Contact Area and its Relation to Friction?

[Qbranch]

So... the consensus is that contact area has little or nothing to do with friction? But it does?

Ok... i'll go get empirical data sometime...

-q

[Paul Copioli]

QBranch (aka Alex) wants a definitive answer and he will get it right now.

Static friction force does not depend upon surface area. Static friction force does not depend upon surface area. Static friction force does not depend upon surface area.

This assumes one major thing: The surface pressure between the two items is low enough to not cause material failure at either surface (wheel/tread or carpet).

You must make your wheel width wide enough to not rip up the carpet and not yield your rubber (at least, too much). You should design your wheel width to not fail either material. Once you have done that, the width doesn't mean squat.

Alan is correct about dynamic friction: surface area plays a bigger role.

I have posted numerous times on this and the width does not matter.

I will not argue with any of you about this. I am as certain as can be on this issue ... believe me.

[Richard Wallace]

Originally Posted by Paul Copioli

QBranch (aka Alex) wants a definitive answer and he will get it right now....

You must make your wheel width wide enough to not rip up the carpet and not yield your rubber (at least, too much). You should design your wheel width to not fail either material. Once you have done that, the width doesn't mean squat. ...

I will not argue with any of you about this. I am as certain as can be on this issue ... believe me.

Word. Thanks, Paul. Definitive answers are a good thing. (emphasis mine)

[Alan Anderson]

Originally Posted by Paul Copioli

Static friction force does not depend upon surface area. Static friction force does not depend upon surface area. Static friction force does not depend upon surface area.

You can change this slightly and avoid certain potential counterarguments:

Static friction force depends only on the coefficient of friction and the "normal force" (weight, for horizontal surfaces).



There are boundary conditions for some combinations of materials where the coefficient of friction can change based on pressure, and since pressure depends on area and force, changing the area can affect the friction force, but the static friction force still depends only on the coefficient of friction and the normal force.

[Aren_Hill]

So if width doesnt mean squat quick question. in st louis us (1625) had to face 217 and 148 at times 217 had 6 wheel drive with im guessing 1inch wide tires and 148 had 6 wheel drive with 2 inch wide ifi traction wheels. both with what i believe to be identical tread. yet we could push 217 easily and 148 we tied head on. we had a 3 speed 4 wheel swerve drive with 1.75" wheels covered in lower cof blue nitrile roughtop from mcmaster. any explanation? my next years plans already inclue 2.5"wide wheels at the moment

[Guy Davidson]

Gear ratios maybe? If you had a three-speed transmission, there's a good chance your lower gear was lower than theirs. That would result in more torque being delivered to your wheels, and hence a greater pushing power.

[Paul Copioli]

Head on, 1625 did not push 217 easily. I can show you the video evidence. However, we were easily pushed sie to side for reasons completely separate from surface area and frictional force. Our side bumpers were located such that we (inadvertantly) were giving our opponents the ability to transfer our weight to them in a side pushing match, which lowered our normal force and increased our opponents normal force.

Besides, who says our robot was optimized for max pushing force last year? We had a single speed transmission that was not optimized for pushing.

The fact that your team could push the T-Chickens last year has nothing to do with the fact that surface area has nothing to do with static friction.

[meaubry]

Please keep in mind that what is written into fomulae is often NOT directly transferrable to robot drive systems in a way that can be accurately and completely calculated.

Yes, you can make general assumptions regarding friction and the effects of one material vs. another, etc. - but, some arguements relative to traction, best drive system configuration, best wheel type and material, treads vs wheels, and other drive system decisions - are best left to experimenting and lessons learned in the real world.

Therefore, overly concerning yourself about static friction will only address one element of the problem. Robots are usually not designed for static friction.

Something I learned a long time ago relative to contact area and friction. This applies more to mechanisms that are designed to slide, not grip.
If the opposing materials are too smooth (maximizing the contact area) they will react opposite of what you would expect and want them to do. They have more difficulty sliding over each other. Sliding is accomplished easier when the contact surface is a little rough, giving up a little contact surface is productive in some cases.

As to my own experiences in drive systems relative to this question, I would have to say that our robots with more contact area produced better traction against the carpet. When comparing the robots using wheel chair wheel (smooth) vs. treaded pnumatic tires - the treaded pnuematic tires won hands down. The differences could be attributed to a combination of both different material and more surface contact as the pnuematic tire actually increases in contact area as they are pushed against due to the forces subjected to.

Will you get to a point where increasing contact area no longer makes a significant difference? Yes and No - it all depends on what you are attempting to do with it. Andy explained it well in his response. But at the same time, would the Beatty Beast have been such an immovable object without all of those file cards??

Just my thoughts - good topic,
Mike Aubry

[Joachim]

The rule that static friction does not depend upon contact area is a first-order approximation only. If a first-order rule is good enough, go ahead and use it.

[Lil' Lavery]

Contact area has no significant impact on static friction in the FIRST world. As mentioned earlier, Normal Force can be defined as Pressure multiplied by Area, but in this scenario Pressure is equal to Normal Force/Area. So your pressure will decrease at the same rate your area increases, resulting in your Normal Force remaining the same. As such, your traction will remain the same. Granted you want to have large enough wheel that the tread won't fail (nor the carpet).


As for the debate on PID traction systems, couldn't it be accomplished by comparing the data from the encoders in the drive to an accelerometer? If the acceleration of the wheels is greater than the acceleration of the robot, wouldn't it determine that they are spinning out?
I speak purely out of speculation, and I don't have any real experience developing PID systems, but it seems like that solution could be possible to me.

This is an interesting thread.

Short disclaimer: I have yet to take physics, so anything I write is a result of observation.

Anywhoo. I think I understand what people here are talking about, yet I'm still confused. If everyone says that the contact area doesn't matter, then can someone explain how this makes sense: I built two robots. Both were with the kitbot chassis. Both were driven off of Banebots transmissions. Both had two driven wheels, and two casters. One bot had two andymark kit wheels per axle for driving, and the other had only one per axle. I added extra weight on each in order to make the weight exactly equal. I wired both motors to a single battery and switch. Then I put the two drive bases head to head, set so they would drive directly into each other, on an area of FIRST carpet. When turned on, the base with 2 wheels per axle could overcome and outpush the other base every time. We did this 12 times, each time changing to a new, fresh battery.

Anyone care to help me out here? I guess that physics and math and stuff say this shouldn't work, but it did. So I'm confused.

[JesseK]

If the acceleration of the wheels is greater than the acceleration of the robot, wouldn't it determine that they are spinning out?

This is valid if you're moving in a single direction. However, consider from a standstill when you're trying to turn. You'd have to use a gyro to make that determination.

Thursday night another 1885 mentor was under the impression that traction control isn't needed at all. I see the greatest advantage during climbing ramps; he says our "drive straight" code takes care of that since it uses the gyro & encoders to keep us going in a straight line. Not sure how to argue it with him that it's hard to not over-compensate with the gyro technique, but perhaps he's correct and it's enough for the time being. No student is advanced enough in programming yet to be able to take interest in it, but maybe we can throw it in there late in the season.

Low-gear + high-torque situations are the next advantageous place to use it.

[lukevanoort]

Originally Posted by 114ManualLabor

Anyone care to help me out here? I guess that physics and math and stuff say this shouldn't work, but it did. So I'm confused.

Where did you put the weights to equalize the mass of the two 'bots? I ask because if there was more weight directly over the double-wheel's driven wheels it would behave as you describe.

Originally Posted by lukevanoort

Where did you put the weights to equalize the mass of the two 'bots? I ask because if there was more weight directly over the double-wheel's driven wheels it would behave as you describe.

Both had the weight added directly over the axles.


I think for the upcoming season, I'm sold on tracks. I get this from watching an outback system push anything it came against across the field. There might not be an easy explanation, but I've seen tracks outpush wheels every time.

[Guy Davidson]

Originally Posted by 114ManualLabor

but I've seen tracks outpush wheels every time.

Are you sure it wasn't just different gear ratios? If both are driven by the same motors, the tracks having a lower gear-ratio in their gearbox would make a big difference.