Contact Area and its Relation to Friction?

[Andy Baker]

Originally Posted by Al Skierkiewicz

Q,
Andy Baker or Raul Olivera will have a good handle on this question. You may want to PM them or use the email option directly in case they haven't seen this thread.

In 2003, we TechnoKats did a bunch of friction tests to compare the coefficient of friction between various types and sizes of rubber on the HDPE ramp. We also compared the forces required to push a robot on carpet depending on the type of drive base it had (wheels vs treads).

We found that there is a very, very slight (under 5%) advantage to having a larger surface area on the plastic. This could have been due to a number of things... but each time we pulled on a larger area sample, it took a bit more force to move our object.

We also found this same slight advantage (about 5%) when comparing treads to wheels on carpet. A treaded robot was a bit harder to move.*

In both cases, there is not an ideal flat-to-flat surface interaction. I believe that the mechanical interaction between one surface and the other creates this slight advantage for larger surfaces. It's easy to see between wheels and carpet, and it's at a smaller scale between rubber and HDPE.

* - This test was enough to prove to me that treads on a FIRST robot were NOT worth the effort. We did treads in 99, 01, 02, 03, but not ever since. We were very surprised at the results and did not see the 5% advantage justify the effort to do treads. Even though there is this 5% increase in the friction coefficient, a track system has more efficiency losses (maybe as high as 5% more, to offset the friction advantages) and uses more hardware to make the system weigh more.

I'll contend that a wheeled FRC robot will push with as much force (within 3%) as a tank tread robot, as long as some conditions are met:
1. The robots weigh the same
2. The robots have similar Cg locations
3. The tread material for the wheels and treads are the same
4. There is a PID control system for the wheels (and the treads, to make it fair) so they don't slip

All of these friction comparisons above were between STATIC situations. During a static comparison (when the wheels or treads are not spinning or moving), then the friction is very close. I believe that in FIRST, a wheeled robot with good treads on the wheels will hold it's ground very well until the wheels start spinning and DYNAMIC traction starts to come into play. Possibly the reason why treaded robots push around wheeled robots at times in FIRST is because wheeled robots get into dynamic friction situations by not having a traction limiting program, like PID control.

This is definitely an area where more testing is needed. I am sure that others out there have opinions on this, and I am eager to see what people say.

Andy B.