1983 Swerve Module Concept (Feedback Wanted)

FSM_MASTER-1.bmp (734 KB)

This swerve module was designed to be manufactured without the use of a CNC mill. The model is parametric, but the shown instance is 7.5" long x 6.25" wide x ~10.5" tall, with a projected weight of 7.6 pounds (using a full CIM and a non-lite versaplanetary with a 550).

The central mounting plate has three verstions with different mounting configurations: holes on the left and right; left and center; and center and right. The left center configuration is pictured

All feedback, question, and suggestions are greatly appreciated. Thanks!

FSM_MASTER-2.bmp (1.1 MB)

FSM_MASTER-1.bmp (734 KB)

FSM_MASTER-2.bmp (1.1 MB)

FSM_MASTER-1.bmp (734 KB)

FSM_MASTER-2.bmp (1.1 MB)

I like the look!

A couple questions:
What’s going on with the really long screw (?) bore in the bevel shaft on the fork and on the pivot shaft?
How is the pivot control gear made?
What CAD software did you use - I don’t recognize the look of those screenshots
What’s the fillet radius on your plates - it seems too small for the endmills we use, so is it water-jetted?

I’ll offer that it may be easier for people to give feedback if you post the CAD files as well as screenshots, since they can look in more detail.

  1. the holes in those two shafts are just for lightening. since those shafts experience no significant forces, and the machining of the shafts would be done with a lathe anyway, I added those holes to save, if I remember correctly, about 0.5 oz between the two.
  2. The pulley is 3D printed out of a polycarbonate-ABS mixture.
  3. Autodesk Inventor (the full section screenshots are done using ray tracing)
  4. the fillets have a radius of 3/32, but we are using a water-jet.

Yeah, Should have thought of including the cad in the op. here is is


Looks good!

Things I recommend through experience:

  1. You can make the top plate ribs a little smaller in width to save some weight (same with the wheel module).
  2. Unless there is a specific reason, you can reduce the number of parts/hardware by flipping the hex bearings on the wheel module so the flanges are on the inside of the module. Just add some spacers on the shaft to prevent the bearings from falling into the module. Theoretically, if you are constraining the shaft via bolts on the end of the hex shaft in the wheel module, you can leave the bearings the way they are and omit the small retaining screws holding the bearing against the plate.
  3. Is there a reason for the belt serpentine bushings for both rotation and wheel power? Tooth wrap is important, but I would think that you can get away with removing at least the rotation belt bushing.
  4. The screws which retain the bushings that constrain the planar movement of the module will be under a lot of load. Those threads will act as a file on the soft bronze bushings (I assume) overtime. You may want to look into switching those out with shoulder bolts or somehow find a way to put in a small ball bearing rather than bushing (can reduce to 3-4 at that point).

Overall, looks nice!

  1. The bushing (although actually at the moment it is a bearing on a shoulder bolt) on the wheel module belt is to reduce overall height, and the one on the rotation belt is to make the whole module more compact by making the belt “dodge” the CIM shaft, which it would otherwise intersect with.
  2. Good point, thank you! Ill see if I can find a combination of fasteners that would work better. Do you think a radial needle bearing on a shoulder bolt would work?

Needle bearings are expensive and finicky about their loading. And when they fail, they fail into a lot of pieces that then fail other things. I’d avoid them. They’re also expensive.

You won’t see high surface speeds at that interface. You might see “weird large forces” from hitting things when you are driving. Your bronze/oilite bushing is well suited to soaking up that abuse, it’s definitely what I wish I thought of and an idea I’ll steal in the future. Chris is pointing out a wear issue that might creep up on you after a few hours of practice. Smart bolt selection will let you avoid it… Or you can find a way to package ball bearings in there :slight_smile:

I see. would stacking two ball bearings be a viable option? With the current parameters of the module, the width of the bearing needs to be 3/8, but as there are no ball bearings with that width on McMaster that also have a reasonably small housing diameter, it seems like I might want to stack two bearings of 3/16 width around a shoulder bolt to achieve the same effect (McMaster-Carr these are the ones I’m looking at). The point that bronze bearings are soft and would wear down makes mar nervous about using them on the theoretically even but practically grooved surface that they would be contacting on their outsides (these irregularities would arise fact that that surface is comprised of multiple circles stacked on top of each other. See the 3/4 section view in OP).

On the subject of needle bearings, if they are finicky, is it acceptable to use roller needle bearings on shoulder bolts for belt tensioning, and should I be worried that I am using thrust needle bearings for the primary carriage azimuth?

I would recommend using radial ball bearings for the same reasons s-neff shared. You can double stack, however, there may not be a huge advantage to that, you just need to make sure the dynamic loading can be handled by the bearings. You can stack or have more around the inside of the hub, I would personally stack rather than add more due to DFA efficiency. Needle bearings are useful in many different scenarios, but if you can, try to use ball bearings as they usually can handle much higher loads in a smaller package. I’ve seen needle bearings fail, and it’s not pretty.