Statics › Section F · Equilibrium of structures › Topic F.4
The method of joints
Every pin in a truss is a tiny particle in equilibrium. Solve them one at a time, passing each answer to the next joint, and you can find the force in every member of a determinate truss.
Joints as particles
Isolate one pin of a truss and look at what acts on it: the forces from the members that meet there, plus any load or reaction at that joint. All these forces pass through the pin, so they are concurrent. A moment equation about the pin is automatically satisfied and tells you nothing, which leaves the two equations of 2D particle equilibrium: \(\Sigma F_x = 0\) and \(\Sigma F_y = 0\). Each joint can yield at most two unknowns. Because a member carries the same force at both ends, a force found at one joint becomes known at the joint at its other end.
The procedure
1. Check that the structure is a truss and that it is determinate, using \(2j = m + r\) and triangles. 2. Remove zero-force members by the two rules, to save work. 3. Look for a solvable joint: one with at least one known force and no more than two unknowns. If none exists, first find the support reactions from a free-body diagram of the whole truss. 4. Solve that joint with \(\Sigma F_x = 0\) and \(\Sigma F_y = 0\). 5. Carry the results to neighbouring joints and repeat until every member is known. 6. If you found the reactions, the last joint has spare equations: use them as a check.
Drawing a joint diagram
Draw the joint as a dot. Show loads and reactions in their actual directions. Draw every unknown member force as tension, an arrow pulling away from the joint along the member, labelled with the member's name such as \(F_{AD}\). Solve, then read the sign: positive is tension, negative is compression. When you carry a solved force to the next joint, keep its signed value and again draw it pulling away from that joint; the negative sign takes care of compression. Redrawing the arrow as well counts the sign twice. Record each member force with T or C in a table.
Resolving member directions
Each member's direction comes from the joint coordinates, not from a protractor. For a member from joint P to joint Q, find the run \(\Delta x\), the rise \(\Delta y\) and the length \(L\). A tension force \(F\) pulling on P has components
For a member rising 3 m over a run of 4 m, \(L = 5\) m and the fractions are 0.8 and 0.6. At the far joint Q the same member pulls the other way, so both fractions change sign. Fractions keep every equation exact, with no rounded angles.
Worked example: a roof truss, reactions and supports
A roof truss has a pin at A(0, 0), a roller at C(8, 0), a bottom joint B(4, 0) and an apex D(4, 3), in metres. Members: AB, BC, AD, CD and BD. At D act 18 kN down and a wind load of 4 kN to the right; at B, 6 kN down. 1. Whole truss: \(\Sigma F_x = 0\) gives \(A_x = -4\) kN. Moments about A, counterclockwise positive:
so \(C_y = 13.5\) kN and \(A_y = 24 - 13.5 = 10.5\) kN. 2. Joint A: AD has fractions (0.8, 0.6). \(\Sigma F_y\): \(10.5 + 0.6F_{AD} = 0\), so \(F_{AD} = -17.5\) kN (C). \(\Sigma F_x\): \(-4 + F_{AB} + 0.8(-17.5) = 0\), so \(F_{AB} = 18\) kN (T). 3. Joint C: CD has fractions (−0.8, 0.6), so \(F_{CD} = -22.5\) kN (C) and then \(F_{BC} = 18\) kN (T).
Worked example: a roof truss, finishing and checking
4. Joint B carries the 6 kN load and members BA, BC and BD. Horizontally, \(-18 + 18 = 0\), which checks. Vertically, \(F_{BD} - 6 = 0\), so \(F_{BD} = 6\) kN (T). 5. Joint D is now a check, since every force on it is known. AD pushes D with \(17.5 \times (0.8, 0.6)\), which is (14, 10.5) kN, CD pushes it with \(22.5 \times (-0.8, 0.6)\), which is (−18, 13.5) kN, BD pulls it down with 6 kN, and the loads are (4, −18) kN. Horizontally \(14 - 18 + 4 = 0\) and vertically \(10.5 + 13.5 - 6 - 18 = 0\). Both balance, confirming AB = BC = 18 kN T, AD 17.5 kN C, CD 22.5 kN C and BD 6 kN T.
Worked example: a wall bracket without reactions
A truss is pinned to a wall at A(0, 0) and E(0, 2). Joints B(2, 0) and C(4, 0) lie on the bottom chord, D(2, 2) on the top. Members: AB, BC, CD, DE, BD and AD. A 9 kN load hangs at the tip C. 1. Joint C has a known load and only two unknowns, so no reactions are needed. CD has fractions \((-1, 1)/\sqrt{2}\). \(\Sigma F_y\): \(F_{CD}/\sqrt{2} - 9 = 0\), so \(F_{CD} = 9\sqrt{2} \approx 12.7\) kN (T); \(\Sigma F_x\) gives \(F_{BC} = -9\) kN (C). 2. Joint B: BA and BC are collinear and B is unloaded, so \(F_{BD} = 0\) and \(F_{AB} = -9\) kN (C). 3. Joint D: vertical balance gives \(F_{AD} = -9\sqrt{2}\), about −12.7 kN (C), then horizontal balance gives \(F_{DE} = 18\) kN (T). 4. Check: the wall pushes (18, 9) kN at A and (−18, 0) kN at E. The two 18 kN forces form a 36 kN·m couple, balancing the opposite couple of the 9 kN load and the 9 kN at A: \(9 \times 4 = 36\) kN·m.
Reading the results
The signs tell a structural story. In the roof truss, the sloping rafters AD and CD are in compression, pushing outwards and down on the supports, while the bottom chord AB–BC is a tie in tension that stops the supports spreading apart. The wind load makes the rafters unequal. In the wall bracket the pattern flips: the top chord DE is in tension and the bottom chord in compression, because a cantilever's top is stretched and its bottom squeezed. A tie that comes out in compression usually signals a sign error.
Solving all joints at once
Instead of moving joint by joint, you can write both equations at every joint, \(2j\) equations in all, and solve them together for the \(m\) member forces and \(r\) reactions. For a determinate truss this is a square linear system that a computer solves in one step, as structural software does. It needs no clever starting joint, but it hides the physical reasoning and is impractical by hand, so sequential solving remains the way to learn and to check computer output.
Common mistakes to avoid
First, starting at a joint with three unknowns: two equations cannot solve it, so find the reactions or pick another joint. Second, writing a moment equation at a joint, which gives no information because the forces are concurrent. Third, using the wrong direction fractions, such as swapping 0.8 and 0.6, or forgetting that they change sign at the member's other end. Fourth, carrying a compression force to the next joint as a positive number. Fifth, skipping the final check joint, which catches most slips.
Key takeaways
The method of joints treats each pin as a particle under concurrent forces: two equations per joint, ΣFx = 0 and ΣFy = 0. Find the reactions first unless a joint is already solvable, then move from joint to joint, never facing more than two unknowns. Draw unknown member forces as tension, so a negative answer means compression, and carry signed values forward. Resolve members with run and rise over length. Use the last joint as a check. Computers solve all the joint equations at once.
Flashcards
Flashcard
Equilibrium of each pin, treated as a particle
Test your understanding
Question 1 of 5
Why can't a moment equation help you at a single truss joint?
Question 2 of 5
At joint A, the pin reaction is 10.5 kN upward and member AD rises at 3 vertical to 4 horizontal (length 5). Assuming tension, ΣFy gives 10.5 + 0.6 F_AD = 0. What is the force in AD?
Question 3 of 5
A member found to be −12 kN at one joint is carried to the joint at its other end. Using the tension-assumed convention, how should it be entered there?
Question 4 of 5
A truss joint has three unknown member forces and no load. What should you do?
Question 5 of 5
In the roof truss worked example, which statement describes the bottom chord AB–BC?
