Dropouts for hub motors

[friendly1uk]

I keep seeing horizontal dropouts on bikes made by newbies. I have aired my concern a few times, and peeps seem to miss my point. Is it not as glaringly wrong as putting your shoe's on the wrong feet?

edit: I'm not passing along my opinion here on purpose. I want to see if other people go straight for the obvious. I do believe it's obvious too.

 

[wesnewell]

Vertical dropouts are usually shallow and don't leave enough room for proper fit. It's almost impossible to change wheel alignment in them. I'll take horizontal dropouts over vertical any time. These are the horizontal type I'm talking about.
http://sheldonbrown.com/gloss_dr-z.html#dropout

 

[cal3thousand]

I don't see anything obviously wrong with horizontal dropouts. Care to elaborate?

For me, they seem easy to build torque plates

 

[friendly1uk]

got about 10 seconds....

I'm surprised by sheldons idea of horizontal. To me, it is parallel with the ground.

It's all about the axle. You take one, skim it's sides then drill or channel it, and call it a motor shaft. It's not as strong in all directions anymore. With load limits like 80kg you need to get it right. I will see about some pics later........ gotta fly!

 

[cal3thousand]

got about 10 seconds....

I'm surprised by sheldons idea of horizontal. To me, it is parallel with the ground.

It's all about the axle. You take one, skim it's sides then drill or channel it, and call it a motor shaft. It's not as strong in all directions anymore. With load limits like 80kg you need to get it right. I will see about some pics later........ gotta fly!

OK, I get it, so the orientation of the axle places stresses in a direction where it has been already compromised due to the removal of material for the flats when in a horizontal dropout.

 

[alsmith]

Surely the type of dropout has more to do with the original drive system that was fitted to the bike than anything else?

 

[neptronix]

I completely agree with you, but horizontal dropouts are designed for single speed compatibility without a tensioner, and are easily made safe with a torque plate.

I have 2 amateur-built frames like this, but it's never been a prob for me, since i don't go light on the torque plates at all ...

 

[friendly1uk]

Phew, I don't need to load mspaint lol

I don't see how torque plates can help? They would hold the end of the shaft, beyond the failure point which is on the inner edge of the dropout. Torque plates would protect the shaft from external influences, like a car hitting you from the side. I don't see them helping when you land after some air. In fact, upon landing you may wish the shaft to flex a small amount, pivoting about the inner edge of the dropout. Ultimately loosening your nuts. But if you stop the flex, the energy goes elsewhere. It is more likely to shear off than flex and return.

Single speed chain slack adjustment is a good call, but the angled dropouts called horizontal on sheldons site would do. They allow rearward movement without turning the shaft so far it is at it's maximum disadvantage.


Nothing I say here is fact, I'm just trying to knock the ball back n forth in my usual argumentative fashion. I would love to know more about when these things do fail though. I have vertical dropouts and won't drop off a curb without almost stopping and rolling down. This is because I can see it is poorly made and carries this 80kg load limit. I wouldn't dream of giving someone a croggy. I do see peeps take such motors and talk about giving them a beating, but that is not really data. I would feel a lot happier with rear suspension, which would make a huge difference to the shock loading. It is calculable, but phew... My head hurts just thinking about the math. Be nice to see more examples of what will and won't work.

Edit: I should find some failure posts to examine.

 

[Punx0r]

Just a thought: If you have an M12x1.5 thread on your axle then the minor diameter is ~10mm. So it's just as weak in both directions

The axle may then even be weaker if vertical the flexing will be at a thread root, which will act as a stress riser and promote cracking. When horizontal you just have the influence of the thread root at each edge. I'm sure it would still crack, but may take longer.

I believe some larger motors have an M14 thread which will give you a minor diameter of ~12mm.

 

[11spokes]

I agree

 

[dogman dan]

I think more motors have had their axles twisted off, than broken off.

Landing huge drops with a hubmotor is not what the motors were designed for. I think if you huck big air, you got the wrong kind of ebike.

 

[kdog]

Gotta love this forum...
I'm helping a mate with his build. He wants to use a SS frame and I was considering this exact issue late last night with some new build inspired insomnia...
This build will be a commuter ( no jumps etc) and a Mac motor)
F1-Other than your concern( which is understandable), are there any known failures attributed to this?
Cheers
Kdog

 

[mountain biker]

what about mountain bikes with front shocks from Walmart.26" Hyper Havoc Full Suspension Men's Mountain Bike, Black. does this effect the torque plates.? for a front drop out.?

 

[John in CR]

Friendly 1,

I thought about it the first time you brought it up and dismissed it as an insignificant difference. That's because the 80kg is nothing compared to forces from the torque of the motor at such a small radius. Even a puny motor can have 100Nm of torque, and 100Nm at a radius of 5mm is 20,000 newtons of force or 2039kg.

All the broken axles I've seen on the forum with good enough closeup pics clearly showed the failures a being from a twisting force not a downward force. The only exception was someone who drilled their axle out so much while fitting larger phase wires that the axle looked like thin wall tubing where it broke on the wire side.

While you were so worried about putting shoes on the wrong feet, you didn't realize it was some little kids shoes you were trying to put on. 8) Don't feel too bad about though, there are guys doing time consuming physical mods to their motors to make them cool better, when proper controller tuning to match their motor, voltage, and load would make it so much of that heat they're trying to dissipate wouldn't be created to begin with.

John

 

[friendly1uk]

Thing is John, big numbers aside, if it busts under motor power alone then it was never any good anyway. I'm interested in the builds where motor power is 90% of the way towards axle failure, and then you want to hit a pothole that will chuck you in the air. A load that is much harder to calculate, but it's your numbers plus mine that should be busting things. If it's motor power alone doing the snapping, your doing something else wrong. Something for a different thread.

You know patting a grown man on the back with comments like 'don't feel too bad' is never likely to be taken well on an internet forum. It reads more like you have issues. Obviously it was your mistake not mine so I'm not too bothered. Still... I'm glad I found this thread again.


punx0r your maths had me questioning myself for a moment lol. The flat sides are machined lower than the base of the threads. I have never actually measured any of it, but a quick look says more.

As you see, the threads are cut away entirely, and then a bit more. I'm having a change of heart though having studied the photo a while. If the flats were only as deep as the base of the threads as in your mathematical approach, it could be weaker. This flat area of perhaps 5mm wide wouldn't exist. The flat area would be very narrow and form a peak where stress would be concentrated. Skimming below the thread base has removed a high stress area.

I'm no longer convinced that turning the axle would be like night and day. However stress analysis is not in my job description. My ears are still wide open for debate. I think they perhaps have it about right though.

 

[John in CR]

:? You start the thread saying everyone is missing your point that is as obvious as putting shoes on the wrong feet, yet you have a problem getting some of that attitude back, especially when it's YOU who has missed the obvious...and continue to do so. ie Proper axles capable of handling the forces of the motor are strong enough to mount them with the flats in any orientation. Sure it's obvious which direction is stronger relative to the vertical loads, but it's irrelevant. Hell, even the shape of the cut for the transition from round to flat is drastically more important than the orientation of the dropouts.

 

[amberwolf]

Since AFAICT the actual problem the OP has with the dropouts isn't the dropouts, but the orientation of the axle flat being parallel with the ground being a potential failure point due to bending loads there, I'd like to reference here an axle in a used *front* kit I got, which is very noticeably bent.

http://www.endless-sphere.com/forums/viewtopic.php?f=4&t=64279

However, since it was in a front dropout, it was *not* oriented horizontally, but rather nearly vertically. Unfortunately, as I never saw it on the bike itself, I don't know if the bend is forward or backward facing vs how it was mounted on the fork.



The bike itself was not crashed or damaged AFAICT from the single pic I saw of it with the motor mounted before the seller took it off, and the seller (whom I spoke with for a while when I picked it up) didn't have any damage on the bike to repair from any crash that could've bent this axle.

The motor is an old Crystalyte X5304 front 26", still in it's original spokes and rim AFAICT, and there is no significant damage to the rim either (which one would expect to result from something that bent the axle this badly).

So while I don't know *what* bent this axle, it most definitely *is* bent and in the direction of the axle flats, as opposed to the threaded area, so it does follow (probably) that the axle *is* weaker in this direction. (it is worth noting that I've been told the old X5xxx series stuff have weak axles in general)


However, there are no (visible) cracks in the axle at the interface between uncut section and flatted section, as of yet.


Even so, it is likely that when I use this motor (probably on CrazyBike2's rear) it will have it's axle replaced with as large a diameter hollow pipe axle as I can fit in it, clamped and pinned to horizontal "dropouts" made just for it. I haven't found it yet, but Farfle made a hollow axle for something some time ago, and that's what this is to be based on.

 

[dogman dan]

The real problem is quite clear, and I've been saying it for years. The design of the axle of the typical hub motor sucks. Along with the alloy and temper of the steel used.

The wire comes out thorough the axle, whether through the center, or through a notch on the side. It does weaken it for sure, and usually helps with the twist off failures. As John points out, those forces get very high in a higher powered motor.

What we need is hub motors with much larger diameter bearings, allowing large space for wires, cooling fluids, whatever, to enter the hub as well as a much larger flat area for the torque plate or arm. The axle is then solid, round, and 10mm, made from a decent alloy. If you really wanted it, it could even be QR, with a wing nut on the torque arm bolt.

While the current design is more or less ok for a 500w commuters bike, we need much better design when the bike will ride off road, with 1500w or more of power, and be bolted to a conventional frame dropout.

A direct drive motor could easily be designed to stand the most wicked hucks. Likely a much stronger motor cover would be needed, but that's just a matter of choosing a better alloy. Not expected to be cheap, just as a good DH fork is never cheap. This would be a motor for the elite rider, not the average joe on a budget.

In the end though, a more likely solution we will see is higher powered mid drives, combined with much stronger chains and cassettes, or IGH hubs.