The Moment Distribution Method: Solving Indeterminate Beams Without Equations
The moment distribution method, published by Hardy Cross in 1930, transformed the practice of structural analysis for continuous beams and frames. Before it, solving highly indeterminate structures required assembling and solving large systems of simultaneous compatibility equations — a process both tedious and prone to arithmetic error. Cross’s insight was to replace that global solution with a physical iteration: lock every joint, compute the resulting fixed-end moments, then unlock one joint at a time and distribute the imbalance to adjacent members. The procedure converges to the exact elastic answer without solving a single equation simultaneously.
Why Indeterminate Structures Need More Than Equilibrium
A simply supported beam is statically determinate: the three equilibrium equations — sum of horizontal forces, sum of vertical forces, and sum of moments — yield all reactions and internal forces. A two-span continuous beam over three supports adds one unknown reaction, making it once indeterminate. That extra unknown requires a compatibility equation: a statement about how the structure deforms. As spans multiply, the system of equations grows rapidly. The moment distribution method sidesteps this by working iteratively with joint moments, balancing each joint in turn until residual unbalances become negligible.
Key Parameters: Stiffness, Distribution Factor, Carry-Over
Stiffness factor K is the rotational resistance of a member end. For a prismatic member of length L and second moment of area I, the far-end-fixed stiffness is K = 4EI/L. If the far end is pinned, the near-end stiffness reduces to K = 3EI/L because some of the applied moment is shed through the pin without generating a reaction moment.
Distribution factor (DF) at a joint is the fraction of any unbalanced moment that a member absorbs when the joint is unlocked:
DFmember = Kmember / ΣKat joint
The distribution factors at any joint always sum to exactly 1.0, confirming that the full unbalanced moment is absorbed by the connected members in proportion to their rotational stiffness.
Carry-over factor (COF) accounts for the moment induced at the far end when the near end is rotated. For a far end that is fixed, COF = 0.5: half of the moment distributed at the near end is carried over. For a far end that is pinned or free, COF = 0, because no moment can develop there.
Fixed-End Moments
Locking every joint creates a fully fixed-end condition on each member, generating fixed-end moments (FEMs) that are standard results from beam tables. Two cases that appear most frequently:
Uniform load w over a span L, both ends fixed: FEM = ±wL2/12
Concentrated load P at midspan, both ends fixed: FEM = ±PL/8
The sign convention assigns clockwise moments positive. At an internal joint, moments from adjacent spans are summed; the algebraic difference is the unbalanced moment that the distribution process corrects.
The Iteration Procedure
| Step | Action |
|---|---|
| 1 | Compute FEMs for all spans under the applied loading |
| 2 | Compute DF and COF for all member ends at each joint |
| 3 | Sum FEMs at each free joint; compute the unbalanced moment |
| 4 | Distribute the unbalanced moment (reversed in sign) using DFs |
| 5 | Carry over half of each distributed value to the far end of each member |
| 6 | Return to step 3; repeat until residuals are negligible |
| 7 | Sum all values in each column to obtain final end moments |
Convergence is rapid for typical structures. Three to five cycles reduce residual unbalanced moments to below 1% of the initial FEMs. For hand calculation, four cycles are generally sufficient for engineering accuracy.
Numerical Example: Two-Span Continuous Beam
Consider a continuous beam of two equal spans, each L = 6 m, carrying a uniform dead load of 20 kN/m. Both end supports are pinned; the interior support is a continuous column. Flexural stiffness EI is uniform throughout.
Fixed-end moments for each span: FEM = wL2/12 = 20 × 36 / 12 = 60 kN·m. At the left end of span 1 and the right end of span 2, the FEMs act on pin supports, so they are immediately released (carried-over moment = 0 because COF = 0 for a pinned far end). At the interior joint, span 1 contributes +60 kN·m and span 2 contributes −60 kN·m; the unbalanced moment is +120 kN·m. With equal spans and equal EI, each member has the same stiffness factor at the interior joint, so DF = 0.5 for each. Distributing: each member receives −60 kN·m. After carry-over and immediate release at the pinned ends, the interior support moment converges to −60 kN·m — exactly matching the classical three-moment equation result.
The moment distribution method is physically transparent: stiffer members attract larger fractions of the unbalanced moment. This insight — that stiffness drives moment distribution — is the clearest conceptual framework for understanding why moment concentrates at rigid connections and stiff supports. No matrix output communicates this as directly.
Scope and Limitations
Moment distribution applies directly to prismatic or non-prismatic members in flexure: continuous beams of any number of spans, non-sway frames, and — with an iterative sway correction — laterally loaded frames. Its limitations are those of linear elastic analysis: it does not capture material nonlinearity, large deformations, or geometric buckling effects. For sway frames with multiple stories, the sway correction becomes cumbersome compared to direct stiffness methods. Modern software uses matrix stiffness analysis for such structures, but engineers who understand moment distribution have a reliable mental model for checking whether software output is in the right ballpark — a check that remains valuable regardless of the tool used to generate numbers.