pic: 6 CIM swerve CAD from 2471

[Bryce2471]

[cdm-description=photo]40806[/cdm-description]

[Andrew Lawrence]

The pocketing in the tubing loses a lot of strength that could cause the module (and by extension your frame) to bend. If you want to lighten the module, mill down the tubing walls from 1/8" to 1/16". You'll still lose weight without losing the structural integrity of the solid tubing.

[Bryce2471]

The pocketing in the tubing loses a lot of strength that could cause the module (and by extension your frame) to bend. If you want to lighten the module, mill down the tubing walls from 1/8" to 1/16". You'll still lose weight without losing the structural integrity of the solid tubing.

Great idea! I'll do some stress analysis on this, and see where it stands. Although I doubt it will be stronger in the directions that I'm specifically worried about for those side walls. considering the way that it would be integrated with a frame.

[kk052]

hmm how will this cope if one wheel needs to spin faster than another?

[cadandcookies]

hmm how will this cope if one wheel needs to spin faster than another?

Do forgive me if this is a basic question, as I haven't dealt with a swerve drive since 2011 and my memory of that is a bit hazy, especially in terms of controls, but in what situation would this be necessary with a swerve? Doesn't the ability to turn the modules negate the need for moving one wheel (or set of wheels) faster than the other, as is necessary with a traditional skid-steer?

[kk052]

Do forgive me if this is a basic question, as I haven't dealt with a swerve drive since 2011 and my memory of that is a bit hazy, especially in terms of controls, but in what situation would this be necessary with a swerve? Doesn't the ability to turn the modules negate the need for moving one wheel (or set of wheels) faster than the other, as is necessary with a traditional skid-steer?
if you want to pivot around a specific pint you need to move the wheels close to that piont at a different speed than the wheels farther, however you may be able to get away with basic movements by just direction, but the wheels would skid a lot.

[EricH]

if you want to pivot around a specific pint you need to move the wheels close to that piont at a different speed than the wheels farther, however you may be able to get away with basic movements by just direction, but the wheels would skid a lot.

If there are two of these, you could do that between the sides fairly easily.


But I can't think of many cases where a swerve would need to pivot around a specific point (that is, center of rotation) that isn't the center of the robot. Depending, of course, on the swerve, and what's on top of it. There have been some very good swerve drives that could not pivot at all. (They had other means of getting aligned with targets and all that--or not, as the case may be.)

[kk052]

But I can't think of many cases where a swerve would need to pivot around a specific point (that is, center of rotation) that isn't the center of the robot. Depending, of course, on the swerve, and what's on top of it. There have been some very good swerve drives that could not pivot at all. (They had other means of getting aligned with targets and all that--or not, as the case may be.)

we have used a off center pivot on our code before, (it would rotate around the two front corners dependent on which button is pushed) witch was insanely useful to pivot the robot around other robots, you could also use an off center pivot to "lock" the distance on to whatever goal the game has. off center pivot's has some uses, but not needed, just helpfull.

[cadandcookies]

if you want to pivot around a specific pint you need to move the wheels close to that piont at a different speed than the wheels farther, however you may be able to get away with basic movements by just direction, but the wheels would skid a lot.

Thank you. Now I remember 1717's swerve in 2012 and the pivot maneuvers they had. I think the usefulness of pivots would be highly variable on the game though.

[asid61]

What's the weight with what is shown in the picture?
Where are the turning encoders going?

Cool design. I likw the integration with the frame a lot. I would be a little worried about collisions, but you would need to test on that.

[Bryce2471]

What's the weight with what is shown in the picture?
Where are the turning encoders going?

Cool design. I likw the integration with the frame a lot. I would be a little worried about collisions, but you would need to test on that.

The weight for everything in the picture, should be about 15.5 lbs.
The encoders are a little hard to see in the picture, but they are there. they are mounted to the top of the upper gearbox frame, facing downward.
You're right that I will need to test this design to make sure that it holds up, but i'm pretty sure that collisions will be a non-issue.

[BBray_T1296]

It is hard to see, but I am concerned that the belt may get caught between the gear and the middle CIM pinion and get chewed up. Is there a flange on the underside of those pulleys? or is there some other way to insure the belt does not ride a little too low?

[Bryce2471]

It is hard to see, but I am concerned that the belt may get caught between the gear and the middle CIM pinion and get chewed up. Is there a flange on the underside of those pulleys? or is there some other way to insure the belt does not ride a little too low?

You're concern is warranted, and in the CAD there are some washers that hang over either side of those pulleys to keep the belt in place. Although, I'll admit that if I were to try and build this design, I would replace all those pulleys with fanged versions. I CADed this in a very short amount of time, when I didn't have good internet access, so I used pulleys that T already had CAD for.

Assuming that the robots center of rotation is somewhere along its mid plane. (I've found a reason of it to be anywhere else) Then a robot with this drive base would be able to translate without any skidding, or rotate without any skidding, but it would have a limited amount of skidding when rotating and translating at the same time. I say "limited", because the module is only 20" long and the other side length of the robot would most likely be about 35".