A frequently misunderstood aspect of using bolts to clamp things together is “preload,” which is the tension applied to the bolt itself by tightening the nut (as opposed to the actual load on the bolted joint). A sufficient preload will help avoid bolt fatigue as joint loads vary. Fluctuating joint loads will affect the clamping force on the bolting components (bolt and nut) rather than changing the tension on the bolt itself. (This is not absolutely true, but it’s accurate enough for most bolting discussions.) As a rule of thumb, bolt preload should exceed the maximum load the joint is expected to experience by 15% or so. In the nut turn example below, the maximum joint load is expected to be 435 pounds.
For this to work properly, the joint must be stiffer than the bolt. The shank of high-tech bolts usually are necked down to the same diameter as the root of the thread. As long as the shank is not smaller than the thread root, it won’t be any weaker than the thread, and therefore overall strength will be unaffected, but it is much more flexible (less stiff) than the original nominal size shank.
There are two ways to measure bolt preload: 1) with a torque wrench and 2) by measuring the angle the nut has turned. Of the two, the latter is more accurate since friction plays a significant - and more importantly, unknown - role when using a torque wrench.
Torque = K × preload × diameter
K, the so-called “Nut Factor” (more about nut factor in a future issue), usually varies between 0.07 and 0.30, and is very sensitive to a number of influences, ranging from temperature to thread condition, and even to how rapidly the nut is tightened.
Measuring the angle the nut has turned is simply measuring how much the bolt is stretching, equal to the pitch (distance between threads) times the number of turns. Using this formula requires that the components being bolted not compress much (or compress a known amount), and that the "spring rate" of the bolt is known.
Nut Turns = preload ÷ (spring rate × pitch)
For example, if the "spring rate" of a 1/2-13 B7 bolt is 66,000 pounds per inch, and you need 500 pounds of preload, you will need to stretch the bolt 500 ÷ 66,000 = 0.0076 inch. At 13 threads per inch, this equates to 0.0984 turns, or about 35° past snug. (0.0984 x 360o = 35.4o).
If more than one bolt is used in a joint, and those bolts are closer together than about four diameters, the preload on one bolt will affect the preload on the other bolts by compressing the joint. This effect is known as "crosstalk," and then all bets are off. Joints that are significantly less stiff than the bolts, such as joints involving gaskets, suffer much worse from crosstalk. The best way to control crosstalk is to use a carefully thought out tightening sequence (usually for circular patterns, diametrically opposed alternating bolts), and to tighten the bolts in small steps. Even so, it's an inexact science, and for this reason offering any joint make-up advice that could be interpreted as engineering consultation (with its attendant liability) should be avoided.