6941 offseason swerve using modified WCP modules

Hi! To prepare for the 2020 season, my team decided to build a swerve drivetrain. We are a relatively new and inexperienced team, so we modified the WCP SS swerve modules(the old version without falcons) instead of designing new modules from scratch. Thanks a lot to 1323, WCP and 2910 for making this project possible!

Here are the main modifications to the module that we made to simplify the project:

-Azimuth bearing changed to SA035 X contact bearing because they are easier to source in China and are also possibly more suited to take moment loads than radial bearings

-modification of main azimuth gear and plates to accommodate SA035 bearing. After modifying the azimuth gear, we actually had to have a machine shop mill the gears from 7075 round stock instead of from vexpro gears.

-redesigning shafts so that they could be machined from vexpro hex stock. This made machining easier, but also meant much more backlash as the hex no longer is oversized. To deal with this, we added retaining compound to fill in the gaps.

-We didn’t want to directly make the leap to treaded billet wheels. And so, we changed the wheel to 4’’ Hi-Grip which is cheap and simple. However, this change led to most of the redesigns. We had to change the large bevel to fit the 1.875’’ bolt pattern, and also make the wheel forks, small bevel, and spacers accommodate this change.

-When the machining of aluminum wheel forks messed up, we designed new forks optimized for printing and made them in-house to save time. Astonishingly, the printed forks seemed to be more rigid.

-Changed azimuth encoder to direct 1:1 after building the first module to increase accuracy. We were happy to find out a few weeks later that the newest WCP falcon swerve has the exact same change!


After completing the first module, we attached it to a recycled practice bot frame along with three caster wheels. It looked pretty messy and could only strafe while spinning uncontrollably.

When the second module got assembled too, we built a new frame out of extrusion and attached both modules, allowing full swerve motion. We plan to get two more modules completed and also build an everybot on top of the swerve to test scoring before kickoff.

We are still very new to both swerve and FRC, and unsure whether the decisions we made in this project are sound, so we’d love to get as much advice as possible!

6 Likes

Super cool! I’ve been taking a look at a similar design this month, and I’m encouraged to hear you had success with 3D printed forks. I really think with plastic forks the design of a swerve module could be brought down to sub $200 (before motors, wheels or encoder) to be cost competitive with a COTS tank-drive single speed gearbox.

1 Like

Thanks! I am glad to know our project encouraged you. The printed forks are indeed significantly cheaper than aluminum ones. But still, beating tank drive in terms of cost seems to be pretty hard, with new pinions allowing usage of single-reduction gearboxes for various speeds. Does your team plan to use more printed parts such as 3d printing azimuth belt pulleys to further reduce the cost of swerve?

I’m not sure what our team will plan to do :slight_smile: This would be a summer 2020 project if anything. But yeah, the idea would be to use 3DP parts for both the fork, the azimuth pulleys, and the encoder mount. The only machined parts would be a CNC aluminum base plate, and a laser cut plastic top plate. The design also uses a minimum amount of gears, bearings, shafts, and other hardware. I think $200 (USD) can be achieved.

Current (very unfinished) design at right (WCD swerve at left, SDS MkII in the middle)

2 Likes

Wow! I really like the minimization of parts. The module looks very simple, effective and compact. After 2020 season maybe we would attempt a similar layout too :slightly_smiling_face:

2 Likes

1640 has been redesigning our swerve too. The main steering pulley is 3d printed along with the wheel forks. Used PETG for most parts and a gear and drive pulley out of a PC alloy carbon fiber.
The 2 plates are cut on our router out of .25 G10. Yes, we did make special provisions to control the G10 cuttings and dust. We have only been test driving a week. We have repeatedly smashed into this years ball coral at full speed and fully weighted chassis. Most all of our module is 3d printed or non-metallic except for shafts, stand offs, and gears. So far it’s holding up. We will post cad soon.

2 Likes

Great to hear this, I really look forward to seeing your newest design. 1640’s trip to China this year helped us learn a lot, and is one of the main reasons why we were inspired to try swerve.

Here’s a video of the drivetrain in action. Although driving it is quite intuitive, handling three degrees of freedom simultaneously still makes driving it feel a bit harder than driving WCD.

We also almost finished the everybot structure, so if we are lucky soon we could try some scoring and gain more understanding of how swerve should be driven.

3 Likes

I printed an entire SDS MKI pod over the summer and was surprised on how rigid it actually is. I modified holes and used 3D printed pins as fasteners as well so EVERYTHING was 3D printed. I think if printed out of nylon you can even get away with printed the large bearing around the pulley.

I haven’t been able to test it for real but it is promising to think that turning pulleys as well as wheel forks, and nylon bearing would hold up for a competition as 3D printed parts. Possibly longer!

Here is a STEP file to the module just in case anyone is interested


Swerve-V7.step (46.6 MB)

2 Likes

This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.