Help please, Lightweight motor for Hovercraft

Hi Everyone,

As the title suggests, I'm working on a design for an electric hovercraft and need suggestions on sources for a lightweight motor and generator.

I'm talking about a small craft for 1 to 3 persons, not a model, or a large craft. Now a completely electric hovercraft isn't going to perform very well, so I'm only looking at making the lift system electric. One standard way of making a hovercraft is to use separate engines for lift and thrust. Another is to use a single engine with mechanical drives to the lift and thrust fans. That's what I'm using at the moment, but it has the problem that when you reduce the thrust, you also lose lift. So what I'm looking at is an electric motor to drive the lift fan and a generator (which might be the same device as the motor?) driven from the engine.

Basic parameters are:
Lift fan: 4 to 5 kW at 3,000 to 5,000 rpm, so I'm hoping a direct drive motor would be possible.
Engine operating range: 3,000 to 12,000 depending on the craft.
Weight of course is absolutely crucial.

I haven't said anything about batteries yet, because I'm still working out whether I actually need them, or a big capacitor, or what. Nor about the controller, because I think that might not be the most difficult part. Also at this stage it helps to not have the voltage defined.

What do you reckon? Is there a lightweight device out there that I could use one of as the generator and another as the motor. Or is it best to have two different devices?

Thanks

Nick

----------------------
PS. Since I'm fairly new here, I ought to introduce myself a bit. I live in South West England; I'm an electronics engineer; I've been flying airplanes for 20 years; I've been racing hovercraft for 10 years and I ride an electric bike.
----------------------

Welcome to here.
Sounds like an intresting build. I think you'd end up with more weight trying to use batteries, unless you don't mind a very short run time. But running from a generator should be possable. the Etek, Etek-RT, and Mars motors all run in about the range you're looking for.

Gas engine to generator to electric motor efficancy losses will probably be higher than mechanicle driveline losses, but that might not be a problem.

Hi Drunkskunk,

Yes the aim is not to use batteries for the power source, so I may not need much weight there. And to be honest the real weight comparison should be not with the existing mechanical drive train (toothed belt, drive shaft, right angle gearbox) but with a hypothetical variable ratio one. So a slightly heavier but more versatile electric transmission would have an advantage.

Thanks for the motor tips. I will check them out - do you have any links?

Nick

I would take a look at neu motors.

http://www.neumotors.com/20061222/2200_series__.html

Thanks Drewjet,

I seem to keep turning up 3 kg motors that are too low power and very powerful ones that are way too heavy. Looks like Neu have a good range that I will study.

Keep the ideas coming, please. Anyone got any experience of LEMCO motors?

Nick

Lemco motors are pretty heavy.

I wonder if you could use multiple RC model outrunner motors for the lift fans and one larger one to use as a generator. If all the motors and generator are 3 phase brushless motors with permanent magnets, you could just tie all the wires together and it would be sort of like an electric transmission. No controller. No batteries. The motors would try and turn at the same rpm as the generator. Ideally, you'd want the generator part to be built into the flywheel on the engine to eliminate any belts etc. I'm not really sure this would work, but I don't see why not.

Hi Fechter,

An electric transmission is really what I'm after, but I want to have a variable input speed, so I reckon I will need a controller, even if I can do without batteries.

Multiple motors is an interesting idea. Can a set of brushless motors be run off one controller? You couldn't do it on a road vehicle as the wheels get out of phase, but driving fans you possibly could.

What about several motors driving the same fan? If they have through shafts that should be easy enough.

Nick

I have driven to motors off one controller. It does take having identical motors to do it. I have also seen multiple motors driving one propshaft.

you can run 2 motors off 1 controll, but if one motor is under higherload, it can cause some phase problems, and possably hurt the controller. its better to have them attached to a common shaft. There are a number of RC motors that are capable of 3000 watts, but the RPMs are higher.

I think what you're trying to do has been partialy done before. Electricppg.com is a powerglider company that uses an electric fan with an RC motor. I bet for racing, you would want the gas motor powering the main fan, but its proof that the idea is atleast partly doable.

Thanks guys,

The PPG comparison is an interesting one - similar ballpark but I think the PPG peak power requirement is a bit higher. I shall look into RC motors (links, suggestions?). I notice that in that PPG picture they have a reduction drive and my guess is the prop is running about 3000 rpm. If its possible I would prefer the simplicity of a direct drive.

Nick

For thrust purposes a large prop spinning slowly is much more efficient.

If you want to go multiple motors it is hard to beat http://www.hobbycity.com out of China.

How about the Plettenberg Predator 30 or 37?
http://www.plettenberg-motoren.com/UK/Motoren/aumo_ueb.htm
http://www.icare-rc.com/plettenberg_predator.htm

There's a company called Hacker that makes an A200 brushless outrunner. That motor is a direct bolt in replacement for a 150cc nitromethane RC aircraft motor.
Unfortunatly, there site is down right now while they revamp there system, So its hard to give specs and such.

I'm not going to pretend to understand hovercrafts, but I would guess your lifting fan is more about total air volume and velocity, and not so much about fan thrust? If thats the case, could you use several smaller fans?

Thanks guys,

Those RC motors look very interesting. I didn't realise they went up to such powers. Has anyone here used them - are those figures reliable; are those continuous ratings? What's the catch?

Drunkskunk, you're right in that thrust and lift are different requirements. There are craft that use one fan for both, but the best performance comes from using separate ones. For lift, the pressure generated across the fan is very low - its mainly to overcome the distribution losses, but you do need a reasonable volume, and hence power, to operate on anything other than glass smooth surfaces. Large craft may well have multiple lift and trust fans, but for the small size I'm looking at very few have done it that way.

Nick

I haven't personaly used the bigger stuff, but I've seen them fly. My personal experiance is with the 400 watt size. 49 grams of motor produces a little more than 1/2 a horsepower, and thats enough to propell a 3 pound plane up to unbelieveable speeds.
The plane below is a menece at my airfield. I rutienly chase, catch, and pass big motored planes with it. Since electric motors make full power from 1 rpm on up, these motors have far better acceleration that an equivilent Gas motor. this plane has beaten a porshe down an 8th mile road.

Drunkskunk said:

the 400 watt size. 49 grams of motor produces a little more than 1/2 a horsepower

Click to expand...

At that rate, 4 kW should be 490 g. So why does a 400 W bike motor weigh 2 kg?

These power to weight ratios make the electric option well viable for the hovercraft. But one thing I'm learning is that with electric motors, power is a variable feast. There seems to be this idea that you can always give it more volts, get more rpm and therefore claim more power. Is there a standard definition, or is this even worse than defining the power output of audio amplifiers?

Nick

Tiberius said:

There seems to be this idea that you can always give it more volts, get more rpm and therefore claim more power. Is there a standard definition, or is this even worse than defining the power output of audio amplifiers?

Nick

Click to expand...

Thats true. you can keep adding voltage and get more power up to a point were the voltage starts arking inside the motor, or the motor can't disapate the heat anymore, or the rpms climb so high it all flys apart.
Generaly, the specs on the RC motors are close to it's true max, as having a spinning blade equiped device flying over the heads of innocent bystandars is a bit of a legal liability, and they want to cover them selves with accurate information. As for Bike motors, its important to note that most DC brushless motors are still wound by hand, and there is some variance between motors. The manufactures rate them at a minimum of what they will do. Most will be capable of double their rated power.


The weight diffrence has everything to do with RPM. my 400 watt plane motor spins at 30,000rpm. my 500 watt bike motor does 300rpm. 1/100th the speed. Very roughly, thats an increase of 1% weight for 1% decrease in RPM. To make a motor work at the same power output but lower RPM, you need more magnets, more motor poles, and more iron mass for the magnetic field to work against.

The whole spinning blade, bystander question is an issue with hovercraft as well. In racing situations we have to specify a full containment duct or cage in case blades break off when there is a collision or a rollover. There is also the question of reliability.

So am I right in thinking that a chunky bike type motor could be run at double the quoted power, but with an RC one I should stick to the quoted numbers? I agree with the insulation breakdown and flying apart points, but isn't there also the more practical problem of finding a controller to operate at very high voltages?

Nick

Kinda. Most of us have to mod our own controllers if we want to go over 72V. Either the cheap controllers won't handle it, or the controllers that will are too pricey. Most R/C controllers seemed to be designed for crazy high amperage (e.g. 150A @ 48V), since they have a high rpm/V ratio.

BTW, it's rarely dielectric breakdown that kills an overvolted motor. They can usually take quite a bit of voltage. It's the brush arcing that destroys it. It will hold an arc and send power to a coil even after it's past a magnet pole, so almost all the power going to that coil is generating heat. Then the insulation melts, and the windings start shorting, and it starts drawing huge currents, and then your power wiring catches fire, and then your batteries explode, etc.

For any electrical motor the maximum efficiency operating point and maximum power density operating point are usually different. For a full scale hovercraft I would guess you should worry more about operating near max efficiency point (keeping weight within reason). High end RC motor manufacturers (Neu, Pletti etc.) usually provide efficency curves. Looking at those you can determine what power (and best rpm) you can get from a particular motor without too much efficiency degradation. Best xmission efficiency would be in the ballpark of 0.9 (motor) x 0.9 (generator) x 0.9 (controller) = 0.73.

On running multiple motors from the same controller - I've tried that on RC models but would not recommend for a full scale for safety/reliability reasons. If for any reason two motors run out of sync there is a good chance of frying the controller.

How much variation in gas engine rpm do you want to compensate for in the controller ? This would dictate the best solution. If it is in the ballpark of 20-30% you can use RC ESC (motor controller in RC slang) designed for RC helicopters - they feature an integrated governor mode. This will compensate for generator voltage variation due to changing rpm. Take a look at Castle Creations Phoenix HV-110 controller (good for over 5kW). Then adding a 3-phase bridge to a generator feeding a big 50V capacitor will complete the circuit.

If you need much wider rpm range to cover you may have efficiency issues with this simple scheme and a more elaborate electronics would most likely be required. Another way around it would be using variable pitch prop on the thrust engine - should greatly reduce engine/generator rpm range.

curious said:

How much variation in gas engine rpm do you want to compensate for in the controller ?

Click to expand...

Hi curious,

A good question. I'm looking at several configurations, each with different answers. Given that some small craft work with fixed transmissions to both lift and thrust fans, it could be argued that no compensation is needed. However, what's happening there is that people are relying on the fan characteristics, and the lift system absorbs excess power at high rpm.

In practice then, the minimum I would want is to vary the gas engine from 2/3 to max rpm - 1:1.5 ratio. Ideally it would be nice to run 1/3 to max - 1:3 ratio.
As you say, a variable pitch thrust fan could be used, but even then I would want the lift system to be able to cope with a range of engine rpm.

Nick

curious said:

you can use RC ESC (motor controller in RC slang)

Click to expand...

Hi Guys,

I'm trying to get my head around these Electronic Speed Controllers. The RC motors seem to be brushless and sensorless.

Do the ESC's simply switch the supply around between the wires, and vary the rate to control the speed? This would be like the 3 phase variable speed motor control for an AC motor. Or do they phase the switching to match the motor position and have a separate power control with PWM? That would be like a controller for a brushless motor with sensors.

The fact that they stand a range of input voltages would make me think the latter, but if its only a narrow range that would point to the former.

Nick

I'm sure someone can crunch the numbers, and I don't want to discourage a fun sounding project, but my gut tells me no RC motor will work, an Etek either. Even the smallest one man hovercraft use 50 to 65 horsepower engines. Unless you get close to that type of power you're just not going to ened up with what you said you were aiming for. Don't let me stop you, I'd love to see one built, I just think that the weight of motor and batteries is going to limit the ability to build a hovercraft that gives pleasing performance.

Nick -

Most RC ESCs for brushless motors are are of sensorless type. They use motor back-emf signal measurement to synchronize winding commutation with the rotor phase. Only at the start-up they operate the motor without feedback like a stepper. Once say a 5% of the nominal rpm is reached there is enough back-emf to close phase control loop. On top of the phase commutation the bridge mosfets also get high frequency PWM signal from the throttle (and/or governor loop). One thing to remember is that RC ESC throttle input is not a simple analog value like in EV controllers but a special form of PWM signal (different from mosfet PWM mentioned above). A simple encoder is needed if ESC throttle input is not fed from standard RC RX.

Now I would be somewhat concerned about real world efficiency of electric transmission. The 0.73 figure is kind of optimistic best case scenario. The idea is certainly doable as a proof of the concept but wouldn't a second small 5kW gas engine be a simpler, cheaper and more efficient solution ?

curious said:

Nick -

Most RC ESCs for brushless motors are are of sensorless type. They use motor back-emf signal measurement to synchronize winding commutation with the rotor phase. Only at the start-up they operate the motor without feedback like a stepper. Once say a 5% of the nominal rpm is reached there is enough back-emf to close phase control loop. On top of the phase commutation the bridge mosfets also get high frequency PWM signal from the throttle (and/or governor loop). One thing to remember is that RC ESC throttle input is not a simple analog value like in EV controllers but a special form of PWM signal (different from mosfet PWM mentioned above). A simple encoder is needed if ESC throttle input is not fed from standard RC RX.

Click to expand...

Thanks for the explanation, curious. That was how I was hoping they worked. Is that throttle control the same PWM signal used for servo positions in RC; I am at least familiar with that scheme?

Nick