Welding down a torque arm

RTLSHIP

10 kW
Joined
Jul 24, 2008
Messages
521
I'm wondering if anybody has tried spot welding down a torque arm to the axle of a front hub motor. I have a motor that still works but has some striped threads. There are a few concerns: 2 different kinds of steel and the high heat applied to axle could melt insulation.
Yet it makes more sense than replacing the axle. Of course, the torque arm, if ideally installed, would prevent spin outs
to a high probability. This seems like the most practical and cost effective way to go. The tire and spokes could still be replaced and wiring side of hub could still be accessed.
 
spinningmagnets said:
I've seen some bonded with 3M's DP-420
Is that an epoxy? True welding would work to the point of having to break the torque arm parts. I know that the stripped threading side can be re-threaded but the axle (on that side) will have less overall volume and a new, smaller bolt.
 
I have a steel bike frame but prefer to weld torque arm right on axle itself.
 
Why mess up the axle of the hub motor?

I just take an angle grinder to some 1.5-2" x 5" x 1/4" flat bar steel to try and get a snug fit on the axle, figure out which way the dropout slot is going, make the axle notch. Then where the hose clamps would go I notch it the width of hose clamp. I usually put one at the end (notch), and one closer to the axle with no notch to make a snug tight fit against the fork itself.

If I were to make it look nicer I would try thicker steel to make a threaded tap/dye hole for a clamping torque arm, then 1/4" steel up the fork to clamp, the transition of metal thicknesses would be a bolt. Entire contraption would be an L shape.

Or just buy the fancy torque arms from Grin/em3ev/bmsb/luna if your selling the bike and want to have a cleaner look.
 
markz said:
Why mess up the axle of the hub motor?

I just take an angle grinder to some 1.5-2" x 5" x 1/4" flat bar steel to try and get a snug fit on the axle, figure out which way the dropout slot is going, make the axle notch. Then where the hose clamps would go I notch it the width of hose clamp. I usually put one at the end (notch), and one closer to the axle with no notch to make a snug tight fit against the fork itself.

If I were to make it look nicer I would try thicker steel to make a threaded tap/dye hole for a clamping torque arm, then 1/4" steel up the fork to clamp, the transition of metal thicknesses would be a bolt. Entire contraption would be an L shape.

Or just buy the fancy torque arms from Grin/em3ev/bmsb/luna if your selling the bike and want to have a cleaner look.

The axle is already messed up on this hub. And the torque arm slot is a bit worn. I just want to know if it has ever been done and worked.
 
Since you don't want to weld on your bike, fabricate a nice thick .25" or thicker clamping type torque arm. Almost none of my builds have the axle nuts installed on the hubmotor axle. My clamping dropouts hold the motor wheel on running power from 11kw-32kw peak input with regen braking on every build. None ever get loose or require tightening, and the clamping bolt that closes the open end of the dropout makes it literally impossible for the wheel to come off during operation in any condition. Before I used only custom built frames I did the same with my torque arms.
 
John in CR said:
Since you don't want to weld on your bike, fabricate a nice thick .25" or thicker clamping type torque arm. Almost none of my builds have the axle nuts installed on the hubmotor axle. My clamping dropouts hold the motor wheel on running power from 11kw-32kw peak input with regen braking on every build. None ever get loose or require tightening, and the clamping bolt that closes the open end of the dropout makes it literally impossible for the wheel to come off during operation in any condition. Before I used only custom built frames I did the same with my torque arms.
I'm running 48 v on a 750 watt hub here. I don't need to hold down that much torque. I was also considering welding down the torque washers on both sides. I just have to make sure the hub covers still come off.torque 2.JPGtorque 1.JPG This 10 x 14 mm torque arm is more like 11 x 15 . That's from spinout damage. I vise griped the slot to make it compress back to original
 
RTLSHIP said:
I have a steel bike frame but prefer to weld torque arm right on axle itself.

If you have a steel frame, then why not just beef up your dropouts with 1/4" steel plates? No torque arm required.
 
E-HP said:
RTLSHIP said:
I have a steel bike frame but prefer to weld torque arm right on axle itself.

If you have a steel frame, then why not just beef up your dropouts with 1/4" steel plates? No torque arm required.
can you show me a pic of that? I do have steel dropouts and filed it to fit the axle, but once I rode it the axle actually seemed to
widen the dropouts. Either that or I unwittingly overfiled the dropouts. These dropouts are 1/4" (or 2/8) already. I was able to use a small vise to move it back where it should be. First, I have to fix the torn wiring which is not difficult. Next, a machinist is going to rethread the worn side and give me a smaller bolt. No worries yet as this is a back up or second kit. Part of the problem could be looseness on the worn axle side. When the axle gets loose it can bend/smear things easily.
 
RTLSHIP said:
... the axle actually seemed to widen the dropouts. Either that or I unwittingly overfiled the dropouts. These dropouts are 1/4" (or 2/8) already. I was able to use a small vise to move it back where it should be. First, I have to fix the torn wiring which is not difficult. Next, a machinist is going to rethread the worn side and give me a smaller bolt. No worries yet as this is a back up or second kit. Part of the problem could be looseness on the worn axle side. When the axle gets loose it can bend/smear things easily.

If the slot in the dropout was "widened" and then corrected by bending the sides back in place then the steel in that area has been softened and needs to be retempered (or it was too weak to begin with). That makes a very strong argument the dropouts need reinforcement or that torq arms are a necessity. If it is a steel frame then the optimal solution is to remove the existing dropouts, fabricate new dropouts from stronger, thicker material and weld them in place.
 
RTLSHIP said:
John in CR said:
Since you don't want to weld on your bike, fabricate a nice thick .25" or thicker clamping type torque arm. Almost none of my builds have the axle nuts installed on the hubmotor axle. My clamping dropouts hold the motor wheel on running power from 11kw-32kw peak input with regen braking on every build. None ever get loose or require tightening, and the clamping bolt that closes the open end of the dropout makes it literally impossible for the wheel to come off during operation in any condition. Before I used only custom built frames I did the same with my torque arms.
I'm running 48 v on a 750 watt hub here. I don't need to hold down that much torque. I was also considering welding down the torque washers on both sides. I just have to make sure the hub covers still come off.torque 2.JPGtorque 1.JPG This 10 x 14 mm torque arm is more like 11 x 15 . That's from spinout damage. I vise griped the slot to make it compress back to original

Those types of torque arms are terrible and always deform. If you use regen, which should always be a goal, then all thru hole torque arms start loose and only get worse, messing up the axle in the process. That's because only a press on fit can make a tight fit, but then you couldn't get it on or off.

You're talking about welding, so you can do metal work. Just make a good torque arm that is both thick and has a bolt for a clamp tight fit on even a damaged axle.

You don't want to weld on an axle, as you're just as likely to make it brittle at that high stress point, which could lead to a truly catastrophic failure.
 
John in CR said:
Those types of torque arms are terrible and always deform.

No name Chinese versions, yes. They have soft metal and loose fit, in my observation so far.

Grin TAs and carefully handmade ones seem to work just fine.

Just make a good torque arm that is both thick and has a bolt for a clamp tight fit on even a damaged axle.

I don't think ten percent of the hub motors I've seen have enough axle stud length for thick clamping torque arms. It's usually not an option.

Box end wrenches work great. Closed eye torque arms can work every bit as well, if they fit closely and are made of good material.

Really, the root problem is Chinese manufacturers persisting with a half-assed junky design (and folks buying it anyway). It's neither expensive nor difficult for a manufacturer to do it right-- like every coaster brake hub already does-- but it's difficult and expensive for end users to cope with the consequences of them doing it wrong.
 
Even cheap torque arms typically work fine as long as axle nuts are nice and tight. Once regen is introduced all through hole torque arms are proven inadequate in the long run unless you go with a ridiculously expensive solution like Nord Lock nuts that prevent axle movement more than the torque arms. It's the alternating forces at the axle during acceleration and braking that cause the problem due to the soft steel used for hubmotor axles, and thousands of pounds of force acting on them at such a small radius. By their very nature there is looseness in the fit of a through hole torque arm, and the alternating forces will rock the axle back and forth causing deformation of the axle and/or torque arm, usually noticeable with a clicking sound when alternating between acceleration and braking, along with loosening of the axle nuts.

We're talking about the mechanism by which the motor and wheel are attached to the bike, and once it gets loose and spins the axle can climb right out of the dropout. Those who treat that connection with as little regard as a regular bike wheel that has no spinning forces at the axle, are making a potentially fatal mistake and giving similar advice borders on negligence. Why do that when making a custom solution that clamps properly to the axle is so simple?
 
E-HP said:
RTLSHIP said:
can you show me a pic of that?

Can you post a pic of your frame's dropouts?
yes, but but it will take a day or so, as camera is at work and bike is at home. By the way the fork drop outs are 1/8" thickness not 1/4" I misspoke.
 
John in CR said:
Even cheap torque arms typically work fine as long as axle nuts are nice and tight. Once regen is introduced all through hole torque arms are proven inadequate in the long run unless you go with a ridiculously expensive solution like Nord Lock nuts that prevent axle movement more than the torque arms. It's the alternating forces at the axle during acceleration and braking that cause the problem due to the soft steel used for hubmotor axles, and thousands of pounds of force acting on them at such a small radius. By their very nature there is looseness in the fit of a through hole torque arm, and the alternating forces will rock the axle back and forth causing deformation of the axle and/or torque arm, usually noticeable with a clicking sound when alternating between acceleration and braking, along with loosening of the axle nuts.

We're talking about the mechanism by which the motor and wheel are attached to the bike, and once it gets loose and spins the axle can climb right out of the dropout. Those who treat that connection with as little regard as a regular bike wheel that has no spinning forces at the axle, are making a potentially fatal mistake and giving similar advice borders on negligence. Why do that when making a custom solution that clamps properly to the axle is so simple?
I've never been in danger as the forks were always steel and had a torque arm. A key issue is tightness of the nut. A crooked spring washer will help. So will retreading.
 
Chalo said:
John in CR said:
Those types of torque arms are terrible and always deform.

No name Chinese versions, yes. They have soft metal and loose fit, in my observation so far.

Grin TAs and carefully handmade ones seem to work just fine.

Just make a good torque arm that is both thick and has a bolt for a clamp tight fit on even a damaged axle.

I don't think ten percent of the hub motors I've seen have enough axle stud length for thick clamping torque arms. It's usually not an option.

Box end wrenches work great. Closed eye torque arms can work every bit as well, if they fit closely and are made of good material.

Really, the root problem is Chinese manufacturers persisting with a half-assed junky design (and folks buying it anyway). It's neither expensive nor difficult for a manufacturer to do it right-- like every coaster brake hub already does-- but it's difficult and expensive for end users to cope with the consequences of them doing it wrong.
good point. I agree there are a manufacturing design issues. But I have another bike that runs fine. It has a different controller and axle has not been damaged yet. Maybe the regen of some controllers, as John CR suggests, creates more axle movement in multiple ways. I even use a torque arm with filled in steel strapping and no luck.1 or 2 pcs of strap. troque arm.JPG
 
RTLSHIP said:
I even use a torque arm with filled in steel strapping and no luck.1 or 2 pcs of strap. troque arm.JPG

If you use shims, they should be placed on the flats, not the curved sections.

I use a shim with my Grin torque arm to make it a near press fit. I use a piece of razor blade broken off to be the same width as the flats. I slip the sharp blade end in between the arm and axle, after the arm is in place, and use a screw driver and hammer to pound it in place. The torque arm has no movement once the shim is in. The blade is hardened steel, so it's not going to deform like soft steel. Another idea is to used a old set of feeler gauges, like are used to set valve clearances on car engines. You can then choose the perfect thickness for a shim, to the closes 1/1000th of an inch. I take more time to remove the wheel, using a screw driver to slowly pry the TA off, but at least I know I have no axle movement.
 
RTLSHIP said:
E-HP said:
RTLSHIP said:
can you show me a pic of that?

Can you post a pic of your frame's dropouts?
yes, but but it will take a day or so, as camera is at work and bike is at home. By the way the fork drop outs are 1/8" thickness not 1/4" I misspoke.
That's pretty thin, but I'm more interested in how flat the area is.
 
E-HP said:
RTLSHIP said:
I even use a torque arm with filled in steel strapping and no luck.1 or 2 pcs of strap. troque arm.JPG

If you use shims, they should be placed on the flats, not the curved sections.

I use a shim with my Grin torque arm to make it a near press fit. I use a piece of razor blade broken off to be the same width as the flats. I slip the sharp blade end in between the arm and axle, after the arm is in place, and use a screw driver and hammer to pound it in place. The torque arm has no movement once the shim is in. The blade is hardened steel, so it's not going to deform like soft steel. Another idea is to used a old set of feeler gauges, like are used to set valve clearances on car engines. You can then choose the perfect thickness for a shim, to the closes 1/1000th of an inch. I take more time to remove the wheel, using a screw driver to slowly pry the TA off, but at least I know I have no axle movement.
I just put that onto give folks a photo idea. I do put it on both flats and curve. What I'm using is steel strapping that is used to band boxes down to a pallet. shim pic.JPG here is my ride to work wheel
 
RTLSHIP said:
E-HP said:
RTLSHIP said:
I even use a torque arm with filled in steel strapping and no luck.1 or 2 pcs of strap. troque arm.JPG

If you use shims, they should be placed on the flats, not the curved sections.

I use a shim with my Grin torque arm to make it a near press fit. I use a piece of razor blade broken off to be the same width as the flats. I slip the sharp blade end in between the arm and axle, after the arm is in place, and use a screw driver and hammer to pound it in place. The torque arm has no movement once the shim is in. The blade is hardened steel, so it's not going to deform like soft steel. Another idea is to used a old set of feeler gauges, like are used to set valve clearances on car engines. You can then choose the perfect thickness for a shim, to the closes 1/1000th of an inch. I take more time to remove the wheel, using a screw driver to slowly pry the TA off, but at least I know I have no axle movement.
I just put that onto give folks a photo idea. I do put it on both flats and curve. What I'm using is steel strapping that is used to band boxes down to a pallet. shim pic.JPG here is my ride to work wheel
EHP, you are referring to the razor blade backing unwraped? You make it conform?
 
It's oh so easy to fabricate a nice thick clamping torque arm for that. Then 3 smallish welds to secure it in place leaving it easily removed later to restore the fork to original form if desired, but it would still function like a clamping dropout so the wheel is easy on and off to change change tires.

I learned my lesson with quite similar forks and motor in the first meter of my first hubmotor ride. I painstakingly made dual through hole torque arms with too thin stainless. They fit snugly but the axle spun as the torque arms cut through it like butter and climbed out of the dropouts. The only thing that saved me from a face plant was that I was barely moving and had secured the torque arms well enough to prevent the wheel from completely separating from the bike, but now I can imagine all of the worst case scenarios possible and build all of my bikes accordingly.
 
John in CR said:
Even cheap torque arms typically work fine as long as axle nuts are nice and tight. Once regen is introduced all through hole torque arms are proven inadequate in the long run [...]

Seems like you've identified the easily preventable part of the problem.

You don't have to use regen, even if you can. If you use a geared hub motor with a one way clutch, you can't even if you want to. Both my hub motor bikes are direct drive, but I don't use regen because I don't want to wreck my bike.

Why do that when making a custom solution that clamps properly to the axle is so simple?

Here's the cable side axle end of the Jump bike Bafang hub motor that represents today's best value in a hub motor:
IMG_20210503_100035475~2.jpg

It's the shortest axle end I have here to take a picture of, but it's not uniquely short. There are lots of similarly short axle studs, especially on bikes with through-the-axle cables where part of the axle is slotted for cable relief.

Once you install a C-washer to allow cable clearance, and the fork tip, and maybe another C-washer outboard to clear lawyer lips, and enough threads of engagement in the axle nut, how much is left for a clamping torque arm? There isn't even enough for a cheesy Chinese torque arm.

You could find enough room for clamping fork tips by leaving off the axle nuts entirely, but then it's no longer a simple thing to do.
 
Back
Top