I have noted that it occurs if a black-oxide steel alloy screw has been installed for awhile. I seem to remember that tightening then quickly removing a steel alloy screw from aluminium alloy threads does not have nearly the break-free torque. Maybe the steel removes the aluminium oxide which reforms after awhile? There is a characteristic "smoke" and odour when the screw / bolt breaks free.

Overcoming the static friction?

Cylinder head bolts are usually in tension, they stretch which is why you don't use them again

Roy

ALL bolts are in tension and stretch; that's how they work! Really critical bolts – con rod bolts for example – are tightened by measuring how much they have stretched

You're thinking of torque-to-yield bolts, that are deliberately tightened to near their failure point which means they shouldn't be reused.

Why not? Seems to me that tightening a nut is trying to draw the bolt through the hole, and tightening a bolt into a static thread is trying to force the head in. The only thing that stops it unwinding from its own elastic tension is the frictional force between the thread flanks that is augmented by that tension to exceed the rotational vector derived from the helix angle.

So I'd think that any bolt that's in tight enough to resist vibration is in tension. Even if it's loose and the nut's just hanging on the thread, it's in slight tension – come to that its own weight can put it in tension.

Well I have seen cases where bolts are used in shear. The bolt is retained by a castle nut which is tightened just for the purpose of not falling off and secured by a cotterpin or tab washer. All the load the bolt carries is in shear not tension in these cases I mention.

That's common on cable linkages. But the cotter pin is doing the retaining, not the nut which is more of an adjustable spacer.

To measure the torque of a retaining bolt (or stud and nut assembly) the usual method employed by engine development engineers is to mark the radial position of the bolt (or nut) relative to a surrounding fixed point, then to 'break off' the fastener and then to re-tighten to 'mark' – this will give the torque load of a fastener assembly. Where an engine cylinder head is concerned the torque loads will vary according to differences in gasket collapse, local casting collapse, either through thermal cycling or casting imperfections, and also due to thread 'pull'. There are now, however, systems which allow the torque induced in the bolt (or stud) to be read by recording the 'pitch'when the exposed end of the bolt (or stud) is struck, this without the need to disassemble. This uses the similarity between a bolt and a guitar string in tension to produce a 'pitch' when excited – a higher 'pitch' indicates a higher tension. This system requires a correlation exercise to remove all associated variables such as casting material, cavity differences, and other damping inducing effects which may be present. So, it will only work in a production environment for a specific application and / but with a corresponding error bandwidth due to casting production tolerances.

As mentioned previously by another respondent, the best way to tension conrod bolts is to measure the 'stretch' of the bolt by a non rotating point anvil micrometer. Carillo rods using APC fasteners have a 'stretch value' defining their fitting tension – a typical value may be something in the order of 0.0048".

.

Torque wrenches are for tightening bolts, not for undoing them. On a typical application such as a cylinder head, the bolts will be done up to a lower torque on the first time round, and then to the full tightness finally. The final tightening should be done in one continuous movement per bolt. If the tightening is interrupted just before the final torque is reached, the bolt will not be fully tightened when the torque wrench is taken to the set torque.. As soon as the tightening stops, the friction increases greatly before the bolt starts to move further. This is why it is common for taking some heads down to a lower torque, and then finishing by turning through a predetermined angle. On some older engines, the head was retightened after running to working temperature and then cooling. Then the way to retighten, was to slacken off each bolt, one at a time, before final tightening in the original sequence, to eliminate the "stiction". It is important to adhere to instructions regarding lubrication, or not.

One of the jobs at work was to bolt together the propeller hubs of C130J Hercules, before final machining. The 18 threequarter inch UNF bolts were tightened with a torque wrench in a predetermined sequence to stretch them from 0.012" to 0.015". Torquing alone was not considered accurate enough.

.

Thanks for the link, Chris … I do see what you mean now

But basically what you have there is a hinge: Which is fine if you want or need to allow movement.

MichaelG.

A couple of primers on bolted joints which discuss bolt shear:

Machine Design: What’s the Difference Between Bearing, Shear, and Tear-Out Stress?

Engineer's Edge article Bolt or Pin in Single Shear Equation and Calculator

Edited By James Winkler on 04/12/2019 10:40:33