Payloads ... would they affect drone's Dynamics?

Joined
Nov 27, 2016
Messages
295
Reaction score
60
Location
Chicago, IL
I've seen some 'mods' where drones are fitted with various add-ons, be it a tracking GPS, an FLIR camera, or else. Hence my question:

1. Is the drone (P3S or else) smart enough to adjust thrust from each of the propellers to adjust vehicle attitude/pitch/etc ?

I'm guessing the props/speeds are tuned for the vehicle's CG somewhere in the middle. But, what if you attach an accessory, and the vehicle balance is off? Does the vehicle's control algorithm take care of such 'deviations' ?

I assume so as I've done some tests with my P3S (without propellers), and as hold and tilt the drone, I can hear some of the motors spin faster to recover and bring the vehicle to a 'horizontal' position. This done with just a triaxial accelerometer as in other consumer electronics

Now, that was the easy part. How about the attachment/payload hurting the 'dynamics' of the bird (eg, non symmetric drag forces during forward travel... are there built-in controls to counter this?

You can tell, the more I look at my drone, the more fascinating it is ... great job DJI!
 
Last edited:
These drones are aeronautical nightmares! They are really flying rocks, no glide ratio, etc. so adding things doesn't really affect them that way. It's the added weight that is the issue. All the lift is generated by the props and the more weight, the more lift required and the harder the props and motors have to work, which drains the battery faster. So the net effect is reduced flight time.
 
  • Like
Reactions: ROD PAINTER
This will help answer some of your questions...

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Sent from my XT1585 using PhantomPilots mobile app
 
The software will work the same way because as the drone becomes unbalanced the feedback back to the software algorithm are the same. It is still limited on its capabilities (thrust) so it will recover / react more slowly and take longer overall.


Sent from my iPhone using PhantomPilots
 
I've seen some 'mods' where drones are fitted with various add-ons, be it a tracking GPS, and FLIR camera, or else. Hence my question:

1. Is the drone (P3S or else) smart enough to adjust thrust from each of the propellers to adjust vehicle attitude/pitch/etc ?

I'm guessing the props/speeds are tuned for the vehicle's CG somewhere in the middle. But, what if you attach an accessory, and the vehicle balance is off? Does the vehicle's control algorithm take care of such 'deviations' ?

I assume so as I've done some tests with my P3S (without propellers), and as hold and tilt the drone, I can hear some of the motors spin faster to recover and bring the vehicle to a 'horizontal' position. This done with just a triaxial accelerometer as in other consumer electronics

Now, that was the easy part. How about the attachment/payload hurting the 'dynamics' of the bird (eg, non symmetric drag forces during forward travel... are there built-in controls to counter this?

You can tell, the more I look at my drone, the more fascinating it is ... great job DJI!

Yea, control is very complex.
Consider this:

Depending on the model of Phantom, they have 12 or more degrees of freedom (DOF) in their sensor suite...
(I say '12 or more' cause some models have redundant sensors)
3-axis Accelerometer,
3-axis Rate Gyro
3-axis Magnetometer
2-axis GPS (altitude (z) is not thought to be used actively)
1-Barometer
(The IMU is generally defined to contain all but the GPS & magnetometer)

There are six DOF for an aircraft in flight...
The trajectory has three DOF and its attitude along the trajectory has three DOF.

Then a quad (4 motors) has 4 DOF in it's control output.

Further complicating the control problem is the number of control inputs (the four motors) is fewer than the number of degrees of freedom in flight, resulting in an under-actuated system.
The MC is 'working' very hard to maintain all 6 in-flight DOF with just 4 motors!
 
Last edited:
This is becoming interesting, and gonna bring me back to my graduate school days ...

For a system as complex as this aircraft, it is not sufficient to think of the degrees-of-freedom as those associated with position only. In fact, in strict controls terminology, each coordinate and its rate of change (speed) are both state variables. Thus, if we decide to write the equations of motion of a drone, it'll be 12×12 state matrix. In such setting, I agree with you that we have here 12 (not 6) degrees-of-freedom.

Now, when it comes to the various sensors used, I would not call those DOFs, to me they're feedback signals one must use to tweak the control/input signals.

When it comes to controllability of the vehicle, and as you stated, there are only 4 independent inputs (speeds of the 4 propellers) and its very hard to imagine a 12 DOFs system controlled by 4 inputs. What this says is that we'll end-up with 'dependent' DOFs. For example, we can't have the vehicle move forward/laterally without it pitching/rolling i.e., forget lateral motion of the vehicle and keeping it perfectly horizontal. Helicopters can do this because their blades have variable pitch (cyclic+collective). Fixed-rotor machines cannot

Now, if we really want to make this more complex (and we must), we need to add the 3 rotations of the camera relative to the aircraft to bring the total to 15 DOFs. But there, we also get 3 actuators ... though those do very little in terms of the dynamics of the whole aircraft. :)
 
The DOF quantity and designation is not mine but rather the definition in the context of physics.

"In physics, the degree of freedom (DOF) of a mechanical system is the number of independent parameters that define its configuration. It is the number of parameters that determine the state of a physical system and is important to the analysis of systems of bodies in mechanical engineering, aeronautical engineering, robotics, and structural engineering.

"The position and orientation of a rigid body in space is defined by three components of translation and three components of rotation, which means that it has six degrees of freedom."
 
Love the video. This goes to prove that DJI has done a lot in terms of active control of the aircraft, and often such prowesses are 'not' visible.

Now, such 'extreme' stability, is that available only on the Phantom 4 Pro or also on previous generations?

Seems like the P4P is kind of like an X-29 on steroids.
 
Last edited:
The DOF quantity and designation is not mine but rather the definition in the context of physics.

"In physics, the degree of freedom (DOF) of a mechanical system is the number of independent parameters that define its configuration. It is the number of parameters that determine the state of a physical system and is important to the analysis of systems of bodies in mechanical engineering, aeronautical engineering, robotics, and structural engineering.

"The position and orientation of a rigid body in space is defined by three components of translation and three components of rotation, which means that it has six degrees of freedom."

Thank you kindly. Not to beat this to death but ... I think I know what a DOF is (I have a PhD in Mechanics/Dynamics :)). While I agree fully that only 6 DOFs are sufficient to completely and uniquely locate a rigid body in space at any given time, this is 'not' sufficient if we are to 'control' in real time the dynamic maneuvers of such rigid body. Hence, people with Controls background speak instead of 'State Variables'. As such, our vehicle has 12 such 'states', the 6 DOFs and how rapidly they're changing (speeds). Now, it could very well be that some of such states are irrelevant to dynamics, but the number 12 is a 'worst-case' scenario. From a dynamics point-of-view, some of the DOFs are related/dependent through so-called non-holonomic constraints ... Think of it this way: It takes 3 coordinates to locate a boat on the water (x,y and orientation) and yet the boat can only move forward (or reverse) as it cannot rotate in place, nor slide sideways. It can only move in one direction and thus, it has only one degree-of-freedom. Thus, we need 3 coordinates to completely/uniquely locate a boat on the ocean, and yet such boat has only one degree-of-freedom ... Ok, enough said I guess.

Here is one thing I have not 'resolved' yet ....

1. To move left/right, neighboring propellers spin faster/slower
2. Same thing for forward/reverse motion
3. However, to rotate in place, I'm thinking diagonally props would move faster/slower ... and yet, when I do a free body diagram and analyze the vertical forces from the propellers, none of such forces applies a torque about the vertical axis of the vehicle.

Unless rotation has more to do with 'circulation' in fluid mechanics ... something I need to dig a bit deeper into.

Anyone has a good reference on the aerodynamics/controls of quadcopters?

Cheers :)
 
Why is it there are so many BS-ers here?

My son, not yet a HS graduate, with a simple Google search, found there are 6 'motions' which define a ship's position:

The Linear-s are:
Heave, Sway, and Surge

The Rotational-s are:
Pitch, Roll, and Yaw

http://icter.sljol.info/article/download/2847/3771/
 
Last edited:
Why is it there are so many BS-ers here?

If the thread is of no interest to you, then save yourself some time and move on ... ;)

It seems you edited your answer and 'improved' it ... it's alright, we all make mistakes and hopefully learn from them. As for the knowledge you get from google search, if that satisfies your curiosity (thirst for knowledge) then so be it, but you must accept that some of us are interested in more than just superficial answers.

Happy flying or whatever is it you do, that brings you immense joy.
 
Last edited:
As you can see, I'm interested in the subject.
But not your CV.
 
Guys..Guys
have some dip.gif
 
I attached about a pound of weight in the form of a plastic sampler and 300 ml of water with 10 feet of fishing line to my P4, and slung it aroind pretty good. The bird handled it fine. It had to work a little harder = reduced battery life. I was very impressed at how well it kept its flight characteristics.
 
Thank you kindly. Not to beat this to death but ... I think I know what a DOF is (I have a PhD in Mechanics/Dynamics :)). While I agree fully that only 6 DOFs are sufficient to completely and uniquely locate a rigid body in space at any given time, this is 'not' sufficient if we are to 'control' in real time the dynamic maneuvers of such rigid body. Hence, people with Controls background speak instead of 'State Variables'. As such, our vehicle has 12 such 'states', the 6 DOFs and how rapidly they're changing (speeds). Now, it could very well be that some of such states are irrelevant to dynamics, but the number 12 is a 'worst-case' scenario. From a dynamics point-of-view, some of the DOFs are related/dependent through so-called non-holonomic constraints ... Think of it this way: It takes 3 coordinates to locate a boat on the water (x,y and orientation) and yet the boat can only move forward (or reverse) as it cannot rotate in place, nor slide sideways. It can only move in one direction and thus, it has only one degree-of-freedom. Thus, we need 3 coordinates to completely/uniquely locate a boat on the ocean, and yet such boat has only one degree-of-freedom ... Ok, enough said I guess.

Here is one thing I have not 'resolved' yet ....

1. To move left/right, neighboring propellers spin faster/slower
2. Same thing for forward/reverse motion
3. However, to rotate in place, I'm thinking diagonally props would move faster/slower ... and yet, when I do a free body diagram and analyze the vertical forces from the propellers, none of such forces applies a torque about the vertical axis of the vehicle.

Unless rotation has more to do with 'circulation' in fluid mechanics ... something I need to dig a bit deeper into.

Anyone has a good reference on the aerodynamics/controls of quadcopters?

Cheers :)
What happens if you add rotational torque to your analysis? Should be able to get your theoretical AC spinning nicely then. That's exactly how it works in practice (opposing pairs with respect to the phantoms and other quadcopters).
 
I attached about a pound of weight in the form of a plastic sampler and 300 ml of water with 10 feet of fishing line to my P4, and slung it around pretty good. The bird handled it fine. It had to work a little harder = reduced battery life. I was very impressed at how well it kept its flight characteristics.

That is just mind-boggling. I wonder if there is a P3 owner brave enough to try the same feat on their machine. I'm not that brave ... i.e., haven't even flown my bird yet (too cold outside, and still have lots of reading to do).
 
What happens if you add rotational torque to your analysis? Should be able to get your theoretical AC spinning nicely then. That's exactly how it works in practice (opposing pairs with respect to the phantoms and other quadcopters).

I still have not convinced myself where the torque comes from. If it's internal between each motor and its propeller, then it cancels out. If the torque is from the propellers interacting with the surrounding air, ... then I need to scratch my head a bit more. I've seen some published work that claims that the propellers deflect, and that is what gives rise to a horizontal force component, but I can't wrap my head around that either ... yet.

Sorry, slow evening (i.e., long day at work) here ... what is AC ?
 
I still have not convinced myself where the torque comes from. If it's internal between each motor and its propeller, then it cancels out. If the torque is from the propellers interacting with the surrounding air, ... then I need to scratch my head a bit more. I've seen some published work that claims that the propellers deflect, and that is what gives rise to a horizontal force component, but I can't wrap my head around that either ... yet.

Sorry, slow evening (i.e., long day at work) here ... what is AC ?
AC for aircraft..... try solving for one prop at any distance from a point and you will get rotational force in opposition to the prop rotation. The furthest described blade path has the greatest influence. Now you have me considering the math.
 
  • Like
Reactions: JB63

Members online

No members online now.

Forum statistics

Threads
143,054
Messages
1,467,297
Members
104,919
Latest member
BobDan