Joint Stiffness Ratio of a Bolted Joint

This is the single most misunderstood relation in mechanical design, and understanding it changes how a bolted joint is drawn.

Tighten a bolt and two things happen at once: the bolt STRETCHES and the clamped members COMPRESS. They are two springs sharing a load, and they are not in series — they are in parallel, acting against each other. Now apply an external tensile load trying to pull the joint apart. The joint stretches a little; the bolt, already stretched, stretches slightly more; and the members, already squashed, un-squash by the same small amount. The extra force in the bolt is that small deflection times the bolt's stiffness. The relief in the members is the same deflection times theirs. The load divides in proportion to stiffness, and the bolt's share is $C = k_b/(k_b + k_m)$.

For a steel bolt through steel plates, C typically lands between 0.1 and 0.3. The bolt sees only ten to thirty per cent of the external load. The other seventy to ninety per cent merely reduces the compression already in the members. A properly preloaded bolt in a stiff joint barely notices the working load it is holding — its tension moves a few per cent while the joint stays closed — and that is not an accident of the design, it is the point of preloading.

The consequence for fatigue is decisive. A bolt in a cycling joint sees an alternating stress proportional to C, so a joint with C = 0.2 gives the bolt a fifth of the stress amplitude that the same load would apply to an unpreloaded bolt. Analysing a bolt as though it carried the whole external load is not being conservative; it is using the wrong model, and it leads people to fit a bigger bolt when what the joint actually needed was more preload or a stiffer stack. It is also why bolts in cycling joints fail from LOOSENING far more often than from being undersized: lose the preload and the joint separates, C stops applying, and the bolt suddenly takes everything.

Gaskets invert all of this. A soft gasket in the stack is a soft member spring, k_m collapses, C climbs toward 1, and the bolt takes nearly the whole external load on every cycle. That is why serious pressure joints confine the gasket in a groove with metal-to-metal contact outside it, so the members stay stiff and the gasket only seals. The stiffness of the members themselves is the harder number to compute, because the compression under the bolt head spreads out into the plate rather than staying in a neat cylinder; the standard treatment models it as a hollow cone or frustum of about thirty degrees half-angle around the hole. NASA Reference Publication 1228, the Fastener Design Manual, works that geometry through in full and is freely available.