Trusses: Method of Sections
To find one member in the middle, cut the truss in two.
Before this Lesson
You want the force in one bottom-chord bar of a six-panel bridge truss, three panels in from the support. By joints, you start at the support and walk in: four joints, six equations, and any slip at the first joint rides along to the last. All that for one number. There's a shorter road, one equation long, and it only asks you to take a saw to the bridge, on paper.
Saw the truss in two
- From Trusses: Method of Joints: a member's force lies along its axis. Draw an unknown pulling away, so a positive answer is tension and a negative one is compression.
- From Beam Reactions: a moment equation can't see a force whose line passes through the moment point.
Draw a line across a truss so it splits into two pieces, and look only at the left one. The whole truss is sitting still, so the left piece is too. What holds it there is the bars you just cut, which were pulling or pushing on it all along. Replace each cut bar with an arrow along its axis, and forces that were hidden inside the truss become external forces on the piece.
The Handbook's Plane Truss: Method of Sections entry says just that, in one sentence. Everything below is how to use it well.
Here's why a cut beats a joint. At a joint, every force passes through one pin, so you get two equations. A cut piece is a rigid body with its forces scattered around it, so you get three: , , and , the Handbook's Equilibrium Requirements written out for a plane. Three equations settle three unknowns, so the cut can pass through at most three members of unknown force. Cut a fourth and you're one equation short. The Handbook doesn't state that limit, so carry it in yourself.
The routine:
- Treat the whole truss as one body and find the reactions.
- Cut through the member you want and at most two other unknown members.
- Keep whichever piece has fewer forces on it, and draw all of them: the reaction, every joint load on that piece, and the cut members pulling away.
- Write the one equation that sees only the member you want.
Step 4 is where the thinking goes, and it depends on the member.
One moment equation for a chord
A flat truss has two long lines of bars, the top chord and the bottom chord, with a web of verticals and diagonals between them. Cut through one panel and you've cut two chords and a diagonal, and the diagonal meets each chord at a joint.
Say you want the bottom chord. Go to the joint where the diagonal meets the top chord. Both of those forces run through that point, so neither has a moment about it, and the bottom chord is alone in the equation. It's the trick you used at a support to get a reaction. The chord's arm is its perpendicular distance from that joint, which is the truss's depth. It isn't a member's length or the panel width. And that joint can sit on the piece you threw away, since all you need is its location.
Which joint you go to depends on which way the diagonal runs, and that's what sets the common patterns apart. A Pratt truss has diagonals sloping down toward midspan. A Howe truss has them sloping up toward midspan. A Warren truss zigzags, usually with no verticals at all. The Handbook doesn't name these patterns, so know them by sight.
The web takes what the chords can't
Now you want the diagonal. The moment trick needs the other two cut members to meet, and two parallel chords never do. But they give you something better. Horizontal chords have no vertical part, so sum the vertical forces on the piece, and the web member is the only unknown left. It supplies whatever vertical force the reaction and the loads leave unbalanced. For a diagonal, that's its force times rise over length.
A vertical member takes the same equation, but you have to reach it with a slanted cut, since a straight one would run right along the bar. Lean the line the way the diagonals lean, so it crosses one chord, the vertical, and the other chord. Lean it the other way and you cross five bars. Then notice which piece the vertical's bottom joint landed on, because its load goes with it.
A truss is a beam with holes in it
Look back at what those two equations did. The reaction and loads on the piece tried to spin it, and the chords answered with a pair of opposite forces. They also tried to slide it vertically, and the web answered. A solid beam does exactly this at every cross section. Sag a beam between two supports, and its top squeezes while its bottom stretches. The Handbook's Shearing Force and Bending Moment Sign Conventions entry calls that compression in the top fibers and tension in the bottom fibers. A truss gathers that squeeze and stretch into its two chords, and leaves the shear to the web.
The two chords act as a couple, with an arm of the truss depth . That makes a chord's force the bending moment over the depth:
Here is the moment of everything on one side of the cut, about that chord's own moment center, the joint where the other two cut members meet. For CD in the Pratt truss, that was , and . The couple isn't perfect, which is why each chord needs its own . The same cut's top chord carries 54 kip, and JD's horizontal part makes up the 6 kip difference. The web carries the shear , the net vertical force on the piece, which was 8 kip there. The Handbook prints neither idea for trusses. They're your two section equations in beam language, and Internal Shear and Moment in Beams treats that language properly.
It also gives you signs without any algebra. On a simply supported truss under gravity, the top chord is in compression and the bottom chord is in tension. For a diagonal, take a cut in the left half and look at the left piece. The reaction beats the loads on it, so the piece has a net upward force, and the web has to hold it down. A diagonal sloping down toward midspan can do that by pulling, so it's in tension. That's a Pratt truss. The right half is the mirror image, so what counts is the slope toward midspan, not left or right. That assumes symmetric loads. With a lopsided load, cut the truss and ask the same question.
When the chords slope
All of that leaned on horizontal chords. Tilt the top chord, as in a pitched truss, which peaks at midspan, or a camelback, whose top chord rises in several straight runs, and its force gets a vertical part. Now holds two unknowns, the diagonal and the top chord, and it can't give you either one. The slope rule for signs goes too, because a sloping chord can carry more or less than the whole shear.
So go back to moments. The chords aren't parallel anymore, so run their lines out past the truss, and they cross at a point in empty air. That's fine, since a moment center doesn't have to be on the body. Both chord forces have lines through that point, and the diagonal is alone in the equation.
Two small moves finish the job. Find the crossing point from the top chord's slope. Then slide the diagonal's force along its line to where it reaches the bottom chord, which doesn't change its moment. There, its horizontal part runs straight through the crossing point, so only its vertical part has an arm.
How much load can it take?
Turn the question around. A chord has a limit, and you want the largest equal panel load that stays within it. There's nothing new here. Every force in a truss is proportional to its loads, so write the same moment equation with in place of a number. The chord force comes out as a multiple of , and you set its size equal to the limit.