Flange Bolt Torque Calculator

A bolted flange does not seal because the bolts are tight; it seals because the bolts are stretched. That stretch creates a clamp force — the preload — which squeezes the gasket hard enough to seal and holds the joint together against internal pressure and external loads. The whole point of tightening is to put a specific, known preload into each bolt. Torque is merely the tool we reach for because it is easy to apply and measure with a wrench.

Preload is set from a target stress: pick the bolt stress you want, multiply by the bolt's tensile stress area, and you have the target tension F = σ·As. The stress area is slightly smaller than the nominal shank area because it accounts for the threads, and it is the right area to use for both stress and preload. Choosing the target stress is a joint-design decision — high enough to seat the gasket and resist the service loads, low enough not to yield the bolt or crush the gasket — typically landing somewhere between 40 and 70 percent of the bolt's yield strength.

Once the target preload is fixed, the question becomes how much torque delivers it, and that is where friction enters. The same target preload can need very different torques depending on lubrication and surface condition, which is exactly why preload, not torque, is the quantity that actually matters. Torque is a means; preload is the end.

The long-form torque equation splits the applied torque into three pieces. The first is the pitch term, the genuinely useful part that climbs the thread helix and stretches the bolt — and it is tiny, usually under 15 percent of the total. The second is the thread-friction term, the torque spent sliding the loaded thread flanks past each other. The third is the bearing-friction term, the torque spent dragging the turning nut face (or bolt head) across the joint surface. Add them up and the two friction terms typically account for around 90 percent of everything you put in.

That breakdown explains the famous rule of thumb that only about 10 percent of tightening torque becomes bolt tension. It also explains why the friction coefficient is the single most influential and most uncertain input in the whole calculation. A dry, slightly rusty bolt might sit near μ = 0.2, while the same bolt with a good anti-seize drops toward 0.1 — and that change alone can halve the torque needed for the same preload. Get the friction wrong and the preload is wrong by the same proportion.

The nut factor K in T = K·F·D is simply this whole equation collapsed into one number, K = T/(F·D). It is convenient for field use because you can hand a technician a single torque value, but it hides the friction sensitivity. When a published K table and a friction-based calculation disagree, it is almost always because they assume different lubrication. The honest way to read any nut factor is as a statement about the assumed friction condition, not a fixed property of the bolt.

A single torque number is necessary but nowhere near sufficient for a reliable flange. ASME PCC-1, the industry guideline for pressure-boundary bolted-joint assembly, treats tightening as a controlled process. It covers joint and flange-face inspection, gasket selection and handling, lubrication, the tightening pattern and increments, and the verification passes — because the way a flange is assembled matters as much as the final torque.

The reason is elastic interaction. As you tighten one bolt, you compress the gasket locally, which relaxes the bolts already tightened nearby. Tighten in a haphazard order and the joint ends up with wildly uneven bolt loads and a gasket squeezed harder on one side than the other — a recipe for leaks. The cross or star pattern, applied in stepped increments (often roughly 30, 60, then 100 percent of target, followed by one or more rotational passes at full torque), spreads the load evenly and lets the gasket settle before the final preload is locked in.

PCC-1 also pushes toward measuring or controlling preload more directly where it counts: calibrated torque with a specified lubricant, turn-of-nut angle control, or hydraulic tensioning on large studs. The torque figure from a calculator like this one is the starting point for the procedure — the target each pass works toward — but the standard exists precisely because reproducible, leak-free joints come from the whole method, not from one number applied once.