Team 1296 Mecanum Bump Test

We tested our mecanum wheels over a 4 X 4 to see how they worked. So far it looks good. Let us know what you think. http://youtu.be/jl8pHpCjQtY

A few things you’ll want to change:

-ballast the robot to full weight
-a ‘real’ barrier overall dimensions at 4"x6"
-metallic bump material with radiused corners (I think this will really impact the wheels’ traction)

Otherwise it looks good, plenty of ground clearance!

Thanks for the input. We are thinking the extra weight will help but are worried about the “real” radiused metal bump. The bump being an extra 2" wide should not be a problem though as the opening in the middle of the robot is 21". The bot will have a nice ramp lowering device but we wanted to make sure it goes over the bump just in case.

We had quit a few folks say that it would not go over and some who thought it too dangerous (including me) but those super-duty mecanums from AM are really tough and the angled front edge helps - so much for the common “mecanums can’t go over the bump” theory!

Im interested to see it with all the extra weight, because its going to change the force vectors on the angled wheel mounts.

True, but it will increase traction <fingers crossed>!

I’m looking forward to more weight too, should make it easier :slight_smile: more traction!

Polymers as traction surfaces generally lose traction efficiency as normal force is increased. Search CD or any tire test data if you don’t believe me. I would anticipate some specific traction (tractive force per unit weight) loss as you add on weight. You will have more total tractive force, but relatively more mass to move.

I’d be worried about repeatedly ramming my mecanum wheels into the bump… The AM HD’s are quite nice, but they’re not indestructible!

The angled surface on the wheel assembly hits the bump first - it helps a little. And we don’t plan to make a living going back and forth, just did not want to NOT be able to.

Why would we not believe you?

Depends on the nature of BOTH surfaces of course but I’m sure you are basically correct and we have looked at the data. If we double the mass I expect a 25% reduction in the friction coefficient for urethane rubber. So I’m thinking the effect is (at worst) linear for the small increase in weight (and probably not that bad) and the bot will push its way over the bump at full weight - we’ll see soon.

Right on – I couldn’t see that on the video!

Get to it Keith!

What I am worried about is that the wheels with lose its strength after hitting the steel bump a million times. Also you got over a wooden bump do you think there might be a difference between the steel and the wooden because they are two completely different materials?

We are thinking the wheels will hold up but the metal bump is worrisome.

Man that robot looks sexy!!

It was meant as the ‘royal you’ as every time this topic of polymer load/pressure sensitivity comes up someone inevitably goes: “well that’s not what my physics teacher says!” and I might be a little tired of making the same points over and over :slight_smile:

As a physics teacher, I use the disclaimer that the line “…but it’s really more complicated than that!” is implicit at the end of every explanation.

While that may be true for “frictional traction”, it’s not necessarily true on carpet where a considerable amount of the tractive force is actually a normal force between projections on the wheels and fibers on the carpet. The reason brekoflex type belting and some mechanums get such good traction is the amount of embedded overlap when edges on the wheel surface dig down into the carpet, which is a function of weight but more importantly an inverse function of contact area. Our previous testing and data has shown this.

Some teams have run experiments showing that wider wheels, i.e. ones with more ground contact, get better grip. A big part of this is that the ‘embedded overlap’ interface you talk about becomes saturated at a certain contact pressure (the carpet and tread have finite depth) and as pressure increases above this threshold the coefficient of friction drops off. Even before the interface is totally saturated the coefficient of friction generally drops off slightly with load because the carpet/tread presses together more firmly, making it more difficult for the tread to embed further.

This analysis is based on my knowledge of tire-road interfaces and what I’ve learned from reading many CD threads on the subject.

One thread with related discussion is here: http://www.chiefdelphi.com/forums/showthread.php?t=98664

Is that a normal 3.5" thick 4x4? Or is it a special 4" thick 4x4? :slight_smile:

If I were you, I’d build a short section of “real” barrier, put it on carpet, and try it. You might find that it’s a bit harder on the robot than it looks in the video with the 4x4.