 

[Steve Carr]

It will be your final flight. You can count on it. I'm sure it has been tried with other brands, all unsuccessfully.

 

[mcbutler]

Hey All - question for anyone who either has experience with this - I can't seem to find an answer or video.

Barring any violation or concern of regulations/laws, if an individual were to take their TH up to 1500ft+ altitude but needed to rapidly descend, it would seem the fastest way would be to cut power to the motors; allowing the device to begin a free-fall.

In the simulation software, when you fly the Q, you can restart the motors and the craft will correct itself and respond to throttle commands, however, I am wondering if this is a situation that would happen in reality?

I've seen posts on the DJI boards that "people have done this" but when pressed for videos or proof, they simply reply with "just google for a video on YouTube" - I have, and I can't find anything. I've tried 14 different keywords but the closest is a CSC video from about 20' where the individual was testing whether the system would actually shut off if it knew it was mid-air. It did. The next closest are simple wreck videos of drones failing mid-flight... etc. None that pose an intentional power cut mid-flight with attempted power restore and in-descent stabilization.

On my DJI, I would suspect this wouldn't work because the IMU doesn't show interactive function once you land and power down. However, when I've hooked directly to my TH, I notice that nearly all of the sensors remain live/active and appear to be functional.

If I were rich and carefree I'd be tempted to try it out myself. This isn't a question to challenge legality or rationale of such a maneuver, more just curiosity around the general physics and capability of the craft itself.

You could use some sort of arrangement to test this non-destructively. Something along the line of securing a piece of elastic from high tree limb. Attach the elastic to the H and stretch the elastic from the H to the tree above till it almost lifts the H from the ground. Now start the H, take off to min height. Elastic will still be under tension so no risk of prop fouling. Now do the dirty and carry out your test. Maybe you could lift it up with more powerful drone and at an altitude cut the power, let the other drone take the strain whilst you see if it restarts.
Lots of options, all complicated though.
Another one would be perform the test over a mature wheat field, this stuff is very good at catching RC models ;-)

 

[SeamusOD]

Done it lots of times with naze32, flip32 and apm boards, only done from about 200ft but self corrects no problem.
So long as something in the firmware dosnt stop the H starting the motors when its tumbling it would work as well, can you start the motors if you hold the H upside down ?

 

[PatR]

Flying an APM controlled X-8 I've used throttle to stop the motors from about 300' but that was set up where the motors are stopped but the system still armed and engaged. They don't fall flat, but tumble or invert. It can work out well for some systems but I would be reluctant to test it out with an H. My X-8 had lots of thrust reserves built into the system to assist with arresting the descent or fly with two motors out. I don't believe the H has proportionally as much.

 

[BobW55]

I agree, would work in theory, BUT what do you do if she is upside down and powers back up. The aerodynamics of the air frame, she will tumble like crazy. I am not risking my $1000 H to test it out. I would love to see the video of it being done. Now if you could bring the motors close to idle, she would fall faster and possibly remain upright.
On my 686 that is how I fast descend. It stays right side up but falls like a brick,

 

[prathbun]

... can you start the motors if you hold the H upside down ?

I believe that it goes into an emergency shut-down when it's upside down. Did a hand-catch last week and the **** thing refused to go to idle, so I flipped it. Motors shut off almost instantly, accompanied by much beeping and red flashing lights.

Of course, if it's in free-fall it won't know whether it's upside down or not...

 

[Murray Martz]

When I got my USB SIM for the H, the first thing I did was to try that at 400 feet, it didn't work no matter how many times I tried it. The SIM has pretty realistic dynamics, but again, it is only a SIM. I don't feel I would trust trying it in the real world. Plus, depending on the height, at some point it would hit terminal velocity, and would just become a mini rocket. One thing to note was during the SIM, the H started to tumble instantly each time the kill was hit.

 

[PatR]

That's what the X-8 did. Multirotors are not even a little bit aerodynamically stable.

 

[OmegaOmega]

It will be your final flight. You can count on it. I'm sure it has been tried with other brands, all unsuccessfully.

Hi Steve - It's been done with other brands, successfully, but I can't find any documentation with relation to Yuneec's products. Additionally, Yuneec turns off the motor (at least when at a hovering idle) if the device is tipped more than 90 degrees (or flips over). The DJI (at least from a year's old firmware) does not.

Here's a video of a DJI flyer arresting motors at around 5000ft and then restarting - you'll note that the craft seems to maintain an upright (although terrifyingly unstable) position, allowing him to restart motors. (Fast forward to 12:20)

 

[OmegaOmega]

You could use some sort of arrangement to test this non-destructively. Something along the line of securing a piece of elastic from high tree limb. Attach the elastic to the H and stretch the elastic from the H to the tree above till it almost lifts the H from the ground.

I'm thinking high elevation - 1000ft+

 

[OmegaOmega]

Flying an APM controlled X-8 I've used throttle to stop the motors from about 300' but that was set up where the motors are stopped but the system still armed and engaged. They don't fall flat, but tumble or invert. It can work out well for some systems but I would be reluctant to test it out with an H. My X-8 had lots of thrust reserves built into the system to assist with arresting the descent or fly with two motors out. I don't believe the H has proportionally as much.

This is exactly what I'm wondering. I know the unit goes into a "find-me" alarm mode when it's turned upside down during flight, so assuming that the craft stays right side up, my curiosity would be on the physics side of whether recovery is possible. I've seen multiple fall videos of DJI, and more often than not, it seems like they stay upright during a fall (assuming no impact/crash); but I'm not sure how the TH handles it, since it's a completely different design.

I have no idea how to calculate the true surface area, but if you estimate it off Yuneec's dimensions, and calculate against a fully loaded drone weight, you'd theoretically get a fall velocity of maybe 80ft/s. Given that the drone can ascend at 5m/s (or roughly 11mph) you would potentially be decelerating at around 16ft/s - however, the drone exceeds these thrusts when maneuvering, so I'm not sure if it would be constant or if the software would push the motors.

If someone were at 3,000ft altitude, you'd have maybe 40+/- seconds to compensate. The bigger question is, does the TH have enough power to do so?

 

[OmegaOmega]

I believe that it goes into an emergency shut-down when it's upside down.

This was my experience as well. My Autel doesn't shut off and is similar to this:

 

[OmegaOmega]

I agree, would work in theory, BUT what do you do if she is upside down and powers back up. The aerodynamics of the air frame, she will tumble like crazy. I am not risking my $1000 H to test it out. I would love to see the video of it being done. Now if you could bring the motors close to idle, she would fall faster and possibly remain upright.
On my 686 that is how I fast descend. It stays right side up but falls like a brick,

Given that the Yuneec's ST-16 controller gives you so much access to the controls, do you know if one would be able change either the rate of descent or amount of power sent to the props when holding the left stick all the way down? Or is this particular function hard-coded/locked out?

 

[Mojoe35]

I would think if you still have line of sight and need to descend quickly, just turn off the GPS and put the H in a shallow dive. It will definitely fly faster with GPS off. I would hate to attempt turning off the motors and then not being able to restart in time!

 

[OmegaOmega]

I would think if you still have line of sight and need to descend quickly, just turn off the GPS and put the H in a shallow dive. It will definitely fly faster with GPS off. I would hate to attempt turning off the motors and then not being able to restart in time!

Ah - I didn't know turning off GPS changed the allowances for ascent/descent?

 

[OmegaOmega]

Aha, just found this, @Mojoe35 - looks like non-GPS mode allowed this guy to get up to 60mph. I can't seem to find any videos on comparing ascent/descent though - but amazing to see you can get up to nearly 60...

 

[PatR]

Okay, for our aero physics major, Omega, which falls faster, a pound of feathers or a pound of lead? Just to be different I'll toss in a pound of gold. Surface area is not all there is to the aerodynamic equation. There are also multiple shapes that influence drag as well.

As for the reason for needing an emergency descent, anyone using a trained spotter would most likely never have an emergency that required a terminal descent. Regarding the DJI videos, I wonder how many failures there were for each success? Trashed 10 but one made it. Post that video! They are known to pad their stats quite a bit.

The code used in the H is largely designed to protect users from themselves. Going by some of the posts in this forum that was a wise decision. They do enough free work under the guise of warranty as it is. I can't imagine how much that would increases if people thought it OK to turn the motors off while in flight. There's enough of those doing it now because they have yet to learn how to land...

 

[prathbun]

I think you might be a little wide of the point of the post, PatR...

 

[Rayray]

The post is pointless, so any reply can not be less pointless.
May as talk about flying pigs, and the associated aerodynamics.