In machine assembly it is often required that bolts used to fasten machine parts be installed with specific design preloads. Because it is inconvenient to measure preload directly, preload specifications are usually based on some more easily measured quantity with which the level of preload may be correlated. Most often this quantity is the torque to be applied to the bolt at installation. Studies by Blake and Kurtz [1] and Heyman [2] have shown that when bolts are torqued into place, the fraction of applied torque which translates into useful preload is small and widely variable. This is so because the large majority of applied torque is absorbed in overcoming friction in the bolt’s threads and at the underside of the bolt’s head. Consequently, even though the torque to install different bolts may be identical, small variations in frictional conditions from one installation to the next can result in large variations in preload. The unreliability of torque as an indicator of preload has been the motivating factor behind the development of a number of alternate methods of measurement [2–5].