Statics › Section F · Equilibrium of structures › Topic F.3

How trusses work

A truss carries a heavy load across a long gap using surprisingly little material, by arranging slender members into triangles so that each one is only pushed or pulled. This topic sets up the model engineers use, shows how to check that a truss can be solved, and how to spot members that carry nothing.

What a truss is

A truss is a structure of long, slender, straight members connected only at their ends. Bridges, roofs, pylons and crane booms use trusses to span long distances while staying light and strong. A planar truss lies in one plane and is built from triangles; a space truss is three-dimensional, built from tetrahedra. Joints are labelled with letters and each member is named by its end joints, so member BF runs from B to F.

Why triangles

Pin four bars into a square and push one corner sideways: the square racks into a parallelogram without any bar changing length. Pin three bars into a triangle and it cannot change shape unless a bar stretches or shortens, because three side lengths fix a triangle completely. A truss made of triangles is therefore rigid on its own, even off its supports. Real trusses join members with bolted, riveted or welded gusset plates, making stiff joints that are hard to analyse exactly, so engineers start from a simpler model.

The simple-truss model

The model rests on four assumptions. Members are joined by frictionless pins. Loads and support reactions act only at the joints. Each member's weight is neglected, or split in half and applied at its two end joints. Members are rigid. Under these assumptions every member touches the rest of the structure at exactly two points, so every member is a two-force member and carries a pure axial force, tension or compression, along its length. This model suits preliminary design; later stages check joint stiffness and the stresses that set member sizes.

Tension, compression and buckling

The axial force decides how a member is built. A tension member is pulled; even a slender rod or cable handles tension well, because pulling straightens it. A compression member is pushed, and a long slender one can buckle: it bows sideways and collapses at a load far below what the material could take in a short block. Compression members are therefore made stockier or braced partway along, which is why one bridge can mix thin rods in tension with chunky sections in compression.

Counting equations and unknowns

Each pin of a planar truss gives two equilibrium equations, so a truss with \(j\) joints has \(2j\) equations. The unknowns are the \(m\) member forces and the \(r\) support reactions. The truss can be determinate only if

2j = m + r

If \(2j > m + r\), there are too few members and the truss is unstable: it will collapse. If \(2j < m + r\), it has redundant members and is statically indeterminate, solvable only by using member deformations. The count is necessary, not sufficient: members must also form triangles. A space truss gives \(3j\) equations instead.

Simple trusses and redundancy

A simple truss starts from one triangle and grows by adding one new joint and two new members at a time. The first triangle has 3 joints and 3 members, and each step adds 1 joint and 2 members, so \(m = 2j - 3\). With the usual three reactions, a pin and a roller, this satisfies \(2j = m + 3\) exactly. The price is no redundancy: if one member fails there is no other load path, and neighbouring triangles fold in turn. Structures like this are called fracture-critical. A well-known US case is the Silver Bridge over the Ohio River, an eyebar-chain suspension bridge without redundancy: a crack in a single eyebar brought it down in 1967, killing 46 people.

Named truss patterns

Many truss layouts carry their designer's name. In a Pratt truss, patented in 1844, the diagonals slope down towards the middle of the span, so under ordinary downward loads most diagonals are in tension and most verticals in compression. The Howe truss, patented in 1840, reverses the diagonals, putting them in compression and the verticals in tension. The Warren truss, patented in 1848, uses alternating diagonals that form a row of triangles, often with no verticals at all. All three are simple trusses.

Zero-force members

Under a given loading some members carry no force, and two rules find them by inspection. Rule 1: if only two non-collinear members meet at a joint with no load or reaction, both are zero-force members. Rule 2: if exactly three forces act at a joint, counting members, loads and reactions, and two of them are collinear, the third is zero. Its component perpendicular to the collinear pair has nothing to balance it. Apply the rules repeatedly: once a member is known to be zero, drop it, and new joints may then qualify. A zero-force member is zero along its whole length.

Worked example: is this truss determinate?

A Pratt bridge truss has bottom joints A(0, 0), B(3, 0), C(6, 0), D(9, 0) and E(12, 0), and top joints F(3, 3), G(6, 3) and H(9, 3), in metres. Its members are the bottom chord AB, BC, CD, DE; the top chord FG, GH; end diagonals AF and HE; verticals BF, CG, DH; and diagonals FC and HC. A pin at A and a roller at E give \(r = 3\). 1. Count: \(j = 8\), \(m = 13\), so \(2j = 16 = m + r\). 2. Arrangement: it grows from triangle ABF by adding C, G, H, D and E with two members each, so it is a stable simple truss. 3. Remove FC: \(m + r = 15 < 16\), and square panel BCGF can rack, so the truss is unstable. 4. Add a second diagonal BG instead: \(m + r = 17\), indeterminate by one.

Worked example: finding zero-force members

Use the Pratt truss from the previous card with a single 20 kN load at C. 1. Joint B has no load and three members: AB and BC are collinear, so BF is a zero-force member by Rule 2. 2. Joint D is the same, so DH is zero. 3. Joint G has no load; FG and GH are collinear, so CG is zero. 4. Joint F now has AF, FG and FC, none collinear, so neither rule applies and those members do carry force. 5. A full solution confirms it: FC carries 14.1 kN tension, FG 20 kN compression, and BF, CG and DH exactly zero. Add 10 kN at B and BF carries 10 kN tension: zero-force describes a load case, not the member.

Common mistakes to avoid

First, deleting zero-force members from the design. They brace compression members against buckling and carry load when the loading changes, as when a lorry crosses a bridge. Second, applying Rule 2 at a joint that carries a load or reaction not along the collinear pair: that makes four forces and the rule fails. Third, trusting \(2j = m + r\) alone, when a badly arranged truss can pass the count and still be unstable. Fourth, forgetting to count each support's reactions: a pin gives two, a roller one. Fifth, over-eliminating members until a load has no path to a support.

Key takeaways

A truss is a framework of slender straight members joined at their ends in rigid triangles. The simple-truss model makes every member a two-force member in tension or compression; long compression members can buckle. A planar truss is determinate when 2j = m + r and its members form triangles; fewer members make it unstable, more make it indeterminate. Simple trusses satisfy m = 2j − 3 but have no redundancy. Zero-force members follow from two rules and still matter.

Flashcards

Flashcard

Truss

Slender straight members joined only at their ends, arranged in triangles

Bridges, roofs, pylons.

Test your understanding

Question 1 of 5

A planar truss has 9 joints and is supported by a pin and a roller. How many members must it have for 2j = m + r to hold?

Question 2 of 5

A planar truss has 6 joints, 8 members, a pin and a roller. What does the count tell you?

Question 3 of 5

Why does the simple-truss model make every member a two-force member?

Question 4 of 5

Joint K of a truss has no external load and three members: KL and KM are collinear and KN is not. What is the force in KN?

Question 5 of 5

A member is a zero-force member under one load case. What should the designer do with it?

← Previous topicNext topic →