Hot Take on My Drivetrain Design

Hello everyone,

I just designed this chassis as an offseason project. Based on what you all can see, would this be effective in competition (IE not get eviscerated by a defense bot)? I understand the aspect that geometry plays into this (parts with 3D geometry are stronger than 2D plates), but I feel like I’ve done a reasonable job placing standoffs and brackets to counteract this problem. If there’s any good way to test this in SolidWorks, I would love to know.

Would love to see some feedback. Thanks everyone.

5 Likes

Hi! I’m not sure this is what you meant but I feel like it should be asked:

Can you defend your choice as to why this custom drive over a kitbot chassis?

8 Likes

Our programmers didn’t like the dropped center wheel of the KOP during autonomous.

This was also supposed to be a fun project for the team to do that goes a little bit outside the box of what we are normally able to make so doing things custom seemed like a good idea.

2 Likes

Unfortunately, this is a problem you will simply have to deal with. The drop-center is integral to having a robot that can drive properly, as it eliminates wheel “scrub” from the corner wheels. Center drop can be reduced by using omni-wheels on the corners, but a robot with perfectly in-line wheels(even omni wheels) will have a very hard time turning, which will make autonomous even harder.

2 Likes

Fair enough. Thank you for the input.

I will also add that a center drop does add a slight “suspension” to the drive. For example, in the 2020/21 game, where there were 1” bumps in the field, if you have no drop, when you hit the barriers, the front 2 wheels will immediately pop up at the same time, giving only your back wheel a chance to push the whole robot. With a center drop, when the robot hits the bump, it will rock back on the rear 2 wheels and give slightly more pushing power to clear the bump smoothly.

This is false. You can run a 4WD with omni-wheels and turn just fine; why do you think a 6WD with omni wheels would require a center drop?

OP: Your design looks fine, though most of the important constraints have little to do with how the drive performs and a lot to do with how easy it is to build, and it’s hard to judge that without knowing how you intend to manufacture this.

18 Likes

A 6wd with omnis needs a slight center drop to make sure that the center wheels are actually in contact so you maintain traction (iirc 254 did 1/16 in 2018)

4 Likes

That depends specifically on your choice of wheels; you’ll have to nudge the center up or down to account for differences in the effective diameter of the wheels, since it’s extremely hard to source an omni-wheel and a traction wheel with identical diameters.

5 Likes

Yeah, but generally speaking you’ll end up needing to have the bottom of your center wheel slightly lower than your corner wheels, otherwise it isn’t guaranteed to be in contact

1 Like

“Contact” here isn’t really a dichotomous thing. The real question is, “how much do you want to load the center wheel relative to the outer wheels?” This depends on your design goals.

4 Likes

It is very kop-like, but missing the features of having bends in the plates, which do impart a lot to the strength of that design. If you have the ability to make long sheet metal bends that’d be an improvement.

1 Like

could you provide an example where loading the outer wheels would be better then the center ones that makes a significant difference to justify not having a drop?

2 Likes

Ceteribus paribus, I’d default to loading all three wheels equally. Whether I’d need a drop (or a raise) to do this would depend on my choice of wheels.

why is that your default? if this was all theoretical then i would get it but we live in a world where machining isn’t perfect. floors defiantly aren’t perfect and you would assume that there’s no other obstacles you would have to drive over which most games have.

Because equal loading means equal wear and a decreased risk of failure.

This isn’t a decision that can be made in the abstract. What, particularly, do you need the robot to do? What does that say about weighting the wheels? Decide that, and then figure out what amount of center adjustment is necessary to achieve it.

OP, I like the drivetrain. My only issue is with all the L-brackets, which are always kind of finicky to get to fit tightly. You might want to look into tab-and-slot connections to make sure everything fits well. @Allison_K might have some pictures of fitting a captive nut into into a tab and slot design, which is cool.

This could be a minor complaint, but if you look at this design from the top, it’s made of rectangles. Rectangles are easy to make, but not particularly rigid. The bellypan is doing some good work to keep the big center rectangle in shape, but the outer rectangles that make up the wheel wells aren’t all that well supported, and might be susceptible to deforming like this:

If you attach those two plates with a flat plane (tab & slot attachment?), you can keep that assembly as rigid as the center part is. (As a bonus, it’ll also help to fix @ngreen’s concern with the sheet metal bends).

@oblarg and Co, this conversation is technically interesting, but only tangentially relevant to OP’s drivetrain. Maybe bring it to a new thread?

10 Likes

I got you.

Section 3.5

3 Likes

Understandable. I’ve also seen some teams use a lathe to mill down the center wheel. This makes it smaller than the outer ones, effectively eliminating the drop. This might serve as a cheaper option.
(Or at least a prototyping alternative to see if you like no-drop before committing)

Such a good document. I can’t believe that I always forget about it!

Page 49, for the captive nut idea.

3 Likes