Who makes their own wheels?

In an effort to collaborate with my teaching comrades, I think a wheel project would be good. It is a great way to connect with a physics teacher with Newtons Laws of motion and overcoming the moment of inertia.

If any one has done this please offer advice.

We can use our Velox CNC to mill them out of aluminum and finish on a Lathe. We can also 3d print using PLA, ABS, and Polycarbonate.

We used 2.5" 3D printed idler wheels on our track system this year. They held up better than expected and were very light weight. Our test wheels didn’t have bearings and the later ones used 4 skate bearings to help support the hub. Without the stacked bearings the wheels would waller out quickly. We also used some large 6" 3D printed idlers on the prototype. They used two flange bearings and held up fine.

We replaced the large printed idlers with turned aluminum. I’d recommend turning your blanks and then machining the features. I blanked the idlers at work and then had the students mill the bearing pockets. The lathe at the school is pretty tired. I uploaded the video of the process. The students had to mill the soft jaws before milling the wheel.

We also modified some COTS idlers, these videos may give your some idea of the secondary operations.

If we need to ever make a lot of wheels we’d likely mill out a pattern and cast the blanks several at a time from scrap. Material cost on large wheels can add up quickly as round bar is typically a few bucks a pound. I think the large idlers were about 7 lbs each in material. We made four.

A method that I’d like to try would be to mold some from castable plastic resin in a silicone mold. Any teams casting them in plastic?

We made custom hubs for 10" Harbor Freight pneumatic wheels this year from stacked laser-cut plywood plates. Worked great and were much lighter than the original hubs. I can find a picture/ render if there’s interest.

Also experimented with making CNC-routed wood wheelsbefore the season.

The version 2 of the wheel is downright cool. What a great basic cnc project. I would think machining them from something like birch plywood would prove very sturdy.

If some draft was designed in (or profiled with a tapered end mill) the wooden wheel could then be used as a pattern for casting in resin or even aluminum.

I agree the wood wheels are downright cool.

In the years we have done West Coast drive with traction wheels we have used custom wheels. We have our sponsor machine them for us but we could make them ourselves if we so desired.

One thing you should look into is milling cutters where the shank is smaller than the cutting flutes, look up T-slot cutters. This might let you make it all in one go on your router if you are clever, it’s how our sponsor machines them.

The most important feature you should look at how to machine is the cross slot for installing the tread on the wheel.

Great information RoboChair, I had not thought of T-Slots for this purpose.


We used to make custom wheels, but there are so many good options now I don’t see why 99.9% of teams should. We likely won’t ever again.

I was just thinking that’s such a great quote… Then clicked on the link and laughed.

So 2056 doesn’t teach students or collaborate with other teachers in the offseason? I agree that making custom wheels during the build season is not an ideal expenditure of time.

However, I thought my job was to educate students. Which I do year round. If my students know how to design and make a wheel, they will understand how to select the best wheel for our application.

I guess we could go the route of copying what the best teams do, but I prefer my students to understand why we do things.

You can argue that the best of both worlds is to have your students design and make something that doesn’t directly have more COTS options and is more analogous to the type of things they would design in season.

Also, it probabaly wasn’t intentional but your tone seems to really dismiss system level design. Effective use of COTS parts is real engineering and exists in all industries. It’s really an arbitrary line anyway, in terms of a robot scoring points is a wheel really all that different from a bolt? A connector? Etc…?


Since there are so many good COTS options for drivetrain wheels, another suggestion might be looking into making custom wheels out of a softer rubber to use as intake or shooter wheels.

I know 125 cast their own urethane wheels for their 2012 shooter, there should be some pics up on CD of that process.


I like the idea of making shooter wheels. Currently on the COTS market there isn’t much to be said for wheels with a large moment of inertia but remain grippy. Sure, there are the McMaster options, but those tend to be rather expensive.
Maybe something made of aluminum with holes to put steel rods into? This would also let you try balancing the wheel.

While we’re on the discussion (and since I have one hanging on my wall)-

195 made a custom shooter wheel for 2013.

The wheel was ~12" in diameter, weighed 4.5lbs and was direct driven by a CIM.

It started as a 14"x14" block of aluminum that was Wire EDM’d to have a spoke profile.

The block was then placed on a CNC mill and the ‘wheel’ was milled out using a .5" end mill. This left the 14"x14" block with a 13" hole cut out of it.

The rim of the wheel was then sandblasted to allow for the urethane to hold while at speed.

The outside of the block was used as the mold for the urethane cast. The wheel was replaced back inside the block and a .5" dowel pin was used to make sure the wheel was centered in the block. 60A durometer urethane was poured into the mold and was let to sit.

The overall process was a lot of work, but it worked well as a flywheel.

Mountain Dew can for scale.

In 2012 Skunkworks made a flywheel using a method similar to what you are proposing. We cut an aluminum hub that was .25" thick and of 6" diameter iirc. We added .25" steel rings with matching OD and 5.75" ID, sandwiching the aluminum hub. The assembly had considerable mass, but the mass was where we wanted it to maximize moment of inertia. We covered the wheel with a section of inner-tube from a motorcycle tire to give it a more desirable surface.

I would consider constructing a custom flywheel a worthwhile endeavor, but in my experience an accurate shooter has much more to do with proper control than superior mechanical design.

We 3D printed and then cast hubs for urethane tube to use as intake wheels in 2015.https://scontent-yyz1-1.xx.fbcdn.net/t31.0-8/10256793_10155208355065576_3280327124933048631_o.jpg

Uploaded a video of another flywheel I build for a physics project with the method I outlined previously. Might help make sense of my garbled description.

I don’t think one aspect is more critical than the other. You need both, period. 469’s shooter in 2012 was not the best mechanically (double axle, very low MOI, low compression), and as hard as we tried, no amount of code was going to make up for that.