A 3D printed friction drive

[Gryphe]

Good day, everyone.

I have spent the past few weeks researching friction drives, mainly being attracted to the simplicity of the designs. I also happen to have a 3D printer. Add both together and you get a crazy idea turned into an actual, useful project.

Included with this post are screenshots of the actual drive. It is a fairly weak system using a 42 mm Turnigy SK3 350 kv motor and a 4S LiFePo4 battery but it gives me the assistance I need while still allowing me to pretend I'm within the legal EU limits. (No, I'm not.) That and I didn't want to invest a huge amount of money into something that might or might not work. 200 kilometers later the results are clear; it works.

The drive can be removed quite easily by using a handle which imitates the function of a lock nut, leaving just the clamp itself.

The material used was Colorfabb XT, a PET plastic hybrid which has so far proven to be capable of handling a sizable amount of abuse. It's tough, but not brittle.

Seeing as how the 3D printer community is all about sharing (but don't understand friction drives the way you folks do here) I'd like to offer the source files to those interested. Just tell me what 3D format you prefer and I'll look into it. I'm currently using Rhinoceros 3D for all my modeling but seeing as it's not extremely popular I'm not entirely sure what format is commonly used nowadays.

Last but not least, you'll have to excuse the quality of the last picture. My phone has a pretty awful camera and my workplace is a mess after printing and testing 10+ prototypes.

 

[nieles]

hi,

nice friction drive!

could you make a video of it engaging and disengaging?

 

[Gryphe]

You ask, you receive. You can clearly see the pivoting system doing its job here. Frankly, I'm quite proud of it.

Naturally it almost immediately disengages since the rear wheel is lifted from the ground.

[youtube]KyUUFpLcbuk[/youtube]

 

[nieles]

nice smooth motion, i like it a lot!!

is there a spring "helping" the motor engage?

 

[Gryphe]

No, just a pair of skateboard bearings which keep the mechanism working smoothly. The "tabs" on the arm restrict the movement.

I'm currently designing a fan for the motor after pushing the motor/ESC to its absolute limits.

 

[tahustvedt]

Seems to work really smoothly. Looks great. Is there a spring or just gravity holding the motor away from the tire?

I can't help to think the seat tube clamp be more secure if it was oriented upside down from now, since the motor now wants to pull up on the mount.

Anyway. Sweet design to work well from the first go.

 

[Gryphe]

Just take a look at the images; these should give you a general idea of the mechanism.

I divided the exploded view into three seperate groups; the locking handle, the pivoting arm and the clamp itself.

View attachment 1

Since the design's main focus was to get it working on my own bike I haven't looked into making the pivoting limits adjustable.

 

[Masure]

Very nice !

You talk about 10 prototypes, what kind of problem did you have with the first tries ? Would be nice if you could provide step files or anything compatible with CAD software.

 

[Dman12]

That is nice, I am working on my 3D printer these days, it can be a cool project to try once I'm done.
Can we get a video in action?
also can you share the STL files? Thanks

 

[Gryphe]

A video will be tough but I'll see what I can do.

I'll get those STL files posted as soon as I've tested the latest revision of the arm. (I was starting to notice some wear in a specific region so I'm printing an improved version at this very moment.)

As for those previous prototypes, well, first I had to find out just how much plastic I needed to keep things from falling apart. After I defeated that specific challenge I had to invent a reliable pivot arm.

 

[Gryphe]

Alright, with the tests complete I hereby release the STL files. You can download them here. These are production files so they might include a few additional meshes. (Mostly used to assist with the printing process.)

Feel free to do whatever you feel like with them. Just mention me if you start to release adjusted versions of the drive. (Not guessing to be likely, but still.)

This version includes the latest version of the arm, which now has an added hole for a cable tie (and tackling the aforementioned wear and tear issue).

 

[\/ampa]

How much time need the 3D printer to print the complete friction drive ?

 

[Gryphe]

About 5-6 hours of printing, using ~30 meters of 1.75mm filament.

 

[Dman12]

Thanks for the files. I might use them for some tests sometime in the future.
also lurking thread.

 

[Gryphe]

I'm currently awaiting a new shipment of plastic. (I've gone through a lot in the past few weeks.)

It should arrive tomorrow and will allow me to print and test a new variation of the drive, relying on a roller rather then the can of the motor. (I'm worried about motor wear.) The roller will be printed as well, though will most likely require covering of some kind. Maybe glue some sand paper on it?

Considering each new prototype costs me around ~3 Euro of plastic it's fun to try new ideas, learn from them and use the knowledge to design the next prototype.

Here, have a sneak peak of my current (and probably nowhere near final) prototype.

 

[darth_elevator]

This is a great thread. Great idea. And great results -- 200km and many more km I hope. And thanks for sharing the STL files. I'm going to see if I can get my friend with a 3D printer to make one for me.

 

[Gryphe]

Speaking of those 200, I passed the 300 earlier this day. (I commute to work by bike)

Seeing as how my new shipment of plastic failed to arrive today I put some of my leftovers together to check out the new prototype. So far I'm quite pleased with the results. The new pivot mechanism is extremely robust and should last for a very, very long time. The roller will of course require further testing. As I printed tonight's prototype with a low quality, but fast printing rate this version isn't ready for road testing. A proper version will follow as soon as my shipment arrives.

 

[darth_elevator]

....I passed the 300 earlier this day....

Even better!

 

[Gryphe]

Seeing as how my roller design needs more time at the moment I instead designed a can-based mount. Feeling fairly confident I rode the ~15 km to work this morning. It hasn't fallen apart yet, so that's always a good sign.

Main features include but are not limited to;

Using a hose clamp to fix it to the frame. (Much stronger while also much easier to adjust.)
The pivot mechanism is now located in the center, ensuring a nicer, safer and tighter fit.
A more efficient fan design to keep everything from melting down. My motor barely felt warm this morning.

No files yet. I got a serious amount of testing and (possibly) fine tuning to do.



View attachment 1

 

[nieles]

what about combining the two designs (the last can-based design and roller design)

there are two options, move the motor shaft (so it protrudes on the other side of the motor) and mount a bearing on the right side mounting flange.
Or put a bearing in the fan and use a short axle stub in the mounting flange.

 

[Gryphe]

Hiya,

I considered those options but past prototypes have shown these have a tendency to cause a lot more tear and wear, strangely enough.

Now, I tested a working roller design yesterday and though the roller was working the entire mechanism somehow jammed and exploded into a thousand pieces so I'm staying with the can-based approach for the time being.

The STL files for the can-based prototype mentioned in my last post have now been fine-tuned and can be found here.

EDIT: I realized I hadn't posted an actual, real life picture of the current state of my friction drive, so here you go.

How waterproof are brushless motors in general? Will it be destroyed if it stands outside for 8 hours? Opinions seem to vary wildly in the RC world. Otherwise I might have to go and invent an outer case.

 

[Dman12]

Looking good.

by the way, why don't you print the various tests out of ABS and save the expensive colorfabb filament for the working designs?

 

[tahustvedt]

Very nice. I ike to see the evolution.

Brushless outrunners can probably operate submerged in water, but they have many parts that will corrode. The magnets can corrode, the can will corrode if the protection is scratched, but the stator is probably the worst. I guess you'll have to try and see how it goes. You can install bearings with rubber seals to keep water out. Metal seals don't work well against water.

 

[Kepler]

Very well done. The design is looking more and more refined. I really like the simple seat tube attachment method. Also the cooling fan is a great touch.

Do you plan to incorporate any spring tensioning into the pivot to neutralize the motor weight for better engagement? Also do you plan to incorporate adjustable stop screws on both ends of the pivot travel.

 

[Gryphe]

With the 42mm SK3 motor I've been using weight hasn't been much of a issue; the plastic provides a bit of a cushioning effect and with the repeated abuse tests it's gone through there have been no signs of cracks. With the 50mm that will arrive soon I expect this to become more of a challenge, as it's double the weight. Parts have already been printed. (The 42mm was getting too hot on extended rides.)

One of the earlier prototypes had adjustable endstops but seeing as there's currently a slim chance of someone else starting to use my designs I've all kept it custom for my personal bike, seeing as a solid piece of plastic provides a much more robust part. An adjustable version will follow. Just not now.

This is what I printed last night. It incorporates the same pipe clamp concept as with the last design and also features additional reinforcement techniques. The arm is 15mm thick.