Alternative motor shaft design and dropouts concept

[Kodin]

I've been thinking about it for a while, and I'm wondering why no one has considered designing a motor with custom dropouts and a shaft that can transmit torque more effectively to the frame. Looking at options, there's keyways, some sort of hex shape, or some sort of spline. I'm divided right now on hex or keyway as splines get expensive quick, and really seem overly-complex for the cost. Here's my idea. On high-powered systems we're already deploying custom cut torque arms anyway, so why not do something along the lines of an item we can all recognize:

Take a torque arm and cut a hex into it. Then take a large diameter shaft, and cut a tight-tolerance hex into the shaft where the torque arm would engage. Then make that torque arm clamping by welding tubes to the open ends and put a bolt through it. Gives you more diameter and more engagement points to work with. You can then cut a thread into the outside section of shaft for a nut, but the strength comes from the hex, and you aren't limited to a 10mm thickness anymore.

I'd assume from a manufacturing standpoint it'd be pretty quick to re-tool for this; the only concern is varying tolerances for diameter. Other option is keyways, however that gets a bit more challenging unless I can figure out a good method for that while keeping it somewhat easily removable...

Best design I can come up with off the top of my head is this:

Figured it'd have some sort of slide-in backing plate to keep the shaft from sliding off the keyway that's engaged by the bolt as well.

...Thoughts?

 

[Hillhater]

How is that hex design any better than existing "flats" in a clamping drop out ...which is already used by many. . ?

 

[Kodin]

Mostly in cross-section thickness. Most of the motor shafts out there are designed to fit into standard bike dropouts. I think that's a serious limitation that can easily be mitigated by using something like a combo torque-arm/dropout that mounts up to the existing dropouts, clamps to the frame like a torque arm, and extends out to provide new larger diameter dropouts. Either down or back from the existing frame. I wouldn't consider it a "universal" solution, but may be more flexible than what's currently available. Not to mention, we're seeing more purpose-built ebike frames lately, and a new standard makes sense as motors continue to become more and more powerful.

 

[Cowardlyduck]

I like where this is going and definitely think something needs to change.
Clamping drop-out's are great, if you have them or can mount them, but not all bike's/design's allow for them, and really that's an aftermarket solution to a manufacturing problem. The manufactures should all agree to make it so that it's not a problem to begin with by designing something adequate.

I think a bolt driven through the middle of the axle would be the best option.

No idea if this would actually work, but it would prevent any play even if there is slop between the torque arm and axle flats which from what I've seen is often the case. It would also prevent the motor flying out the back without the need for clamping dropouts....this actually happened to me.

I welcome any manufactured solution to this ongoing problem as I've yet to see the hub motor manufactures even so much as acknowledge it as a problem.

Cheers

 

[Alan B]

Best is to remove the torque reaction from the axle and handle it separately. As Justin has done, as drum brakes do, as some IGH's do.

 

[Kodin]

Quite a novel idea if you don't use a chain at all like a motorcycle, but for the cog side you're limited to a certain shaft size for freewheel and cassette hubs. I've been trying and trying and trying to get a cassette to fit with a decent shaft size on my motor and finally realized I can simply take a 9-speed DNP freewheel and shove a bearing on the end and be done with it. Gives me the largest shaft diameter I can find that way, since cassettes can't get more than about 14mm through them before they no longer turn. Hell, through-axle hubs are only 12mm for rears. I got the hex idea from Justin's custom 20mm through-axle creation, but that won't always work with every motor as many simply don't have a large enough diameter bore through the stator for a shaft that large. I DO like the idea of a through-bolt for the axle; definitely simplifies connection, but still allows for a bit of twist in the mount if you use regen. Maybe a combination of through-bolt and key for the ultra-paranoid? Looking at a ~1/2" inch keyway length, (12mm just for the hell of it,) on a 14mm shaft, (largest you can fit in a cassette hub that I've found,) this calculator says you need a 5mm square key installed. That's totally doable. And that's for just one side connection at 100 nm. I was thinking two keys, one on either side, with 1/2" thick dropouts and keys. If you go to freewheel you can go even larger.

 

[greg_p]

Why not using traditional motorbike-like fixture. It is working for big power since years, you can set this by yourself and it is rock solid.

You set first right side to adjust chain tension, and then you go left side to manage wheel alignment.

 

[Kodin]

Because those don't have to lock the shaft from spinning against the frame; only lock the shaft to the frame. Hubmotors transmit all their torque through their axles. A motorcycle style mounting arrangement will still allow the shaft to spin. That's why you need torque arms on pretty much 90% of high-powered bikes out there.

 

[Hillhater]

You guys dont seem to have grasped the basic principle of a clamping drop out.
None of the above designs are actually clamping the axles very effectively.
You have to design it such that the clamp can flex onto the axle, and the bolt need to be positioned to exert maximum clamping load .
but i agree with Alan B that its time we got the torque off the axle and the wireing also out of it too.
Big dia bearing on one side allowing torque reaction hook up and sensible wiring feed.

 

[Kodin]

I was just informed of Taperlock bushings. Basically a bushing with a taper on the outside, a slot cut into it to be able to compress onto a shaft, and a keyway. Weld or bolt a hub into a frame, screw down the bushings onto the shaft, done. Further, to pull the whole wheel, remove the grub screws, install one to extract the bushing, pop it, slide off the shaft, repeat other side, and do a slide left, tilt, slide right, motor/wheel removed.

[youtube]fT4axK3haQ0[/youtube]

(Video is better at explaining than I am)

 

[friendly1uk]

250w is almost the world standard, and needs no special frame or motor design. Allowing both manufactures and people at home to mix and match existing frames with existing motors, giving the maximum choice of possibilities. Why would anyone want to change this?

Motorcycles are a different forum, but I recognise the confusion. These hub motors are generally the same but bigger, and require careful frame choice or special frames. That exist. It's a niche product arena though that you don't want ebikes all falling into. If you want to go bigger still you can change hub motor arrangements to unload the axle but really a mid drive is more suitable.

Most alternate designs add stress points. The hex idea adds 3 to the dropout while thinning the axle on 6 sides not 2 and the flats become shorter so harder to hold.



I have lost more posts than I have made recently. This is my 3rd try. I just know I have been logged out...
Correction 4th try now

 

[dogman dan]

There is a wonderful and elegant solution, but it costs money so nobody does it.

I seriously doubt China respects a patent, so that's not the reason they don't copy the Heinsmann brushed hub motor axle design.

In the Heinsmann, the axle is much thicker where it exits the core on one side. It's literally about 35 or 40 mm diameter. This allows a nice long flat, which is cut into the center of the axle, allowing a torque arm to slot into that key. the TA has a 10 mm round hole in it, for the regular 10 mm round axle to poke through. the TA is on the inside edge of the dropout in this design.

It's highly effective, and the motor fits all bikes easily front or back, with no need for c washers or other adaptations to fit the bike.

But that huge sealed bearing costs at least a few bucks more than a smaller one, so we are still stuck with dumb ass 14 mm axles that don't really fit any bike, trying to hold them from rotating by a tiny flat. It's moronic!!!

FWIW, the picture above of the motorcycle does look like it has something similar, see the TA on the inside. Then on the axle at the nut, round axle.

 

[cycborg]

Best is to remove the torque reaction from the axle and handle it separately. As Justin has done, as drum brakes do, as some IGH's do.

but i agree with Alan B that its time we got the torque off the axle and the wireing also out of it too.
Big dia bearing on one side allowing torque reaction hook up and sensible wiring feed.

In the Heinsmann, the axle is much thicker where it exits the core on one side. It's literally about 35 or 40 mm diameter. This allows a nice long flat, which is cut into the center of the axle, allowing a torque arm to slot into that key. the TA has a 10 mm round hole in it, for the regular 10 mm round axle to poke through. the TA is on the inside edge of the dropout in this design.

It would be great if designs moved in this direction, so we could stop fiddling with imperfect torque arms or custom clamping dropouts. As Alan mentions, Grin is implementing this in their design.

friendly1uk, do you know what that motor is in the photo?

 

[Kodin]

Ah, interesting. The original Grin design was using hexes, looks like they've moved to a spline for better holding capability.

 

[gogo]

friendly1uk, do you know what that motor is in the photo?

That's an Enertrac hubmotor. http://www.enertrac.net/index.php

 

[dogman dan]

Figures Justin would be working on that. Always something in R&D with him. That motor obviously has the larger bearing size I was talking about. Hex, square, slot, whatever it is doesn't really matter as long as the TA has more to grab onto, and the axle isn't weakened. Just anything but the current design, a axle with half the threads cut off making the nut weak, then depending on nut pressure for 75% of the hold.

The current typical design of motor axles is pure shit.

That through axle design could surely be adapted easily to a bolt axle bike as well. It would simply be a through axle, with a 10 mm shaft on each end, for regular QR skewers.

 

[Cowardlyduck]

Figures Justin would be working on that. Always something in R&D with him. That motor obviously has the larger bearing size I was talking about. Hex, square, slot, whatever it is doesn't really matter as long as the TA has more to grab onto, and the axle isn't weakened. Just anything but the current design, a axle with half the threads cut off making the nut weak, then depending on nut pressure for 75% of the hold.

The current typical design of motor axles is pure shit.

Yup, I've been trying to say that for some time...I didn't imbed the video above but I will now as it's more relevant.
[youtube]DcegHq0HCbk[/youtube]
This is what happens when you try to 'make do' with

a axle with half the threads cut off making the nut weak, then depending on nut pressure for 75% of the hold

.

Cheers

 

[dogman dan]

FWIW, I've posted a similar comment about once a year, ( hub motor axles suck) since 2009 or so, when I had a Heinzmann, and immediately saw the way all bike hubmotors should be done.

Because of the oversize axle at the bearing, lots of room for the wires with nearly no risk of pinching them, and still space for a very effective external temp sensor.

The actual axle material was a lot harder too, on this made in Germany motor. So you could really crank down the nuts with no worries too. 8)

 

[Kodin]

Well, to start I'm actually going to leave out the keyways and keys as an experiment. 14mm right side, 25mm left side, simply tightened down like a collet on both sides to the spec of the fasteners. I'll mark the shaft position rotationally and see if it wants to move by trying to abuse it some. If it shows any signs of looseness, I'll add keys, and if not, hell, I just found a non-keyed mount. Turns out there's LOTS of collet-style shaft mountings out there capable of very high torque both with and without keyways, so definitely something I want to explore more...