Best rear torque arm design/layout?

[BAZ]

To beef up the dropouts on my ally frame, I've purchased a few off-the-shelf torque arms from ebay, but unfortunately the dropouts in my frame are shouldered in an unusual way and so all the ones i've tried (eyelet/jubilee clip) fail to line up in a good way.

So i've gone the homebrew torque arm route, I've got a bit of a background in 3D printing so have been designing (and test printing) something that can hopefully be laser cut out of steel plate in the next few weeks.

Unfortunately my frame has a lack of mounting holes leaving only the caliper IS points on one side and nothing on the geared side. I will most likely have to secure the arms to the frame through jubilee clips/hose clamps.

I've had a browse through the torque arm picture thread. I'm limited for tools and equipment and so the more complex axle clamping designs are out of my skillset. It seems some people set their torque arms to push the chainstay under acceleration and others mount them up against the seatstay with a clamp. With the seatstay option, the force seems to be transferred along the clamp to the frame, rather than directly like with the chainstay. This is then all reversed under regen braking, which I won't be running.

The issue I have is this, which torque plate setup is better? I should say I have a 1500w hub motor.

It is my understanding that solid mating with the axle is paramount, no point having a torque arm if it's not contacting the axle well.

With a chainstay torque arm, ensuring the axle is in full contact and tight against the torque arm will have to be done with rubber or metal shims between the torque arm and the chainstay. The accelerating force will then be transferred through the shims into the frame.

With the seatstay torque arm, using a clamp lets you pretension the arm to the axle, providing good axle contact. However this also means all the accelerating force will then be transferred through the clamp into the frame and they tend to be very thin metal.

I'm not sure which option is best.

So I've currently designed two arm options, one for each method i've spoken about. Is there a better solution? Are my designs way off?

Chainstay:

Seatstay:


p.s If any you know places in the UK that could laser/water cut my designs out of 5mm stainless let me know!

 

[teklektik]

 

[SniperGaulois]

If you want to save time on machining metal, and you want a very tight fit to your custom frame, then you can use carbon fiber. Re-cycle one or two of the torque arms, and just mould the axial end into a carbon fiber blob that you wrap to the piece of the frame that you want.

Here's an idea for your bike:

1/ I would keep clear of the derailler and the disk brake bolts, I would instead use the yellow bar which is going like this / from the axle...

2/ I would carbon fiber some female bolts onto the / yellow bar aluminium, 1-2 of them, so you have an attachment point on there to fasten the torque arm. Job time: 30 minutes, drying time, 1 day. Others use hose locks, and there are discussions about if they hold well, probably they are the easiest way to go.

3/ I would wax and pvc tape all the surfaces and the female bolts and make a torque arm which fits just fine and which you can take on or off with a couple of bolts.

The carbon fiber is just fiber soaked in epoxy, bonded to sanded aluminium. it's somewhere in between wet string and plastascene.



I love my torque arm which i made from standard hardened carbon steel. It's 5-6mm thick and I can take the back wheel on and off without taking off the torque arm because it's a long slot. The axis slot is long and I can get the wheel out when the bike is upside down, but it's well fixed when it's riding.

 

[Tolkano]

What program did you use to design the torque arms?

 

[BAZ]

What program did you use to design the torque arms?

Fusion 360, free yearly license to CAD enthusiasts.

If you want to save time on machining metal, and you want a very tight fit to your custom frame, then you can use carbon fiber. Re-cycle one or two of the torque arms, and just mould the axial end into a carbon fiber blob that you wrap to the piece of the frame that you want.

This is really interesting, i've never used carbon fibre as a material before. How easy is it to work with?

I'm currently modifying my design to match teklektik's suggestion, although still I think it requires a much more accurate modelling to ensure tension and mating on the axle vs a jubilee/clamp which you can tension after fitting correctly.

 

[flat tire]

The original design looks great.

Carbon torque arms might not be a great idea. Especially not made out of a mass of loose fiber.

Furthermore you shouldn't bond carbon directly to aluminum as stated above or you'll get galvanic corrosion. You need to use an intermediary like flox or fiberglass.

 

[wturber]

From what I've read of Justin's testing, the main thing you want to do is used thick enough metal. Use 1/4" steel and you'll be fine. I used 1/4" mild steel for both convenience (can get it at any hardware store) and because I'd rather not have a harder steel possibly deform the threads on my motor axle. While it is nice to have a nice tight fit around the axle - especially if you are using regen braking - using a thick enough metal is probably more important since it not only provides more material to resist plastic deformation, but also spreads the load over a wider section of the axle.

I like Teklektik's modification to your brake side torque arm. Your cluster side torque arm design looks fine to me.

 

[DanGT86]

The brake flange is the best possible way to do it. That flange is capable of handling way more torque than your motor would ever output and connecting the axle to the brake flange would provide additional rigidity to the frame in general.

If you use regen at all then you want the tightest possible fit between the axle and the slot in the torque arm. Drive torque and regen are opposing directions so they have a rocking effect that will try to work the nut loose. For this reason I think the best option is a split clamping torque arm where a bolt squeezes the 2 pieces of metal around the axle. That is the very opposite of slop. The next best thing is the tightest possible fitment and a nordlock washer set under the nut. All that being said I have never actually seen a nut wiggle loose with off the shelf torque arms from ebikes.ca. I'm just lucky I guess. its pretty well documented.

If you are jobbing it out to a shop to laser or waterjet I would actually consider having them leave the distance between the flats about 0.1mm small and file it yourself for the tightest possible slip fit. Don't bother with stainless unless you really need the anti corrosion properties. It's not as strong as regular steel and terrible to work with if you want to drill or cut it later.

Nice work on the printed arms by the way. What an awesome way to prototype.

 

[LockH]

DO regret very much anybuddy watt ignores history... including "torque arms":
http://www.ebikes.ca/product-info/torque-arms.html

 

[John in CR]

If you use regen braking, then torque arms that clamp onto the axle are the only acceptable type using anything above very modest power. All non-clamping types will eventually deform or worse, deform the axle, resulting in a loose fit and rocking motion of the axle in the dropout. Any exceptions require frequent inspection of axle nut tightness. I prefer never having to worry about such inspection, and just put air in the tires and remember to charge the battery.

 

[markz]

You asked for the best
Of course mine are

 

[999zip999]

? Markz get out your 4in amgle grinder a plate of steel cut the axle slot tight and weld two nuts on then drill one . Insert bolt and Titan.

 

[BAZ]

Quick update to this. I've revised the caliper side arm to better match teklektik's recommendation. It's taken a few adjustments to get the hole to be a snug fit and align right but I think i'm pretty much there.

Good idea on undercutting the axle and filing it down to be tight. I've bought a small file set for this.



Just need to make the curve a bit less fat and then I can get it cut!

I'm still trying to find a lasercutter/watercutter in the UK that's willing to do such a small order/one offs. A few have got back to me but they have a minimum order of £150 or so

 

[teklektik]

Just need to make the curve a bit less fat and then I can get it cut!

Perhaps something similar to this using simpler geometries and keeping the lines parallel to the existing frame where possible for cosmetics.

I'm still trying to find a lasercutter/watercutter in the UK that's willing to do such a small order/one offs.

Big Blue Saw is popular in the US - they might be worth a look... (I've never used them myself, but...)

 

[BAZ]

Another update, got them cut out of 5mm steel for £15 a piece. Will need a bit of filing as I undersized the holes but looks great!

 

[teklektik]

:thumb:

For other folks in a similar situation - who did those up for you?

 

[SniperGaulois]

Good information about Galvanic Corrosion and carbon fiber epoxy on aluminium. That's not a factor if you use a lock and key system to brace the aluminium. It is only a worry if you carbon fiber onto a flat surface like a bond, only as a glue. If you wrap CF around aluminium, it's set in stone.

For you guys who have awesome bikes that you ride until the frame fatigues, it's very useful. Just wrap CF around your old bike and you can recycle very expensive frames with a restricted stress zone or a bad weld. Because, if you braze aluminium welds, you have to temper the entire frame at 800 degrees for a day and at 400 for another day. That's how aluminium frames are prepared. Carbon fiber is a low temperature fix for aluminium frames.

When we can 3D print carbon fiber tools from the web, it can be the fastest way to customize mismatched torque arms. If you 3D print a mould and stuff it denseley with the epoxy, it can be fun too!

This is a 2 hour bike fix which is way over the top, it you could lift a car on it. https://endless-sphere.com/forums/download/file.php?id=206072&t=1

 

[onemorejoltwarden]

Torque arm fit /design question:
Cheapskate that I am I got some V_1 style NBPower torque arms from AMZ ($15 freeship)
They are not very snug on the 10mm axle flats. ~0.5mm slop
Do the Grin/EbikeCa ones fit better,snugger, as a rule?
Are generic copies of the early Grin TA's problematic?
I could put a small shim next to the flat, but that would be tacky,
or get a friend with a welder to put some slag in the hole and file it to fit.

"250 watt" 20"front Dahon with 11mm steel dropout slot, sensorless non regen
unk small controller that I doubt exceeds 20A
Thanks

 

[Tolkano]

What's the full process for designing something like this?

You sketch it out in cad then print it out on paper then keep printing on paper until its close, then plastic then finally water jet cut in steel?

Hardly anyone is doing custom bike specific torque arms like this and there's not many resources around that explains a noob like myself how to actually do it

 

[bww129]

I don't want to hijack this thread, but since this is the same subject it makes sense to post here instead of starting a new thread. I've fitted a Crystalyle H3525 to an aluminum frame bike along with two Grin V4 torque arms (two piece) and would like to eventually add regenerative braking capability but am not sure if they will be enough after reading the many threads about torque arms, regen, axle bolt tightening, etc. The controller is a 40A C7240-GR and the battery is a Luna Wolf V2 52V. Please see the linked pictures of the TA install from my build thread:
https://endless-sphere.com/forums/download/file.php?id=272058
https://endless-sphere.com/forums/download/file.php?id=272059

Both TAs are a pretty tight fit on the axle as they needed to be wiggled and tapped into place with a hammer. The left side TA (motor wire side) has both pieces almost at 90 degrees to one another, and it makes nice flat contact with the bottom tube (even though it angles in slightly toward the crank) and is secured with two hose clamps. This is ideal for taking off since it will put the load on the bike frame, but for slowing down it will try to pull the TA away from the frame putting the load on the hose clamps. Unfortunately this is the same situation on the drive side because the TA can't be mounted any other way. The drive side TA is also worse in that the angle between the pieces is closer to 135 degrees AND a spacer between the two TA pieces was required to position the TA better inline with the angled bottom tube.

Do you guys think this setup will last given the ~2kW power levels? Or is there any way to easily improve it? Fabricating a custom TA would obviously be ideal, but that might be out the question for the time being.

 

[motomech]

Or for those lacking machining skills (or just plain lazy), one can just buy the V4 T.A. from Grin.

https://www.ebikes.ca/shop/electric-bicycle-parts/torque-arms/torqarm-v4.html

Has worked on every application I have tried it on.

 

[amberwolf]

The arm posted about immediately prior to your post, by bww129, *is* that torque arm.

 

[bww129]

And not just one but two Grin V4 torque arms since they fit on both sides. After reading all the doom and gloom about regen working axle nuts loose unless you have a clamped axle setup it seemed worth the extra investment to help prevent issues.

Guess I'll just see how this setup holds up over time.

 

[Drunkskunk]

And not just one but two Grin V4 torque arms since they fit on both sides. After reading all the doom and gloom about regen working axle nuts loose unless you have a clamped axle setup it seemed worth the extra investment to help prevent issues.

Guess I'll just see how this setup holds up over time.

Having built many ebikes, and road 10s of thousands of miles on ebikes over the last 15 years or so, I can say your setup looks pretty solid.

If you want a more critical opinion, you could spread out those hose clamps to the first and third hole on both sides. it would be stronger that way for torque in both directions. That's just min/maxing what you already have, though.

If you have a potential failure point, it's on the hitch side. Well, it's the hitch. Putting a stack of different materials with different strengths, hardnesses, and thermal expansion rates under a single nut and then putting torque on each of them at different angles and rates is eventually going to cause a problem. Exactly the kind of problem torque arms were designed to catch. Carry a wrench.