What we really want is the compression force on the joint. Most of the “torque” we apply is just overcoming friction in the threads. Only 10-15% of the torque is applied to the joint. Plus torque wrenches are not as accurate as believed, And we can get into arguments about lubrication, surface roughness, corrosion, and so on but the fact is that tension accuracy just isn’t very good.
In electrical contacts the metal is not totally flat and it is covered in oxide. Squishing the joint cracks the oxide layer at the high spots forcing metal on metal contact called alpha spots. Further force smears the aloha spots so they grow in size as well as forming new ones but it’s a “diminishing returns” issue. The aloha spots are literally cold welds. So as force is removed we don’t uniformly lose contact…there is hysteresis. Less than 10% of the area if contact with bus bars actually makes contact.
One thing that this makes clear though is that if you aren’t mushing the copper so hard that it is highly distorted out of shape it’s not tight enough. Many times it bird nests so bad in the ends you have to cut it off and strip a fresh area to land it again.
Mechanically there is something called “dislocation theory” that shows that even though the metal crystals are solid at room temperature they will naturally relax and stress relieve themselves, permanentjy stretching the fastener. Just check for YouTube videos. The surface can also smear a little, allowing some loosening and what is called “fretting”.
Sure metallurgy plays a part, too. The stress/strain relations are different depending on the alloys but really we don’t care what is in the joint. We are torquing and stretching the bolt/screw/box. This puts tension in the electrical contact. The torque is on the steel or aluminum threads, not the copper. That is what the steel washers are for…keep the steel from digging in.
Fortunately this is all known. Electrical joints as far as I can tell are seriously over-rated and we know that because failures are not very often.
There is a web site, boktscience.com, that goes over this in depth. Also the NASA fastener technical manual is quite good. Among other things it explains why you will find “helical spring washers” in a Grainger catalog but not “lock washers” (there is no such thing).