pic: Swerve Module Concept (View 1)

A concept coaxiel swerve module with 3x3 casing and exterior mounted driven and driver sprockets.

All feedback, question, and suggestions are GREATLY appreciated.

When you say 3x3 casting I am just wondering how you would do that… Like do you have a sponsor that is willing to do it; etc.; Or am I miss understanding you

And also why would you chose to do an outside drive with the chain

And how are you going to be rotating the module… It looks like there is a sprocket under the bottom bronze bushing if so are you sure there is enough clearance.

Along with Jeff’s questions, how tall is each module?

What I meant by 3x3 casing is that the housing for the wheel is made from 3"x3" (1/4" W) aluminum extrusion with a piece of 1/4" sheet welded to the top to form a rectangular prism with an open bottom. I placed the sprockets on the outside to both reduce the height and width of the casing in an attempt to save weight. I believe Team 1625 did something similar in 2K8 and I moved the sprocket outside based on advise of one of their members.

The module is steered by a sprocket mounted above the upper bronze bushing. It is not show as it is part of another assembly.

It is 6.6 inches from the bottom of the wheel to the top of the flange of the bottom bushing which is in contact with the frame.

It appears that all of your mounting, to the robot frame, is located at the top of the module. Wouldn’t this cause high stress at the bottom of that mounting if your were to put on high traction tires?

It would. I have tried to make the module as short as possible to reduce the length of the arm that would be applying force to the frame. It is my belief that worst case scenario would be being pushed when the wheels are facing perpendicular to the direction of the pushing. My biggest concern was breaking the wheel or bending the casing. To address this problem I chose to use the AM traction wheels with a custom hub, and use pocketing to reduce as much weight instead of cutting through the entire wall.

You might consider putting a ring near the bottom of the module which would effectively reduce the lever arm from 3+ inches to under 1/2 inch.

Assuming that a robot can push with 180 LBS force (150 LBS * 1.2 CoF) and that your wheel will not move (is sideways against a lip) … probably as close to worst case scenario as we need to get … in your current design you’d see ~ 630 INLB force on the mounting, and if you were to put a ring near the bottom you’d see ~90INLB.

BTW, I’m not picking on the design … I really like it. I’m just concerned that it will fail at the mount to the frame. Gut feeling

What wheels are you planning to use on this design? It looks like you’d be able to thin it down (and use less power while steering the modules) is you used Colson wheels… But switching your wheel choice means examining design priority: Do you want to push, or out maneuver? Crab makes a robot able to run circles around many other robots, but depending on ratios and wheel type, can also be a decent pushing bot.

I’d love to see the rest of your chassis for this design, it looks quite solid.


I appreciate the concern and the suggestions. I completely understand that you’re not picking on my poor little module :stuck_out_tongue: I have considered adding a ring like you suggested. Possibly out of delrin or something of the like. It seems like there is a definite divide on Delphi between the 111 style and the 118 style module.

From what I have seen on Delphi and in person, there are numerous swerves without support rings that have worked just fine. I’m planning on CADing up a version with the ring to see how I would have to alter the frame and module to make it work, but I find that I am of the persuasion that prefers the no-ring approach. I would like to here more about the pros and cons off adding a ring if you would humor me with the discussion…


I am using the 4" AM Traction Wheel but modifying the hub that comes with it as well as adding a custom hub to change it from a 3/4" roller bearing to a 1/2" keyed shaft. The hub I sketched up goes all the way through the wheel. I didn’t want to blow spokes like I have seen happen with some swerves, so I picked the heftiest wheel I could for the cost and added as much support through the custom hub that I could.


The colson wheel would be much stronger than even the andymark wheel with a hub and also have slightly less traction to place less stress on the module. You would also not have to replace tread and they are much cheaper. IMO colsons are the perfect wheel for a crab module, especially with a custom hub.

I’d also go along with the suggestion for colson performa wheels. In nonswerve applications they’ve been our best wheel yet for defensive driving. 148’s coaxcrab in 08 was also highly impressive with the wheels. This thread might help you out just a bit if you want more info on wheel selection…

Thanks for the tip on the Colsons. They look cheap and reliable. I just started working on a version with them.

However, after looking on the “New Stuff” page on AM I think I will go with the 4" performance wheel that are coming out/came out as I would not have to make a custom hub and they appear hefty enough to stand up to the side loads that could be encountered.

Both designs have tradeoffs and their use truely depends on design intent.

Since I’m of the pro ring design, I’ll put forth what I see as it’s biggest pros and cons:

Pro: Less stress on the whole module – with the ring low (near the ground) you have a significantly shorter lever arm and therefore much lower stress (see above post). This will allow you to make your side plates thinner (and lighter) while increasing their robustness.

Con: More framework. The frame must contact the module at 2 rings that are not close to each other. That means more framework and possibly a heavier chassis.

Ring designs tend to have less ground clearance as well, but with the side gear that low in your design I don’t see that as much of an issue.

If anyone else would like to chime in, I’d like to hear y’alls opinion as well.

The “ring” can be put at any height you want really so ground clearance is very tweakable.

And building a frame to go with the “ring style” using 1x1x.0625 box tubing works very well weight wise as opposed to pure top mounting which would require heftier frame members