Traction Questions

[themccannman]

Originally Posted by glennword

IIRC, I learned in physics that the force of friction is dependent upon the coefficient of friction and the normal force, and surface area has nothing to do with it. Given, that equation only holds true for hard surfaces, not "squishy", yielding surfaces like carpet and roughtop tread.

I think you've answered your own question. Classic newtonian friction calculation doesn't work for irregular, soft, non smooth surfaces. The best way to figure this stuff out is test it yourself with tread and carpet. Try driving your robot into a wall and measuring the motor's current draw with different size wheels, this should give you a pretty accurate predictive model of the relationship between wheel width and effective traction.

[Ragingmain]

At the risk of being alienated, since I am new, I will offer my empirical data for this discussion. My time as a racer, motorcycles.

Look at any funny car, drag car, drag motorcycle. They ALL have WIDE tires. Even with rubber compounds being different they all have one thing in common....WIDTH. They would not be able to get the TRACTION needed to propel their vehicles if they were not WIDE. Skinny tires would not do it. If skinny tires worked they would use them as they are cheaper and weigh less. For these type of sports, motor sports, it is all about power to weight ratio. The problem is how to get all that power to as large as surface area, TRACTION, as possible to get down the track the fastest. The same applies for tractor pulls or 4wd trucks. You don't see the skinny tires of a tractor on the back for a reason. It is reasonable to say the sport of robotics, in this case, is no different. We have a fixed weight and yes the force per tire goes down with more tires but that is also dependent on how the weight is distributed.

Of course turning is a different issue all together.

Flame on.

[RunawayEngineer]

It seems like a monthly thing that this problem shatters the illusions of a physics student.
Increasing contact with the carpet does increase traction in certain scenarios. But until someone does an empirical study across several FRC-relevant scenarios, we will all be left to our own intuitions and experience as to what the nature of the relationship is.

[Caleb Sykes]

Originally Posted by Ragingmain

At the risk of being alienated, since I am new, I will offer my empirical data for this discussion. My time as a racer, motorcycles.

Look at any funny car, drag car, drag motorcycle. They ALL have WIDE tires. Even with rubber compounds being different they all have one thing in common....WIDTH. They would not be able to get the TRACTION needed to propel their vehicles if they were not WIDE. Skinny tires would not do it. If skinny tires worked they would use them as they are cheaper and weigh less. For these type of sports, motor sports, it is all about power to weight ratio. The problem is how to get all that power to as large as surface area, TRACTION, as possible to get down the track the fastest. The same applies for tractor pulls or 4wd trucks. You don't see the skinny tires of a tractor on the back for a reason. It is reasonable to say the sport of robotics, in this case, is no different. We have a fixed weight and yes the force per tire goes down with more tires but that is also dependent on how the weight is distributed.

Of course turning is a different issue all together.

Flame on.

You seem to be trying to pick a fight, so my suggestion to you would be to make sure your preconceptions are well grounded before you bury yourself in a hole. I think Mark McLeod's signature might be relevant here:

deduction is limited by knowledge, and knowledge is limited by preconceptions

With that in mind, Are you confident that the sole reason these vehicles use large wheels is to gain traction? Because that is what your post implies.

[Mark McLeod]

I like these answers as to why dragsters use wide tires.

[Ether]

Originally Posted by Mark McLeod

I like these answers as to why dragsters use wide tires.

I like Erck's answer.

Smith and Peterson, not so much.

[Ragingmain]

Not trying to pick a fight at all. Just bringing in a discussion about a subject which there seems to be much passion.

I have a mechanical engineering degree for starters so I have been through all the proper training/schooling for this discussion.

I also know that racers of any kind, or even car manufactures who make high powered vehicles, always spec their tires to be wide to allow for better traction. Also any type of belt driven sheaves on high powered machines are wide for two main reasons. Strength and the larger the surface area the less likely to slip. The diameter of the sheave has a direct correlation to this as well.

I agree about the responses to the question. The first two are silly.

Since our robots do not do burnouts prior to a match the section about negative static friction coefficeint does not apply. This section also speaks to the type of material being used. In our application this is not a concern as well. We dont leave rubber down the playing field

A couple of things I got from the posted article

Friction is surface-area independent in only a few ideal examples. The real world is more complicated. Especially for tires that are made of rubber

"You want to choose a width, height, and tire compound that gives the best friction for the duration of the race."

"Increasing tire diameter and tire width increases the contact area."

It does speak to downward force due to a wing attached yet this does not apply either as we have weight limits and cannot go fast enough in the limited space even if we did have a mechanism to give downward force.

So the only parts of this article which directly apply to us is the width, height, compound and contact area.

I know I could never win a race if I had a skinny tire on my motorcycle.
If even everything else were the same, the burnout, the compound, the track conditions, same powered vehicles. The one with the RIGHT width tire will always win.

Numerous real world examples are before us as to why width, using ruber wheels, does in fact increase traction.

I guess one more way I can put it is this. If you had to move a piece of 4x8 plywood and could not carry it and the only two options you had were to
1. Lay it on the ground flat and try to push it?
or
2. Lay it on its edge and push it?

Same amount of weight. one has a bigger contact patch then the other. Which one will be easiet to move?

[Deke]

Originally Posted by Ragingmain

I guess one more way I can put it is this. If you had to move a piece of 4x8 plywood and could not carry it and the only two options you had were to
1. Lay it on the ground flat and try to push it?
or
2. Lay it on its edge and push it?

I would choose #1, the edge bites in the ground when try to move it. The only thing that is harder about #1 is having to get low on the ground, but I imagine trying to move it on a tabletop since the height is the same. It slides easier on its flat side in my experience, but that is subjective.

[Jared]

Originally Posted by Ragingmain

I guess one more way I can put it is this. If you had to move a piece of 4x8 plywood and could not carry it and the only two options you had were to
1. Lay it on the ground flat and try to push it?
or
2. Lay it on its edge and push it?

Same amount of weight. one has a bigger contact patch then the other. Which one will be easiet to move?

If I'm pushing on a carpet, number 2 is much easier than number one. If I push on a smooth floor, then the second may be easier. It depends on the surface you're pushing on.

[sanddrag]

I've always wondered about the traction of a larger diameter wheel compared to a smaller diameter wheel of the same width. I'd speculate it is more, since the curvature of the wheel approaches a straight line as the diameter approaches infinity. On squishy surfaces such as rubber and carpet, I can definitely see this making a difference.

It would be cool if AndyMark and VEX would step up to the plate to put this debate to rest once and for all, with some real scientific testing and published results

In the end though, the real limiting factor in pushing is the internal resistance of the battery.

[T^2]

Originally Posted by sanddrag

I've always wondered about the traction of a larger diameter wheel compared to a smaller diameter wheel of the same width. I'd speculate it is more, since the curvature of the wheel approaches a straight line as the diameter approaches infinity. On squishy surfaces such as rubber and carpet, I can definitely see this making a difference.

I'd speculate it might be less, because a tighter curvature on the wheel will spread open the geometry of the tread surface, allowing more carpet between them.

[themccannman]

Originally Posted by Jared

If I'm pushing on a carpet, number 2 is much easier than number one. If I push on a smooth floor, then the second may be easier. It depends on the surface you're pushing on.

Precisely, it depends on the surface, hence why modeling this can get complicated, different surfaces interact in unique ways.

Here's some reading that went way over my head that may help understand friction between rough surfaces: https://workspace.imperial.ac.uk/med...20surfaces.pdf

Originally Posted by sanddrag

It would be cool if AndyMark and VEX would step up to the plate to put this debate to rest once and for all, with some real scientific testing and published results.

Why make them do it? Most teams have the resources to test this ourselves. Perhaps we'll do a study of this and publish our results.

Originally Posted by sanddrag

I've always wondered about the traction of a larger diameter wheel compared to a smaller diameter wheel of the same width. I'd speculate it is more, since the curvature of the wheel approaches a straight line as the diameter approaches infinity.

You can simulate this without buying huge wheels, just use a pneumatic wheel and test it at varying levels of inflation, the under inflated wheel will likely have better traction in accordance with your theory.

[sanddrag]

Originally Posted by themccannman

Why make them do it? Most teams have the resources to test this ourselves. Perhaps we'll do a study of this and publish our results.

Because aside from banebots and colsons, they have all the wheels.

[EricH]

Originally Posted by themccannman

Why make them do it? Most teams have the resources to test this ourselves. Perhaps we'll do a study of this and publish our results.

Cyber Blue tried that about 3-4 years ago, IIRC. I don't recall seeing any results, though I do think I heard a rumor about "'interesting' data" or something of that nature.

[Karthik]

Some testing from 2003:

http://www.chiefdelphi.com/media/papers/1381
http://www.chiefdelphi.com/media/papers/1382