Redundancy: flying on 2 props

Joined
Feb 4, 2014
Messages
166
Reaction score
0
After seeing prop flyoffs and motor failures, it got me wondering...

Wouldn't it be possible to at least control your descent on only two motors? So if one motor died or one prop flew off, the IMU could easily detect the tumbling and start turning motors off one by one until the Z axis became stable. You'd only have to turn a maximum of 3 motors off until you found the right combination that gave you two working motors on opposite sides of the quad. Sure, the body would spin but at least you could lower the spinning body and still fight gravity. Seems like something that could be done in firmware:

(1) Uncontrolled tumbling detected
(2) Switch off motor N
(3) Check IMU to see if switching off motor N caused tumbling to stop on Z axis
(4) If not, increment N to next motor and go back to (2), else go to 5
(5) If Z axis is now upright, disable yaw, pitch, and roll and slowly descend to ground

Mike
 
Yeah, like that! Although I have a hard time believing even IF the motors/props had the power that it could recover from 2 or 3 prop/motor failures as they claim.

But c'mon DJI... we want our prop-loss-recovery software upgrade! ;) Of course I have a feeling they'll want us to all buy Naza V3 for that. :roll:

Mike
 
not to mention that they are using light weight quads with no gimbal and other stuff hanging off the bottom. That changes everything. Trying to recover a prop loss with all that weight might take more height/power than the phantom can afford. Not to mention that you hear about a prop/motor issue once in a blue moon. Not worth DJI's investment into that programming for something that is barely going to be used.
 
It's possible on a hexrotor with DJI's flight controllers (and I've had it happen and can confirm it works great). The physics involved to pull it off on a quad may have been beyond what DJI wanted to deal with.
 
I've been thinking about the need for a system for recovering from a prop or motor failure with a quadcopter which doesn't have the redundancy of a hexa- or octocopter even before I bought my Phantom since such mechanical failures can never be entirely prevented. It shouldn't be that hard.

In case of a prop failure the loss of air resistance and the resulting increase in RPM/drop in current combined with the corresponding loss of attitude should be detected by the electronics in a fraction of a second. The opposing prop should then rev down and the other two rev up to produce extra lifting force. Because these two spin in the same direction the quad will start to rotate along its Z-axis in the opposite direction. That rotation will create a stabilising 'gyroscopic' force same as it does with a frisbee. The sensors can determine when the third prop is 'below the horizon' (x-y plane) and accelerate shortly to lift that side up towards the horizontal plane. As the 'rotor disc' comprised of the four arms of the quad rotates that third propeller will travel around the circumference where it will alternate between being above and below the 'horizontal' plane. The flight controller can speed it up when below and slow it down when above the horizontal plane to achieve a 'stable' flight. The two other props can vary their speed to remain stable along the other horizontal axis as they do in normal flight. On top of that some steering should then still be possible. Depending on the weight and the thrust of the remaining propellers the quad can then still hover or it will descend but at a much lower speed than it will do in an uncontrolled fall.
 
Noël said:

That's interesting and I've seen that too, but it adds cost and weight and I'm not sure it would even work in an uncontrolled fall. The 2 prop solution does not add any weight or cost. Just requires a bit of programming: programming that really shouldn't be that difficult to the people who code for the Naza. I'm quite certain that even at the "normal" 1200g weight, two props could still make it hover. And if not hover, at least a controlled descent. And as pointed out, even if the quad isn't balanced in the center, the shell will be spinning to counter the weight differential. The only real requirement is that it be facing more up than down when the thrust from the two props kicks in.

More important than "saving your quad" is the fact that this little bit of programming would increase safety!

Mike
 
Hello,
I'm a newbie with RC multicopters and just got my P2 and H3-3D few weeks back. Having a fun time flying over houses and traffic, but it got me thinking one day - what if something went wrong (motor failure, etc). How would the P2 recover from that and how significant of a danger that would pose to the people and vehicles around where the P2 is being flown.

Working in an industry that takes pride in safety, that thought really bothered me and I have not flown in a public area ever since.

Am very very keen to find out if there is any way to improve safety and redundancy upon failure (like how commercial aircraft can still operate with one engine failure). After all there should be no difference between commercial aircraft and drones, both are in the air and flying close to populated areas (unless you are filming in remote locations, then the risk would be damage/loss of your UAV).

Also the high rates of flyways (and the safety aspect associated with it) - anyone has a good solution to this (besides not flying near comms towers and power pylons)?

Keen to follow on latest developments on this issue...

Cheers and fly safe!
 

Recent Posts

Members online

Forum statistics

Threads
143,094
Messages
1,467,583
Members
104,977
Latest member
wkflysaphan4