torque wrench calibration.
an accepted exercise in "calibrating a guess"
Since the relationship of applied torque to actual bolt stress is highly variable, calibration of torque tools serves little to no useful purpose. Fundamentally, it doesn't matter if a wrench produces "x" torque with an accuracy of +/- 3.0% if the outcome of the torquing procedure results in, for example, a bolt stress variation of +/- 35.0%.
Even in an unlikely instance where actual friction conditions are the same as the designer had initially assumed, the calibration may be misleading and could provide a dangerous false sense of security; If the tool isn't used under the same controlled environmental conditions as in the calibration lab, internal friction factors would be different. Furthermore, the more a tool is operated, the less relevant the calibration is. Consider a hydraulic torque wrench calibrated at 21°C:
- the internal lubricant begins to break down upon actual use
- if the tool is operated in an environment warmer or colder than 21°C, the lubricant will have a different sliding coefficient
- if the tool is operated in an environment warmer or colder than 21°C, the fit between internal components may change
- the more the tool is cycled, the hotter it gets and the less viscous the lubricant becomes
Where accuracy of bolt stress is critical, all detrimental internal and external friction variables must be addressed. The only way of doing this effectively is by verifying the outcome of the torquing procedure. Therefore, any individual inconsistencies can then be corrected by applying whatever input force is required to produce the required stress.