Friction Drive Mountain Bike

[motas]

Hey all,

Thinking a friction drive might be the ticket for my rigid bikepacking/commuting mountain bike.

Has anyone done any testing on a friction drive with mountain bike tread?

I'll be using a fast rolling tread up to a 2.6" width which should help increase friction and reduce slippage. I'm also thinking a larger roller with a nice coarse knurling will help.

Cheers.

 

[amberwolf]

THere's some info in one or more of these
https://endless-sphere.com/forums/search.php?keywords=tire*+friction&terms=all&author=&sc=1&sf=titleonly&sr=topics&sk=t&sd=d&st=0&ch=300&t=0&submit=Search

and there's other info in some of these, but most of them are not relevant:
https://endless-sphere.com/forums/search.php?keywords=friction+drive+tire*&terms=all&author=&sc=1&sf=msgonly&sr=posts&sk=t&sd=d&st=0&ch=300&t=0&submit=Search

This is a narrowed down verison of that with 'mountain":
https://endless-sphere.com/forums/search.php?st=0&sk=t&sd=d&sr=posts&keywords=friction+drive+tire%2A&sf=msgonly

and another for knobby
https://endless-sphere.com/forums/search.php?st=0&sk=t&sd=d&sr=posts&keywords=friction+drive+tire%2A&sf=msgonly

 

[motas]

Thanks Amberwolf,

I've already been searching everywhere for threads on the topic but haven't dug anything up yet. I've also been reading through a couple of the LONG build threads on friction drives but the closest I've found as far as tyres go is random cheap mountain bike tyres.

Was hoping there might be someone experimenting with new ideas for friction drives, or at least someone else who's thought about it.

I'm thinking a large diameter and wide roller to suit the wide tyres will increase the friction, and spread the wear on both the tyre and the roller. With the larger diameter the RPM will need to be reduced with a small belt drive to the motor. This makes gearing and motor choice pretty flexible as well.

 

[amberwolf]

Tire diameter has nothing to do with the speed you go with a friction drive.

It's entirely up to the motor speed and the *roller* diameter.

Seems counterintuitive, but it doesn't work like "gearing" does, even though you would expect it to, because effectively the roller is directly rolling on the road surface itself, as far as "gearing" ratios go.

The problem with any knobby is that it's not continuous surface. Depending on design of tire and roller, and roller diameter, the roller may begin to slip as it leaves the surface of one knob set and goes to the next. If the knobs are not gridded, but instead overlap, that's less of an issue.

One option is hybrid tires like the CST City and the similar Kenda Kross, which have a center slick area but very knobby edges. Another, that Kepler used, was the Schwalbe Hurricane.

Depends on the terrain you must cross and whether you really *need* knobbies to do it.

I wish more people would post their failures when building things, instead of just the final success, because it would help show what doesn't work.

Most likely you'll have to do the testing of different tires and rollers, if there's nothing in the existing threads that helps. (there is a LOT there, but it requires quite a lot of reading to find it).

 

[motas]

Tire diameter has nothing to do with the speed you go with a friction drive.

It's entirely up to the motor speed and the *roller* diameter.

Seems counterintuitive, but it doesn't work like "gearing" does, even though you would expect it to, because effectively the roller is directly rolling on the road surface itself, as far as "gearing" ratios go.

The problem with any knobby is that it's not continuous surface. Depending on design of tire and roller, and roller diameter, the roller may begin to slip as it leaves the surface of one knob set and goes to the next. If the knobs are not gridded, but instead overlap, that's less of an issue.

One option is hybrid tires like the CST City and the similar Kenda Kross, which have a center slick area but very knobby edges. Another, that Kepler used, was the Schwalbe Hurricane.

Depends on the terrain you must cross and whether you really *need* knobbies to do it.

I wish more people would post their failures when building things, instead of just the final success, because it would help show what doesn't work.

Most likely you'll have to do the testing of different tires and rollers, if there's nothing in the existing threads that helps. (there is a LOT there, but it requires quite a lot of reading to find it).

You are correct, but I was referring to a large diameter roller, not a large diameter tyre. I am an engineer (well, technically next week) but it did take me minute to get my head around the gearing. You have to divide the RPM by the wheel radius from roller to wheel, but then multiply it again from wheel to ground, which always equals a factor of 1.

The large diameter roller means there is a bigger contact patch, and better able to reach the next bit of tread. But direct driven by the motor would not have enough torque, so I will need gearing between the motor and the roller. A large diameter and volume tyre also makes it a bigger contact patch, and they're typically run at lower pressures <30PSI which makes the contact patch even bigger again.

I'm thinking a hard pack tyre something like this will give me good dirt traction without too much knobiness:

It will mostly be riding dry dirt and some rockiness. I doubt I'll ever ride in mud.

I guess I'll just have to try it and see how I go!

 

[amberwolf]

Probably the best way.

Mud isn't conducive to friction drives, anyway; if the bite of the roller is deep/hard enough it might be able to push past chunks in the tread, but if the mud has a lot of clay it's still going to make the roller slip, and fill up both the knurls and the knobs pretty quick.


The only friction drive I built myself was using skate wheels as rollers; they worked well enough under the dry conditions we usually have here, but sucked when wet or dirty, etc. I used the Kenda Kross at the time, tried others but nothing worked as well as the smooth center strip those had.

 

[motas]

Well, the friction drive only needs to have more friction than the wheel has with the ground. On wet clay this isn't a very high benchmark. Either way I won't be going anywhere fast though, so lucky I don't like riding in rain or mud!

I'm also interested to know, is there any consensus about whether a rotating mount or a sliding mount is better? A rotating mount is neater and easier, but from what I've read the sliding mounts work very smoothly. I was told once the best way to make a sliding mechanism, is to fake it with rotation :lol:

 

[amberwolf]

So far, the stuff Kepler has done has been very successful, so if I were to build another friction drive, I'd do it like one of his versions. AFAICR they are all rotation types.

(I have one of his older ones, probably v1, that's going to go on a road bike as a backup bike at some point, but I can't use a bike like that for most of what I need to do).

My first used a cable-pull from a friction shifter to engage the drive. The second left the drive always engaged.

 

[motas]

The frame it's going on has handy threaded pannier mounts right where the wheel is on the seat stay. I'm thinking a custom seat tube clamp which ties into the pannier mounts as well will give a very solid pivot point for the friction drive.

Another advantage of gearing between the motor and roller is I can easily add a one way bearing or freewheel and leave it always engaged with almost zero drag so don't need any form of disengagement.

 

[motas]

Here's a concept of what I am thinking.



It's an aluminium frame which replaces the seat tube clamp, and bolts onto the pannier mounting points for additional support. The 10mm pivot is mounted on bearings so it can move freely. A spring will most likely be required to make initial contact with the tyre, although some people say the inertia on startup is enough.

Motor is a 1200kW Turnigy brushless: https://hobbyking.com/en_us/turnigy-aerodrive-sk3-4250-350kv-brushless-outrunner-motor.html

Roller is a hollow steel cylinder, with a 50mm diameter. It has a slight curve to it which increases the surface area with the tyre, but does create a little additional friction as the speeds vary slightly. It has a coarsely knurled finish for better traction. It is mounted to two clutch bearings on an aluminium driveshaft so it can spin freely when the motor is off.

Power is transferred between the motor and roller through a timing belt. With a ratio of 1.5:1 it has a top speed of 40km/h on 6S. This can be adjusted with different pulleys for more torque at a lower speed. For example for commuting it may be geared to a higher speed, but for off road bike packing it can be geared down to handle a loaded uphill.

 

[SlowCo]

The amount of power you'll be able to usefully transfer to the tire and thus to the ground will be limited with a friction drive in your situation.
I don't know if you've already seen this thread ( https://endless-sphere.com/forums/viewtopic.php?f=28&t=96769 ) but it seems almost the same amount of engineering work and components for a better result. It provides decent assist through all the gears without being dependent on weather and grip on/from the tire. Just a thought and (better?) alternative to a friction drive on a knobbly MTB bike/tire.

 

[motas]

I did look into a mid drive. I really like the idea of it, but I haven't found a good way to keep it compact and neat, and integrate with decent components (eg GX 12 speed). The issue always come down to the gear reduction needed to match the pedaling cadence.

The best option is to cut the frame and weld in a mount for a proper mid mount, but I don't want to do that on this bike as there are times I want it to be a regular bike. And at that stage I'd probably just build a whole new frame.

I don't mind the idea of a high pivot style chain line around a 3rd sprocket, but then the gear reduction has to be done in a single stage in limited space. I could mount this off the bottle mount on the seat tube and it would be removable with a bit of work. There are potential issues with lack of chain wrap. Might be an option to look into though.

The good thing with a friction drive is it sits up out of the way, uses existing mounts, and solves the massive gearing reduction by using the wheel. There's examples of 1000W+ friction drives with much smaller tyres and rollers, and people who have built them seem to think they are a suprisingly good option.

 

[SlowCo]

I understand the attraction of a "simple" friction drive on a road bike. With Keplers race bike as the best executed build I've seen so far:

[youtube]Miy1V0YjwBQ[/youtube]

I think his minimalist design with the RC motor directly driving the tire with very clever geometry to engage is on another level. Less is more :thumb:

And he even did a full suspension MTB version: https://endless-sphere.com/forums/viewtopic.php?f=28&t=96581


So go for it and update this thread with your build. I'll be following your progress.

 

[Grantmac]

A ways out and will depend on frame design but worth a look:
https://endless-sphere.com/forums/viewtopic.php?f=28&t=103429

 

[motas]

A ways out and will depend on frame design but worth a look:
https://endless-sphere.com/forums/viewtopic.php?f=28&t=103429

If I end up going mid drive, that's very similar to what I'd do, but within the main triangle rather than in front of the BB. I understand they've done that to maintain chain line on full suspension, but for a hard tail it makes not much difference.

The issue with this is to get a maximum 100 RPM cadence with a 32 tooth sprocket. With a 9 tooth output of the mid drive, you need ~15:1 reduction off the motor, which is tough to package compactly, especially with limited width.

I understand the attraction of a "simple" friction drive on a road bike. With Keplers race bike as the best executed build I've seen so far:

[youtube]Miy1V0YjwBQ[/youtube]

I think his minimalist design with the RC motor directly driving the tire with very clever geometry to engage is on another level. Less is more :thumb:

And he even did a full suspension MTB version: https://endless-sphere.com/forums/viewtopic.php?f=28&t=96581


So go for it and update this thread with your build. I'll be following your progress.

I really like his as well. I wish I could fit a motor just above BB height but I don't think I can so next best spot is on the seat stays, which at least gets it up out of the way.

The main change I want to make is a seperate roller for more control over friction and gearing, as well as a bit more toughness. The other way would be to custom make a motor housing as a roller. It'd definitely be doable if you could find the right motor to strip the stator from, or even the right laminations to wind yourself and would be a very neat option.

 

[amberwolf]

Most of the RC motors I've seen disassembled here on ES have bells that you could replace with a kurled roller.

 

[motas]

Most of the RC motors I've seen disassembled here on ES have bells that you could replace with a kurled roller.

That could definitely be a good option, certainly reduces the weight and complexity of the system. The two downsides are a lack of gearing, which means kV must be kept to an absolute minimum for the larger roller, and lack of one way bearings so the motor will drag with the wheel. It would be simple enough to flip the motor off the wheel if the battery runs flat though, and I'm not sure how bad drag is anyway.

The main thing I'd want to check is whether I can put an extended driveshaft in so the roller/bell can be supported by bearings and the frame on both sides. You wouldn't happen to have any pictures handy of a disassembled motor would you? I'll see if I can find some somewhere.

 

[motas]

Found one. It would be ideal to have the drum supported on both ends, with the wires exiting through a hollow shaft. However this is probably impractical in such a small motor. It would be possible to have a bearing large enough to fit over the roller itself, but weight would be an issue. Two small bearings that the roller rolls on could be the best option. Otherwise just need to make sure the roller is strong enough to only be supported at one end which it probably is.

It looks simple enough to extend the shaft so at least the shaft can be supported on both ends though, so that's good news.

 

[amberwolf]

Dual-ended support of the bell has been done on some of the friction drive motors, and some of the other RC-motor-driven middrives. I don't recall a hollow axle, but there have been several that use a large ID bearing on one end of the shaft, with a spacer between it's ID and the shaft, and then holes in that spacer to pass the phase/etc wires thru.

I don't have any links to specific instances of these, but I know they're around here somewhere.

It would be simple enough to flip the motor off the wheel if the battery runs flat though, and I'm not sure how bad drag is anyway.

Drag may be worse the higher the votlage the motor generates vs the battery voltage vs the controller's auto-rectifying ability, so the faster you go above the motor's "natural speed" at that particular voltage, the more drag there will be.

If you're not above that speed, then the drag is going to be mostly the magnet pull against the stator iron (simplification), though any currents generated within the windings may also create drag (see above). Currents generated within any part of the spinning motor may sap energy from the rotation, and that's the drag you feel.



ElectricGod, Thud, AussieJester, probably SpinningMagnets, and a number of others I can't recall ATM have posted threads / pics of various RC motors, including upgrading / rewinding / etc. I coudln't make a search that finds any significant number of these, though.


Oh, and LightningRod's Big Block / etc stuff may also have info you can use, as well as another big middrive development thread whose name escapes me.

 

[motas]

Ah the spacer is an excellent idea, thanks!

The same bearing could be used on both ends of the roller as the primary mount to the chassis, with a small mount inside for rotation forces only. That would be very robust. Weight penalty would depend on how narrow of a bearing is available.

I need to do some research on what motors would be suitable though. It would need an exceptionally low kV.

 

[Hillhater]

.

The main thing I'd want to check is whether I can put an extended driveshaft in so the roller/bell can be supported by bearings and the frame on both sides. You wouldn't happen to have any pictures handy of a disassembled motor would you? I'll see if I can find some somewhere.

If you dig a little deeper into outrunner motors, you will find that some are equipped with not only extended shafts, shaft bearings at both ends, and a “skirt” bearings on the open end of the outer can (bell)
https://hobbyking.com/en_us/hxt-63-74-200kv-brushless-outrunner-eq-5240.html?___store=en_us

 

[motas]

.

The main thing I'd want to check is whether I can put an extended driveshaft in so the roller/bell can be supported by bearings and the frame on both sides. You wouldn't happen to have any pictures handy of a disassembled motor would you? I'll see if I can find some somewhere.

If you dig a little deeper into outrunner motors, you will find that some are equipped with not only extended shafts, shaft bearings at both ends, and a “skirt” bearings on the open end of the outer can (bell)
https://hobbyking.com/en_us/hxt-63-74-200kv-brushless-outrunner-eq-5240.html?___store=en_us

Looks like it's discontinued :/

This is the best one I've found so far but haven't done a lot of research yet:

https://hobbyking.com/en_us/turnigy-sk8-6364-110kv-sensored-brushless-motor-14p.html

That would give me around a 60km/h top speed which is a bit excessive, but it is rated to 1750W so probably no big deal to be running it at lower RPM. I was thinking somewhere around the 1kW mark would probably be decent enough.

You wouldn't happen to have any pictures of a skirt bearing would you? Would be interested to see how that works inside the motor.

Cheers.

 

[Hillhater]

This is typical of the skirt bearing as used in those Turnigy motors.
Notr the bearing in the alloy holder around the stator end hub
Im sure some similar motors are still available in various sizes.

 

[amberwolf]

You wouldn't happen to have any pictures of a skirt bearing would you? Would be interested to see how that works inside the motor.

Some of these posts have some pics of skirt bearings in various motors, if it helps:
https://endless-sphere.com/forums/search.php?keywords=%22skirt+bearing%22&terms=all&author=&sc=1&sf=all&sr=posts&sk=t&sd=d&st=0&ch=-1&t=0&submit=Search

 

[motas]

Thanks guys those are very helpful!

Definitely plenty of slim bearings available with a large ID & OD.

After going down the mid drive rabbit hole (again) my thinking has split in two, I am thinking a friction drive will be a smaller & cheaper project, and if it fails I may go down the mid drive path after.

For a friction drive I am thinking I will steal the stator from the best fitting motor I can find. I'll completely ditch the shaft and bell, and bolt the stator directly to one side of the friction drives frame. I'll then use large diameter skirt bearings to support each side of a roller, with the magnets from the bell attached inside so it is also the rotor. The advantages are obviously that it will be very robust, simple and compact. The disadvantage is if the motor dies it needs to be replaced with another of the exact same diameter, or a new roller is required to suit a new motor diameter.

My mid drive concept is to use a cylcoidal reduction unit mounted to the front of the seat post, with a small freewheeling cog attached to the existing chain line. An additional roller may be needed to increase chain wrap around both the motor cog and the cranks. This assembly could also include a battery and controller for a neat little package. I haven't seen many builds using a single chain line, which is either a gearing problem or there's something I'm missing.