[FTC]: Is this Mecanum wheel configuration acceptable?

[ImMoMo]

Hey!

So I'm a college mentor for my old team, and I'm helping out with FTC.. so it's pretty late into the our build season, and we don't have a substantial drive train design..

So using what little CAD I learned in my intro CAD class in college, I decided to sort of mock up a way of connecting the NeverRest motor to the Mecanum wheel..

The students want to go towards the Mecanum wheel route, so we're going to try to CAD and assemble a similar structure, using
- NeveRest motors
- Andymark Mecanum wheels
- Tetrix Wheel Hub

It seems like it all fits well in CAD, but I would like another opinion if these parts would be compatible.

[GeeTwo]

Can you increase the brightness and contrast? I can see what I assume is a mecanum wheel because you mentioned it, and a couple of other shapes before I enhance it.

After doing both adjustments, there are a couple of questions that come quickly to mind:

First, is this picture inverted? If not, you will have VERY little ground clearance, and the bolts securing the gearboxes to the c-channel will be in tension rather than compression. Don't worry so much about the robot holding the wheels up as that the wheels hold the robot up.

Second, consider whether the gearbox can handle the shear force of supporting its share of the robot weight. In the very best case, 1/4 of the robot weight will be applied at the center of the wheel to the axle. In a more realistic low-impact case, it will be more like 1/2 of the robot weight. In a high-impact case (e.g. something similar to STRONHOLD, where you are hitting vertical walls and catching air), you will want to multiply by another factor of 3 or 4 or more. As I look at the NeveRest40 (picked one at random), I do not see what sort of lateral torque it will support. If the gearbox is not up to supporting this torque, the simplest solution is to place another support member and bearing at the end of the shaft, past the wheel, and support a significant portion of the robot weight on those members.

[Billfred]

Originally Posted by ImMoMo

Hey!

So I'm a college mentor for my old team, and I'm helping out with FTC.. so it's pretty late into the our build season, and we don't have a substantial drive train design..

So using what little CAD I learned in my intro CAD class in college, I decided to sort of mock up a way of connecting the NeverRest motor to the Mecanum wheel..

The students want to go towards the Mecanum wheel route, so we're going to try to CAD and assemble a similar structure, using
- NeveRest motors
- Andymark Mecanum wheels
- Tetrix Wheel Hub

It seems like it all fits well in CAD, but I would like another opinion if these parts would be compatible.

They say that if you say "AndyMark" three times we show up. Or sometimes just once...

Some considerations I see:

1) Those parts will all mate, but we'd suggest the Nub over the Tetrix hub for the D-bore. Set screws aren't a good look.
2) On one hand, cantilevered wheels will give you some flex that is desirable in mecanum drives. On the other hand, supporting both ends is beneficial as weights increase (especially for prolonging motor life). Our TileRunner chassis is certainly one way to accomplish that (and speed the motors back up, to account for mecanum inefficiencies), whether you buy it or just draw inspiration from it. A second Nub on the other side of the wheel and D-shaft into a bearing block would accomplish this too. And as Gus says above, make sure it'll actually clear (or just flip the drivetrain over).
3) You don't mention which NeveRests you're using. 20s are on the ragged edge in direct-drive applications for FTC-weight robots; teams using 40s during Fight Night had no problems, while teams using 20s were breaking gearheads (admittedly, under higher-than-usual voltages and abuse).

[ImMoMo]

Originally Posted by GeeTwo

Can you increase the brightness and contrast? I can see what I assume is a mecanum wheel because you mentioned it, and a couple of other shapes before I enhance it.

After doing both adjustments, there are a couple of questions that come quickly to mind:

First, is this picture inverted? If not, you will have VERY little ground clearance, and the bolts securing the gearboxes to the c-channel will be in tension rather than compression. Don't worry so much about the robot holding the wheels up as that the wheels hold the robot up.

Second, consider whether the gearbox can handle the shear force of supporting its share of the robot weight. In the very best case, 1/4 of the robot weight will be applied at the center of the wheel to the axle. In a more realistic low-impact case, it will be more like 1/2 of the robot weight. In a high-impact case (e.g. something similar to STRONHOLD, where you are hitting vertical walls and catching air), you will want to multiply by another factor of 3 or 4 or more. As I look at the NeveRest40 (picked one at random), I do not see what sort of lateral torque it will support. If the gearbox is not up to supporting this torque, the simplest solution is to place another support member and bearing at the end of the shaft, past the wheel, and support a significant portion of the robot weight on those members.

Sorry about the weird image, still familiarizing with SolidWorks, I will attached a non rendered image.

Yes, the model is flipped, so the wheels will and motor will be on the bottom side of the structure.

I believe we will be using a NeveRest 20, or possibly a Matrix motor to drive Mecanum wheels. I wasn't really concerned with the weight impacting the motors too much -- well not until you mentioned it (: I think we may end up supporting the other side of the wheel, like you suggested. It doesn't seem favorable given the limited time, however it most likely will happen!

[ImMoMo]

Originally Posted by Billfred

They say that if you say "AndyMark" three times we show up. Or sometimes just once...

Some considerations I see:

1) Those parts will all mate, but we'd suggest the Nub over the Tetrix hub for the D-bore. Set screws aren't a good look.
2) On one hand, cantilevered wheels will give you some flex that is desirable in mecanum drives. On the other hand, supporting both ends is beneficial as weights increase (especially for prolonging motor life). Our TileRunner chassis is certainly one way to accomplish that (and speed the motors back up, to account for mecanum inefficiencies), whether you buy it or just draw inspiration from it. A second Nub on the other side of the wheel and D-shaft into a bearing block would accomplish this too. And as Gus says above, make sure it'll actually clear (or just flip the drivetrain over).
3) You don't mention which NeveRests you're using. 20s are on the ragged edge in direct-drive applications for FTC-weight robots; teams using 40s during Fight Night had no problems, while teams using 20s were breaking gearheads (admittedly, under higher-than-usual voltages and abuse).

Thank you for mentioning the Nub, I was looking for something like that for a while now. Doesn't the Nub also require a set screw? It seems like it provides the same function as the hub though. I think the Nub will be beneficial for supporting the shaft, like you suggested. I'm thinking we add another member to support the end of the shaft using the method you suggest, my only question is - what kind of bearing block/bearing would fit the 6mm d-shaft that will be output from the other side of the wheel?

We will try to use 20's, but if we can't get those ordered on time, I'm thinking we go with the Matrix motors.. they seem similar to the 20s/40s.
I don't predict too much strenuous motor action..

[GeeTwo]

Originally Posted by ImMoMo

Doesn't the Nub also require a set screw?

Just looking at the web pages, it seems that the Tetrix hub has a round bore and set screw, but there are Nubs available with an actual D-bore and set screw. This means that on the D-bore Nub, the set screw is only there to keep the shaft from sliding along the axis; on a round-bore hub, it is also required to transfer torque.

[Billfred]

Originally Posted by GeeTwo

Just looking at the web pages, it seems that the Tetrix hub has a round bore and set screw, but there are Nubs available with an actual D-bore and set screw. This means that on the D-bore Nub, the set screw is only there to keep the shaft from sliding along the axis; on a round-bore hub, it is also required to transfer torque.

Gus nailed it.

As for bearings, AndyMark sells a 6mm ID bearing that could be pressed into a plate or frame member.

Matrix motors have a different bolt pattern than the Tetrix and NeveRest motors (consider the third photo here; ours uses the six M3 threaded holes on the nose), so do plan accordingly. All three of our current NeveRest motors have the same bolt pattern (the NeveRest 3.7 will be different). But yes, their 12V motor would be around the same speed as a NeveRest 40.