This topic just keeps on coming up and it is reasonably important.
The Workshop Manual states to following
There are several alternatives to moly greases now and some of the ceramic compounds work just as well and cause much less mess.
There will always be a debate about how tight a bolt of a nut should be and the advice that is offered seems variable.
In some instances we are advised to overtighten bolts (Rocker shaft bolts would be an example) and in other cases to undertighten.
The real answer is that to correctly seat a wheel and to prevent the nut from working looses a specific preload is normally needed.
When we torque up a nut we are trying to estimate the preload from the torque applied.
The type of lubrication, the thread condition and the cleanliness will all have a significant impact on this relationship and is why torque measurement is often fairly unreliable.
If we consider a typical nut and bolt then the torque needed will be due the friction of the nut faces(50%), the friction in the thread(40%) and when this friction is overcome the bolt will develop a preload (10%)
There is also a good deal of scatter associated with torque measurements and variations of up to 30% on preload are commonly reported.
The more modern tightening methods use a small value of applied torque to seat the two parts followed by an angular rotation to ensure the correct level of stretch and hence preload. This method significantly reduces preload and modern digital torque wrenches normally provide this feature.
