This discussion is getting quite focussed on the real issues. I think it is obvious that the quality of machining, the "fit & finish" of a nutted axle system is crucial to the amount of strength it has. I have a couple of vintage NOS BH track hubs that appear to have whitworth threads on them. Regardless of the name, the threads are a fine machine grade. Turning the nuts onto the axles requires a firm twist but it is smooth and consistant. You can tell that the tolerances on the mating surfaces of the threads are perfect. Couple that, with the high quality of steel being used in the british bike industry at the time and you've got a one tight rear wheel on your bike. I think I would be hard-pressed to strip those nuts with anything less than a 12" wrench.
Any Raleigh Sports bike or roadster made in the early 60's or before, reflects this quality. You will rarely find a stripped nut on these bikes and these axles and nuts may have been loosened and re-tightened dozens of times. If the threads on these axles/nuts were not perfect, I should think that the holding strength would drop off severely. Add some softer steel to the equation and you've got the junk we see in the department stores around the world.
Warren Young Toronto
>Don Ferris wrote:
>
>> As far as the strength
>>of a bolted connection, as with anything else, the strength of the assembly
>>is only as great as its weakest link. If the case of nuts and bolts, the
>>strength of the nut's threads in shear could/would/should equal or exceed
>>the the tensile strength of the bolt/axle. Whether or not this is the case
>>in all nuts and axles as it applies to all nutted hubs, I can't comment,
>>I've never tried to tighten one to failure. You can make the case that not
>>all hub nuts are properly designed, but the same is true for skewers. A
>>wash. Obviously an axle that has flats on it, ala Sturmey's, etc., the
>>threads in the nuts have a reduced area to purchase and would be far weaker
>>unless the nut was threaded deeply.
Sheldon wrote
>There are a couple of differences you're overlooking. The threads in
>a QR skewer are adjusted under no load, so there are no frictional
>heating issues weakening the threads. A thread can _hold_ a greater
>stress than it can generate by being tightened without damage.