Calibrated Wrench
This method is sometimes known erroneously as "torque control". The Research Council stipulates that each day, for each different diameter, length, and grade of bolts, a representative sample of three bolts must be selected to calibrate a wrench.* The bolts are tensioned in a Skidmore-Wilhelm bolt tension calibrator on the site to measure (i.e. "calibrate") how much torque it takes to get the bolts up to the correct tension.** The wrench is then set to cut out at that torque (the wrench becomes "calibrated"), and then all similar bolts that day are installed to that torque after the joint is snugged first. Rotation during the tightening process must be limited to a specific value. On the next day another set of torques is similarly established and the wrench(s) calibrated again. And so on. Tables of torque by bolt size and diameter are not applicable, and this entire method is not allowed in Canada. Actual results of tension in bolts produced by this method are acknowledged to be highly variable, even when this method is followed religiously, which is rarely the case.
*By "representative sample" they mean bolts in the same condition as the ones that are about to be tightened.
**That is, to a tension 5% higher than the specified minimum.
