Proof Strength, Proof Stress, and Accuracy

As far as I know a proof stress is a stress that a manufactured item is subjected to before being approved for shipment, an end of the line quality control. The prro stress should not cause any permanent deformation. So a cable that might break at 20 tons is tested at 10 tons and sold as a ten ton cable. It has a safety factor of 2 in this case, so the buyer knows that every one of these cables has been pretested to 10 tons and does not peramently deform.

Same for bolts etc., in their field

A reference shows that a proof stress is usually to a small permanent deformation,

http://en.wikipedia.org/wiki/Proof_stress

[2] Just remember one thing, designers & draftsmen can kill a good design concept or job that is to be manufactured by ridiculous tolerances, too many have never stood in front of a lathe, mill, suface grinder or cylindrical grinder and have no real idea of the difficulties of holding the tolerances they specify, set CAD systems to default to 2 or 3 decimal places [1/10,000" or microns ] & suface finishes that are not functionally required.

This is probably made worse today by having CNC machinery that manufacturers specify as having micron capabilities, [ which many do !] I always reckon to add an zero to a dimension that isn't needed, triples the cost !!

I ALWAYS had designers give me a hard copy rough-out of the finished CAD design so I could go over it and correct & pencil in changes to dimensions or query them to be justified as well as some aspects of the actual designs regards the problems that would be raised in the actual manufacture stages.

Most of this should be done in the preliminary design meetings BEFORE they begin. DON'T OVER TOLERANCE !!!! SAVE YOUR COMPANY MONEY !!!

having spent the last 20 yrs in the machinery rebuilding and repair business(master millwright and master electrician) i have increasingly come across ,especially in bigger corps the setting of ridiculous tolerances as a way to force bargain or attempt to completely sidestep paying for a rebuild... i did a centerless grinder with a guy that the customer called out a runoff repeat tolerance of .000001 on a slide with a modified CNC type control. we had to come up with numbers to plug into the program to try and account for the flex of the tailpiece on each movement of the slide (withdraw slightly and adjust in /out) it looked like a mad scientist workshop w/ papers and scrwaled notes every where but we finally managed to get it to repeat @ .000002... and when the customer came in to see, it they were stunned and two of their engineers admitted that they never expected us to make it work but would always call that out and then jack people up for deep discounts when they couldn't achieve them... i was ready to strangle him