On 07/02/2018 16:08, newshound wrote:
On 07/02/2018 02:12, Roger Hayter wrote:
Brian Reay wrote:

On 07/02/2018 00:22, Roger Hayter wrote:
Dave W wrote:

On 06/02/2018 21:42, Chris wrote:

Hi,

Say I do a nut or bolt up to (for example) 120ft/lbs as checked by a
decent torque wrench in calibration, can that same nut/bolt be
undone by
the same amount of force, or does it take more - and if so, how
much more
and why?

I would guess more, because when doing it up it must be sliding, but
when undoing you have to overcome the coefficient of static friction,
which is always more than the coefficient of dynamic friction.

An exception, in the case of some cylinder head bolts, is the use of
bolts torqued beyond their elasic limit.


The idea of those bolts is simply to ensure a more predictable clamping
force than can be obtained by measuring the torque applied. A torque
wrench doesn't measure the clamping force, at least not directly. Other
factors, dirt etc, can cause higher/lower torques to be applied by the
wrench to achieve the same clamping force.

The bolts, at least when new, are designed to 'go over' they limit
reliably. At that point the clamping force should be as expected.

That isn't ex Uni, it was explained to me by a tank mechanic. (Not the
water kind.) He was working on an M1A1 engine.

They're still very probably going to need less torque to undo than they
did to do up.

I disagree. They are torqued up until they yield, but they will work
harden so that when you stop, they are a bit stronger than when they
were "new". When you undo them, you still have static friction to
overcome, plus any contribution from corrosion or other mechanisms which
increase the adhesion between the surfaces.


Roger hasn't considered the 'stress / strain' curve for, in this case, a
bolt being tightened.

The stress (tension/load), which acts along the length and provides the
clamping force- doesn't suddenly change in gradient at the elastic limit
(Yeild point). It changes, but only slightly. The bolt is then in the
plastic range, it won't return to its original length if undone. But the
stress either side of the Yield Point doesn't change dramatically. If it
did, the idea of deformable bolts would be flawed. If you continue to
tighten, then you will reach a peak on the curve, the Ultimate Tensile
Strength. Go further, and the stress decrease but, oddly, you will
eventually reach the Fracture Point where, unsurprisingly, the bolt snaps.