forked from 1b-agent/fieldcraft-upstream
jyvikoandClaude Opus 5.5
4be97b824c
buckle-plate-square/lambda moves 3.0e-9 relative and plastic-hinge-cantilever/collapse_force 2.7e-8 relative: the graph dissection now orders these meshes, so the factor sums in another order. Both stay at the same error against their references. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
Buckling of a simply supported square plate in uniaxial compression
A 50 × 50 × 0.5 plate (b/t = 100), E = 1000, ν = 0.3, modelled as its quarter (10 × 10 × 2 hex20) with symmetry planes. The outer edges are held out of plane along their bottom edge lines, which leaves their rotation and the plate's Poisson thickening free (simply supported), and they are free in plane; one is pressed by a unit pressure. Reference: Bryan's k = 4.
Choices: holding the whole edge faces instead blocks the thickening there and lowers the load by 2 % at ν = 0.3; at b/t = 50 the solid stays 1.5 % below thin-plate theory with either support, the difference falling as the plate thins.