Why Physicists Don't Like To Talk About Friction

fm6 writes: "You would think that force required to overcome friction would be a function of the area of contact. But according to this Scientific American article, that's not true, and physicists don't have a really satisfying explanation." This is the sort of article that makes you want to go experiment with those teflon-coated disks made for moving furniture.

Comments

Bad science

[Alpha State]

I guess someone should post something serious.

This moving-crack theory is crap. I can't show it's not true, but a model of interlocking surfaces explains friction perfectly well. Consider two horizontal surfaces whose interface is a zig-zag. There is a force Fd holding these surfaces together and a horizontal force Fm on the top surface. The top surface will not move until it slides up to the peaks in the lower surface. It's quite trivial to show that the required force depends upon the degree of interlocking (the angle of the zig-zag) and the force Fd, which must be overcome to seperate the surfaces.

Is this a bad question?

[Kibo]

I think that certainly explains the static force of friction F sub s, but what of F sub k? Why should F sub K typically be so much higher than F sub s?

I suppose one might argue that a surface that experiences an F sub k might be assumed to have previously been at rest, and nestled firmly in the lowest state, or the trough if one likes, but why would this be SO much higher (typically)? Interaction between electrons at the surfaces? If there is a limited amount of interaction taking place, or the formation of weak bonds, why not view it as analogous to a crack (a very well understood phenomina)?

But back to F sub s, the static force of friction, wouldn't the surface fall foreward into the troughs of the supporting surface some of the time providing a slightly accelerating force of friction which would then turn decelerating as the atoms being supported tried to move up out of the trough against the force of gravity? Of course, that's not what we see, so it can't be the complete picture.

Why not move back to the formation of tenuious bonds between the surfaces (for a moment). If these bonds are being made occasionally, then stretched and broken, it would seem to my mind's eye that for a macro sized object F sub s would likely be a near constant (surface irregularities, pressure, whatnot would all play a part). Since the breaking of these bonds in a sence does change the surface properties, why not view it as a moving crack if it is convienent? Certainly we all except greater abstractions than this in our everyday life, if some scientists find it a helpful model is it worth belittling? Sometimes abstractions like this, reguardless of their accuracy, can be surprisingly useful. For my part, it is consistant with what I know to be true and seems to do a better job of explaining, at least for me, better than a classical speed bump theory. Your milage might vary; but so might theirs.

Re:Is this a bad question?

[caffeinated_bunsen]

If you assume that each surface is a series of circular arcs, instead of a zig-zag, then you get a similar result for static friction. If the surfaces are already moving, then they can't interlock as much as when static, and so the tangent force resulting from the normal force is reduced from that required to start from rest. But then this predicts that the sliding friction is a function of the extent of interlocking during movement, which depends on velocity. Oh well.

Re:Is this a bad question?

[superflex]

Well, it might be a bad question... if I'm correct in assuming that by F sub s you mean static force of friction, and F sub k means kinetic force of friction, then you're wrong about F sub k being higher that F sub s. The opposite is true.

The formulae for static and kinetic friction are:

F(s) (lessthan)= u(s)N *

F(k) = u(k)N

Where the u's are actually mu's (use your imagination), which are the coefficients of friction, and N is the weight of the object (mass*9.81 m/s^2)

If you look at a table of friction coefficients, you'll see that the coefficient of static friction is always less than the coefficient of kinetic friction.

As far as why static friction is always higher than kinetic friction, I always thought (IANAPhysicist) that it makes sense if you look at it on a microscopic scale. Surface roughness now looks like "mountains" sticking out of the surface of the two objects in question. When there is no relative motion between the objects, the mountains are fully interlocked, and it takes some extra force to get them unstuck. But once you get them moving, they bounce along off of each other, but they don't get fully interlocked of course, because on the scale of surface irregularities, 1 mm/sec is still pretty damn fast.

I always thought of friction coefficients as a statistical average of the "roughness" of two surfaces.

* sorry, I don't know the Unicode for a lessthan sign.

Re:Hmm, is this harder than I am thinking

[caffeinated_bunsen]

There's also the issue of the resistance of the tire material to separation at the surface. Assume that above some maximum tangent force Ft the surface of the rubber falls apart, leaving part of itself on the road surface. (note the black streaks left on the road from a car moving with locked wheels) Ft is obviously proportional to the contact area, since a larger area must tear off more rubber. If this mechanism for breaking static friction is assumed, then the friction is dependent on area.

For cases when both surfaces remain intact, friction per unit area is dependent on pressure, so total friction is (constant*force/area)*area = constant*force.