How Trusses Work: Tension, Compression, and the Method of Joints

A truss turns a long span into a triangle grid, and triangles don’t change shape without changing the length of their sides. That means every member does just one job: it’s either being stretched (tension) or squeezed (compression) — no bending. Putting material in pure tension/compression is why a truss can span far with very little weight.

Reading a truss

For a simply-supported truss carrying gravity load, one part of the pattern is always true:

• Top chord → compression. The top is being shortened, like the top of a sagging beam.
• Bottom chord → tension. The bottom is being stretched.

The web members (the verticals and diagonals) depend on the truss type:

• Pratt — diagonals in tension, verticals in compression (efficient in steel, where tension members can be slender).
• Howe — the reverse: diagonals in compression.
• Warren — alternating diagonals, tension then compression, often without verticals.

Some members carry no force at all in a given load case — “zero-force members” — present for stability or to brace longer members, not to carry that particular load.

The method of joints (the sign trick)

To find each force, engineers isolate one joint at a time and balance the forces (ΣFx = 0, ΣFy = 0). The convention that makes it readable:

• A member in tension pulls away from the joint.
• A member in compression pushes toward the joint.

Assume every unknown is tension; if the math returns a negative number, it’s actually compression.

Try it interactively

Truss solvers are already a solved problem on the web, so rather than rebuild one we point you to good free ones — drag loads and supports and watch each member turn red (compression) or blue (tension):