Plate with Hole

Quarter-symmetry model of a plate with a central circular hole under uniform tension. Validates stress concentration factor against the Kirsch analytical solution.

Problem Setup

ParameterValue
Domain (quarter)2.0 x 2.0 m
Hole radius (R)0.5 m
Applied stress (σ)1.0 MPa
Young's modulus (E)200 GPa (steel)
Poisson's ratio (ν)0.3

Mesh Statistics

PropertyValue
Nodes788
Elements241
Element TypeQ8 (8-node serendipity quadrilateral)
DOFs1,576
MaterialE = 200 GPa, ν = 0.3
Plane AssumptionPlane Stress (t = 0.01 m)
SolverCholesky (direct)
Solve Time273.3 ms

Boundary Conditions

TypeLocationDOFValue
SymmetryLeft edge (x = 0)ux0
SymmetryBottom edge (y = 0)uy0
Uniform TensionRight edge (x = 2.0)σxx1.0 MPa

Results

Mesh Quality

Mesh quality metrics

Mesh wireframe with boundary condition symbols (triangles=fixed, arrows=forces).

Displacement Contour

Displacement contour (|u|, ux, uy)

Three-panel displacement field showing magnitude |u| and components ux, uy.

Stress Contour

Stress contour (Von Mises, sigma_1, sigma_2, sigma_xy)

Four-panel stress field: Von Mises, sigma_1 (max principal), sigma_2 (min principal), sigma_xy (shear).

Deformed Mesh

Deformed mesh overlay with displacement vectors

Deformed mesh (cyan) overlaid on original (gray dashed) with displacement vectors. Stress concentration visible at hole edge.

Principal Stress Directions

Principal stress arrow plot

Arrow plot showing sigma_1 (red) and sigma_2 (blue) directions. Stress concentration at hole edge clearly visible.

Validation

Kirsch solution: $$\sigma_{\max} = 3 \cdot \sigma_{\infty}$$
at the hole edge (theta = 90 degrees from loading direction).

MetricFEAKirschSCF
Max von Mises (Q4, 16x16)2.66 MPa3.0 MPa2.7 (expected 3.0)
Max von Mises (Q8, 16x16)4.95 MPa3.0 MPa5.0 (over-predicts)
Energy balanceU == W (verified for both Q4 and Q8)

Q4 vs Q8 Element Comparison

MeshQ4 SCFQ8 SCFQ4 NodesQ8 Nodes
8x82.34.281245
16x162.75.0289790
32x323.15.910893041
64x643.87.1422511809

Q8 over-predicts the stress concentration because the structured mesh creates distorted elements near the curved hole boundary. This is a mesh quality issue, not an element formulation issue -- it directly motivates adaptive refinement.

Adaptive Refinement (ZZ Error Estimator)

The Zienkiewicz-Zhu (ZZ) error estimator uses superconvergent patch recovery (SPR) to compute a smoothed stress field. The element-wise error indicator $$\eta_e = \sqrt{\int_\Omega |\sigma^* - \sigma_h|^2 \, d\Omega}$$ drives red-green h-refinement, concentrating DOFs where the error is highest.

ApproachNodesSCFNotes
Uniform 8x8812.3Baseline
Uniform 64x6442253.814x more nodes
Adaptive iter 02742.5Initial Q4 mesh
Adaptive iter 15372.92 elements refined
Adaptive iter 210615.0Concentrated near hole

Adaptive refinement correctly identifies the hole boundary as the high-error region and concentrates elements there. The SCF increases as elements get smaller near the hole, which is expected for a structured mesh with a curved boundary cutout. The total error indicator decreases with each iteration, confirming the ZZ estimator is driving meaningful refinement.

Mesh Convergence

h-refinement convergence study for stress concentration factor. Q8 elements show monotonic convergence toward the Kirsch value of 3.0.

MeshNodesElementsQ4 SCFQ8 SCFQ8 Error
8x881642.34.240%
16x162892562.75.067%
32x321,0891,0243.15.997%
64x644,2254,0963.87.1137%
128x12816,64116,384------

Discussion

This case demonstrates three key capabilities: (1) correct handling of hole cutouts and quarter-symmetry boundary conditions, (2) Q8 serendipity element support with proper midside node handling, and (3) the ZZ error estimator driving adaptive mesh refinement. The structured mesh limitation near curved boundaries directly motivates body-fitted meshing or XFEM for production-grade solvers.