Why Bridges Don't Fall: Forces in a Truss

Stand under a railway bridge as a freight train passes overhead and you will feel, in your chest, the weight of several hundred tons being held up by what looks like nothing — a lattice of steel bars, most no thicker than your forearm, bolted into triangles. The bridge does not bend. It barely vibrates. The trick is not in the steel itself, which is unremarkable, but in the geometry: the truss.

A truss is a frame built from straight members joined at their ends to form triangles. The triangle matters because it is the only polygon that cannot change shape without changing the length of one of its sides. Push down on the top of a four-sided frame and it collapses into a parallelogram; the corners pivot, the sides stay the same length, and the frame folds. Push down on a triangle and nothing moves unless you actually stretch or crush one of its three sides. This rigidity, achieved without heavy materials, is what makes the truss the workhorse of structural engineering.

When a load lands on a truss — a train, a snowfall, the weight of the bridge itself — that load travels through the members as one of two forces. A member in tension is being pulled at both ends, like a rope in a tug of war; its job is to resist being stretched. A member in compression is being squeezed at both ends, like a column under a roof; its job is to resist being shortened. In an idealized truss, every member experiences only these axial forces, directed along its length. None of them are asked to bend. This is the deep economy of the design: bending is expensive, requiring deep beams and heavy material, while pure tension and compression can be carried by slender bars.

Consider a simple Pratt truss supporting a deck. The load on the deck pushes the bottom chord downward, and the bottom chord responds by stretching horizontally — it is in tension. The top chord, curving over the load like the rim of a bow, is squeezed shorter — it is in compression. The diagonal and vertical members between them carry the load up and out toward the supports at each end, alternating between tension and compression depending on their angle. By the time the force reaches the abutments at the ends of the bridge, it has been split into thousands of small axial pulls and pushes, each handled by a member sized for exactly that job.

This division of labor is why a truss can use less material than a solid beam of the same span and still carry more load. A solid beam under a load bends, and bending stresses its top and bottom fibers heavily while leaving the material near its center almost idle. A truss is, in effect, a beam with the lazy material carved out, leaving only the fibers that are actually doing work — the top chord in compression, the bottom chord in tension, and the web of diagonals transmitting shear between them.

The accounting is exact. An engineer analyzing a truss can isolate any joint, write down the forces meeting at it, and require that they sum to zero in every direction — because the joint is not moving. Apply this rule at every joint and the unknown forces in every member fall out of the algebra. This is called the method of joints, and it is the first real calculation a structural engineering student performs. Done correctly, it tells you not only whether the bridge will stand but exactly how hard each bar is working.

A truss that fails almost always fails for a reason the analysis would have caught: a member sized for tension that ended up in compression and buckled, a connection that could not transmit the force the bar was carrying, a load case the designer did not anticipate. The triangles themselves rarely betray the engineer. They do what triangles do — they hold their shape — and pass the burden, member by member, to the ground.