pic: 4143 offseason motor in wheel swerve

4143’s offseason build. Inspired by Team Neutrino 2012 and Pwnage 2014.

Looks interesting. Where is that ring gear coming from? If it’s custom, it could take a long time to make, even with good manufacturing sponsors.
Second thing I can see is the massive coaxial-style top piece. How is it attached to the cim/wheel assembly? Also, whydid you choose to place gear at the top as opposed to a sprocket or pulley?
I assume the wheel is custom. Is it riding on a bearing or a bronze bushing or a plastic bushing?

Last thing: speed and weight. What are you geared for, and how much does the shown stuff weigh? You should try for 15fps+ to maximize usage of your 4 cims. As for weight, it’s really your call, but I like my swerves to be less than 9lbs as a general rule of thumb.

Very nice swerve drive. I like the use of a planetary there. I recommend QTC gears for the ring gear, or even a AM gem ring gear (because it’s only $10 and it’s hex shaped!). Also, check out this article from QTC on internal gearing: http://www.qtcgears.com/Product-Technical-Information/Internal-Gears-Technical-Information.htm
It explains the three types of interferance you should seek to avoid, and the number of teeth you need on the planet gears to avoid them.

Our first test swerve chassis from last year wore out. 3.5" colson wheels became 3", the frame was bent, and rivets became loose from the all the driver practice.

This design will test a number of new ideas, including a plus-shaped frame, stiffer swerve pivots, and custom internal ring gear wheel. This is also the first year our team will have a Tormach cnc mill dedicated to us.

The design was born after seeing Team Neutrino’s motor in wheel swerve module on display at IRI in 2013. Pwnage improved on it again last year.

As designed now, the wheel diameter is 4.4" with a 74 tooth ring gear. All available cim output gears can be used (11, 12, 13, and 14). I hate quoting top speed because some teams use cim free speed and some take off an efficiency factor.

More pictures, cad, and section views coming later today.

Can you please explain how it works? It looks really cool, but I am not entirely sure how it works or even what it is.

Sorry, picture got approved after I went to bed and couldn’t get back to until now.

Section View:


Step file:

It’s a swerve drive; you rotate the wheel to face in different directions in addition to spinning it with a CIM or other motor. So you can rotate the wheel sideways to translate left and right as well as forward and backward.

This is one with a CIM actually inside the wheel, acting as an axle. It lowers the center of gravity and is really cool anyway.

Okay, I’m looking at the cad.
Main thing I’m wondering about is the internal gear/ wheel assembly you have going on. How are you planning on making that? Do you have the correct broach for it?

Interesting chassis design. Will it bend in the middle? I like how it saves weight and space on the 2x1.

You are correct the wheel is an assembly in the cad. This is really one piece. We have parameterized Creo involute gear files. It was being lazy that the wheel was modeled separately and ‘glued’ to the gear. It is designed to be made as one piece out of 4" OD .5" wall 6061 pipe.

There is a machining process called gear shaping. It is not too common any more but is still about the only way to form a gear up against a shoulder. This article describes the history of process. Here is the first random video I found.

There are in fact many different ways to shape gears, as well as hob them. For internal gears it is advisable to shape instead of hob, and their profile makes it very difficult to shape them properly (espcially in your usage context). The style of shaping you linked to is difficult for gears that go up to a shoulder. I would only be comfortable using an internal gear rotary broach to broach up to a shoulder, and you would have to make that tool yourself.

Your sponsors may not have the correct technologies to produce internal gears. Any mill can make external gears, but internal gears are far more difficult to create, and even more difficult if they have shoulders. Make sure your sponsor can make these gears before going any further. If they cannot, just bolt a pre-bought internal gear to the end of the assembly instead.

I do like how it’s designed to be made out of 4" stock tubing though. Very clever.

We used MITCALC Internal Spur Gears software to generate a tooth profile. They had a free 30 day trial a while back. We generated one tooth profile and then patterned the tooth around the gear. The internal gear was wire EDM’d.
You could possibly have them waterjet or high-def lasered. I would recommend bolting on the gear to your part, not integrating it into the design. Stock Drive Products also sells internal spur gears but the 1/4" thick ones are brass and the stainless/aluminum ones are only 1/8" so you’ld have to stack them up to get enough tooth width.

Waterjet cutting of gears can definitely work, you just have to be aware of some of the limitations of the tool.

If you’re planning to waterjet cut the gears, you might want to take a look at this page, which lists some of the limitations of waterjet cutting gears, including kerf and surface finish.

Progress is being made. It should be moving in a week or so…


I’m not sure if this picture will work:

Are those herringbone gears on the top!? Holy cow you guys got gears up in there.
It looks wonderful! You might want to take some scotchbrite to the module so it displays better.
Why gears to turn the module over chain or belts?

How is the drive CIM wiring connected to the controller? Unless you’re limiting the rotation, you’re going to need slip rings.

A mercotac slip ring is visible protruding from the top of the module.

The herringbone gears are 3d printed in gray plastic. Very hard to machine, very easy to print. You see them a lot on 3d printer extruder designs.

Our pyramid climber had so much chain we kind swore off chains. The memories of hunting for fallen master link clips is fading slowly.

First movements of a new swerve drive are usually not pretty. This one wasn’t too bad.

It weighs in at 52lbs right now, with control system, without battery. A proper battery tray and bumper mounts might take it to 55lbs. It is a little heavy, but we used a lot of bearings. I’m not sure this design will ever see competition, but it was fun to make.

The video has a real good view of the steering system.

Looking good!
Why not competition? Too much machining?

Yes, we made this prototype because I wanted to see if we could make internal gears. Having made it, I’m not sure there are that many advantages over simpler designs. Perhaps I’ll change my mind after we test it more.

BTW the gears were made with old Fellows shaper cutters found on ebay. It’s possible the cutters are older than I am.

We’re releasing the native Creo models and drawings. Included in them are mostly parametric involute gear designs. They have been pretty useful for 3d printing gears.