pic: Jumpin' on the Swerve Train

[Chris Hapstack]

loading photo ...

[Chris Hapstack]

CAD can be found here!

I definitely had fun incorporating a team number into the lightening holes in the side rails. Better view here.

[Andrew_L]

Don't get me wrong - this looks super cool, but unless your team is extremely prepared for it and has done it in the offseason, "jumpin' on the swerve train" is probably the best way to shoot yourself in the foot during build season. And this is coming from one of the world's biggest proponents of the swerve train.

Onto the more design specific stuff:

-How much does the drivetrain weigh?
-How much does a single module weigh?
-How complex is it to machine?
-What made you choose this specific form factor?

[Crimsonyde]

This is definitely something i'd love to see in 2017.

[Chris Hapstack]

Originally Posted by Andrew_L

Don't get me wrong - this looks super cool, but unless your team is extremely prepared for it and has done it in the offseason, "jumpin' on the swerve train" is probably the best way to shoot yourself in the foot during build season. And this is coming from one of the world's biggest proponents of the swerve train.

Onto the more design specific stuff:

-How much does the drivetrain weigh?
-How much does a single module weigh?
-How complex is it to machine?
-What made you choose this specific form factor?

I should clarify that the swerve train I'm referring to is more of the off-season, design-exercise type. Seems like it was the most popular thing to do this off-season. I'm more of a 6WD drop center guy myself, but figured it would be fun to try my hand at some swerve design while I still had free time. I have no idea if we'll be doing swerve this season, but figured I would post this anyway.

As for the tech specs, the entire base shown here is 43.5 pounds and a swerve module is 8.4 pounds. Definitely on the heavy side, but I was shooting for robustness. The frame would all be waterjet-cut, and the swerve modules themselves have 3 team-manufactured parts: a 3D-printed encoder mount, a (set of 4) standoffs done on a lathe, and a tube that requires 2 mill setups. By form factor do you mean "length vs width" of the base?

Do you think adding frame-mounted bearings to the tops of the Revolution Pro 2 vertical drive shafts would add stability in a helpful way, or just over-constrain things?

[Andrew_L]

Originally Posted by Chris Hapstack

I should clarify that the swerve train I'm referring to is more of the off-season, design-exercise type. Seems like it was the most popular thing to do this off-season. I'm more of a 6WD drop center guy myself, but figured it would be fun to try my hand at some swerve design while I still had free time. I have no idea if we'll be doing swerve this season, but figured I would post this anyway.

As for the tech specs, the entire base shown here is 43.5 pounds and a swerve module is 8.4 pounds. Definitely on the heavy side, but I was shooting for robustness. The frame would all be waterjet-cut, and the swerve modules themselves have 3 team-manufactured parts: a 3D-printed encoder mount, a (set of 4) standoffs done on a lathe, and a tube that requires 2 mill setups. By form factor do you mean "length vs width" of the base?

Do you think adding frame-mounted bearings to the tops of the Revolution Pro 2 vertical drive shafts would add stability in a helpful way, or just over-constrain things?

Form factor as in why did you do the 221 style swerve instead of the Aren Hill style swerve?

As evidenced by the post above, I'm not too much of a fan of the 221 form factor anymore - it used to be the best method, but ever since Aren had decided that moving sideways wasn't a waste of time I think his form factor is a much better solution. I'd look into that style if you can.

[Chris is me]

Originally Posted by Andrew_L

Form factor as in why did you do the 221 style swerve instead of the Aren Hill style swerve?

As evidenced by the post above, I'm not too much of a fan of the 221 form factor anymore - it used to be the best method, but ever since Aren had decided that moving sideways wasn't a waste of time I think his form factor is a much better solution. I'd look into that style if you can.

Aren has a long, long history of moving sideways.

The 221 form factor is still a reasonable choice - there's nothing WRONG with it, but the concept design Aren posted in... 2014? does reduce part count, module height, etc.

With the way he is transferring power from outside the module via a belt and tucking the CIM into the chassis, there would be extra weight in the power transfer stage that would need to be added from the coaxial input to the offset output, which probably cancels out the extra weight of the bevel intermediate shaft. You also have to source a bevel gear for the output whereas you can use unmodified COTS bevel gears from Vex for this design. It's not really a huge disadvantage or anything to use this style of module.

Really the main argument against the 221 style module design is that you can just go buy a 221 module if you want to use it. If you're building your own and it's similar to a COTS product, it's not the most efficient use of resources.

Great design! I like the chassis design and the sheet metal work.

[Chris Hapstack]

Originally Posted by Chris is me

Really the main argument against the 221 style module design is that you can just go buy a 221 module if you want to use it. If you're building your own and it's similar to a COTS product, it's not the most efficient use of resources.

Great design! I like the chassis design and the sheet metal work.

The availability of a COTS swerve module was 100% the reason I went with the 221-style.

Also, thanks!

[marshall]

This is cool and I'll add that 900 was considering a very similar style design (also based on the 221 modules):

https://workbench.grabcad.com/workbe...Eehe_yum5yhJ04

[Chris Hapstack]

Originally Posted by Andrew_L

Form factor as in why did you do the 221 style swerve instead of the Aren Hill style swerve?

Also, I think my mind is still in the "robot must fit under the low bar" mode from Stronghold, so I probably subconsciously wanted the chassis to be short...

[Chris Hapstack]

Originally Posted by Chris Hapstack

The swerve modules themselves have 3 team-manufactured parts: a 3D-printed encoder mount, a (set of 4) standoffs done on a lathe, and a tube that requires 2 mill setups.

I forgot a couple things: A CIM shaft-to-magnet coupler for the drive encoder, also done on a lathe. Plus you'd have to pocket one of the gears a little bit for VersaPlanetary fastener clearance.

[bstew]

Cool design!

Originally Posted by Chris Hapstack

Also, I think my mind is still in the "robot must fit under the low bar" mode from Stronghold, so I probably subconsciously wanted the chassis to be short...

Another Aren Hill swerve design like this may better fit that goal.