My first torque meter, 50yr ago, incorporated a dial indicator, spring, arm & thrust bearing. Hi-tech, but hysteresis (stiction) made it useless. Quickly replaced w/conventional, much better.
Most people that go back to basics, use handbook data for yield stress, which is 0.2% strain offset. For steel, this corresponds w/60,000psi stress level. A major zero shift in most cases. Much better to use 1/2 of the handbook value.
Tensile stress in wire, especially in stretch winding, can exceed torsion stress. Another reason fatter, longer wire is more forgiving.
Cross-talk (tension unwinding twist) could be significant, but can be measured.
Next to hollow tube, circular cross section is worst from deflection/stress ratio. Better is + or - ie flat strip or x section. Have seen torque screwdrivers with ribbon shaped active sections.
Had considered making a custom strain gage torque transducer. Could automate torque/turns energy release with good accuracy. Much simpler & cheaper would be tying in to a cheap digital scale. Difficult to automate, but ...
Food for thought,
H
_________________________________________________________________
With Windows Live, you can organize, edit, and share your photos.
http://www.windowslive.com/Desktop/PhotoGallery
Received on Tue Sep 01 2009 - 01:11:57 CEST