Pinched ring — diametrical deflection#
Classical thin-ring benchmark from Castigliano’s 1879 derivation. A circular ring of mean radius \(R\), bending stiffness \(EI\), loaded by two equal and opposite point forces \(P\) acting along a diameter deflects by
The line of action of the loads shortens by \(\delta_{\text{diam}}\); the perpendicular diameter lengthens by the Castigliano companion value \(\delta_{\text{perp}} = P R^{3}(4/\pi - \pi/4) / (E I)\).
Problem#
Parameter |
Value |
|---|---|
Mean ring radius |
0.1 m |
Cross-section |
10 mm × 5 mm (radial = 5 mm) |
Young’s modulus |
200 GPa |
Poisson’s ratio |
0.30 |
Applied point load |
10 N |
Expected |
(one loaded point’s inward motion) = 3.57 × 10⁻⁵ m |
femorph-solver result#
Ran by tests/validation/test_pinched_ring.py using the
BEAM188 kernel on a quarter-symmetry chain of straight
2-node beam segments discretising the arc θ ∈ [0, π/2].
BCs at the two cut ends enforce the full-ring symmetries:
θ = 0(loaded point(R, 0)):UY = UZ = ROTX = ROTY = ROTZ = 0.UXis free to register the inward deflection; the in-plane rotationROTZ = 0enforces the x-axis symmetry of the cross-section.θ = π/2(symmetry point(0, R)):UX = UZ = ROTX = ROTY = ROTZ = 0.UYis free.
Half of the load (P/2) is applied at the loaded node
inward along -x; by quarter-symmetry the full-ring
diametrical deflection equals twice the loaded-node |u_x|.
Refinement |
Elements (quarter arc) |
|
Error vs Castigliano |
|---|---|---|---|
Coarse |
20 |
3.570 × 10⁻⁵ |
−0.02 % |
Medium |
40 |
3.573 × 10⁻⁵ |
+0.08 % |
Refined |
80 |
3.574 × 10⁻⁵ |
+0.10 % |
Converges to within 0.1 % — the small positive drift reflects the slight extensional stiffness the 2-node BEAM188 introduces (Castigliano’s derivation assumes pure bending with zero axial strain). Inside the published 5 % engineering tolerance.
Cross-references#
Source |
Reported |
Problem ID / location |
|---|---|---|
Closed form (Castigliano) |
3.57 × 10⁻⁵ |
Timoshenko & Young 1968 §79 |
Roark’s Formulas for Stress and Strain |
3.57 × 10⁻⁵ |
8th ed. Table 9.2 Case 13 |
femorph-solver (finest) |
3.574 × 10⁻⁵ |
|
MAPDL Verification Manual |
≈ 3.57 × 10⁻⁵ |
VM-38 pinched-ring family |
Abaqus Verification Manual |
≈ 3.57 × 10⁻⁵ |
AVM 1.5.x pinched-ring family |
NAFEMS R0011 |
≈ 3.57 × 10⁻⁵ |
Pinched-ring benchmark (quarter-symmetry) |
Source#
Problem class:
femorph_solver.validation.problems.PinchedRing.
Backing regression test:
tests/validation/test_pinched_ring.py — asserts
|u_x| matches the closed form to within 0.5 % on every
refinement, not just the finest.