Optimal Wheel Placement for 4-wheel drive

Given a generic “box” robot being driven by 4 CIMs on 4 standard, rubber tread wheels, what are the optimal positions for the 4 wheels to be in to allow for maximum driving power while still being able to easily turn in place? Or could someone point me to a set of equations/graphs to determine the optimal distance?

Thanks for any help!

See this whitepaper:

If you’re really set on doing 4 wheel drive, you can try replacing two wheels (front or back pairs) with omni wheels. It makes turning very easy, but it also means that you’ll turn if your hit from the side a bit easier. The only real way that 4 wheel drive robots with all traction wheels can work is with wide robots where the front and back wheels are close together. Even then it’s not all that great (Except for Lunacy. Then it really didn’t matter where you had the wheels :p)

It definitely mattered where you had the wheels…

Are each of the 4 wheels independently driven or are the front and rear wheels on each side chained (or belted) together?

It makes a difference in the analysis of turning forces.

If the wheels are independently driven with equal torque, a simple model with an analytical solution is possible.

If the wheels are chained, it can be solved using nonlinear constrained optimization. This can be done in Excel or a CAS such as Maxima, Octave, SciLab, Matlab, etc.

In the case of independently driven wheels with equal coefficient of friction in all directions and center of mass located at center of geometry, a very simplified analysis is possible. PDF.

Same as Scenario2 above, but the center of mass is located aft of the center of geometry. PDF.

Same as Scenario3 above, but the coefficient of friction is different in forward/aft vs sideways directions. PDF.

4-wheel skid-steer with front and rear wheels chained together on each side. Maximum coefficient of friction occurs in the Y direction. Minimum coefficient of static friction occurs in the X direction. For any other direction, elliptical interpolation between uy and ux is used to compute the effective static coefficient in that direction. The static coefficients for the front wheels are not necessarily the same as for the rear. Center of Mass is located aft of Center of Geometry. An analytic solution for the free-body static force diagram is not possible, so contrained optimization is used. PDF. . . Excel solver. . . Maxima solver. . . Solver help file.

Due to changes made by the web host, many of the links in the above post are now broken.

So I collected everything together in one place and posted it here.

Hi All:
In the past 2 years my team has built a 4 wheel, wide robot. It’s been great. However this year we want to build a long robot to support the weight of the totes. Since we won’t be pushed, I’m considering using 4 or 8 Omni wheels on a long robot.

I’ve searched for threads on Omni, H-drives and Slide drives. Most of these robots were squarish.
Does anyone have drive experience with all Omni’s on a long tank drive robot?
Do we think it will turn well with the KoP motors and transmissions if all wheels are at the same height?
Will robot CG fore/aft placement have a greater affect on 4 or 8 wheel Omni drives than scrub wheels on a tank drive?

Thanks for your help.

I’m not fully following your question. Are you looking to use the Omni wheels to achieve holonomic motion? Or just to achieve quicker turning in a scrub steering system?

Both. First want to be able to turn. Once we can make that happen then get the robot to move sideways.