Looking for featherweight geared hubmotors, >10kW

[kubark42]

Silly question perhaps, what's a redrive?

Some kind of speed reduction unit.

please dont forget that motor bearings are designed for axial loads. aka: rotating, not pulling/pushing.

a prop directly attached without thrust bearings will basically pull the rotor out from the stator and kill the bearings en if they do stay together. not something you want to happen in the air, or like...at all.

I had this same conception, and was surprised to find out that that's not what happens. It turns out that the forward thrust is a very small component of the overall loads, and standard roller-bearings can resist both axial and radial forces. I have several dozen airplane motors, from 5W for drones to >40kW for manned flight and every one of them uses standard roller bearings. Likewise for my 25hp 2-stroke plane engine where I just replaced the bearings.

P.S. Does anyone else have problems with E/S email notifications? I just got an alert there was a new message and I see I missed several days of the thread.

[flippy]

drone engines are NOT the same as a full on 5ft wide prop on a plane. the forces are immense. and a plane carries meat bags, drones dont.

and most drones have deep groove ball bearings, those can handle -some- axial loads. still, drones have very low runtimes so bearing life is generally longer then most other things on a drone as they usually end up crashed before the bearings give up.

and your petrol engine HAS axial bearings. every ICE motor has them. you just probably missed them as they are often integrated on a surface.

[everythingisawave]

The problem you're having (torque vs watts vs size vs rpm) has been discussed in some detail around here before, but I couldn't find the threads in a minute's search. If I run across one later I'll link it for you.

Basically, the problem is watts = torque x rpm, and it takes a certain amount of physical windings and magnets to produce a certain amount of torque, more or less, given a certain rotor diameter. And motors are really efficient only in a part of their RPM and torque range--each one with it's own curve.

So if you want more torque, you either make the motor larger diameter (larger moment arm, I think it is called), or you make it more massive with more copper and magnets. Or some compromise between the two.

If you want more torque but you cant' make it big like that, you spin a smaller one really fast instead, and then downconvert that speed into torque externally (losing some of the power in the conversion).

The little I have read and seen of EV-aircraft of all types end up using the latter option. Some of them don't say why, but those that did usually were that the weight and volume of a high-rpm motor plus the reduction unit was less than that of a direct drive system massive enough to do it. And that the fast motors / reductions were tyipcally COTS and DD systems were not. Some that said why did it because their props needed the high RPM anyway, rather than the torque, so they just directly used the motors without reductions.

But I know nearly nothing about how to figure out what an aircraft needs for this kind of thing so while I can help "keep it real" for some things (I'm good at pattern analysis, deductions, stuff like that, and have a lot of general knowledge tucked away in here somewhere, when I can find it), I am not sure I am qualified to advise which actual systems will do what you want. :/

One thing that is motivating going to higher speeds is sheer power density. Torque requires flux, and for a given size motor this is severely limited by material properties like iron saturation, whereas it's easier to increase DC link voltage and just increase the maximum speed. Dielectric ratings of power devices are relatively high, and the high end EVs are going to higher and higher DC voltages. On the other hand, there are some high torque motors that provide great power density like the YASA motors, although nothing as light as you need. CityAirbus seems to be going the high torque low speed route too.

https://www.airbus.com/innovation/zero- ... irbus.html

[meddesign]

That sounds pretty close to a Sur-ron motor. Not a hub motor, but they have a good track record.

I could not find specifications ....I send them an e-mail.....do you have the specifications?....

Check out this page:
https://www.sur-ron.com/us/power.html

Keep in mind the power rating of an electric motor is largely determined by the controller and heat dissipation. With a lot of air, you can run the motor at a higher power level without overheating. The motor on my Sur-ron is "rated" at 4kW but I can run it at over 10kW for fairly long bursts.

thank you....I did look and there was more data....but not what I would call a spec. sheet...with weight and power output continuous/max etc....there was some reference on one motor of up to 18kw.....I have seen video of Chinese stuff where they blip the engine for a second and say look XXXX max power......the reviews I saw had 5 stars for delivery of a new engine after the first one burned up......so why did the first one burn up?.....The USA distributor or what ever he is....said he did not have that information and I should contact China....this is really not very professional even though they do have a really slick set of web pages. please remember I can not pull over to the side of the road if my motor fails......"a fairly long time or up to xx power" is not any way to design an aircraft engine

[Elektrosherpa]

please dont forget that motor bearings are designed for axial loads. aka: rotating, not pulling/pushing.

a prop directly attached without thrust bearings will basically pull the rotor out from the stator and kill the bearings en if they do stay together. not something you want to happen in the air, or like...at all.

There is not a thing called "motor bearing".
The right term is ball bearings, and there are various forms of these, for various purposes:
https://www.nskeurope.de/de/products/wh ... types.html

So this statement above is way to simplified.

[flippy]

There is not a thing called "motor bearing".
The right term is ball bearings, and there are various forms of these, for various purposes:
https://www.nskeurope.de/de/products/wh ... types.html
So this statement above is way to simplified.

its not simplified. i have repaired hundreds of DD motors. exactly ZERO of them had any other bearing in them then the cheapest chinese basic deep groove ball bearings like a 6004 2RS. those sizes are rated on a axial load capacity of 0.25 of basic load rating wich usually is not a lot, especially not for the bottom barrel chinesium bearings found in the motors we get from factories. and the bearings are not mechanically fixed. they can just slide out if moderate force is applied. i never have to use more then a simple light tap of a deadblow hammer to pop a bearing out of the housing. how the fudge do you expect a DD motor to keep from rubbing the bearings out when you only give them an axial load 100% of the time?

a 10kW power output will roughly generate 800~1000 newtons of thrust with a good prop. how do you think a bearing desinged for radial loads and -some- axial loads will survive the loads a propellor and the axial force of pulling a plane will last? dont you think putting on a axial bearing to take those forces might not at least be prudent?

[meddesign]

There is not a thing called "motor bearing".
The right term is ball bearings, and there are various forms of these, for various purposes:
https://www.nskeurope.de/de/products/wh ... types.html
So this statement above is way to simplified.

its not simplified. i have repaired hundreds of DD motors. exactly ZERO of them had any other bearing in them then the cheapest chinese basic deep groove ball bearings like a 6004 2RS. those sizes are rated on a axial load capacity of 0.25 of basic load rating wich usually is not a lot, especially not for the bottom barrel chinesium bearings found in the motors we get from factories. and the bearings are not mechanically fixed. they can just slide out if moderate force is applied. i never have to use more then a simple light tap of a deadblow hammer to pop a bearing out of the housing. how the fudge do you expect a DD motor to keep from rubbing the bearings out when you only give them an axial load 100% of the time?

a 10kW power output will roughly generate 800~1000 newtons of thrust with a good prop. how do you think a bearing desinged for radial loads and -some- axial loads will survive the loads a propellor and the axial force of pulling a plane will last? dont you think putting on a axial bearing to take those forces might not at least be prudent?

https://www.skf.com/us/products/rolling ... cid-493604

if you type in your bearing it will give the dynamic load rating.....for the basic 6004 it is 8kN....8000 N so .25 of that is 2000 N or about 450 LB...now there are 2 of these I would assume in your motor so 900 lb.....with my propeller I am hopping for 80 lb of thrust out of my 20 hp ICE....these thrust numbers have been measured with my dia. and pitch prop. with the same engine and rpm .....I think we have some wiggle room here. but thank you for your post... I think it does offer some insight to the quality of the Chinese motors.

[flippy]

axial load calculations is done with static load, not dynamic load.
because the value is so low they dont advertise it and you need to calculate it yourself. but dont forget that its the LIMIT of what it can take. not how long it can take that load....

and only 1 bearing is acutally doing the work when its butted against the hub. the other bearing is technically floating and does nothing apart from keeping the axle in place, it does not have any axial loads. so all the axial load is held by 1 chinesium bearing and the actual 1 side the the motor that generally is just bolted on with a couple cheap stainless M4 bolts. not very confidence inspiring when you are a couple thousand freedoms up in the air and your prop and 50% of your engine yeets itself from its designated place back down to earth...

[meddesign]

axial load calculations is done with static load, not dynamic load.
because the value is so low they dont advertise it and you need to calculate it yourself. but dont forget that its the LIMIT of what it can take. not how long it can take that load....

and only 1 bearing is acutally doing the work when its butted against the hub. the other bearing is technically floating and does nothing apart from keeping the axle in place, it does not have any axial loads. so all the axial load is held by 1 chinesium bearing and the actual 1 side the the motor that generally is just bolted on with a couple cheap stainless M4 bolts. not very confidence inspiring when you are a couple thousand freedoms up in the air and your prop and 50% of your engine yeets itself from its designated place back down to earth...

if it is going to fly it needs to be light.....the issue is will it work?.....and how do we calculate it or test for it or find the honest vendor with good test data....and my prop is a model prop because I currently have a 3W-200 Giant RC scale ICE motor ....larger props turning slower do give much higher thrust numbers.....

[flippy]

i think you can spare a couple hundred grams for a proper axial bearing and solid mounts. just take a good shit before you fly to compensate.

[kubark42]

drone engines are NOT the same as a full on 5ft wide prop on a plane. the forces are immense. and a plane carries meat bags, drones dont.

The forces in airplanes have a way of being surprisingly non-intuitive. The thrust forces are not immense, and in fact they are surprisingly small. (See below.)

The main forces the bearings have to resist are the gyroscopic forces which arise when the plane changes direction, esp. if there is sharp turbulence. In a direct drive setup, these forces are by far the dominant forces on the shaft bearings. Once we spec the bearings to survive these, the axial thrust force is just a few percent additional load. The calculations are straightforward and published on bearing websites.

and most drones have deep groove ball bearings, those can handle -some- axial loads. still, drones have very low runtimes so bearing life is generally longer then most other things on a drone as they usually end up crashed before the bearings give up.

and your petrol engine HAS axial bearings. every ICE motor has them. you just probably missed them as they are often integrated on a surface.

I think it's important to not confuse plain bearings with deep-groove ball bearings. You're right that four-stroke engines typically have plain bearings( specifically journal bearings) and these cannot resist axial forces. As a result, the engines must have a thrust bearing (as shown in the photo). However, the two-stroke powerplant which carries this meatbag up uses ball bearings only. When I tore the entire unit down last summer I replaced them so I'd consider myself familiar with the subject.

I also disagree that drones necessarily have short lifetimes. Consumer drones perhaps do, but the larger scale ones, such as we built, were tested for thousands of hours. (We had a non-stop two-week long flight of our tethered drone.) My Joby JM-1, with its roller ball bearings, was developed for both drones and full-scale, and is currently being used for manned aviation tests.

a 10kW power output will roughly generate 800~1000 newtons of thrust with a good prop.

I am not aware of any airplane prop which can make that much thrust with that little power.

• My 18.3kW engine (two ball bearings) with a 46.5 x 35" wooden GSC prop makes 600N of static thrust.

• My 16kW Geiger motor (two ball bearings) with a 55" x 30" carbon-fiber Magcad prop will make 600N of static thrust.

• My 12kW Joby JM-1 (two ball bearings) could not make 800N of thrust even when pushed to 17kW peak.

• My 8kW T-motor U15-II (two ball bearings) makes 150N static thrust.

• My Emrax 188 (never disassembled so don't have firsthand experience with the bearings) could make that much thrust, but it's a 52kW motor.

Maybe if you played around with ecalc.ch you could find a 10kW motor with prop which could produce that much static thrust-- although I am somewhat doubtful--, however the prop would be so flat as to be completely impractical in forward flight. It's useful to know that somewhere around 67% pitch vs diameter the prop will be stalled in a full-speed static test and so there is a sharp upper limit on how much static thrust we can get out of a prop designed to provide high dynamic thrust.

how do you think a bearing desinged for radial loads and -some- axial loads will survive the loads a propellor and the axial force of pulling a plane will last? dont you think putting on a axial bearing to take those forces might not at least be prudent?

Adding the axial bearing can actually make things worse, especially with alignment. It could pull the roller bearings out of alignment or force them to bear the opposite thrust load.

Sometimes you'll see angular contact bearings in motor drawings, but I personally have yet to see one in the wild.