SA-06-02 Chapter 06 · Matrix Methods of Structural Analysis

Stiffness Method (Displacement & Direct Assembly)

In this chapter (4 topics)
On this page

1. Introduction & Executive Overview

The Stiffness Method—also universally designated as the Displacement Method or Direct Stiffness Method (DSM)—is the foundational computational engine underpinning modern structural engineering and Finite Element Analysis (FEA) software (including ETABS, SAP2000, STAAD.Pro, and ANSYS). While the Flexibility (Force) Method selects internal redundant forces (DsD_s) as unknowns, the Stiffness Method treats the unknown joint displacements—translations and rotations (DkD_k)—as primary unknowns.

The formulation begins from a kinematically locked base system, where every free degree of freedom is artificially clamped (Δ=0\Delta = 0). External member loads generate fixed-end holding forces and moments ([PF][P_F]). The artificial clamps are then systematically released, allowing the structure to deform into its true equilibrium state under the action of net unbalanced joint forces [Pnet]=[P]−[PF][P_{\text{net}}] = [P] - [P_F].

   Actual Indeterminate Structure (Dk active DOFs)
                      │
                      ▼
   Kinematically Locked Base System (All DOFs Δ = 0)
   - Apply member span loads → Generate Holding Forces [P_F]
                      │
                      ▼
   Joint Deformation & Equilibrium ([K][Δ] = [P] - [P_F])
   - Direct Assembly of Member Matrices: [K] = Σ [k_element]
   - Invert Global Stiffness: [Δ] = [K]⁻¹ · [P_net]
                      │
                      ▼
   Member Action Recovery
   - Internal Shears, Moments & Axial Forces: {p} = [k]{d} + {p_F}

Core Axioms of the Stiffness Formulation

  1. Kinematic Completeness: The total number of primary unknowns is strictly equal to the Degree of Kinematic Indeterminacy (DkD_k) of the structure.
  2. Equilibrium Equations at Joints: The governing equations express dynamic or static force and moment balance at each unconstrained degree of freedom (∑Fx=0\sum F_x = 0, ∑Fy=0\sum F_y = 0, ∑Mz=0\sum M_z = 0).
  3. Physical Meaning of Stiffness Coefficient (kijk_{ij}): kijk_{ij} represents the force (or moment) developed at coordinate ii when a unit displacement (or rotation) is introduced at coordinate jj, while maintaining all other degrees of freedom strictly fixed (Δj=1\Delta_j = 1, Δm=0\Delta_m = 0 for all m≠jm \neq j).
  4. Symmetry & Positive Definiteness: By Betti-Maxwell reciprocal theorem, kij=kjik_{ij} = k_{ji}. For any kinematically stable structure, the global stiffness matrix [K][K] is strictly symmetric and positive definite ([X]T[K][X]>0[X]^T [K] [X] > 0 for all nonzero displacement vectors [X][X]), guaranteeing a unique, non-singular solution with strictly positive diagonal terms (kii>0k_{ii} > 0).

2. Fundamentals & Physical Mechanics

2.1 Degrees of Freedom & Kinematic Indeterminacy (DkD_k)

In matrix structural analysis, nodes (joints) define the discrete connectivity of elements. Each node possesses unconstrained degrees of freedom depending on structural topology:

  • 2D Truss Joints: 2 translational DOFs per joint (u,vu, v). Rotations are unrestrained and zero-moment pins.
  • 2D Beam / Frame Joints: 3 DOFs per joint (u,v,θu, v, \theta). If axial inextensibility is assumed, axial translations (uu) are suppressed, reducing beam joints to transverse displacement and rotation (v,θv, \theta).
  • 3D Frame Joints: 6 DOFs per joint (u,v,w,θx,θy,θzu, v, w, \theta_x, \theta_y, \theta_z).

The global degree of kinematic indeterminacy is:

Dk=ndof⋅j−rs−mrelD_k = n_{\text{dof}} \cdot j - r_s - m_{\text{rel}}

where jj is the total number of joints, ndofn_{\text{dof}} is the degrees of freedom per joint (3 for planar frames, 2 for planar trusses), rsr_s is the number of support constraints, and mrelm_{\text{rel}} is internal releases (such as internal hinges where additional relative rotation occurs).

2.2 Physical Interpretation of Stiffness Coefficients

To visualize kijk_{ij}, consider a multi-degree of freedom beam system:

  • When a unit vertical displacement (Δ1=1\Delta_1 = 1) is applied at DOF 1 with Δ2=Δ3=⋯=0\Delta_2 = \Delta_3 = \dots = 0:
    • k11k_{11} is the reactive vertical force at DOF 1 resisting the unit displacement (direct stiffness, always >0> 0).
    • k21k_{21} is the reactive moment developed at DOF 2 to keep DOF 2 from rotating during the translation of DOF 1 (cross-coupling stiffness).
    • k31k_{31} is the reactive force at DOF 3, etc.
  • Dimensional homogeneity of stiffness coefficients:
    • Force per unit displacement: [k]=kN/m[k] = \text{kN/m} or N/mm\text{N/mm} (when both ii and jj are translations).
    • Moment per unit rotation: [k]=kN⋅m/rad[k] = \text{kN}\cdot\text{m/rad} or N⋅mm/rad\text{N}\cdot\text{mm/rad} (when both ii and jj are rotations).
    • Force per unit rotation OR Moment per unit displacement: [k]=kN/rad=kN[k] = \text{kN/rad} = \text{kN} or kN⋅m/m=kN\text{kN}\cdot\text{m/m} = \text{kN} (cross coupling between translation and rotation).

3. Mathematical Derivations & Theoretical Formulations

3.1 Prismatic Flexural Beam Element (Euler-Bernoulli Beam)

Consider an isolated 2D flexural member of length LL, flexural rigidity EIEI, with 4 local degrees of freedom ordered as:

{dˉ}=[vˉ1θˉ1vˉ2θˉ2]T\{\bar{d}\} = \begin{bmatrix} \bar{v}_1 & \bar{\theta}_1 & \bar{v}_2 & \bar{\theta}_2 \end{bmatrix}^T

where vˉ1,vˉ2\bar{v}_1, \bar{v}_2 are transverse displacements and θˉ1,θˉ2\bar{\theta}_1, \bar{\theta}_2 are end rotations.

       v1, θ1                                     v2, θ2
        ▲   ╭─╮                                    ▲   ╭─╮
        │   │ ↻                                    │   │ ↻
     (1)●───┴──────────────────────────────────────●(2)
        └────────────────────── L ─────────────────┘

The governing differential equation for flexure in the absence of intermediate span loads is:

EId4v(x)dx4=0  ⟹  v(x)=C1x3+C2x2+C3x+C4EI \frac{d^4 v(x)}{dx^4} = 0 \implies v(x) = C_1 x^3 + C_2 x^2 + C_3 x + C_4

Applying the four hermite boundary conditions:

  1. v(0)=vˉ1v(0) = \bar{v}_1
  2. dvdx∣x=0=θˉ1\left.\frac{dv}{dx}\right|_{x=0} = \bar{\theta}_1
  3. v(L)=vˉ2v(L) = \bar{v}_2
  4. dvdx∣x=L=θˉ2\left.\frac{dv}{dx}\right|_{x=L} = \bar{\theta}_2

The internal shear force V(x)=−EId3vdx3V(x) = -EI \frac{d^3 v}{dx^3} and bending moment M(x)=EId2vdx2M(x) = EI \frac{d^2 v}{dx^2} yield the classical 4×44 \times 4 element stiffness matrix [kˉ][\bar{k}]:

[kˉ]=[12EIL36EIL2−12EIL36EIL26EIL24EIL−6EIL22EIL−12EIL3−6EIL212EIL3−6EIL26EIL22EIL−6EIL24EIL][\bar{k}] = \begin{bmatrix} \frac{12EI}{L^3} & \frac{6EI}{L^2} & -\frac{12EI}{L^3} & \frac{6EI}{L^2} \\ \frac{6EI}{L^2} & \frac{4EI}{L} & -\frac{6EI}{L^2} & \frac{2EI}{L} \\ -\frac{12EI}{L^3} & -\frac{6EI}{L^2} & \frac{12EI}{L^3} & -\frac{6EI}{L^2} \\ \frac{6EI}{L^2} & \frac{2EI}{L} & -\frac{6EI}{L^2} & \frac{4EI}{L} \end{bmatrix}
Mechanical Signatures:
  • Pure Unit Rotation at End 1 (θˉ1=1\bar{\theta}_1 = 1, all other DOFs 0):
    • Near-end moment: mˉ1=4EIL\bar{m}_1 = \frac{4EI}{L}
    • Far-end moment: mˉ2=2EIL\bar{m}_2 = \frac{2EI}{L}
    • Carry-Over Factor (COF): COF=mˉ2mˉ1=2EI/L4EI/L=12=0.50\text{COF} = \frac{\bar{m}_2}{\bar{m}_1} = \frac{2EI/L}{4EI/L} = \frac{1}{2} = 0.50
    • Equilibrium shears: vˉ1=6EIL2\bar{v}_1 = \frac{6EI}{L^2} (upward), vˉ2=−6EIL2\bar{v}_2 = -\frac{6EI}{L^2} (downward).
  • Pure Unit Relative Settlement (vˉ1=1\bar{v}_1 = 1, all other DOFs 0):
    • Restoring shears: vˉ1=12EIL3\bar{v}_1 = \frac{12EI}{L^3}, vˉ2=−12EIL3\bar{v}_2 = -\frac{12EI}{L^3}
    • Induced end moments: mˉ1=6EIL2\bar{m}_1 = \frac{6EI}{L^2}, mˉ2=6EIL2\bar{m}_2 = \frac{6EI}{L^2}.

3.2 2D Pin-Jointed Truss Element & Coordinate Transformation

For an axial bar in local coordinates (axial translation uˉ1,uˉ2\bar{u}_1, \bar{u}_2 only):

[kˉ]=EAL[1−1−11][\bar{k}] = \frac{EA}{L} \begin{bmatrix} 1 & -1 \\ -1 & 1 \end{bmatrix}

In a 2D global frame (X,Y)(X, Y), the element is inclined at angle θ\theta to the positive XX-axis. Each node has two translational DOFs: {d}=[u1,v1,u2,v2]T\{d\} = [u_1, v_1, u_2, v_2]^T. The direction cosines are:

λx=cos⁡θ=X2−X1L,λy=sin⁡θ=Y2−Y1L,L=(X2−X1)2+(Y2−Y1)2\lambda_x = \cos\theta = \frac{X_2 - X_1}{L}, \quad \lambda_y = \sin\theta = \frac{Y_2 - Y_1}{L}, \quad L = \sqrt{(X_2 - X_1)^2 + (Y_2 - Y_1)^2}

The transformation matrix [T][T] relating local displacements to global displacements ({dˉ}=[T]{d}\{\bar{d}\} = [T] \{d\}) is:

[T]=[λxλy0000λxλy][T] = \begin{bmatrix} \lambda_x & \lambda_y & 0 & 0 \\ 0 & 0 & \lambda_x & \lambda_y \end{bmatrix}

Since [T][T] is an orthogonal transformation matrix, the global element stiffness matrix [kglobal][k_{\text{global}}] is obtained via contragredience:

[kglobal]=[T]T[kˉ][T]=EAL[λx2λxλy−λx2−λxλyλxλyλy2−λxλy−λy2−λx2−λxλyλx2λxλy−λxλy−λy2λxλyλy2][k_{\text{global}}] = [T]^T [\bar{k}] [T] = \frac{EA}{L} \begin{bmatrix} \lambda_x^2 & \lambda_x \lambda_y & -\lambda_x^2 & -\lambda_x \lambda_y \\ \lambda_x \lambda_y & \lambda_y^2 & -\lambda_x \lambda_y & -\lambda_y^2 \\ -\lambda_x^2 & -\lambda_x \lambda_y & \lambda_x^2 & \lambda_x \lambda_y \\ -\lambda_x \lambda_y & -\lambda_y^2 & \lambda_x \lambda_y & \lambda_y^2 \end{bmatrix}

3.3 The Direct Stiffness Assembly Algorithm

The global stiffness matrix [K][K] of an assemblage of mm elements is formed by accumulating element stiffness contributions at shared global DOFs:

[K]=∑e=1m[Te]T[kˉe][Te][K] = \sum_{e=1}^{m} [T_e]^T [\bar{k}_e] [T_e]

In practice, assembly is executed via the Code Number Array (Destination Vector): For element ee connecting joints with assigned global DOF numbers (i,j,k,l)(i, j, k, l), the element stiffness term krsek_{rs}^e is added directly to the global matrix location:

Kcode(r),code(s)←Kcode(r),code(s)+krseK_{\text{code}(r), \text{code}(s)} \leftarrow K_{\text{code}(r), \text{code}(s)} + k_{rs}^e

Terms associated with restrained boundary DOFs (supports) are discarded or directed into support reaction recovery vectors.

3.4 Load Vector Assembly & Support Settlement

The governing structural equation is:

[K][Δ]=[P]−[PF]=[Pnet][K] [\Delta] = [P] - [P_F] = [P_{\text{net}}]

Where:

  • [P][P]: Direct nodal loads applied externally at active degrees of freedom.
  • [PF][P_F]: Fixed-end holding force vector at active DOFs due to span loads acting on locked members: [PF]=∑e[Te]T{pˉF,e}[P_F] = \sum_{e} [T_e]^T \{\bar{p}_{F, e}\}
  • [Pnet][P_{\text{net}}]: Equivalent nodal force vector driving the free joint deformations.
Support Settlement & Known Boundary Displacements:

Partitioning active unconstrained DOFs (subscript ff) and specified support settlements (subscript ss):

[[Kff][Kfs][Ksf][Kss]][[Δf][Δs]]=[[Pf]−[PFf][Rs]−[PFs]]\begin{bmatrix} [K_{ff}] & [K_{fs}] \\ [K_{sf}] & [K_{ss}] \end{bmatrix} \begin{bmatrix} [\Delta_f] \\ [\Delta_s] \end{bmatrix} = \begin{bmatrix} [P_f] - [P_{Ff}] \\ [R_s] - [P_{Fs}] \end{bmatrix}

Expanding the first row for unknown displacements [Δf][\Delta_f]:

[Kff][Δf]+[Kfs][Δs]=[Pf]−[PFf][K_{ff}] [\Delta_f] + [K_{fs}] [\Delta_s] = [P_f] - [P_{Ff}] [Kff][Δf]=[Pf]−[PFf]−[Kfs][Δs][K_{ff}] [\Delta_f] = [P_f] - [P_{Ff}] - [K_{fs}] [\Delta_s]

Support settlements [Δs][\Delta_s] act as equivalent loads −[Kfs][Δs]-[K_{fs}] [\Delta_s] added to the net driving vector. Once [Δf][\Delta_f] is solved, support reactions [Rs][R_s] are recovered from the second partitioned row:

[Rs]=[Ksf][Δf]+[Kss][Δs]+[PFs][R_s] = [K_{sf}] [\Delta_f] + [K_{ss}] [\Delta_s] + [P_{Fs}]

4. Engineering Diagrams & Visual Explanations

Figure 1: Direct Stiffness Method — Kinematic Locking & Assembly

The diagram below illustrates the fundamental physical concept: transforming an indeterminate continuous beam into a kinematically locked primary structure by introducing artificial rotation clamps, evaluating fixed-end reactions, and releasing the clamps through global matrix equilibrium.

Stiffness Method Concept and Assembly
Analytical Breakdown of Figure 1:
  1. Actual System vs Clamped Base: At joints B and C, rotational DOFs θB\theta_B and θC\theta_C are free (Dk=2D_k = 2). Introducing rotational clamps locks all joint movements (Δ=0\Delta = 0).
  2. Fixed-End Holding Moments (PFP_F): External span loads acting on the locked members produce standard fixed-end moments (−wL2/12-wL^2/12 and +wL2/12+wL^2/12). At joint B, the clamp carries the out-of-balance holding moment PF1=MF,BA+MF,BCP_{F1} = M_{F,BA} + M_{F,BC}.
  3. Physical Stiffness Columns (k11,k21k_{11}, k_{21}): Applying a unit rotation θB=1\theta_B = 1 while keeping clamp C locked (θC=0\theta_C = 0) mobilizes flexural resistance from both adjoining spans: k11=4EI/L1+4EI/L2k_{11} = 4EI/L_1 + 4EI/L_2. The carry-over moment to joint C produces cross-coupling stiffness k21=2EI/L2k_{21} = 2EI/L_2.
  4. Symmetry & Reciprocity: Rotating joint C (θC=1\theta_C = 1) induces k12=2EI/L2k_{12} = 2EI/L_2, demonstrating exact symmetry k12=k21k_{12} = k_{21} in full compliance with Betti's law.

Figure 2: Element Stiffness Matrices & Coordinate Transformations

The following diagram details the local 4×44 \times 4 flexural stiffness matrix, the geometric orientation of inclined members with direction cosines (λx,λy\lambda_x, \lambda_y), and the end-to-end 6-stage algorithmic workflow for computational Direct Stiffness Analysis.

Element Stiffness and Coordinate Transformation
Analytical Breakdown of Figure 2:
  1. Local Flexural Coefficients: The 4×44 \times 4 matrix explicitly correlates end shears and moments to transverse deflections and rotations. Diagonal terms (12EI/L3,4EI/L12EI/L^3, 4EI/L) represent direct elastic resistance and are strictly positive. Off-diagonal terms demonstrate the classic 50%50\% moment carry-over factor (2EI/L2EI/L).
  2. Coordinate Transformation ([T][T]): The direction cosines λx=cos⁡θ\lambda_x = \cos\theta and λy=sin⁡θ\lambda_y = \sin\theta resolve local axial and shear actions into orthogonal global Cartesian axes (X,Y)(X, Y). Because transformation matrix [T][T] is orthogonal ([T]−1=[T]T[T]^{-1} = [T]^T), energy is conserved during rotation.
  3. The 6-Stage Algorithm: Shows how finite element solvers discretize domains, formulate local matrices, transform coordinates, scatter entries into banded global matrices, apply boundary conditions, and back-substitute to recover internal member actions.

5. Practical Engineering Applications & Software Implementations

                     FEA COMPUTATIONAL PIPELINE
┌──────────────────────────────────────────────────────────────────┐
│ 1. Pre-Processing: Node Coordinates, Member Connectivity, Loads   │
│ 2. Local Element Formulation: [k_local] (Beam, Truss, Shell)      │
│ 3. Coordinate Transformation: [k_global] = [T]ᵀ · [k_local] · [T] │
│ 4. Skyline / Banded Direct Assembly: Global Sparse Matrix [K]     │
│ 5. Constraint Enforcement: Elimination or Penalty Method          │
│ 6. Equation Solver: Cholesky (LLᵀ), LDLᵀ, or Conjugate Gradient   │
│ 7. Post-Processing: Internal Forces, Moments, Stresses & Envelopes│
└──────────────────────────────────────────────────────────────────┘
  1. Computational Dominance over Force Method:
    • In commercial software, the Stiffness Method is universally preferred over the Flexibility Method because DkD_k can be automatically determined purely from nodal coordinates and joint types. In contrast, the Flexibility Method requires selecting released redundant bases (DsD_s), which is notoriously difficult to automate without creating unstable kinematic mechanisms.
  2. Matrix Sparsity & Bandwidth Optimization:
    • In large civil structures (e.g., a 100-story skyscraper with 100,000100,000 DOFs), [K][K] is over 99.5%99.5\% sparse (mostly zeros).
    • Nodes are renumbered using algorithms like Reverse Cuthill-McKee (RCM) to minimize the matrix bandwidth B=(DOFmax⁡−DOFmin⁡+1)B = (\text{DOF}_{\max} - \text{DOF}_{\min} + 1), drastically reducing storage and computational complexity from O(N3)\mathcal{O}(N^3) to O(N⋅B2)\mathcal{O}(N \cdot B^2).
  3. Condition Number & Numerical Stability:
    • The condition number κ(K)=λmax⁡λmin⁡\kappa(K) = \frac{\lambda_{\max}}{\lambda_{\min}} indicates numerical sensitivity. Ill-conditioning occurs when very stiff elements (e.g., shear walls modeled as beams) connect directly to highly flexible members, or when round-off errors occur in double precision calculations.
  4. Internal Hinge Handling:
    • An internal hinge is handled in DSM either by:
      • Adding a separate rotational degree of freedom at each side of the hinge (condensed node).
      • Static condensation of the element stiffness matrix prior to global assembly to zero out the moment at the hinged end.

6. Comparative Analysis Table

Characteristic / Parameter Flexibility Method (Force Method) Stiffness Method (Displacement Method) Slope-Deflection Method
Primary Unknowns Redundant internal forces / moments (RR) Joint displacements / rotations (Δ\Delta) Joint rotations (θ\theta) and chord drift (ψ\psi)
Governing Equations Geometric compatibility equations Static force / moment equilibrium Joint moment equilibrium (∑M=0\sum M = 0)
Number of Unknowns Static Indeterminacy (DsD_s) Kinematic Indeterminacy (DkD_k) Unconstrained joint rotations/sway
Base / Primary System Statically determinate released structure Kinematically locked clamped structure Restrained continuous spans
Matrix Properties [F][F]: Symmetric, positive definite [K][K]: Symmetric, positive definite, banded Symmetric system of scalar equations
Ease of Computation Manual base selection needed; hard to automate Highly systematic; standard for all FEA software Manual pencil-and-paper calculation
Support Settlement Effect Directly enters compatibility vector [Δs][\Delta_s] Enters as equivalent joint load −[Kfs][Δs]-[K_{fs}][\Delta_s] Enters via chord rotation angle ψ=δ/L\psi = \delta/L
Temperature / Misfits Added as free-structure displacement [DL0][D_{L0}] Converted into fixed-end forces [PF][P_F] Modifies fixed-end moments (FEMFEM)
Matrix Relationship [F]=[K]−1[F] = [K]^{-1} (for non-singular unconstrained) [K]=[F]−1[K] = [F]^{-1} Direct scalar precursor to Stiffness Matrix

7. Worked Numerical Examples

Problem 1: Continuous Beam ABC — Exact Dual Verification with Flexibility Method

Statement: A continuous beam ABCABC has span AB=4.0 mAB = 4.0\text{ m} and span BC=6.0 mBC = 6.0\text{ m}. Support AA is rigidly fixed, while supports BB and CC are frictionless rollers. Span ABAB carries a uniform distributed load w=12.0 kN/mw = 12.0\text{ kN/m}. Span BCBC carries a concentrated point load P=60.0 kNP = 60.0\text{ kN} at midspan (3.0 m3.0\text{ m} from BB). Flexural rigidity EIEI is constant throughout. Using the Stiffness Method:

  1. Formulate the global stiffness matrix [K][K].
  2. Assemble the fixed-end force vector [PF][P_F] and net equivalent load vector [Pnet][P_{\text{net}}].
  3. Compute joint rotations θB\theta_B and θC\theta_C.
  4. Determine all member end moments and verify exact duality with the Flexibility Method.
       w = 12 kN/m                    P = 60 kN
  ┌──────────────────┐                    │
  ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼  ▼                    ▼
  ══════════════════════════════╦══════════════════════════════
  ▒▒█                          ▲                              ▲
  ▒▒█ (Fixed)              (Roller)                       (Roller)
   A                            B                              C
  ├──────── 4.0 m ─────────────┼────────────── 6.0 m ─────────┤

Solution:

Step 1: Identify Active Kinematic Degrees of Freedom (DkD_k)

Neglecting axial deformation:

  • Support AA is fixed: θA=0\theta_A = 0, vA=0v_A = 0.
  • Support BB is a roller: vB=0v_B = 0, rotation θB\theta_B is free →\rightarrow DOF 1.
  • Support CC is a roller: vC=0v_C = 0, rotation θC\theta_C is free →\rightarrow DOF 2.
{Δ}=[θBθC],Dk=2\{\Delta\} = \begin{bmatrix} \theta_B \\ \theta_C \end{bmatrix}, \quad D_k = 2
Step 2: Formulate Global Stiffness Matrix [K][K]

Using direct physical stiffness definition:

  • Apply unit rotation θB=1\theta_B = 1 (θC=0\theta_C = 0):

    • Moment at BB from span BABA: kBA=4EIL1=4EI4=1.0EIk_{BA} = \frac{4EI}{L_1} = \frac{4EI}{4} = 1.0 EI
    • Moment at BB from span BCBC: kBC=4EIL2=4EI6=23EI≈0.6667EIk_{BC} = \frac{4EI}{L_2} = \frac{4EI}{6} = \frac{2}{3} EI \approx 0.6667 EI
    • Total moment at DOF 1: K11=kBA+kBC=1.0EI+23EI=53EIK_{11} = k_{BA} + k_{BC} = 1.0 EI + \frac{2}{3} EI = \frac{5}{3} EI
    • Carry-over moment to DOF 2 (joint CC): K21=2EIL2=2EI6=13EIK_{21} = \frac{2EI}{L_2} = \frac{2EI}{6} = \frac{1}{3} EI
  • Apply unit rotation θC=1\theta_C = 1 (θB=0\theta_B = 0):

    • Carry-over moment to DOF 1 (joint BB): K12=2EIL2=13EIK_{12} = \frac{2EI}{L_2} = \frac{1}{3} EI
    • Moment at CC from span CBCB: K22=4EIL2=4EI6=23EIK_{22} = \frac{4EI}{L_2} = \frac{4EI}{6} = \frac{2}{3} EI

Global stiffness matrix:

[K]=EI[53131323][K] = EI \begin{bmatrix} \frac{5}{3} & \frac{1}{3} \\ \frac{1}{3} & \frac{2}{3} \end{bmatrix}

Determinant of [K][K]:

det⁡[K]=(EI)2[(53)(23)−(13)(13)]=(EI)2[109−19]=(EI)2(99)=1.0(EI)2\det[K] = (EI)^2 \left[ \left(\frac{5}{3}\right)\left(\frac{2}{3}\right) - \left(\frac{1}{3}\right)\left(\frac{1}{3}\right) \right] = (EI)^2 \left[ \frac{10}{9} - \frac{1}{9} \right] = (EI)^2 \left(\frac{9}{9}\right) = 1.0 (EI)^2

Inverse matrix [K]−1[K]^{-1}:

[K]−1=11.0⋅EI[23−13−1353][K]^{-1} = \frac{1}{1.0 \cdot EI} \begin{bmatrix} \frac{2}{3} & -\frac{1}{3} \\ -\frac{1}{3} & \frac{5}{3} \end{bmatrix}
Step 3: Compute Fixed-End Holding Moments [PF][P_F] & Net Load Vector

Adopting standard clockwise-positive sign convention for end moments:

  • Span ABAB (L=4.0 m,w=12.0 kN/mL = 4.0\text{ m}, w = 12.0\text{ kN/m}):

    MF,AB=−wL1212=−12×4212=−16.0 kN⋅mM_{F,AB} = -\frac{w L_1^2}{12} = -\frac{12 \times 4^2}{12} = -16.0\text{ kN}\cdot\text{m} MF,BA=+wL1212=+12×4212=+16.0 kN⋅mM_{F,BA} = +\frac{w L_1^2}{12} = +\frac{12 \times 4^2}{12} = +16.0\text{ kN}\cdot\text{m}
  • Span BCBC (L=6.0 m,P=60.0 kNL = 6.0\text{ m}, P = 60.0\text{ kN} at midspan):

    MF,BC=−PL28=−60×68=−45.0 kN⋅mM_{F,BC} = -\frac{P L_2}{8} = -\frac{60 \times 6}{8} = -45.0\text{ kN}\cdot\text{m} MF,CB=+PL28=+60×68=+45.0 kN⋅mM_{F,CB} = +\frac{P L_2}{8} = +\frac{60 \times 6}{8} = +45.0\text{ kN}\cdot\text{m}
  • Holding Moments at Active DOFs ([PF][P_F]):

    • At DOF 1 (Joint BB): PF1=MF,BA+MF,BC=+16.0+(−45.0)=−29.0 kN⋅mP_{F1} = M_{F,BA} + M_{F,BC} = +16.0 + (-45.0) = -29.0\text{ kN}\cdot\text{m}
    • At DOF 2 (Joint CC): PF2=MF,CB=+45.0 kN⋅mP_{F2} = M_{F,CB} = +45.0\text{ kN}\cdot\text{m}
    [PF]=[−29.0+45.0] kN⋅m[P_F] = \begin{bmatrix} -29.0 \\ +45.0 \end{bmatrix}\text{ kN}\cdot\text{m}
  • Applied External Joint Moments [P][P]: No external moments are applied at joints BB or CC:

    [P]=[0.00.0] kN⋅m[P] = \begin{bmatrix} 0.0 \\ 0.0 \end{bmatrix}\text{ kN}\cdot\text{m}
  • Net Driving Vector [Pnet]=[P]−[PF][P_{\text{net}}] = [P] - [P_F]:

    [Pnet]=[0−(−29.0)0−45.0]=[+29.0−45.0] kN⋅m[P_{\text{net}}] = \begin{bmatrix} 0 - (-29.0) \\ 0 - 45.0 \end{bmatrix} = \begin{bmatrix} +29.0 \\ -45.0 \end{bmatrix}\text{ kN}\cdot\text{m}
Step 4: Solve for Unknown Joint Rotations
[θBθC]=[K]−1[Pnet]=1EI[23−13−1353][29.0−45.0]\begin{bmatrix} \theta_B \\ \theta_C \end{bmatrix} = [K]^{-1} [P_{\text{net}}] = \frac{1}{EI} \begin{bmatrix} \frac{2}{3} & -\frac{1}{3} \\ -\frac{1}{3} & \frac{5}{3} \end{bmatrix} \begin{bmatrix} 29.0 \\ -45.0 \end{bmatrix} θB=1EI[23(29.0)−13(−45.0)]=1EI[58.0+45.03]=+103.03EI≈+34.3333EI rad\theta_B = \frac{1}{EI} \left[ \frac{2}{3}(29.0) - \frac{1}{3}(-45.0) \right] = \frac{1}{EI} \left[ \frac{58.0 + 45.0}{3} \right] = +\frac{103.0}{3 EI} \approx +\frac{34.3333}{EI}\text{ rad} θC=1EI[−13(29.0)+53(−45.0)]=1EI[−29.0−225.03]=−254.03EI≈−84.6667EI rad\theta_C = \frac{1}{EI} \left[ -\frac{1}{3}(29.0) + \frac{5}{3}(-45.0) \right] = \frac{1}{EI} \left[ \frac{-29.0 - 225.0}{3} \right] = -\frac{254.0}{3 EI} \approx -\frac{84.6667}{EI}\text{ rad}
Step 5: Member End Actions & Flexibility Duality Verification

Using the member stiffness/slope-deflection relations (Mij=MF,ij+2EIL(2θi+θj)M_{ij} = M_{F,ij} + \frac{2EI}{L}(2\theta_i + \theta_j)):

  • Span ABAB:

    MAB=MF,AB+2EI4(2θA+θB)=−16.0+0.5EI(103.03EI)=−16.0+103.06=−96.0+103.06=+7.06 kN⋅m≈+1.1667 kN⋅mM_{AB} = M_{F,AB} + \frac{2EI}{4}(2\theta_A + \theta_B) = -16.0 + 0.5 EI \left(\frac{103.0}{3 EI}\right) = -16.0 + \frac{103.0}{6} = \frac{-96.0 + 103.0}{6} = +\frac{7.0}{6}\text{ kN}\cdot\text{m} \approx +1.1667\text{ kN}\cdot\text{m} MBA=MF,BA+2EI4(2θB+θA)=+16.0+1.0EI(103.03EI)=+16.0+103.03=48.0+103.03=+151.03 kN⋅m≈+50.3333 kN⋅mM_{BA} = M_{F,BA} + \frac{2EI}{4}(2\theta_B + \theta_A) = +16.0 + 1.0 EI \left(\frac{103.0}{3 EI}\right) = +16.0 + \frac{103.0}{3} = \frac{48.0 + 103.0}{3} = +\frac{151.0}{3}\text{ kN}\cdot\text{m} \approx +50.3333\text{ kN}\cdot\text{m}
  • Span BCBC:

    MBC=MF,BC+2EI6(2θB+θC)=−45.0+13EI(2⋅103.03EI−254.03EI)M_{BC} = M_{F,BC} + \frac{2EI}{6}(2\theta_B + \theta_C) = -45.0 + \frac{1}{3} EI \left( 2 \cdot \frac{103.0}{3 EI} - \frac{254.0}{3 EI} \right) MBC=−45.0+13(206.0−254.03)=−45.0+13(−48.03)=−45.0−16.03=−151.03 kN⋅m≈−50.3333 kN⋅mM_{BC} = -45.0 + \frac{1}{3}\left( \frac{206.0 - 254.0}{3} \right) = -45.0 + \frac{1}{3}\left(-\frac{48.0}{3}\right) = -45.0 - \frac{16.0}{3} = -\frac{151.0}{3}\text{ kN}\cdot\text{m} \approx -50.3333\text{ kN}\cdot\text{m} MCB=MF,CB+2EI6(2θC+θB)=+45.0+13EI(2⋅−254.03EI+103.03EI)M_{CB} = M_{F,CB} + \frac{2EI}{6}(2\theta_C + \theta_B) = +45.0 + \frac{1}{3} EI \left( 2 \cdot \frac{-254.0}{3 EI} + \frac{103.0}{3 EI} \right) MCB=+45.0+13(−508.0+103.03)=+45.0+13(−405.03)=+45.0−45.0=0.0000 kN⋅mM_{CB} = +45.0 + \frac{1}{3}\left( \frac{-508.0 + 103.0}{3} \right) = +45.0 + \frac{1}{3}\left(-\frac{405.0}{3}\right) = +45.0 - 45.0 = 0.0000\text{ kN}\cdot\text{m}
Duality & Equilibrium Check:
  1. Joint BB Moment Equilibrium: MBA+MBC=+50.3333−50.3333=0.0000M_{BA} + M_{BC} = +50.3333 - 50.3333 = 0.0000 (Exact balance!).
  2. Roller CC Boundary Condition: MCB≡0.0000 kN⋅mM_{CB} \equiv 0.0000\text{ kN}\cdot\text{m} (Verified!).
  3. In Topic SA-06-01 (Flexibility Method Example 3), the redundant support moments for this identical structure were solved as: MA=+1.1667 kN⋅m,MB=−50.3333 kN⋅mM_A = +1.1667\text{ kN}\cdot\text{m}, \quad M_B = -50.3333\text{ kN}\cdot\text{m} The results are 100.000% identical, providing mathematical proof of the exact duality between the Flexibility and Stiffness formulations.

Problem 2: 2D Planar Truss Analysis via Coordinate Transformation

Statement: A 2-bar pin-jointed planar truss consists of Member 1 connecting pinned node 1 (0,3.0 m)(0, 3.0\text{ m}) to free node 3 (4.0 m,0.0 m)(4.0\text{ m}, 0.0\text{ m}), and Member 2 connecting pinned node 2 (0.0 m,0.0 m)(0.0\text{ m}, 0.0\text{ m}) to node 3 (4.0 m,0.0 m)(4.0\text{ m}, 0.0\text{ m}). Both members have cross-sectional area A=1000 mm2=1.0×10−3 m2A = 1000\text{ mm}^2 = 1.0 \times 10^{-3}\text{ m}^2 and modulus of elasticity E=200 GPa=2.0×108 kN/m2E = 200\text{ GPa} = 2.0 \times 10^8\text{ kN/m}^2, yielding axial rigidity EA=200,000.0 kNEA = 200,000.0\text{ kN}. Joint 3 is subjected to external loads: horizontal force PX=+40.0 kNP_X = +40.0\text{ kN} (rightward) and vertical force PY=−60.0 kNP_Y = -60.0\text{ kN} (downward). Using the Direct Stiffness Method:

  1. Formulate the global stiffness matrix [K][K] for active DOFs at node 3.
  2. Determine nodal displacements u3u_3 and v3v_3.
  3. Compute internal member axial forces N1N_1 and N2N_2 and verify joint equilibrium.
   (0, 3) Node 1 
     ▲
     │ \
     │   \  Member 1 (L = 5.0 m)
     │     \
   3m│       \
     │         \
     ▼           \
   (0, 0) Node 2 ── Node 3 (4, 0) ──► Px = +40 kN
     ▲  Member 2    │
     │  (L = 4.0 m) │
                    ▼ Py = -60 kN
     ├──── 4.0 m ───┤

Solution:

Step 1: Member Geometry & Direction Cosines
  • Member 1 (Node 1 →\rightarrow Node 3):

    • X1=0.0X_1 = 0.0, Y1=3.0Y_1 = 3.0; X3=4.0X_3 = 4.0, Y3=0.0Y_3 = 0.0
    • Length L1=(4.0−0)2+(0.0−3.0)2=16+9=5.0 mL_1 = \sqrt{(4.0 - 0)^2 + (0.0 - 3.0)^2} = \sqrt{16 + 9} = 5.0\text{ m}
    • Direction cosines: λx1=4.0−0.05.0=+0.80,λy1=0.0−3.05.0=−0.60\lambda_{x1} = \frac{4.0 - 0.0}{5.0} = +0.80, \quad \lambda_{y1} = \frac{0.0 - 3.0}{5.0} = -0.60
    • Axial stiffness factor: EAL1=200,000.05.0=40,000.0 kN/m\frac{EA}{L_1} = \frac{200,000.0}{5.0} = 40,000.0\text{ kN/m}
  • Member 2 (Node 2 →\rightarrow Node 3):

    • X2=0.0X_2 = 0.0, Y2=0.0Y_2 = 0.0; X3=4.0X_3 = 4.0, Y3=0.0Y_3 = 0.0
    • Length L2=(4.0−0)2+(0−0)2=4.0 mL_2 = \sqrt{(4.0 - 0)^2 + (0 - 0)^2} = 4.0\text{ m}
    • Direction cosines: λx2=4.0−0.04.0=+1.00,λy2=0.0−0.04.0=0.00\lambda_{x2} = \frac{4.0 - 0.0}{4.0} = +1.00, \quad \lambda_{y2} = \frac{0.0 - 0.0}{4.0} = 0.00
    • Axial stiffness factor: EAL2=200,000.04.0=50,000.0 kN/m\frac{EA}{L_2} = \frac{200,000.0}{4.0} = 50,000.0\text{ kN/m}
Step 2: Formulate Global Stiffness Submatrices at Node 3

The active degrees of freedom at Node 3 are u3u_3 (DOF 1) and v3v_3 (DOF 2). Nodes 1 and 2 are rigidly pinned supports (u1=v1=u2=v2=0u_1 = v_1 = u_2 = v_2 = 0).

  • Element 1 contribution to Node 3:

    [k1]33=EAL1[λx12λx1λy1λx1λy1λy12]=40,000[(0.8)2(0.8)(−0.6)(0.8)(−0.6)(−0.6)2]=40,000[0.64−0.48−0.480.36][k_1]_{33} = \frac{EA}{L_1} \begin{bmatrix} \lambda_{x1}^2 & \lambda_{x1} \lambda_{y1} \\ \lambda_{x1} \lambda_{y1} & \lambda_{y1}^2 \end{bmatrix} = 40,000 \begin{bmatrix} (0.8)^2 & (0.8)(-0.6) \\ (0.8)(-0.6) & (-0.6)^2 \end{bmatrix} = 40,000 \begin{bmatrix} 0.64 & -0.48 \\ -0.48 & 0.36 \end{bmatrix} [k1]33=[25,600.0−19,200.0−19,200.014,400.0] kN/m[k_1]_{33} = \begin{bmatrix} 25,600.0 & -19,200.0 \\ -19,200.0 & 14,400.0 \end{bmatrix}\text{ kN/m}
  • Element 2 contribution to Node 3:

    [k2]33=EAL2[λx22λx2λy2λx2λy2λy22]=50,000[1.000.000.000.00]=[50,000.00.00.00.0] kN/m[k_2]_{33} = \frac{EA}{L_2} \begin{bmatrix} \lambda_{x2}^2 & \lambda_{x2} \lambda_{y2} \\ \lambda_{x2} \lambda_{y2} & \lambda_{y2}^2 \end{bmatrix} = 50,000 \begin{bmatrix} 1.00 & 0.00 \\ 0.00 & 0.00 \end{bmatrix} = \begin{bmatrix} 50,000.0 & 0.0 \\ 0.0 & 0.0 \end{bmatrix}\text{ kN/m}
  • Global Assembled Stiffness Matrix [K][K]:

    [K]=[k1]33+[k2]33=[25,600.0+50,000.0−19,200.0+0.0−19,200.0+0.014,400.0+0.0]=[75,600.0−19,200.0−19,200.014,400.0] kN/m[K] = [k_1]_{33} + [k_2]_{33} = \begin{bmatrix} 25,600.0 + 50,000.0 & -19,200.0 + 0.0 \\ -19,200.0 + 0.0 & 14,400.0 + 0.0 \end{bmatrix} = \begin{bmatrix} 75,600.0 & -19,200.0 \\ -19,200.0 & 14,400.0 \end{bmatrix}\text{ kN/m}

Determinant of [K][K]:

det⁡[K]=(75,600.0)(14,400.0)−(−19,200.0)2=1,088,640,000−368,640,000=7.20×108 (kN/m)2\det[K] = (75,600.0)(14,400.0) - (-19,200.0)^2 = 1,088,640,000 - 368,640,000 = 7.20 \times 10^8\text{ (kN/m)}^2

Inverse matrix [K]−1[K]^{-1}:

[K]−1=17.20×108[14,400.019,200.019,200.075,600.0]=10−5[2.00002.66672.666710.5000] m/kN[K]^{-1} = \frac{1}{7.20 \times 10^8} \begin{bmatrix} 14,400.0 & 19,200.0 \\ 19,200.0 & 75,600.0 \end{bmatrix} = 10^{-5} \begin{bmatrix} 2.0000 & 2.6667 \\ 2.6667 & 10.5000 \end{bmatrix}\text{ m/kN}
Step 3: Compute Displacements at Node 3

The applied load vector is [P]=[PX,PY]T=[40.0,−60.0]T kN[P] = [P_X, P_Y]^T = [40.0, -60.0]^T\text{ kN}. Since there are no span loads on the truss members, [PF]=[0,0]T[P_F] = [0, 0]^T.

[u3v3]=[K]−1[P]=17.20×108[14,400.0(40.0)+19,200.0(−60.0)19,200.0(40.0)+75,600.0(−60.0)]\begin{bmatrix} u_3 \\ v_3 \end{bmatrix} = [K]^{-1} [P] = \frac{1}{7.20 \times 10^8} \begin{bmatrix} 14,400.0(40.0) + 19,200.0(-60.0) \\ 19,200.0(40.0) + 75,600.0(-60.0) \end{bmatrix} u3=576,000.0−1,152,000.07.20×108=−576,000.07.20×108=−8.00×10−4 m=−0.8000 mmu_3 = \frac{576,000.0 - 1,152,000.0}{7.20 \times 10^8} = \frac{-576,000.0}{7.20 \times 10^8} = -8.00 \times 10^{-4}\text{ m} = -0.8000\text{ mm} v3=768,000.0−4,536,000.07.20×108=−3,768,000.07.20×108=−15730,000 m≈−0.0052333 m=−5.2333 mmv_3 = \frac{768,000.0 - 4,536,000.0}{7.20 \times 10^8} = \frac{-3,768,000.0}{7.20 \times 10^8} = -\frac{157}{30,000}\text{ m} \approx -0.0052333\text{ m} = -5.2333\text{ mm}
Step 4: Member Axial Forces & Statics Check

The axial force NN in a member (tension positive) is:

N=EAL[(λxufar+λyvfar)−(λxunear+λyvnear)]N = \frac{EA}{L} \left[ (\lambda_x u_{\text{far}} + \lambda_y v_{\text{far}}) - (\lambda_x u_{\text{near}} + \lambda_y v_{\text{near}}) \right]
  • Member 1 (Near: Node 1, Far: Node 3):

    N1=EAL1(λx1u3+λy1v3)=40,000[(0.80)(−0.0008)+(−0.60)(−0.0052333)]N_1 = \frac{EA}{L_1} \left( \lambda_{x1} u_3 + \lambda_{y1} v_3 \right) = 40,000 \left[ (0.80)(-0.0008) + (-0.60)(-0.0052333) \right] N1=40,000[−0.0006400+0.0031400]=40,000(+0.0025000)=+100.0000 kN(Tension)N_1 = 40,000 \left[ -0.0006400 + 0.0031400 \right] = 40,000 \left( +0.0025000 \right) = +100.0000\text{ kN}\quad\text{(Tension)}
  • Member 2 (Near: Node 2, Far: Node 3):

    N2=EAL2(λx2u3+λy2v3)=50,000[(1.00)(−0.0008)+(0.0)(−0.0052333)]N_2 = \frac{EA}{L_2} \left( \lambda_{x2} u_3 + \lambda_{y2} v_3 \right) = 50,000 \left[ (1.00)(-0.0008) + (0.0)(-0.0052333) \right] N2=50,000[−0.0008]=−40.0000 kN(Compression)N_2 = 50,000 \left[ -0.0008 \right] = -40.0000\text{ kN}\quad\text{(Compression)}
Statics Equilibrium Check at Node 3:
  • ∑FX=N1λx1+N2λx2=(+100.0)(0.80)+(−40.0)(1.00)=80.0−40.0=+40.0 kN=PX\sum F_X = N_1 \lambda_{x1} + N_2 \lambda_{x2} = (+100.0)(0.80) + (-40.0)(1.00) = 80.0 - 40.0 = +40.0\text{ kN} = P_X
  • ∑FY=N1λy1+N2λy2=(+100.0)(−0.60)+(−40.0)(0.00)=−60.0+0.0=−60.0 kN=PY\sum F_Y = N_1 \lambda_{y1} + N_2 \lambda_{y2} = (+100.0)(-0.60) + (-40.0)(0.00) = -60.0 + 0.0 = -60.0\text{ kN} = P_Y Both equilibrium equations are satisfied to exact floating-point precision.

Problem 3: Propped Cantilever Subjected to Downward Support Settlement

Statement: A propped cantilever beam ABAB has span L=6.0 mL = 6.0\text{ m} and carries a uniform distributed load w=20.0 kN/mw = 20.0\text{ kN/m}. Support AA is rigidly fixed, and support BB is a roller prop. Flexural rigidity is EI=20,000.0 kN⋅m2EI = 20,000.0\text{ kN}\cdot\text{m}^2. During loading, the roller support at BB settles downwards by δ=5.0 mm=0.005 m\delta = 5.0\text{ mm} = 0.005\text{ m}. Using the Stiffness Method:

  1. Formulate the single-degree-of-freedom stiffness equilibrium equation.
  2. Calculate the joint rotation θB\theta_B taking into account both span load and support settlement.
  3. Determine the final fixed-end moment MAM_A and prop reaction VBV_B.
       w = 20 kN/m
  ┌───────────────────────────────┐
  ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼
  ═════════════════════════════════════
  ▒▒█                                 ▲
  ▒▒█ (Fixed)                     (Roller)
   A                                   B ──► Sinks downward by δ = 5 mm
  ├────────────── 6.0 m ──────────────┤

Solution:

Step 1: Kinematic DOF Formulation

Neglecting axial strain:

  • Support AA is fixed: θA=0\theta_A = 0, vA=0v_A = 0.
  • Prop BB has a prescribed settlement Δs=vB=−0.005 m\Delta_s = v_B = -0.005\text{ m}.
  • The only unconstrained degree of freedom is rotation at BB (θB=Δ1\theta_B = \Delta_1).
Dk=1D_k = 1

Stiffness coefficient at DOF 1:

K11=4EIL=4×20,000.06.0=80,000.06.0=13,333.3333 kN⋅m/radK_{11} = \frac{4EI}{L} = \frac{4 \times 20,000.0}{6.0} = \frac{80,000.0}{6.0} = 13,333.3333\text{ kN}\cdot\text{m/rad}
Step 2: Fixed-End Moments from Load and Support Settlement
  • Due to UDL (w=20.0 kN/mw = 20.0\text{ kN/m}):

    MF,ABload=−wL212=−20×3612=−60.0 kN⋅mM_{F,AB}^{\text{load}} = -\frac{w L^2}{12} = -\frac{20 \times 36}{12} = -60.0\text{ kN}\cdot\text{m} MF,BAload=+wL212=+20×3612=+60.0 kN⋅mM_{F,BA}^{\text{load}} = +\frac{w L^2}{12} = +\frac{20 \times 36}{12} = +60.0\text{ kN}\cdot\text{m}
  • Due to Downward Support Settlement at BB (δ=0.005 m\delta = 0.005\text{ m}): Downward settlement produces a clockwise chord rotation ψ=+δL=+0.0056=+11200 rad\psi = +\frac{\delta}{L} = +\frac{0.005}{6} = +\frac{1}{1200}\text{ rad}. The settlement induces fixed-end moments:

    MF,ABsettle=−6EIδL2=−6×20,000.0×0.00536.0=−600.036.0=−16.6667 kN⋅mM_{F,AB}^{\text{settle}} = -\frac{6EI\delta}{L^2} = -\frac{6 \times 20,000.0 \times 0.005}{36.0} = -\frac{600.0}{36.0} = -16.6667\text{ kN}\cdot\text{m} MF,BAsettle=−6EIδL2=−16.6667 kN⋅mM_{F,BA}^{\text{settle}} = -\frac{6EI\delta}{L^2} = -16.6667\text{ kN}\cdot\text{m}
  • Total Holding Action at DOF 1 (PF1P_{F1}):

    PF1=MF,BAload+MF,BAsettle=+60.0−16.6667=+43.3333 kN⋅mP_{F1} = M_{F,BA}^{\text{load}} + M_{F,BA}^{\text{settle}} = +60.0 - 16.6667 = +43.3333\text{ kN}\cdot\text{m}
Step 3: Solve for Rotation θB\theta_B

Since no external concentrated couple is applied at joint BB, P1=0P_1 = 0. The net equivalent driving moment is:

Pnet=P1−PF1=0−43.3333=−43.3333 kN⋅mP_{\text{net}} = P_1 - P_{F1} = 0 - 43.3333 = -43.3333\text{ kN}\cdot\text{m}

Solving the equilibrium equation K11θB=PnetK_{11} \theta_B = P_{\text{net}}:

θB=−43.333313,333.3333=−0.00325 rad=−3.25 mrad\theta_B = \frac{-43.3333}{13,333.3333} = -0.00325\text{ rad} = -3.25\text{ mrad}
Step 4: Compute Member End Moments and Support Reaction
  • Moment at BB (MBAM_{BA}):

    MBA=PF1+K11θB=+43.3333+13,333.3333(−0.00325)=+43.3333−43.3333=0.0000 kN⋅mM_{BA} = P_{F1} + K_{11} \theta_B = +43.3333 + 13,333.3333(-0.00325) = +43.3333 - 43.3333 = 0.0000\text{ kN}\cdot\text{m}

    As required by the roller boundary condition, MBA=0M_{BA} = 0.

  • Fixed-End Moment at AA (MABM_{AB}):

    MAB=(MF,ABload+MF,ABsettle)+2EILθBM_{AB} = \left( M_{F,AB}^{\text{load}} + M_{F,AB}^{\text{settle}} \right) + \frac{2EI}{L}\theta_B MAB=(−60.0−16.6667)+2×20,000.06.0(−0.00325)=−76.6667+6,666.6667(−0.00325)M_{AB} = \left( -60.0 - 16.6667 \right) + \frac{2 \times 20,000.0}{6.0}(-0.00325) = -76.6667 + 6,666.6667(-0.00325) MAB=−76.6667−21.6667=−98.3333 kN⋅m(Hogging moment)M_{AB} = -76.6667 - 21.6667 = -98.3333\text{ kN}\cdot\text{m}\quad\text{(Hogging moment)}
  • Prop Vertical Reaction VBV_B: Taking moment equilibrium about support AA (∑MA=0\sum M_A = 0):

    VB⋅L−w⋅L22−Mhogging=0V_B \cdot L - w \cdot \frac{L^2}{2} - M_{\text{hogging}} = 0 VB(6.0)−20.0(36.02)−(−98.3333)=0  ⟹  VB(6.0)=360.0−98.3333=261.6667V_B (6.0) - 20.0 \left(\frac{36.0}{2}\right) - (-98.3333) = 0 \implies V_B (6.0) = 360.0 - 98.3333 = 261.6667 VB=261.66676.0=43.6111 kNV_B = \frac{261.6667}{6.0} = 43.6111\text{ kN}
Physical Comparison (Effect of Settlement):
  • Without settlement: VB0=38wL=38(20)(6)=45.0000 kNV_{B0} = \frac{3}{8} w L = \frac{3}{8}(20)(6) = 45.0000\text{ kN} MA0=wL28=20×368=90.0000 kN⋅mM_{A0} = \frac{w L^2}{8} = \frac{20 \times 36}{8} = 90.0000\text{ kN}\cdot\text{m}
  • With 5 mm5\text{ mm} downward settlement:
    • Prop reaction decreases by ΔV=3EIδL3=3(20000)(0.005)216=1.3889 kN  ⟹  VB=45.0−1.3889=43.6111 kN\Delta V = \frac{3EI\delta}{L^3} = \frac{3(20000)(0.005)}{216} = 1.3889\text{ kN} \implies V_B = 45.0 - 1.3889 = 43.6111\text{ kN}.
    • Fixed-end hogging moment increases by ΔM=3EIδ2L2×2=3EIδL2=8.3333 kN⋅m  ⟹  MA=90.0+8.3333=98.3333 kN⋅m\Delta M = \frac{3EI\delta}{2L^2} \times 2 = \frac{3EI\delta}{L^2} = 8.3333\text{ kN}\cdot\text{m} \implies M_A = 90.0 + 8.3333 = 98.3333\text{ kN}\cdot\text{m}. The stiffness method accounts for combined transverse load and support settlement in a single unified step.

8. Exam-Oriented Traps, Common Errors & Misconceptions

[!CAUTION] Trap 1: Confusing Holding Forces ([PF][P_F]) with Equivalent Nodal Forces ([Peq][P_{\text{eq}}]) The fixed-end force vector [PF][P_F] represents the forces exerted by the members ON the locked joints. Therefore, the net driving force acting on the free joints is [Pnet]=[P]−[PF][P_{\text{net}}] = [P] - [P_F]. If you add [PF][P_F] instead of subtracting it, the signs of all computed displacements and member moments will be inverted!

[!WARNING] Trap 2: Ignoring Cross-Coupling Terms in Non-Orthogonal Members For an inclined truss bar at angle θ\theta, students frequently take the axial stiffness as simply EALcos⁡θ\frac{EA}{L}\cos\theta or EALsin⁡θ\frac{EA}{L}\sin\theta. The true global stiffness requires transformation [T]T[k][T][T]^T [k] [T], generating coupling off-diagonal terms EALcos⁡θsin⁡θ\frac{EA}{L}\cos\theta\sin\theta. Omitting these cross-terms produces completely erroneous results.

[!NOTE] Misconception 3: "Stiffness Matrix Diagonal Terms Can Be Negative" In any stable physical structure, all diagonal elements kiik_{ii} of the stiffness matrix must be strictly positive (kii>0k_{ii} > 0). A zero or negative diagonal coefficient indicates an unstable kinematic mechanism, unconstrained rigid-body motion, or an algebraic sign error.

[!TIP] Trap 4: Degrees of Freedom Under Axial Inextensibility Assumptions In competitive exams (GATE/ESE), questions frequently state: "Neglect axial deformation". For a continuous beam on non-yielding supports, vertical translations at supports are already zero (vi=0v_i = 0); assuming axial inextensibility suppresses horizontal translation ui=0u_i = 0. Consequently, each intermediate joint has only 1 rotational DOF (θ\theta), dramatically shrinking the matrix order!


9. Competitive Examination Key Points & Memory Triggers

  1. Matrix Size Rule:
    • Flexibility Matrix Order = Ds×DsD_s \times D_s
    • Stiffness Matrix Order = Dk×DkD_k \times D_k
    • Exam Decision Shortcut: If Dk<DsD_k < D_s, choose the Stiffness Method; if Ds<DkD_s < D_k, choose the Flexibility Method.
  2. Carry-Over Factor in Stiffness Terms:
    • For a prismatic beam with far end fixed: COF=k21k11=2EI/L4EI/L=12=0.50\text{COF} = \frac{k_{21}}{k_{11}} = \frac{2EI/L}{4EI/L} = \frac{1}{2} = 0.50.
    • For far end hinged: near stiffness is 3EIL\frac{3EI}{L}, far moment is 0, COF=0\text{COF} = 0.
  3. Properties of Global Stiffness Matrix [K][K]:
    • Symmetric: Kij=KjiK_{ij} = K_{ji} (via Maxwell-Betti Reciprocal Theorem).
    • Positive Definite: det⁡[K]>0\det[K] > 0, all eigenvalues λi>0\lambda_i > 0.
    • Banded & Sparse: Nonzero elements cluster tightly around the main diagonal when nodes are numbered systematically.
  4. Strain Energy in Matrix Form: U=12[Δ]T[K][Δ]U = \frac{1}{2} [\Delta]^T [K] [\Delta] Because strain energy stored in an elastic body is always positive for any real deformation, [K][K] must be positive definite.

10. Previous Year & Model Examination Questions

Question 1 (GATE CE)

A continuous beam ABAB of span LL is fixed at support AA and propped by a roller at BB. A unit rotation θB=1\theta_B = 1 is applied at support BB. The bending moment developed at fixed end AA is:

  • (A) 4EIL\frac{4EI}{L}
  • (B) 2EIL\frac{2EI}{L}
  • (C) 6EIL2\frac{6EI}{L^2}
  • (D) 3EIL\frac{3EI}{L}

Answer: (B) Explanation: From the flexural beam stiffness matrix, applying a unit rotation at end BB induces a near-end moment of 4EIL\frac{4EI}{L} at BB and a carry-over moment of 2EIL\frac{2EI}{L} at the far fixed end AA.


Question 2 (UPSC ESE)

In the stiffness method of structural analysis, the elements of the stiffness matrix kijk_{ij} satisfy which of the following conditions?

  1. kij=kjik_{ij} = k_{ji}
  2. kii>0k_{ii} > 0
  3. [K][K] is singular for a kinematically stable structure
  4. [K][K] is the inverse of the unconstrained flexibility matrix [F][F]

Select the correct statements:

  • (A) 1, 2, and 3 only
  • (B) 1, 2, and 4 only
  • (C) 2, 3, and 4 only
  • (D) 1, 3, and 4 only

Answer: (B) Explanation: For a kinematically stable structure, [K][K] is non-singular (det⁡[K]≠0\det[K] \neq 0 and det⁡[K]>0\det[K] > 0), making statement 3 incorrect. Statements 1, 2, and 4 are mathematically correct.


Question 3 (SSC JE / State AE)

The dimension of a stiffness coefficient corresponding to a moment generated due to a unit linear displacement is:

  • (A) kN⋅m\text{kN}\cdot\text{m}
  • (B) kN/m\text{kN/m}
  • (C) kN\text{kN}
  • (D) Dimensionless\text{Dimensionless}

Answer: (C) Explanation: kij=MomentDisplacement=kN⋅mm=kNk_{ij} = \frac{\text{Moment}}{\text{Displacement}} = \frac{\text{kN}\cdot\text{m}}{\text{m}} = \text{kN}.


11. Summary & Quick Revision Formula Sheet

Entity Mathematical Formula Physical Context / Remarks
Global Equilibrium [K][Δ]=[P]−[PF]=[Pnet][K] [\Delta] = [P] - [P_F] = [P_{\text{net}}] Fundamental governing equation of DSM
Member Rotation Stiffness knear=4EIL,kfar=2EILk_{\text{near}} = \frac{4EI}{L}, \quad k_{\text{far}} = \frac{2EI}{L} Far end fixed; COF = 0.500.50
Member Transverse Shear kshear=12EIL3,kmoment=6EIL2k_{\text{shear}} = \frac{12EI}{L^3}, \quad k_{\text{moment}} = \frac{6EI}{L^2} Restoring actions due to unit transverse deflection
2D Truss Element Stiffness [kaxial]=EAL[1−1−11][k_{\text{axial}}] = \frac{EA}{L} \begin{bmatrix} 1 & -1 \\ -1 & 1 \end{bmatrix} Local coordinate formulation
Direction Cosines λx=X2−X1L,λy=Y2−Y1L\lambda_x = \frac{X_2 - X_1}{L}, \quad \lambda_y = \frac{Y_2 - Y_1}{L} Coordinate orientation angles
Global Element Stiffness [kglobal]=[T]T[kˉ][T][k_{\text{global}}] = [T]^T [\bar{k}] [T] Orthogonal coordinate transformation
Support Settlement Load [Psettle]=−[Kfs][Δs][P_{\text{settle}}] = -[K_{fs}] [\Delta_s] Equivalent nodal vector due to known support sinks
Member Action Recovery {p}=[kˉ][T]{d}+{pˉF}\{p\} = [\bar{k}] [T] \{d\} + \{\bar{p}_F\} Post-processing internal shears and moments
Strain Energy U=12[Δ]T[K][Δ]U = \frac{1}{2} [\Delta]^T [K] [\Delta] Quadratic form, strictly >0> 0

12. References & Standard Code Provisions

  1. Gere, J. M., & Weaver, W. — Matrix Analysis of Framed Structures, 3rd Edition, Van Nostrand Reinhold.
  2. McGuire, W., Gallagher, R. H., & Ziemian, R. D. — Matrix Structural Analysis, John Wiley & Sons.
  3. Hibbeler, R. C. — Structural Analysis, 10th Edition, Pearson.
  4. Reddy, C. S. — Basic Structural Analysis, 3rd Edition, Tata McGraw-Hill.
  5. IS 456:2000 — Plain and Reinforced Concrete - Code of Practice, Bureau of Indian Standards, New Delhi.
  6. IS 800:2007 — General Construction in Steel - Code of Practice, Bureau of Indian Standards, New Delhi.