Sheldon Brown <[email hidden]> wrote in message news:<[email hidden]>...
Quoted message said:Peter Chisholm said:Why NOT use it? What does using it DO that you don't like?
It substitutes the rote repetition of a number somebody wrote on a piece
of paper for the mechanic's judgement which takes into account the
actual tolerances, surface conditions and lubrication condition of the
parts involved.Quoted message said:think, 'crank bolts'....
Yes, I do. If you look at the scripts, they specify the same torque
value for both sides. This is clearly an oversimplification.It is my practice to tighten the left side harder than the right side.
Here's why:When cranks come loose, it is very predominantly the left side that
comes loose. Making the left side bolt extra tight precludes this.Issues of possible overtightening are of more concern on the right
side, because this can cause chainline problems, or, in extreme cases,
chainring clearance problems.Mindlessly using some arbitrary torque number ignores these real-world
differences, as it ignores other aspects of fastener condition/applicatio
n.Torque wrenches are johnny-come-lately tools in the bike industry. I
don't think any bike shop in the world used one before the 1980s, even
though bicycle technology goes back over 100 years before that.Sheldon "Judgement, Not Rote Number Copying" Brown
Dear Peter and Sheldon,
Most fasteners on bicycles are over-engineered by
design, so they neither break off nor loosen when
tightened by hand and Kentucky windage.
After a while, an "experienced" mechanic understandably
mistakes the forgiving nature of so many nuts and bolts
for expertise.
How does the expert know the right tension? Well, it
didn't bust off, it feels pretty tight, and he's been
doing it for years--successfully, with an occasional
need to tighten things up if they turn out to be just
a little bit loose.
The expert mechanic's opinion of his skill is increased
when he fiddles with bearings. Delicate adjustment of
wheel bearings with just enough pre-load for a quick-release
is not something that a torque wrench can do well, so the
mechanic understandably feels that his skillful touch is
superior.
What he forgets is that he doesn't adjust bearings by
skillful touch. He tightens the cone nuts by hand until
he feels resistance--skillful touch. But then he tests
by spinning the wheel and finds that he needs to skillfully
loosen the nut just a bit. Another test spin reveals that
the nut could stand to be just a little tighter---there!
Would anyone want to ride a bike whose bearings had
been adjusted solely by twisting the nuts and never
checked for proper torque by actually spinning the
bearings?
The same is true of wheel-building. Who would want to
ride wheels built solely by how tight the spoke wrench
felt to the mechanic? He tightens a spoke and checks
the result by spinning the wheel until it's roughly
true. Then he tightens the spokes up until some (often
ill-defined) situation occurs or uses a tensio-meter
to check the tension. Various magic incantations follow,
but the truing, plucking, tensio-meter and rim-yield
checking are all torque-wrench-style checking. (It's
noticeable that there is no test for the popular
stress-relief process.)
Such testing is not used on most other bicycle parts
because a quick twist of the wrench is good enough.
Even brief experience is enough to let most people
tighten a nut sufficiently to stay in place, while
few nuts are so delicate that they break even when
torqued to twice the recommended value.
But no matter how proud the experienced mechanic is of
his skillful touch, he usually gives the seat a twist
and a shove after tightening the clamp--and tightens
the clamp if the seat moves. The twist and shove is no
torque wrench, but it's the same idea of testing.
One of the more amusing sights in a motorcycle shop occurs
when one expert mechanic accidentally begins going over a
machine just worked on by another expert mechanic and starts
muttering about things being too tight or too loose.
(Machines worked on by both mechanics have no problems.
They use torque wrenches where it really matters.)
Jesus Christ probably had to listen to one Roman soldier
arguing with another about whether the spikes had been
pounded in far enough, while a third reminded them that
no one had ever fallen off a crucifix and could be slapped
back up in a hurry if he did.
Torque wrenches aren't necessary for putting bicycles
together, but a cheap beam-style torque wrench can often
surprise an experienced mechanic, who may learn that
he's been over-tightening things for years (oops!),
that he's been leaving them almost dangerously loose
(double oops!), or that he's been pretty close to what
the engineers recommend (whew!)--the last being impossible
to verify without a torque wrench, a slight swallowing of
pride, and about ten seconds of practical testing.
Go ahead, spend $20 on a beam-style torque wrench, and see
how close your judgement is to what the designers thought
would be a good heave. It's what most people would expect
from a technical forum. (Luckily, the designers don't expect
as much and over-engineer the parts.)
Carl Fogel