pic: 4 Swerve 9 CAD (FRC 449 Swerve Drive)



4 Swerve 9:
Top speed: 8 fps

Uses the Revolution 2 Swerve Drive Modules

For Drive two modules are driven together (adjacent ones) Uses 4 Cims in ToughBox Mini’s

For Steering the two modules on opposite sides of the frame are steered together. Uses 2 window motors.

The entire system is done using belts, lots and lots of belts.

Any feedback would be great!

You’re better off making your own CIM mount plates and belting those directly to the swerve modules. The fact that you go 12:56 and then use a large pulley : small pulley means you’re not going as fast as you’d like. The difficulty (admittedly) is getting a pulley onto the CIM, yet AndyMark makes a part for that. It’d save you 2 gearboxes worth of weight and increase efficiency.

The turning belt could use some work --but given that you’re trying to keep the center clear and also link opposite corners, some amount of complexity will happen.

Yet by far the biggest concern is the amount of side loading/twisting those swerve mount plates will go through with 140lbs of momentum running around on a field. Is there no way to add a second plate to support the top of the Revolution swerve module column?

He’s using the Toughbox Mini’s, which have the same gearing as a normal Toughbox with a plastic case (14:50 twice for 1:12.75 total reduction). He looks to be going ~2:1 up again, maybe more.

If we guess he’s going 2:1 up in belts, with the 1:2.25 final drive ratio in the Revolution 2, he will go at a peak speed of around 6ft/sec assuming no speed loss.

If he were to mount a pulley directly on the CIM, and a larger pulley on the Revolution crab pod, without the gearbox at all, let’s see what we could get:

-I am using 25 chain sizes since I am more familiar with the available AndyMark offerings for sprockets than the Gates pulley sizes.

CIM sprocket: 10t CIM bore (AM part here)
Crab pod sprocket: 32t plate (assuming with AM hub) (AM part here)

The math now says 12.87 ft/sec with no speed loss. With realistic speed losses for two stages of chain reduction, and generally good build quality (90% efficiency in speed), we can get a more realistic number of 11.58 ft/sec, which is in the ballpark of where you probably want to be.

As for the plate, it looks like he’s using the Revolution Mounting System from Team221, or something very similar. I would still try to mount that crab pod better, possibly mount it to another face (such as the front/back beam). You could alternatively make a moderately thick end beam bracket to replace the corner triangles and crab pod mount with a single piece of plate.

It does look like you’ve put a lot of thought into the manufacturing and mechanical design.

Ah, good catch – I assumed CIMple boxes based upon the drawing and didn’t see the Toughbox Mini’s in the description.

Note that sometimes the 32T flat sprockets have issues with chain interference when mounted to a AM hub. The chains sometimes touch the hub on the well-manufactured hubs, and don’t fully engage the sprockets due to interference on the less-than-ideal ones. As a rule, we’ve done 34T as the smallest allowed size in recent years.

Yes, I forgot about that. We got away with it by turning down the hub to allow good chain engagement.

I can’t find the post but I noticed that your pulley in the middle has the belt on both sides. Your not always guaranteed to line up. Art IV had a good picture explaining it but I can’t seem to find it.

  • Andrew

http://www.chiefdelphi.com/forums/showpost.php?p=976429&postcount=9

This is the post you were looking for.

Thanks for all the feedback so far!

Thanks for pointing this out. Originally we had one mounting plate that connected the swerve modules and the other plates. However, I didn’t want to make the mounting plates without using CNC, and with our limited CNC ability, I thought it would just be easier to use the revolution mounting system. Thankfully we just got a sponsor with greater CNC ability so we’ll look into that. How thick would you recomend making the mounting plate? 3/16 in or 1/4 in?

Regarding the gearing, we are using Toughbox Mini’s with the 10:71 Gear reduction because they are going to come in the 2013 KOP. Along with the additional 2.25:1 reduction with the modules, with 4 in wheels its a little difficult to get a good gear ratio. As of now we have the reduction from gearbox to module at 1:2.4 letting our top speed without losses to be 8 fps. Is this fast enough? Or should we make it a bit faster?

Thanks for pointing this out. We’ll try and adjust our design accordingly.

Thanks, we’ll post a revised edition later.

I think you need a little bit more speed. 8 fps is already a bit slow for FRC. Then you add in the losses and it gets slower. I’d shoot higher, say around 10-12 fps before losses. You won’t be a total speed demon, but that’s a good average speed, even once you factor in the losses.

Our Swerve modules have been at 9.8 Ft. per sec max and are plenty fast. We have discussed a lower top end but are satisfied with this reduction. You have designed a very complicated and limited swerve. Repairs during a competition could be a nightmare. We designed our first swerve with each wheel unit integrated into the frame. Working on it in the pits was time consuming and caused us to miss a match. For 2011 we went to totally self contained wheel modules. If something goes wrong we don’t fix it. We replace it. 4 bolts, 3 pairs of wires and it’s out and replaced in a short period of time. In 2010 we used the window motors (4) for steering. Teams that have used a window motor to run 2 wheels have had problems. After doing swerve for 3 years I can see allot of problems with this design. The are several teams that are successful with swerve. I would review their designs and reconsider you current path.

3/16" plate is probably fine if it’s integrated with the sheet metal gussets (as suggested by others) and therefore supported by the frame on three sides. Maybe someone else with more experience can chime in here; I’m going off of what I’ve seen rather than done.

Don’t be afraid to forgo the KOP drive train in lieu of the voucher. There are some small things you won’t get automatically from the voucher, yet (to me) that doesn’t justify adding complexity to a design. I don’t remember the details specifically (e.g. can one purchase wheels on the voucher instead of the gearboxes), but it’s worth looking into.