Creating a New Model for Friction Based on Experimental Results

[Brewer]

What would be the best way to come up with a new model for friction based on my experimental results?

I have been finding the angle at which a block slides, or topples when the angle it is inclined at is increased. This has been done with blocks that are of different heights (towers of building blocks), and the experimental results don't fit the model for friction.

The initial model is coulombs model
(i.e. F=uR)

The model has been used to work out the angle for sliding and toppling with different amounts of blocks, so I decided to take logs of both the number of blocks and the angle, but then couldn't see how to put this back into a new model.

Any ideas how to look at a new model based my experimental results would be appreciated.

Thanks

 

[Astronuc]

Friction depends on the quality (e.g. roughness) of the interacting surfaces.

If the coefficient of friction does not match the accepted values, and there is always some variation, then perhaps the coefficients are actually different, or one or both surfaces are changing.

If the friction is increasing then one or both of the surfaces may be becoming rougher. If friction is decreasing, then one of both surface may be becoming smoother.

Also, if a lubricant is inadvertently added, then one is affecting the interaction of the two surfaces. Lubricant reduce the shear forces betweent two surfaces.

 

[Brewer]

That kind of makes sense. I have found the coefficient of friction as being the tangent of the angle. My textbook tells me that. For use in my original model I had to use a value for the coeffiecient (henceforth called u). I took this u to be an average value that I had worked out for my experiment.

However the model suggests that u should not change depending on the size of the mass as these cancel out
because F = mg sin x, and R = mg cos x where x is the angle

therefore F = uR
mg sin x = u * mg cos x
which cancels to
tan x = u

Its just bugging me about how to create a new model to fit the experimental results that I have!

 

[rayjohn01]

At least part of the problem is the surface that you happen to be experimenting with.
Friction is poorly undertood -- because it is SO dependant upon the precise nature of the surface which is rarely ideal,
Even for a rough surface it only take 3 points to theoretically to hold them apart -- BUT if you move it , it will be a different 3 points -- with their own properties.
At any particular point you deal with the exact atomic surface , this may include roughness , the environmental gas , surface impurities like skin oil - you name it it is probably there -- so averages and statistics are in order -- do not expect precise results .
Note that if the surfaces were totally clean and flat then they would stick and you could not get them apart.
Ray