Flexibility Method for Indeterminate Structures

Introduction and Context

The flexibility method, also called the force method or consistent-deformation method, is a classical and rigorous procedure for analysing statically indeterminate structures. Unlike the stiffness method, which treats joint displacements as primary unknowns, the flexibility method treats selected redundant forces or moments as the primary unknowns. Compatibility of deformation supplies the additional equations that equilibrium alone cannot provide.

The method is especially valuable for understanding structural behaviour, checking computer-generated solutions, analysing structures with a small degree of static indeterminacy, and developing insight into the relationship between loads, flexibility, energy and displacement. It is applicable to beams, plane and space trusses, rigid frames, arches and other linearly elastic structural systems.

At postgraduate level, the method should not be viewed merely as a hand-calculation technique. It provides the theoretical basis for flexibility matrices, virtual work, unit-load methods, force-based finite-element formulations and several nonlinear or adaptive analysis procedures.

Learning Objectives

After studying this chapter, the reader should be able to:

1. distinguish static indeterminacy from kinematic indeterminacy;
2. select suitable redundants and construct a determinate primary structure;
3. derive and apply the compatibility equations of the flexibility method;
4. calculate flexibility coefficients using integration, virtual work or matrix procedures;
5. analyse indeterminate beams, frames and trusses;
6. interpret the physical meaning of flexibility coefficients and reciprocal relations;
7. identify modelling assumptions, numerical difficulties and sources of error; and
8. use the method critically in professional structural-analysis practice.

1. Fundamental Idea

A structure is statically determinate when all support reactions and internal force resultants can be found from the equations of equilibrium. In a plane structure, the available global equilibrium equations are ordinarily three: ∑Fₓ = 0, ∑Fᵧ = 0 and ∑M = 0. If the number of independent force unknowns exceeds the number of independent equilibrium equations, the structure is statically indeterminate.

The degree of static indeterminacy is commonly written as

where Dₛ is the degree of static indeterminacy, Nᵤ is the number of independent force unknowns and Nₑ is the number of independent equilibrium equations, after accounting for internal releases, mechanisms and possible redundancies in the equations. For complete structures, this simple counting rule must be supplemented by a stability and connectivity check.

The flexibility method removes Dₛ unknown force quantities by selecting them as redundants. The remaining structure, called the primary structure or released structure, must be stable and statically determinate. The redundant forces are then restored as unknown loads, and compatibility conditions require the released displacement components to have their original prescribed values.

For a structure with r redundants, the general compatibility equation is

Here fᵢⱼ is the flexibility coefficient: displacement in coordinate i caused by a unit value of redundant j; Xⱼ is redundant j; Δᵢ⁰ is the prescribed displacement at coordinate i in the actual structure, often zero; and Δᵢᵖ is displacement at coordinate i caused by the applied loads on the primary structure. The sign convention must be consistent throughout.

In matrix form,

where [f] is the flexibility matrix. For a linear elastic, stable structure, [f] is normally symmetric and positive definite when the chosen coordinates are independent.

2. Choice of Redundants and Primary Structure

A redundant may be a support reaction, an internal force, an end moment, a shear force, an axial force or any generalized force that can be removed to produce a valid determinate structure. Good selection is central to an efficient solution.

A suitable primary structure must satisfy four conditions:

• it is stable under the relevant loading;
• it is statically determinate, or has a lower and manageable indeterminacy;
• removal of the redundant does not introduce an unintended mechanism; and
• the released coordinate is clearly defined, including its positive direction.

For example, a fixed-end beam may be converted into a simply supported beam by selecting the fixed-end moment and horizontal reaction as redundants, provided the resulting release is stable. A propped cantilever is often converted into a cantilever by selecting the prop reaction as the single redundant.

Redundant selection is not unique. Different selections produce different intermediate flexibility matrices but the same final reactions and internal forces, provided all compatibility conditions and signs are correctly applied. In computational work, a poor choice may cause an ill-conditioned flexibility matrix even though the physical structure is well behaved.

3. General Derivation from Superposition

Consider a linear elastic structure in which redundant forces X₁, X₂, …, Xᵣ are applied to the primary structure. By superposition, the displacement at coordinate i is

The actual structure imposes a known displacement Δᵢ⁰. For an unyielding support or an unbroken internal connection, Δᵢ⁰ = 0. Therefore,

These equations are called the force-method or compatibility equations. Once the redundants have been obtained, all other reactions and member forces follow from equilibrium and superposition.

The formulation assumes small displacements, linear elastic material response, compatible geometry, and load-independent structural properties. If support settlements, temperature changes or fabrication errors occur, they enter through Δᵢ⁰ or as separate imposed-deformation terms.

4. Flexibility Coefficients by Virtual Work

For beams and frames in which bending dominates, the displacement produced by a real loading system can be calculated by the unit-load method. Let M(x), N(x) and V(x) be the internal bending moment, axial force and shear force due to the actual loading, and let mᵢ(x), nᵢ(x) and vᵢ(x) be the corresponding internal actions due to a unit load applied in coordinate i. Then

E is Young’s modulus, I is the second moment of area, A is cross-sectional area, G is shear modulus, and κ is the shear-correction factor. The integral extends over all members or segments. For slender beams, axial and shear terms are often neglected, giving

For a flexibility coefficient,

The products must use signed internal actions under a consistent convention. Units are important: a translational flexibility has units of displacement/force, while a rotational flexibility has units of rotation/moment. Mixed coefficients have corresponding mixed units.

For a pin-jointed truss, deformation is usually assumed to arise from axial strain. If member k has length Lₖ, area Aₖ and modulus Eₖ, then

where Nₖ⁽ⁱ⁾ is the force in member k caused by a unit load in coordinate i. This expression is applicable when joints are frictionless, members carry axial force only and linear elasticity is valid.

5. Maxwell–Betti Reciprocity

For a linear elastic structure, the displacement at coordinate i caused by a unit force at coordinate j equals the displacement at coordinate j caused by a unit force at coordinate i:

This is Maxwell–Betti reciprocity. It follows from equality of mutual work and is a powerful check on calculations. For example, a rotation at A caused by a unit moment at B must equal the rotation at B caused by a unit moment at A, provided the coordinates and units are defined consistently.

The diagonal terms fᵢᵢ are self-flexibilities and are normally positive. If a computed self-flexibility is negative in a stable linear elastic model, a sign, integration or coordinate error should be suspected. Reciprocity may fail in systems with nonconservative effects, material asymmetry, path-dependent behaviour or inconsistent tangent operators.

6. Energy Interpretation

The strain energy stored in a linearly elastic structure is

For a truss, the bending and shear terms are absent. Castigliano’s second theorem states that the displacement corresponding to a force is obtained by differentiating strain energy with respect to that force:

where Qᵢ is the generalized force associated with coordinate i. This theorem supplies an alternative derivation of flexibility coefficients and is particularly useful when internal-force expressions are available but direct deflection diagrams are inconvenient.

The complementary-energy viewpoint is also important in advanced analysis. The force method is naturally force-oriented, and its stationary conditions can be interpreted as enforcing compatibility through energy minimization under equilibrium constraints.

7. Standard Solution Procedure

A reliable solution follows this sequence:

1. Establish geometry, supports, releases, material properties and sign conventions.
2. Determine the degree of static indeterminacy and check stability.
3. Select independent redundants and release them to create the primary structure.
4. Analyse the primary structure under actual loads to obtain Δᵢᵖ.
5. Apply a unit value of each redundant separately and calculate fᵢⱼ.
6. Assemble and solve [f]{X} = {Δ}⁰ − {Δ}ᵖ.
7. Recover all support reactions and internal forces by superposition.
8. Check equilibrium, compatibility, reciprocity and, where useful, energy consistency.

A flexibility diagram or a table of member actions is often more reliable than repeated freehand sketches. For piecewise members, divide integrations at every point where loading, stiffness or internal action changes.

8. Worked Example: Propped Cantilever

Consider a prismatic cantilever beam of length L, flexural rigidity EI, fixed at A and propped vertically at B. A downward point load P acts at the free end B. Neglect shear deformation and axial deformation. The prop reaction R_B is selected as the redundant, positive upward.

The primary structure is the cantilever AB without the prop. At B, the downward load produces the displacement

The unit upward force at B produces the flexibility coefficient

The actual vertical displacement at B is zero, so

Consequently,

The vertical reaction at A is therefore zero, because the applied load and prop reaction are equal and opposite. Taking moments about A, the fixing moment is zero for this particular loading arrangement. The result is physically reasonable: the prop directly balances the load at the same point, although the beam still carries a bending moment distribution associated with the load and reaction pair between A and B. For a point load applied at a distance a < L, or for a distributed load, the support reactions and fixing moment would not simplify in this way.

The example illustrates an important principle: compatibility, not equilibrium alone, determines the redundant. It also shows why the primary structure must be correctly defined and why the coordinate direction must be retained in the signs.

9. Illustrative Two-Redundant Frame Procedure

For a plane rigid frame with two selected redundants X₁ and X₂, such as a horizontal reaction and a fixing moment, the equations are

With reciprocity, f₁₂ = f₂₁. The determinant is

If D ≠ 0, the redundants are

This explicit form is useful for hand calculations, although matrix factorization is preferable for larger systems. A nearly zero determinant indicates dependent coordinates, a mechanism in the released model or severe numerical ill-conditioning.

10. Support Settlements, Temperature and Lack of Fit

The flexibility method handles imposed deformations naturally. If a support settles vertically by s in the positive coordinate direction, then the corresponding prescribed displacement is Δ⁰ = s. The compatibility equation becomes

Temperature changes may cause free expansion or curvature. For a uniform temperature change ΔT, a member with coefficient of thermal expansion α undergoes free axial strain αΔT. If restrained, the resulting generalized displacement contribution must be included in the compatibility equation. For a temperature gradient producing curvature χ_T, the beam contribution may be written in the virtual-work integral as

subject to the adopted curvature sign convention. Fabrication errors, lack of fit and imposed joint rotations are treated in the same conceptual manner: they are prescribed deformations, not external forces.

11. Matrix Flexibility Formulation

For a structure with many possible force coordinates, the method can be written compactly as

where {q} is the vector of generalized redundant forces and {d} is the vector of compatible generalized displacements. Prescribed displacements impose constraints on {d}. In a force-based finite-element formulation, element flexibility relations are assembled into a global flexibility system, while equilibrium transformations connect member forces to external actions.

The stiffness and flexibility methods are mathematically dual. In a stiffness formulation, generalized displacements are primary and forces are recovered through stiffness relations. In a flexibility formulation, generalized forces are primary and displacements are recovered through flexibility relations. The choice depends on structure type, number of redundants, support conditions, desired outputs and numerical conditioning.

12. Assumptions and Limitations

The elementary method generally assumes:

• small displacements and rotations;
• linear elastic, reversible material behaviour;
• known and constant E, A and I over each segment;
• idealized supports and connections;
• compatibility of connected members; and
• absence of instability, yielding or major geometric nonlinearity.

For reinforced concrete, cracking and creep may make EI load- and time-dependent. For steel structures, yielding, residual stress, semi-rigid connections and second-order effects may invalidate a simple elastic solution. In cable structures or very slender frames, geometric nonlinearity can be dominant. The method can be extended incrementally using tangent properties, but the classical superposition equations no longer apply globally.

Shear flexibility is often neglected in slender beams but may be important in deep beams, short spans, sandwich members and some composite systems. Torsional flexibility must be included when torsional coordinates are present. The analyst should never omit a deformation mode merely because the resulting calculation is inconvenient.

13. Case Study: Redundant Reaction in a Bridge Floor Beam

Background

Consider an illustrative bridge-floor beam continuous over several supports. The beam has nearly uniform steel properties, but one interior support is founded on a compressible bearing. The design team needs the reaction at that support and the effect of a prescribed support settlement.

Problem

A force-method model selects the interior support reaction as the redundant and releases that support. The released beam is stable and determinate. The analysis must include traffic loading, self-weight and a settlement of 8 mm downward at the selected support.

Analysis and approach

The engineer first calculates the displacement at the released support due to actual loads, Δᵖ, using unit-load integration. The flexibility coefficient f is calculated using the same beam model. The settlement is represented as Δ⁰ = −8 mm under the adopted upward-positive convention. The redundant is obtained from

The sign of X indicates whether the real support reaction acts in the assumed direction. The reaction is then restored to obtain bending moments and shears in every span. Independent equilibrium checks are performed, and the calculated displacement at the support is verified to equal −8 mm.

Outcome

The settlement-induced reaction may increase moments in adjacent spans even when external traffic loads are unchanged. If the bearing is flexible rather than rigid, its spring flexibility is added to the structural flexibility. If k_b is the bearing stiffness, its flexibility is 1/k_b, and the compatibility equation uses

for the corresponding coordinate.

Learning points

The case demonstrates that the force method handles support movement directly, distinguishes structural flexibility from support flexibility, and provides a transparent interpretation of load redistribution. It also emphasizes that a support model must reflect actual restraint: treating a flexible bearing as perfectly rigid can produce misleading reactions.

14. Professional Practice, Verification and Technology

Professional analysis should include a clear record of redundant selection, coordinate signs, stiffness assumptions, load cases, deformation components and validation checks. Software output is not a substitute for engineering judgement. A robust workflow compares selected reactions with equilibrium, checks symmetry or reciprocity where applicable, examines deflected shapes and performs sensitivity studies for E, I, support stiffness and settlement.

Computer algebra can evaluate piecewise integrals, while numerical scripts can assemble flexibility matrices and solve multiple load cases. Finite-element programs generally use stiffness-based global solvers, but force-based finite elements and hybrid formulations retain the flexibility perspective. AI-assisted tools may help generate load-case matrices, identify likely redundants or detect anomalous results, but their output requires independent verification. Safety-critical decisions must remain traceable to validated mechanics, applicable design standards and qualified professional review.

Current practice increasingly combines force-based insight with numerical nonlinear analysis, probabilistic modelling and structural health monitoring. Measured displacement or strain data can be used to update uncertain flexibility parameters, but inverse identification is sensitive to sensor placement, noise and model inadequacy. Ethical practice requires transparent assumptions, conservative treatment of uncertainty and compliance with governing building and bridge codes.

15. Best-Practice Checklist

Before accepting a flexibility-method solution, confirm that:

• the structure is stable and its indeterminacy is correctly identified;
• every redundant is independent and has a defined positive direction;
• the primary structure is physically realizable and determinate;
• all relevant deformation modes are included;
• units of flexibility coefficients are consistent;
• fᵢⱼ = fⱼᵢ has been checked where reciprocity applies;
• equilibrium is satisfied after redundant forces are restored;
• prescribed settlements, temperature effects and lack of fit are included;
• the resulting displacement satisfies compatibility; and
• the model assumptions are appropriate for the material, geometry and load level.

Chapter Summary

The flexibility method analyses statically indeterminate structures by selecting redundant forces, releasing them to form a determinate primary structure, calculating load-induced and unit-load-induced displacements, and enforcing compatibility. Its central equation is [f]{X} = {Δ}⁰ − {Δ}ᵖ. Flexibility coefficients can be obtained from virtual work, strain energy, Castigliano’s theorem or member-based summation. Maxwell–Betti reciprocity provides an important verification condition, while the energy interpretation explains why the method is rigorous for linear elastic systems.

The method is particularly efficient for structures with few redundants and is conceptually valuable even when stiffness-based software is used. Correct redundant selection, sign conventions, deformation modelling and independent checks are essential. Support settlements, temperature effects, lack of fit and flexible bearings can be incorporated as imposed-deformation terms. Extensions to matrix flexibility and force-based finite elements connect the classical method with modern computational structural analysis.

Key Terms

Redundant: A force or moment not obtainable from equilibrium alone.

Primary structure: The released determinate structure used as the basis of analysis.

Flexibility coefficient: Displacement in one coordinate caused by a unit force or moment in another coordinate.

Force method: Another name for the flexibility method.

Compatibility: The requirement that connected structural components undergo mutually consistent deformation.

Static indeterminacy: The number of independent force unknowns exceeding independent equilibrium equations.

Unit-load method: A virtual-work procedure for calculating displacement.

Maxwell–Betti reciprocity: The equality fᵢⱼ = fⱼᵢ in a linear elastic reciprocal system.

Self-flexibility: A diagonal flexibility coefficient fᵢᵢ.

Generalized force: A force or moment associated with a generalized displacement.

Generalized displacement: A translation or rotation paired with a generalized force.

Castigliano’s theorem: An energy theorem relating displacement to the derivative of strain energy.

Prescribed displacement: A known movement such as settlement or imposed joint rotation.

Virtual work: Work calculated using a real action and a compatible virtual or unit action.

Flexibility matrix: The matrix relating generalized forces to generalized displacements.

Primary coordinate: A coordinate associated with a selected redundant.

Ill-conditioning: Numerical sensitivity caused by nearly dependent coordinates or very different scales.

Superposition: Addition of responses from separate load or redundant cases in a linear system.

Force-based finite element: A finite-element formulation using member forces or force parameters as primary variables.

Compatibility equation: An equation enforcing the prescribed displacement at a released coordinate.

Review Questions

Short Answer Questions

1. What is meant by a redundant in the flexibility method?
2. Why must the primary structure be stable?
3. State Maxwell–Betti reciprocity.
4. What physical quantity does fᵢⱼ represent?
5. Name two imposed deformations that can be included in a flexibility analysis.

Long Answer Questions

1. Derive the general matrix compatibility equation for a structure with r redundants.
2. Explain how flexibility coefficients are calculated by the unit-load method for a beam.
3. Compare the flexibility and stiffness methods in terms of unknowns, governing conditions and practical application.
4. Discuss the assumptions and limitations of the classical linear elastic flexibility method.
5. Explain the energy basis of the flexibility method using Castigliano’s second theorem.

Analytical / Application Questions

1. A propped cantilever carries a point load at its prop. Select the prop reaction as redundant, derive the compatibility equation and interpret the resulting reaction.
2. Develop the two-equation flexibility formulation for a frame with a horizontal reaction and an end moment as redundants. State the checks required before accepting the solution.
3. A continuous bridge beam experiences support settlement and has a flexible bearing. Explain how both effects should be represented in the compatibility equation and how the final reactions should be verified.

Multiple Choice Questions with Answers

Question 1

In the flexibility method, the primary unknowns are generally:

A. Joint displacements

B. Redundant forces or moments

C. Material strains only

D. Support coordinates

Correct Answer: B. Redundant forces or moments

Question 2

For a linear elastic reciprocal structure, which relation is valid?

A. fᵢⱼ = −fⱼᵢ

B. fᵢⱼ = 0 for all i and j

C. fᵢⱼ = fⱼᵢ

D. fᵢᵢ = 0

Correct Answer: C. fᵢⱼ = fⱼᵢ

Question 3

For a slender beam in which only bending deformation is considered, the unit-load displacement is calculated from:

A. ∫M m/(EI) dx

B. ∫M m EI dx

C. ∫N n/(EA) dx only

D. ∑PL

Correct Answer: A. ∫M m/(EI) dx

Question 4

A suitable primary structure must be:

A. Unstable and highly indeterminate

B. Stable and determinate after release

C. Free of all loads

D. Rigid with no deformation

Correct Answer: B. Stable and determinate after release

Question 5

The compatibility equation for one redundant X is:

A. X = f + Δ

B. 0 = Δᵖ + fX when the prescribed displacement is zero

C. 0 = Δᵖ − f/X

D. f = XΔ

Correct Answer: B. 0 = Δᵖ + fX when the prescribed displacement is zero

Question 6

For a pin-jointed truss, member flexibility contributions are commonly based on:

A. Bending moment squared only

B. Nₖ⁽ⁱ⁾Nₖ⁽ʲ⁾Lₖ/(AₖEₖ)

C. VₖLₖ/(GAₖ) only

D. EI/Lₖ

Correct Answer: B. Nₖ⁽ⁱ⁾Nₖ⁽ʲ⁾Lₖ/(AₖEₖ)

Question 7

A support settlement in the flexibility method is represented as:

A. A prescribed displacement term

B. A reduction in Young’s modulus

C. A unit load in every member

D. A change in the number of equilibrium equations

Correct Answer: A. A prescribed displacement term

Question 8

The diagonal flexibility coefficient fᵢᵢ is called the:

A. Cross-flexibility

B. Self-flexibility

C. Equilibrium coefficient

D. Stiffness determinant

Correct Answer: B. Self-flexibility

Question 9

If the determinant of a two-redundant flexibility matrix is nearly zero, this may indicate:

A. Perfect numerical conditioning

B. Dependent coordinates or ill-conditioning

C. Zero strain energy in every member

D. That compatibility is unnecessary

Correct Answer: B. Dependent coordinates or ill-conditioning

Question 10

Which condition is essential when restoring redundants after solving the flexibility equations?

A. Only local equilibrium must hold

B. The prescribed displacement compatibility must be satisfied

C. All member forces must be equal

D. All flexibility coefficients must be zero

Correct Answer: B. The prescribed displacement compatibility must be satisfied