The other problem is that torque on the shaft and the compression force on the things being clamped are not simply related - you might like to think that the thread ramp angle would convert one to the other but the whole point about threads that are not intended to be worm drives, is that the friction on the thread resisting turning is the dominant term, not the unwinding forces.

(mathematically a thread is just an inclined plane wrapped around a stick...)


So, it depends on the surface finish of the threads in the nut and bolt part, more than anything else.  So the torque to get the thing undone is far from being equal and opposite to the forces originally required to do  it up.

We sort of know this, as an extreme case of rough thread surfaces, when replacing the exhaust pipe on the car, the nut can be completely off load such that the pipe rattles and the joint leaks,  and yet the thing is corroded enough that it still needs several gorrilla-feet of torque to get it to budge, and then only with an awful teeth jarring graunching squeak, or in some cases it is still energetically favourable  to snap the stud off short in the block,  rather than undo the thing, then requiring some cunning with the welder to get it out...

The related effect is that you can clamp a copper wire in a terminal to limit torque, and then warm it and cool it and the thing strains when warm and relaxes when cold, the copper shifts about a bit and afterwards the bolt can be almost any torque, right down to  finger loose.

I'm a great believer in fitting methods with a bit of spring to them (crimps on strands or star washers on bolted lugs and similar) to take up the slack in thermal cycling, but on a normal terminal block, there is no 'stretchy' element, I do not think the bolts compress very much, and only on a cheap one does the terminal tunnel deform, and then it is not reversible.

The other problem is that torque on the shaft and the compression force on the things being clamped are not simply related - you might like to think that the thread ramp angle would convert one to the other but the whole point about threads that are not intended to be worm drives, is that the friction on the thread resisting turning is the dominant term, not the unwinding forces.

(mathematically a thread is just an inclined plane wrapped around a stick...)


So, it depends on the surface finish of the threads in the nut and bolt part, more than anything else.  So the torque to get the thing undone is far from being equal and opposite to the forces originally required to do  it up.

We sort of know this, as an extreme case of rough thread surfaces, when replacing the exhaust pipe on the car, the nut can be completely off load such that the pipe rattles and the joint leaks,  and yet the thing is corroded enough that it still needs several gorrilla-feet of torque to get it to budge, and then only with an awful teeth jarring graunching squeak, or in some cases it is still energetically favourable  to snap the stud off short in the block,  rather than undo the thing, then requiring some cunning with the welder to get it out...

The related effect is that you can clamp a copper wire in a terminal to limit torque, and then warm it and cool it and the thing strains when warm and relaxes when cold, the copper shifts about a bit and afterwards the bolt can be almost any torque, right down to  finger loose.

I'm a great believer in fitting methods with a bit of spring to them (crimps on strands or star washers on bolted lugs and similar) to take up the slack in thermal cycling, but on a normal terminal block, there is no 'stretchy' element, I do not think the bolts compress very much, and only on a cheap one does the terminal tunnel deform, and then it is not reversible.