Hexagonal Chassis + 6 CIM swerve drive

[asid61]

If you are using coaxial swerves, I can't see a reason why not to simply use the drive sprockets to power two wheels at once, then use separate turning motors for each wheel. You can even technically use a single 6 cim gearbox to power any number of modules.

[Tyler2517]

Originally Posted by asid61

If you are using coaxial swerves, I can't see a reason why not to simply use the drive sprockets to power two wheels at once, then use separate turning motors for each wheel. You can even technically use a single 6 cim gearbox to power any number of modules.

If you do this you cant have a true 3D swerve drive (X,Y,rotation). Changing the power to each wheel is how you preform holonomic maneuvers.

[asid61]

Originally Posted by Tyler2517

If you do this you cant have a true 3D swerve drive (X,Y,rotation). Changing the power to each wheel is how you preform holonomic maneuvers.

Oh really? That's annoying. Thank you for the information.

[Tyler2517]

I think at that point it would be a crab drive or something close to it i might have some hidden benefits but a true 4 wheel independent should out perform it. (or 3 wheel independent)

[Tyler2517]

So random question
What happens if you have say a 8 wheel swerve and you powered the wheels in sets of 2. Could you code them to act like any 4 wheel swerve?

[asid61]

Originally Posted by Tyler2517

So random question
What happens if you have say a 8 wheel swerve and you powered the wheels in sets of 2. Could you code them to act like any 4 wheel swerve?

If you have to change speeds for holonomic motion, I don't think you could do that even if it's a multiple of 4.
Not to mention 8 coaxial swerve modules plus the associated turning motors/ gearboxes would be massive. That would weight at least 18bs for the coaxials, 2lb per shifter for 8lbs more, and 8 turning motors at 0.5lbs each plus all the chain and sprockets. So 30 lbs+ just for the swerve modules.
Livable, but I'm more worried about space even with coaxials.

[TheHolyHades1]

Originally Posted by Tyler2517

So random question
What happens if you have say a 8 wheel swerve and you powered the wheels in sets of 2. Could you code them to act like any 4 wheel swerve?

I don't think so. It is an interesting combination, but if I'm not mistaken swerve works on having the ability to control rotation about a point and not a line. You need to power each different point in the rotation at a different speed, and without some sort of differential that wouldn't be possible in swerve.

[Dunngeon]

Originally Posted by Ether

To the extent that the unpowered omnis are supporting weight, they are robbing traction from the driven wheels.

Why not raise the omni's axel center up an 1/8th (maybe less? 3/32?). This would allow the 3 powered wheels to have optimal traction, while still having good stability when being pushed. At the expense of traction.

[Ether]

Originally Posted by Tyler2517

What happens if you have say a 8 wheel swerve and you powered the wheels in sets of 2. Could you code them to act like any 4 wheel swerve?

You could code them like a 4 wheel swerve, be they wouldn't act like a 4 wheel swerve because it wouldn't be kinematically correct. When turning, the wheels would scrub.

[Oblarg]

To anyone having a hard time understanding why you need to power and turn wheels individually for a true swerve:

Take a rectangle. Move it in some manner that involves both a translation and a rotation. Pick some set of arbitrary points (say the corners), and imagine those are your wheels. It should be easy to see that for any generalized movement, those points are all moving in different directions and at different speeds.

[Ether]

Originally Posted by Oblarg

To anyone having a hard time understanding why you need to power and turn wheels individually for a true swerve:

Take a rectangle. Move it in some manner that involves both a translation and a rotation. Pick some set of arbitrary points (say the corners), and imagine those are your wheels. It should be easy to see that for any generalized movement, those points are all moving in different directions and at different speeds.

Correct.

For anyone interested in the math, here are the equations:

http://www.chiefdelphi.com/media/papers/download/3236

[Dragonking]

You could try using 4 cims and 4 mini cims for 4 modules.

[GeeTwo]

Here are 10 pictures of a "semi-swerve" drive I built at about 1/3 FRC scale, using servos to steer and independent motors to drive. The last 8 pictures show how you have to orient the wheels to move various directions, including pivoting around selected points. (These are views from below).

[Ether]

swerve inverse kinematics math quiz:

http://www.chiefdelphi.com/forums/sh...d.php?t=129185