What the hell?! Cromotor axle twisted

[Chalo]

That's an unfortunate side effect of using the axle thread for torque transmission. If they'd put a good torque arm on a square interface on the other side, and tapped the axle for 10mm studs for fastening, then there would have been plenty of room to mill cable slots in what remained of the axle.

Using the threaded portion of the axle to transmit torque is bullshit.

 

[alsmith]

That looks like one of the poorest design solutions possible. Have you emailed the manufacturer direct to explain the problem? Even if they don't help- but they should- the very least it will do is let them know there are problems with the component. It might help to spur them on to a proper redesign.

 

[Lebowski]

That's an unfortunate side effect of using the axle thread for torque transmission. If they'd put a good torque arm on a square interface on the other side, and tapped the axle for 10mm studs for fastening, then there would have been plenty of room to mill cable slots in what remained of the axle.

Using the threaded portion of the axle to transmit torque is bullshit.

What I still don't get is, why are motors like this not build along the lines of a drum brake ?
Use the plate where the brake shoes are as the stator and the brake drum as the rotor.
If you look at the brake-shoe side of a drumbrake, there's no problems here to solidly
attach this to the forks/frame so why do emotors use the axle ?

 

[alsmith]

That's an unfortunate side effect of using the axle thread for torque transmission. If they'd put a good torque arm on a square interface on the other side, and tapped the axle for 10mm studs for fastening, then there would have been plenty of room to mill cable slots in what remained of the axle.

Using the threaded portion of the axle to transmit torque is bullshit.

What I still don't get is, why are motors like this not build along the lines of a drum brake ?
Use the plate where the brake shoes are as the stator and the brake drum as the rotor.
If you look at the brake-shoe side of a drumbrake, there's no problems here to solidly
attach this to the forks/frame so why do emotors use the axle ?

I've wondered this too. The design seemed a bit backwards when I first looked into hub motors. I thought it should be ok for our (UK) legal power of 250W, but when I looked for something with a bit more oomph I was surprised there weren't better options. Well, apart from the central drive from the pedals thing but my bike can't fit that system. It could be so much stronger following a design similar to drum brakes - they handle full braking easily where our hub system often struggles with just part of that when using regen. The only problem would be designing a solid mounting on one of the forks so it would probably have to be welding- not such a problem.

 

[gogo]

What I still don't get is, why are motors like this not build along the lines of a drum brake ?
Use the plate where the brake shoes are as the stator and the brake drum as the rotor.
If you look at the brake-shoe side of a drumbrake, there's no problems here to solidly
attach this to the forks/frame so why do emotors use the axle ?

I think this was discussed in another thread some time ago. You'd end up with an arrangement similar to what markcycle did with the the EnerTrac bringing the torque element through inside a larger bearing. Drum brake plates have a different dynamic loading because they are zero clearance when in operative mode and the forces transmitted are all nicely aligned. Maintaining an air gap between structures requires a whole different support architecture.

 

[warrah]

That's an unfortunate side effect of using the axle thread for torque transmission. If they'd put a good torque arm on a square interface on the other side, and tapped the axle for 10mm studs for fastening, then there would have been plenty of room to mill cable slots in what remained of the axle.

Using the threaded portion of the axle to transmit torque is bullshit.

Chalo, im looking at drawing up a complete redesign for the axle, to give to a machinist. If im getting a machinist to make me a replacement axle, i may as well improve it. Would you be willing to post a sketch? or if not, describe the 10mm stud idea alittle more, i'd like to visualize your idea and seriously consider it. Liking the square torque arm idea, i know exactly where that would go on the axle.

Actually guys, go nuts, tell me any idea you have for a better axle. Im willing to completely redesign my drop outs too. Best idea gets created and tested in real life. I really appreciate the endless sphere communities expertise.

From searching around, im aware that dogmans 'clamping dropouts' are considered one of the best workarounds to the fundemental problem of a 10mm axle made out of mild steel (i kid you not, the machinist did the hardness test), being subjected to motorcycle levels of torque. I was considering upgrading to the clamping design. But instead of dropouts compensating for axle weakness, here we have a chance to do it really right, and make the axle and the dropouts compliment each other.

Some things you might need to know: ive spread the dropouts to 160mm, so you can add some width to the redesigned axle if you want to add areas for torque arms. My frame is 'lowrider hardtail motorcycle' style, so the chainstays and seatstays meet the axle at a similar 45 degree angle (i'll post pictures). Anything else you want to know let me know

 

[wesnewell]

Very rough sketch of how I'd make it.

 

[warrah]

Very rough sketch of how I'd make it.

If you imagine where the dropout would be in relation to the threads, the dropout is blocking the cable in your diagram. They place the wire cutout in line with the flats on the threads so that the cable can exit through your dropout, past the nut.

 

[e-beach]

Very rough sketch of how I'd make it.
View attachment 1

+1 WesNewell design. This is the way I always saw it in my minds eye. Obviously the series wire groove should only be deep enough for the your wires to fit comfortably snug.

 

[e-beach]

If you imagine where the dropout would be in relation to the threads, the dropout is blocking the cable in your diagram. They place the wire cutout in line with the flats on the threads so that the cable can exit through your dropout, past the nut.

Or let the series wires exit the hub and run up the inside of the frame. Fasten them to the chain stay or seat stay for a few inches and then let them migrate to the outside of the frame. Makes for a stronger axle.

 

[warrah]

If you imagine where the dropout would be in relation to the threads, the dropout is blocking the cable in your diagram. They place the wire cutout in line with the flats on the threads so that the cable can exit through your dropout, past the nut.

Or let the series wires exit the hub and run up the inside of the frame. Fasten them to the chain stay or seat stay for a few inches and then let them migrate to the outside of the frame. Makes for a stronger axle.

Youre right of course. If i use some spacers on the wire side of the axle, there will be room enough for the wires to make a sharp turn up the seat stays. Its worth it for the stronger axle. Thanks for the thoughts wesnewell and e-beach.

 

[teklektik]

You might make the axle section that is situated within the dropout thickness to have greater diameter than the threaded portion outside the dropout. If your dropouts are 5/16" thick then make a shoulder section that is a bit thinner (5/16-") and that is 20-25mm tall with the same flatted width as threaded portion (green portion to the sketch in the post above). This is assuming you are reworking your dropouts to be the TAs. Might need to check the disk brake alignment...

 

[wesnewell]

Very rough sketch of how I'd make it.

If you imagine where the dropout would be in relation to the threads, the dropout is blocking the cable in your diagram. They place the wire cutout in line with the flats on the threads so that the cable can exit through your dropout, past the nut.

I don't want the cable to run through the dropouts. If I did, I'd just route it that way.

 

[Chalo]

Chalo, im looking at drawing up a complete redesign for the axle, to give to a machinist. If im getting a machinist to make me a replacement axle, i may as well improve it. Would you be willing to post a sketch?

Here's a quick and crude model I hacked out in Sketchup. The slim cylinders on the ends represent M10x1.0 studs threaded into the axle, and the hexagonal prism outboard of the square interface for a torque arm represents a 30mm locknut with a 10mm thread to tighten onto the stud and hold the torque arm in place. I was thinking the wire channels would be cut with a ball end mill, so they'd have a rounded end too instead of a flat one.

This sketch assumes a 25mm bearing bore, 10mm axle studs, and 5mm wire channels. I don't know what the actual size of the Cromotor's bearings is. And I don't know what the interface to the stator is, so I didn't draw it.

 

[warrah]

There she is. The thicker 25mm bearing side is also the wire side, and the wire channel is big, and needs to be... the chunky phase wires are a tight squeeze through there. So unfortunately there might not be enough remaining meat for multiple m10 studs. But the concept of threaded studs is intriguing to me, do you think the same strength would be had from one or two studs?

teklektik, i was thinking along those lines too, make the dropout part of the axle as wide as possible. But im thinking keep it wide for all the way to the tip, and then either use i giant m20 nut, or ditch the threads altogether. Dogmans clamping dropouts claims not to need them, and presumably a nut would get in the way of chalos stud idea. Maybe nuts arent the way to go

 

[warrah]

Thought i'd post a separate thread for the 'help me redesign cromotor axle', to get as much input as possible as quickly as possible. Using the buses really gets me down, im sure youve all been there at one time or another. Sorry if this is greedy or against endless sphere etiquette. Ideas already posted are equally as valuable, and ive linked the two posts. Im just hoping to generate as much input as possible, get the engineering types interested in this, and maybe some debate so i can learn about the relative pros and cons of different designs. Someones idea will get tested, seems like a fun opportunity!

http://endless-sphere.com/forums/viewtopic.php?f=2&t=53931