Propeller thrust vs. RPM

[sc_deimos]

Hi all,

This is my first post, so please be gentle with me.

I'm in the very preliminary stages of designing a semi-scale electric Osprey project. I have a good idea of the mechanical and layout problems involved, but I'm having no luck in finding propeller thrust information necessary for a VTOL project of this nature. Like the engineers at Alpha Romeo who start with a wheel selection and design a car around them, I'm wanting to start with the propellers and design the plane at its final scale around them.

For those unfamiliar with the Osprey, my initial design requires:
- two 3-blade fixed-pitch propellers (one each of LH and RH screw)
- prop diameter in the range of 14" to 20" (could go a little larger if needed)
- each propeller capable of 1.5-2.0kg thrust (from this you could infer a target AUW of 3.0-4.0kg, the lighter the better of course)

I guess what I'm looking for to answer my questions is a source of propeller graphs for commercially available 3-blade props which plots thrust versus RPM and efficiency versus RPM. Or a program which can generate this data for me.

After that I can start downrating figures taking into account masking effects and such, eventually arriving at an RPM I can find a motor/gearbox combination capable of driving the props at.

This is one of those cases where battery duration/flight time will be taking the back seat. And yes, I know the full-size Osprey uses constant speed variable-pitch props running at about 330RPM. There's only so much you can do in a small light model, though.

Thanks in advance for all your help,
Anthony.

[Peter D Rieden]

<And yes, I know the full-size Osprey uses constant speed variable-pitch props running at about 330RPM.>

There's rather more to it than that. The full-size uses *articulated* props - like helicopter rotors, with both collective and cyclic pitch control, and this leads neatly to the real problem: How do you propose to control this beast?

If you follow the full-size practice and use (say) a pair of model helicopter rotor assemblies you'll then have to find a way of mixing the control functions. This mixing must take into account the fact that (for example) roll control in hover mode progressively changes to yaw control as the rotors are brought around to the forward flight position.

If you're intending to just use throttle control on fixed-pitch props (which don't intermesh) you need to contra-rotate them, which rules out commercial off-the-shelf props because there isn't much of a range available in "handed pairs". You'll still need a pitch-axis control mechanism for hovering and transition, and although differntial throttle initially looks like a simple roll control mechanism on closer examination you find that when you apply roll control the difference in the two motor torques gives you a powerful yaw couple which you have no means to correct. Then there are the gyroscopic couplings...

The power and lift problems are simple, but the control system is an absolute swine to solve. I did a fair amount of work on this concept about 10 years ago, looking to build an OS-91FSR powered model and whilst I came up with solutions to a lot of the problems (like using angled swashplates in the cyclic pitch servo linkages so that the cyclic control went from zero to full as the rotors were rotated to the hover position) the whole thing ended up far to complicated to be practical.

Also bare in mind that the wing needs to be very strong and heavy because the lift forces are at the tips whilst most of the weight is at the roots! I'm sure there are more complex flying machine concepts, but I sure as heck can't think of one!

Best of luck,

PDR

[steve lewin]

Try ThrustHP, available from http://www.bmaps.net/

It's not 100% accurate, particularly for efficiency (i.e. quoted input power) but it will get you close to the sort of figures you're looking for. Of course you'll still need to work out how to turn those big props at the appropriate speed.

Have fun with your very ambitious project

Steve

[sc_deimos]

Originally posted by Peter D Rieden
There's rather more to it than that. The full-size uses *articulated* props - like helicopter rotors, with both collective and cyclic pitch control, and this leads neatly to the real problem: How do you propose to control this beast?
The full-size Osprey is a rather complex beast, mechanically speaking. Boeing engineers may not have heard of the K.I.S.S. principle, but I suspect carrier storage requirements of the Navy and Marines have introduced a lot of complexity. Not only are the props variable pitch (read articulated), they also have knuckles at the blade roots for folded storage. The whole wing then rotates clockwise (as viewed from above) on a fuselage pivot point offset to starboard to store the wing and nacelles on top of the fuselage. See:
http://www.rfcs.webcentral.com.au/osprey/021012-M-XXXXG-001_th.jpg 1024x768 JPEG, 413kb (http://www.news.navy.mil/management/photodb/photos/021012-M-XXXXG-001.jpg)
Of course, neither of these features is planned on this model. :)
If you're intending to just use throttle control on fixed-pitch props (which don't intermesh) you need to contra-rotate them, which rules out commercial off-the-shelf props because there isn't much of a range available in "handed pairs".
Yes, contra-rotating props are on the cards. That's why I asked about LH and RH screws. Having the props intermesh would truncate the fuselage, though. If all else fails I can get props made, but I'd rather go for a commercial solution to ease replacement. I expect to go through many props learning to fly this beast.
The power and lift problems are simple, but the control system is an absolute swine to solve. ... the whole thing ended up far to complicated to be practical.
The drive system complexity of the full-size Osprey is rather involved allowing both rotors to keep running, albeit at reduced power, in the event of a single-engine failure. See:
http://www.rfcs.webcentral.com.au/osprey/v22d_th.jpg 900x643 JPEG, 144kb (http://198.65.138.161/military/systems/aircraft/images/v-22d.jpg)
Such a system on a small model would add weight and complexity unecessarily.
Best of luck,
PDR
Thanks, Peter. I'm sure I'll need it!

[sc_deimos]

Thank you, Steve. I'll check out ThrustHP's site.

[Dick Huang]

sc_deimos
You might want to down load Moto-cal and make it think you have a twin motor model. It will evaluate a family of props to your specification from 0 mph to v max. It's free for 30 days.
Dick Huang:)

[Peter D Rieden]

<The full-size Osprey is a rather complex beast, mechanically speaking. Boeing engineers may not have heard of the K.I.S.S. principle, but I suspect carrier storage requirements of the Navy and Marines have introduced a lot of complexity.>

The design solution really is about as simple as it gets; rotorcraft have certain inherent design complexities and a tilt-rotor only makes them worse. The additional issues of the rotating wing and folding rotors are comparatively simple bits of mechanical design.

<Yes, contra-rotating props are on the cards. That's why I asked about LH and RH screws. Having the props intermesh would truncate the fuselage, though. If all else fails I can get props made, but I'd rather go for a commercial solution to ease replacement. I expect to go through many props learning to fly this beast.>

I doubt this will be feasible. Firstly contrarotation isn't "desirable"; it's essential - countering the rotor torque at the hover would otherwise require a differential tilt control which would be complex to engineer if you want sufficient control speed. Secondly if you want the rotors to intermesh then you'll need to contrarotate them AND have them geared together via a shaft to maintain the blade phasing. This means that you can't use differential throttle (rotor speed) for roll control and you'll need to use some form of rotor pitch control instead. This in turn means you can't use commercial fixed-pitch props!

Differential collective pitch initially looks attractive until you realise that each rotor would then develop a different torque and the counter-rotation would no longer cancel the torques. The result would be a powerful yaw whenever roll-control was used (either adverse or proverse depending on whether the rotors rotate inwards or outwards).

The roll commands also precess gyroscopically to produce pitch inputs, but as the rotors counter-rotate they would pitch up in one side and down on the other; the wing element would need to be VERY stiff in torsion to keep everything pointing the right way. In theory it would be possible to make these gyroscopic moments tilt the rotors against springs in such a way that they exactly cancel the torque imbalance, but the sums necessary to design this are not for the faint-hearted...

<The drive system complexity of the full-size Osprey is rather involved allowing both rotors to keep running, albeit at reduced power, in the event of a single-engine failure.>

I know all this - it's what I do for a living.

<Such a system on a small model would add weight and complexity unecessarily.>

If you're going to intermesh the rotors then you'll need to interlink the two rotors to maintain blade phasing, at which point you may as well use this shaft to drive the rotors and mount a single main drive motor in the fuselage.

However, these really are the minor problems. The biggest challenge is controlling the beast. The Ospray uses a complex synthetic control system which can't easily be replicated in a model. A better example to look to is the Canadair CL-84 programme, in which the control system comprised just a succession of mechanical mixers. Canadair spent ages tinkering with this system to make it work well, and the final achivement was an engineering masterpiece, but it was extremely complex. It also used a tail-rotor to overcome some of the more basic paradoxes!

<Thanks, Peter. I'm sure I'll need it!>

Don't let me put you off; I'm just trying to ensure you appreciate the complexity of the job you're taking on!

Regards,

PDR

[raptor22]

A microchip is probably neccessary-one reason harrier designs are just now becoming a reality for a select few persistend individuals.

--Alex

[migray]

Hi all,

This is my first post, so please be gentle with me.

I'm in the very preliminary stages of designing a semi-scale electric Osprey project. I have a good idea of the mechanical and layout problems involved, but I'm having no luck in finding propeller thrust information necessary for a VTOL project of this nature. Like the engineers at Alpha Romeo who start with a wheel selection and design a car around them, I'm wanting to start with the propellers and design the plane at its final scale around them.

For those unfamiliar with the Osprey, my initial design requires:
- two 3-blade fixed-pitch propellers (one each of LH and RH screw)
- prop diameter in the range of 14" to 20" (could go a little larger if needed)
- each propeller capable of 1.5-2.0kg thrust (from this you could infer a target AUW of 3.0-4.0kg, the lighter the better of course)

I guess what I'm looking for to answer my questions is a source of propeller graphs for commercially available 3-blade props which plots thrust versus RPM and efficiency versus RPM. Or a program which can generate this data for me.

After that I can start downrating figures taking into account masking effects and such, eventually arriving at an RPM I can find a motor/gearbox combination capable of driving the props at.

This is one of those cases where battery duration/flight time will be taking the back seat. And yes, I know the full-size Osprey uses constant speed variable-pitch props running at about 330RPM. There's only so much you can do in a small light model, though.

Thanks in advance for all your help,
Anthony.
Hi Anthony, The tasks we set ourselves hey? I too am in mid development of a vtol Osprey type machine, have you seen the poco from Gress air (17 oz all up weight) its a wingless job and uses a lot of heli parts such as giro's etc, the props they use are off a thing called the dragon flier, two bladed 10.5" left and right handed. I have got some of these blades which need to be sanded on the leading edge and am using two ac25/25/26 brushless motors (about 67 us dollars from aircraft world) which come complete with 5x1 gearboxes, not sure of amperage pull yet but they seem to give loads of thrust. My model will be small about 24" between motors and it will have aileron elevator contol with a single beefy servo doing the tilt thing one continuous glass fibre rod adjoining the motors running through two short
outer tubes which will be glued into the middle of two stubby semetrical wings. Good luck and keep me posted Mike (uk)

[Sparky Paul]

Don Incoll in Australia built an Osprey type with tilt motors..
Here's some of the details..
It flew, but it crashed when the canard collapsed in flight.

[vintage1]

I think making own prop blades is definitely feasible. A CF layup with epoxy.

Gearing two rotors is not impossible either.

Basically if you have access to a lathe, and a computer, and PIC programming skills its probably doable. Defintely want to have a chunk of software between the receiver and the motors and tilt controls as well. And some gyros and stuff.

Motocalc will get you into teh ballpark on thrust/RPM/power I'd say but only into the ballpark.

[migray]

Your joking pal thats way too deep for me!

[mentar]

May I just remind you that Boeing didn't get the control system right itself.
I've heard a few V22s that have crashed and all seem to have crashed due to a software error, which is an indication that the control system is overcomplicated.

The interesting thing is that most of the crashes could have been prevented if there was a backup neural network subsystem, I haven't tested how a dualrotor system would perform but a trirotor one seems to be performing well (in simulations that is). For my trirotor project I am going to try to implement most of the controls as a neural network even though I've always been taught to exclude AI altogether from safety critical systems (I don't want the thing to crash you see) ;)

[sc_deimos]

Don't let me put you off; I'm just trying to ensure you appreciate the complexity of the job you're taking on!


Oh no, not at all. Let's just say the project has "evolved" somewhat. :rolleyes:

For anyone interested...

Arrived two weeks ago were two Hirobo MRB-III 3-blade 30-size heli heads with all associated bearings, collars, linkages, a pair of Shutle ZXX frames to cut-up and a set of LH and RH screw blades. Of course, nobody mentioned that the MRB-III conversion kits actually come with a set of RH blades already, so I now have 3 sets of RH blades. Oh well, I guess this means I only need to order LH blades for the first couple of crashes!

I had always been planning for an onboard control system to save buying a 20-channel-plus radio and spending ages how to drive it before actually programming it to do the task. Now I'm just using a bigger PIC. I realized early-on in the programming that I was going to have to do some funky multi-tasking if I wanted to get 50-frames per second of servo pulses to all 20 (yes, that's right) servo outputs. The simulation ran out of CPU cycles using the built-in 4MHz oscillator so I'm now running with an external 20MHz crystal (sheesh, all that extra weight!) to leave enough breathing room for all of the mixing math.

I've given-up trying to make heads and tails of the motor charts. I'm waiting on an AXI 2820/10 brushless motor and Castle Creations PHX-60 brushless controller so I can do spool-up and current drain tests on a set of ZXX frames. From there I should be able to narrow-down the motor selection somewhat and then finalize current requirements for 10C/15C lithium-polymer battery packs.

30-size equivalent electric heli's seem to get along okay with Hacker B50/15's, so I'm trying motors around that power rating. The AXI should be a little torqueier (is that a word?) being an outrunner. I'd have liked to have gotten hold of a Himax HA-3615-3233 brushless motor to play with but the various agents in Australia only keep limited ranges, and even then only the direct-drive or *43 (4.3:1 geared) variants.

I think my biggest obstacle at the moment is trying to find 12-tooth pinion gears (for 5mm shafts) which have the same pitch as the 77-tooth drive gear on the masts (Hirobo 0402-591, SE Main Gear). They're some weird spacing that's almost 25-pitch, but not. I may have to bite the bullet and replace them with "more standard" 32-pitch gears for an 8mm shaft, however I'm worried about stripping those fine teeth. Particularly if they're nylon. Suggestions anyone?

This will certainly be an interesting build. While I'm waiting for parts to arrive I've just finished an S2G Miniminus and am working on a 200% Plouf. Then, when I finish my ESM F4U Corsair I should have enough workshop space to concentrate on the Osprey again.

[sc_deimos]

May I just remind you that Boeing didn't get the control system right itself.
I've heard a few V22s that have crashed and all seem to have crashed due to a software error, which is an indication that the control system is overcomplicated.

Actually, the only one I'm aware of that crashed due to a "software error" was already having serious issues because one of the hydraulic lines driving the tilt mechanism on a nacelle had been abraded, damaged and burst taking half of the onboard hydraulics system with it. The pilots followed the correct action (by the manual) and pressed the FCS reset button when they saw the warning light on the dash. This only made the flight situation worse because the system went into diagnostics when it detected hydraulic pressure was down. And the pilots pressed the button several times.

That crash was worsened by a software assumption (that the mechanicals were OK, not an error per se), but was caused by poor mechanical design allowing abrasion on the hydraulic lines in a place where they could not be easily inspected by maintenance crew. This area has since gone through a design change.