I think someone said the issue was concentricity; i.e. the MGB clutch arm
is not designed in such a way as to perfectly center the throwout bearing
on the axis of rotation, every time. For the "friction" type of bearing,
this doesn't really make any difference, which explains the
"shortcomings" of the original design in this context. But the roller
bearing assumes concentricity of rotation.
Think about it. If the axis of rotation of the roller bearing, when
applied, doesn't precisely match the center of rotation of the clutch
disc, you get an effect like a tiny bicycle crank. This would cause
vibration, and enormous side stresses on the bearing, which it might not
be designed to support for long.
Of course, if your clutch arm components and bushings were all new, you
might have something close to concentricity, and thus enjoy reasonably
long roller bearing life (as many have reported). But if your clutch arm
was anything like mine (when I last replaced my clutch), with worn-out
bushings and pins and ovaled-out holes, no two clutch pedal applications
would result in the same alignment, and you'd probably experience
catastrophic or premature failure (as has also been reported).
Anyway, I think that is a plausible explanation of both sides of this
situation...
James H. Nazarian had this to say:
>Chris,
>
>I quote British Automotive:
>
> "Before you consider installing aftermarket ball/race type throw out
>bearings,
> be sure to read the following information. If you have already installed
>this type
> of bearing, it is doomed to failure. Presently there are two styles of
>ball/race
> type bearings available:
>
> 1.- A black plastic housing type
> 2.- A steel housing type
>
> Both of these types of bearings suffer from premature "run-dry"
>conditions
>and
> their failure is directly related to being in continual contact with the
>clutch cover
> thrust pad. No.1 will self destruct in a very short time; No.2 will
>survive much
> longer and failure will not be as catastrophic."
>
>Why would a sealed, permanently lubricated roller bearing designed for
>continuous operation with load applied along the axis of rotation, run dry?
>Could it be defective? Or damaged during installation? If so, was it returned
>to and examined by the manufacturer? If so, might we see their analysis or
>reply? If Doug's right, I've been driving several time bombs for the past 15
>years. Maybe Doug will share some of his exhaustive test data and methods
>with
>us.
>
>Before I ask Mr. Moss to fly out to Ohio to help me change all these heavy
>duty
>roller bearings back to the original graphite type, I think I need to see
>some
>proof. Thanks in advance for enlightening us.
>
>Senior Doubting Thomas
>
>AKA Jim
>1960 MGA
>1974 MGB-GT
>and a pool of other octagonal delights
>
>Chris Attias wrote:
>
>> Peter C. wrote:
>> "A few quick thoughts. Roller bearing or carbon.... don't "ride" the
>> clutch. The only time that a clutch can become worn is when your foot is on
>> the pedal. If you don't touch the pedal, the clutch wont wear."
>>
>> This seems to disagree with what I understood about the design of the
>> MGB Borg & Beck clutch assembly. See Doug Jackson's write-up on
>> clutch bearings @ www site:
>> http://www.mgbmga.com/tech/index.html
>>
>> While I realize Doug has some commercial interest in his development
>> of a fix for the problem he finds with roller bearings, he does
>> exhaustive research and shares his experimentation and research
>> freely...(No commercial interest disclaimer...)
>> Chris Attias
>> '64 MGB
>> '84 Alfa Romeo GTV-6
>
>
--
Max Heim
'66 MGB GHN3L76149
If you're near Mountain View, CA,
it's the red one with the silver bootlid.
|