Copter co-axial X8 setup, favor the top motors?

[Andrew Chapman]

Hi all, I was thinking about an X8 setups and was wondering if anyone has tried favouring the top motors, keeping the lower motors only for thrust boost or if the top motors fail?

X8 seems a great config in terms of the redundancy in a more compact unit, but you're throwing away a lot of efficiency from the co-ax motor setup. But it occurred to me that you could leave the bottom motors idling at their MOT_SPIN_ARMED rate until they were needed. e.g. at 50% thrust input it's all about the top motors, but the bottom motors start to ramp up when the top-motor PWM is in the 80-100% range, so that the top+bottom are getting full power at 100% commanded thrust.

In this way you might be able to have a copter that ordinarily performs like a high-efficiency, large prop low-rpm quad, but has the backup redundancy and extra grunt when required.

I don't have an X8 I can test with, so wondering if anyone has thought this through or tested it?

AC.

[Jonathan Challinger]

It'd have some issues with yaw coupling, probably. But the tradeoff may be worth it?

--
You received this message because you are subscribed to the Google Groups "drones-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to drones-discus...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[michel....@gmail.com]

Hi Andrew,

I use X8 copter for filming and feel it rather efficient being smaller than ordinary octa. Surely it's not less efficient than OCTA I used for same purpose before. As I know coaxial reverse rotation propellers are more effective than just two propellers. Even more effective when ducted. Some long range military transportation aircraft use coaxial propellers for that reason.

Lower motor shouldn't be idling. This will cause large power waste because air flow will loose it's energy to spin lower prop instead of making thrust. If you disable lower motors and start upper ones on X8 you'll see lower propeller spin much faster than idle rpm. Air flow is also spinned by upper prop and lower prop works in that spinning air flow. This makes it more effective as energy wasted for air spinning by single prop is partially used by second one.

If you are going to build rather small X8, make prop pairs ducted. This will make it even more efficient. I'm thinking about ducting props on my X8, but it's rather large (up-weight with batteries is about 10kg) and material weight probably will not allow efficiency boost to be significant.

MK

[Holger Steinhaus]

I would expect a largely degraded efficiency, as the slowly rotating prop disturbs the accelerated airstream. I have seen this type of effect while bench-testing single-prop performance of coaxial drives. As soon as the second prop started to windmill (the motor did not have any ESC connected), the thrust of the single prop degraded significantly (by some 30% as far as I remember). After fixing the unused prop, the thrust was close to 90%, compared with a single prop only setup.

Holger

[Jonathan Challinger]

That is simple to solve: use the bottom motors instead of the top.

[Holger Steinhaus]

I did not see a major difference between driven top rotor and windmilling bottom rotor or vice versa. But it's well possible that this observation is limited to large props mounted within a small distance (17" diameter and about 3" distance).

BTW, this behavior is conforming the to usual propeller stream therory - it suggests that 50% of the acceleration happens in front of the prop, and 50% thereafter.

Holger 

[Randy Mackay]

     That’s an interesting idea.  It’s like those cars that shut down some cylinders to save gas.  http://en.wikipedia.org/wiki/Variable_displacement

 

     My guess is that with a slightly modified motors class and *maybe* a change to the throttle curve parameters to keep the thrust linear you could make it work.

 

-Randy

--

[Andrew Chapman]

The 'idle' motor (be it top or bottom) wouldn't be windmilling though, it would be running at some partial throttle setting.

You could run tests starting at 5-10% and working upward to see where the optimal balance was. It would need to be fast enough that it isn't causing any drag from the incoming airflow, but not so fast as to be contributing extra thrust until needed.

And yeah, it would probably be better to idle the top motors, but best to test both ways and see.

AC.

[Robert Lefebvre]

Andrew, have you done any bench testing of this yet?  It seems to me that this theory could be easily and accurately tested on a bench stand. You'd get good data before going to the trouble of building the machine and tweaking the code.  That's what I did before I built my first overlapping prop machine.  

[David Pawlak]

First impression... "Great Idea!!".... but..

I have an X8 configuration with a Pixhawk. Before unertaking it (just for the heck of it) I measured a coaxial configuration with characteristics that duplicate my configuration.

I had heard htat there was a performance loss of about 20% for the coaxial config.so I proceeded to measure.

First, top motor, no prop on the bottom, measured 200, 400 600 800 and 1000 gram thrust. Weight on scale, lifted by motor configuration.

Later Both motors, 400, 800, 1200, 1600 and 2000 grams thrust.

Measuring total thrust diivided by number of motors.... thrust per motor. Target 800 per motor.

Surprise!

Up to about 450 gm per motor there was a loss ( I'm sorry, where I am, I don't have the data) I can get it, but in the order of 5%

After that I had a slight gain!! Don't ask me why. It was a cheap and dirty test. but should be fairly close.

My opinion was.....  "not a big issue".

Supposing my tests were too careless, they were cheap and dirty, there is a lot at play. Whether it's the top motors or the bottom motors that take the load, they are affected by the others. If you put greater pitch on the bottom motors, the top take less load, but the bottom take more. there could be a point, where a special balance between the top and the bottom matches the available pitches. You may be able to buy a prop with a pitch of 4.94 somewhere at a price. Then all of that has to match the performance curve of the motor, which probably means that the top and bottom motors need to be: 1) tailored, or 2) at least different.

You can always use intelligence to balance the top and the bottom, but you will always be compromising the performance of the motor.

Normal operation, top four working, bottom four eliminating drag. (We are already talking about a performance compromise. Consuming to reduce drag).

Failure mode, who cares about the performance cost!! 

So basically we are concerned about the cost of having four extra motors ready for an emergency. Aside from cost of reducing drag we have to concern ourselves with their weight.

So if they are going to be there, we should concentrate on seeing if we can get the pair to work more efficiently all the time... and we get back to extraordinary pitch balaces and custom motor performance curves.

I have a 3DR X8 which is not loaded. Maybe I could configure it for 4 motors no props on the bottom, measure performance, and then  compare that with the eight motor config. It's not the same, but would give an indication of the real coaxial loss. I think it's possible to fly an X( with only four motors,,,, loading the right software.

[Andrew Chapman]

Nothing I've tested yet, just running the idea up the flagpole for now.

What was the result of your overlapping prop tests? Was it a win or did it balance out in the end (larger props give more efficiency, but the overlap takes some away)?

AC.

[Robert Lefebvre]

I showed the typical 20% loss when fully overlapping (coaxial).  With 20-30% overlap, efficiency is similar to a fully separated setup, though it's a bit lost in the standard deviation. However, after I built the machines, my perception was that vibration and noise is up substantially.  And then Forrest actually confirmed that in his own testing.

Final analysis, overlapping is definitely a viable design choice, allowing you to shrink the frame, without the Coaxial loss factor.  But it does cause more vibration that has to be dealt with.  I did solve that problem very well on my H8 frame design.  

As always, there's no free lunch. ;)

I would be surprised if your idea increased efficiency.  First, because running a motor at "idle" is actually very inefficient.  You'd be surprised at how much power it takes.  Second, unless the craft was extremely overpowered, the extra set of motors will always be running anyway.  The most efficient design, is a quad, with the motors running about 70-80% throttle to hover.  So, you're talking about making a quad that hovers at 50% throttle which robs efficiency, while also carrying around the weight of 4 extra motors.  Theoretically, it would probably hover about 40% throttle if you removed the extra motors.  That's not efficient at all.

If you design it to run at the max efficiency point of 80%, then the bottom motors will be running, not idling.

Just my back-of-the-napkin ideas mind you.  It's worth testing.  I'd do it for you myself if I had the time.  I have a nice engine testing fixture you can borrow if you want.

[Robert Lefebvre]

Ducting does not make motors more efficient.  It is always more efficient to simply use a larger rotor.  Where ducting is beneficial is when some other design characteristic precludes the usage of larger propellers, but you have reached a point where you need more thrust from a limited rotor size.  

--

[Gary McCray]

+1 on what Robert said,

Ducts always produce a loss versus larger rotor size, they do permit higher thrusts to be achieved with smaller diameter propellers but do so at the expense of requiring more power and reduced efficiency to provide it.

The extra air friction in the ducts "venturi" is considerable and the loss's are not minor and results both from prop tip and duct wall friction and the heat of compression loss from the venturi effect itself.

For thrust efficiency on props, bigger is always better (of course other considerations place practical limits on this).

Best Regards,

Gary

[Paul Riseborough]

Coaxial counter rotating props with power split between them can be more efficient than a single prop of the same diameter  due to the reduced swirl and therefore wasted energy downstream. Ideally the pitch of the downstream blades should to be slightly higher to take into account the higher flow velocity due to flow contraction behind the first prop.

However two props side by side without overlap and with the power split between them are always going to be more efficient than a coaxial setup as they have twice the effective disc area, and according to momentum theory, the power required to hover at a given weight varies with 1/sqrt(disc area). http://en.wikipedia.org/wiki/Disk_loading.

For the X8 configuration we increase power required due to loss of disc area compared to an octa, but this is partially offset due to the reduction in swirl velocity.

[Philip Rowse]

@Gary & @Rob

Actually, Ducting does make a prop more efficient for Static thrust, so for hovering, Ducted can be more efficient at the same rotor size, power input etc..... the debate gets messy when you try and mix the effects of ducting for higher velocities and low speed.

The theory is around tip vortices, tip speed, supersonic effects at blade tips, but as you move forward the effects of tip vortices reduce, so do the gains. however, there are also gains in ducting the flow for high speed, this is why, even though there was a big push, the UDF for commercial Airliners never got past the testing phase. Keep in mind though, a duct adds inefficiencies if designed incorrectly. on a multicopter, it increases the frontal area of a copter, making forward motion less efficient. (unless you are tilting the rotors for forward flight... but then that opens up another can of worms). its all about the specific application. for efficiency of lifting a large load, a larger rotor will almost always win, but for high speed flight, a smaller diameter, shrouded high speed fan, is normally best. take your pick, high flow, or high thrust..... to get both, requires a bigger motor...

This description does not do this topic justice, it is a complex area, and most likely irrelevant to this topic.

Paul's answer covering disc area about covers it for this x8 application.

[Gary McCray]

Hi Philip,

It wasn't my intention to say that ducting couldn't be used to effectively increase thrust.

Simply that given no constraints on prop diameter (or rotation speed for that matter) that you could achieve equivalent thrust more efficiently by adjusting prop diameter and speed than you could from ducting a smaller prop.

Of course this does require that you also use a motor that is efficient at the appropriate speed / torque for the given prop.

And not even getting into the discussion relating to supersonic speeds at all.

Best Regards,

Gary

[David Pawlak]

Here is my data for the coaxial performance test. As expected, there was a loss up to about 750 gm.

The surprise is the gain after that.... Not sure what's going on there.

[Robert Lefebvre]

David, how exactly did you conduct the test?  You have to be really careful when doing this sort of test using a battery as a power supply.  Especially when you are comparing a two-motor setup versus a one-motor setup.  

I notice your graph is labelled "Current vs. Thrust", but in the table, you should power (watts).

[David Pawlak]

The graphs uses current, the power column is just there for reference, and even then I used an average reference voltage (taking a median between the start voltage and end voltage.)

The only problem is that the latter part of the test, where the voltage would have started to sag and therefore the current start to climb,

I measured the coaxial configuration. So the coax configuration should show higher current consumption.

Anyways... irregardless of the outcome, I needed to use the X8 configuration anyways. Although all this has peaked my curiosity.

I shall have to do some more serious tests. I really would like to see effects of adjusting pitch, upper and lower diameters, and motor curves to find the best fits.