Date: Mon, 13 Dec 1999 10:57:07 -0500
Subject: Clutches
From: Chris L Johnson 

> 1) Buffing, sanding, or knurling the plates of a problematic clutch 
> probably gives about as much improvement as simply taking the clutch
plates 
> out, cleaning them, and shuffling their order. On the fibre plates,
there 
> is too little material to feel safe about much sanding. The metal
plates 
> are so thin that I'm afraid of do-it-yourself dimpling or knurling
because of 
> the possiblilty of distorting them or leaving tiny burrs that will 
> acellerate the wear of the fibre plates. Mild sanding may help, but as
I described, 
> it's effects last no longer than simply taking the plates out and
putting 
> them back in. 
> JET

I'm no tribologist (experts in friction), but I have learned a thing or
two over the years about surface finishes and their effect on mechanics.

With industrial power tools, I found out rotary vane cylinder walls that
were too smooth (less that 16 micro-inches rms) would wear at an
accelerated rate.  Molecular adhesion was suspected.

With 4-cycle engines, super smooth cylinder bores lead to premature
failure because the wiper ring can clean off to much oil each stroke. 
Same priciple applies to 2-strokes.

In prosthetics, I have found that smooth surfaces between elastomers and
plastic parts leads to excessive wear and a propensity for noise.

In trials, I have found smooth disc brakes and glazed pads, or the same
for drum brakes, leads to reduced braking power and more noise.  

Anyway, there are a number of examples where really smooth is not good,
regardless of whether or not the parts have oil or just air in between.

I'm guessing that what happens to oil bath clutches with worn-smooth
surfaces is that, being under spring pressure, if there are no micro
spaces that trap oil, there will be some molecular adhesion between the
plates.  When you scratch the plates or dimple them, bead blast them,
etc., oil gets trapped which floats the plates ever so slightly.  Oil
molecules are conceptually probably like little groups of balls that will
roll between surfaces.  The float would then reduce both adhesion
(sticking and drag) and smooth engagement.

In your e-mail, JET, you said you sanded the metal plates with 400 grit
and the beneficial effect lasted a few weeks.  The reason, I think, is
the scratches were not deep enough.  I've done some checking into this
and found 100 to 120 to be a better grit size for metal plates.  Better
to start out a bit rough and break into the optimum than to lose the
effect in weeks.

Your comment on fiber plates being thin....  They are not cork, per se,
but with many euro bikes at least, are 
sintered metal plus other suff, just like brake pads.  The more metal,
and thinner they will be.  These are usually grainy, so a slight
de-glazing is all they'd probably need.

You also said you expected the Montesa clutch to be problem free because
it was Japanese.  I too, expected that.  The barking clutch and oil
sensitive problems, which I'm sure they've worked out, were indeed a
surprise to me.  The Euros have evolved from generally horrid clutches to
great clutches.  It was just a matter of time.  I remember my KTM
clutches form the mid `80s.  Yuk!  Poor tolerance control and design. 
Way behind the Japanese at that time.  The early Gas-Gas clutches were
drag meisters.  Seems today they are all pretty good.  Bulaco got it
right first time.  Beta has managed to provide an excellent clutch after
many years of grabby performance.  The `99 I rode was nice.

Chris Johnson
Scorpa 250 EZ Fun/Gas-Gas 160/Fantic 305/Yamaha Virago 1,100 & XV920RH
Euro
and now vintage `74 TY 250
chris "at" college-park.com as Director of Engineering at College Park
see www.college-park.com and papazit "at" juno.com (personal address)