QF Solver69 method-family runs from four Gmsh models. 39 finished with passing numerical checks and have a deformation figure; 27 failed and 3 reached the time limit.
Showing 39 of 39 cases with passing numerical checks.
Perforated cantilever plate · 915 elements · 352 nodes
The GIF ramps to the computed state; it is not a load history.
Perforated cantilever plate · 915 elements · 352 nodes
The GIF ramps to the computed state; it is not a load history.
Perforated cantilever plate · 915 elements · 352 nodes
The GIF ramps to the computed state; it is not a load history.
Perforated cantilever plate · 915 elements · 352 nodes
The GIF shows an illustrative phase cycle. Mode amplitude is normalized for display, not a physical displacement.
Perforated cantilever plate · 915 elements · 352 nodes
The GIF shows an illustrative phase cycle. Mode amplitude is normalized for display, not a physical displacement.
Perforated cantilever plate · 915 elements · 352 nodes
The GIF shows an illustrative phase cycle. Mode amplitude is normalized for display, not a physical displacement.
Perforated cantilever plate · 915 elements · 352 nodes
The GIF is a display ramp, not the transient time history. Use the recorded time-step data for the response history.
Perforated cantilever plate · 915 elements · 352 nodes
The GIF is a display ramp, not the transient time history. Use the recorded time-step data for the response history.
Perforated cantilever plate · 915 elements · 352 nodes
The GIF shows an illustrative phase cycle, not physical time integration.
Perforated cantilever plate · 915 elements · 352 nodes
The GIF shows an illustrative phase cycle, not physical time integration.
Perforated cantilever plate · 915 elements · 352 nodes
The GIF ramps to the computed state; it is not a load history.
Quarter-torus elbow · 443 elements · 960 nodes
The GIF ramps to the computed state; it is not a load history.
Quarter-torus elbow · 443 elements · 960 nodes
The GIF ramps to the computed state; it is not a load history.
Quarter-torus elbow · 443 elements · 960 nodes
The GIF ramps to the computed state; it is not a load history.
Quarter-torus elbow · 443 elements · 960 nodes
The GIF shows an illustrative phase cycle. Mode amplitude is normalized for display, not a physical displacement.
Quarter-torus elbow · 443 elements · 960 nodes
The GIF shows an illustrative phase cycle. Mode amplitude is normalized for display, not a physical displacement.
Quarter-torus elbow · 443 elements · 960 nodes
The GIF is a display ramp, not the transient time history. Use the recorded time-step data for the response history.
Quarter-torus elbow · 443 elements · 960 nodes
The GIF is a display ramp, not the transient time history. Use the recorded time-step data for the response history.
Quarter-torus elbow · 443 elements · 960 nodes
The GIF shows an illustrative phase cycle, not physical time integration.
Quarter-torus elbow · 443 elements · 960 nodes
The GIF shows an illustrative phase cycle, not physical time integration.
Quarter-torus elbow · 443 elements · 960 nodes
The GIF ramps to the computed state; it is not a load history.
Quarter-cylinder shell panel · 84 elements · 56 nodes
The GIF ramps to the computed state; it is not a load history.
Quarter-cylinder shell panel · 84 elements · 56 nodes
The GIF ramps to the computed state; it is not a load history.
Quarter-cylinder shell panel · 84 elements · 56 nodes
The GIF shows an illustrative phase cycle. Mode amplitude is normalized for display, not a physical displacement.
Quarter-cylinder shell panel · 84 elements · 56 nodes
The GIF shows an illustrative phase cycle. Mode amplitude is normalized for display, not a physical displacement.
Quarter-cylinder shell panel · 84 elements · 56 nodes
The GIF shows an illustrative phase cycle. Mode amplitude is normalized for display, not a physical displacement.
Quarter-cylinder shell panel · 84 elements · 56 nodes
The GIF is a display ramp, not the transient time history. Use the recorded time-step data for the response history.
Quarter-cylinder shell panel · 84 elements · 56 nodes
The GIF is a display ramp, not the transient time history. Use the recorded time-step data for the response history.
Quarter-cylinder shell panel · 84 elements · 56 nodes
The GIF shows an illustrative phase cycle, not physical time integration.
Quarter-cylinder shell panel · 84 elements · 56 nodes
The GIF shows an illustrative phase cycle, not physical time integration.
Quarter-cylinder shell panel · 42 elements · 56 nodes
The GIF ramps to the computed state; it is not a load history.
Quarter-cylinder shell panel · 42 elements · 56 nodes
The GIF ramps to the computed state; it is not a load history.
Quarter-cylinder shell panel · 42 elements · 56 nodes
The GIF shows an illustrative phase cycle. Mode amplitude is normalized for display, not a physical displacement.
Quarter-cylinder shell panel · 42 elements · 56 nodes
The GIF shows an illustrative phase cycle. Mode amplitude is normalized for display, not a physical displacement.
Quarter-cylinder shell panel · 42 elements · 56 nodes
The GIF shows an illustrative phase cycle. Mode amplitude is normalized for display, not a physical displacement.
Quarter-cylinder shell panel · 42 elements · 56 nodes
The GIF is a display ramp, not the transient time history. Use the recorded time-step data for the response history.
Quarter-cylinder shell panel · 42 elements · 56 nodes
The GIF is a display ramp, not the transient time history. Use the recorded time-step data for the response history.
Quarter-cylinder shell panel · 42 elements · 56 nodes
The GIF shows an illustrative phase cycle, not physical time integration.
Quarter-cylinder shell panel · 42 elements · 56 nodes
The GIF shows an illustrative phase cycle, not physical time integration.
These cases failed or reached the run time limit. Their model inputs and run summaries remain in the downloadable dataset; no failed mesh is shown as a computed deformation.
| Family | Analysis | Method | Outcome | Run message |
|---|---|---|---|---|
| TET4 | Linear static | cg | FAILED | CG requires a verified real SPD reduced matrix. For matrices larger than spd_dense_check_max_dofs, set assume_spd=true only when the model derivation establishes positive definiteness. Use MINRES for symmetric indefinite systems or GMRES/BiCGSTAB for nonsymmetric systems. Run record |
| TET4 | Linear static | conjugate_gradient | FAILED | CG requires a verified real SPD reduced matrix. For matrices larger than spd_dense_check_max_dofs, set assume_spd=true only when the model derivation establishes positive definiteness. Use MINRES for symmetric indefinite systems or GMRES/BiCGSTAB for nonsymmetric systems. Run record |
| TET4 | Linear static | gmres | FAILED | gmres did not converge (iteration limit reached, info=9900). Run record |
| TET4 | Linear static | minres | FAILED | minres residual is abnormal: relative=1.145018e-04, allowed=1.000000e-07. Run record |
| TET4 | Modal analysis | lobpcg | FAILED | Modal eigenpair residual is abnormal: relative=1.193948e-05, allowed=1.000000e-07. Run record |
| TET4 | Material nonlinear static | modified_newton | FAILED | Nonlinear step 1 did not converge in 40 iterations; relative residual=9.552128e-01. Run record |
| TET4 | Material nonlinear static | newton_line_search | TIMED_OUT | Case exceeded its frozen wall-time limit of 60 s. Run record |
| TET4 | Material nonlinear static | arc_length | FAILED | Arc-length continuation reached max_arc_steps before the target load factor. Run record |
| TET4 | Geometric nonlinear static | newton_raphson | FAILED | Geometric nonlinear solve did not converge at increment 1; relative residual=9.093803e-01. Run record |
| TET10 | Linear static | cg | FAILED | CG requires a verified real SPD reduced matrix. For matrices larger than spd_dense_check_max_dofs, set assume_spd=true only when the model derivation establishes positive definiteness. Use MINRES for symmetric indefinite systems or GMRES/BiCGSTAB for nonsymmetric systems. Run record |
| TET10 | Linear static | conjugate_gradient | FAILED | CG requires a verified real SPD reduced matrix. For matrices larger than spd_dense_check_max_dofs, set assume_spd=true only when the model derivation establishes positive definiteness. Use MINRES for symmetric indefinite systems or GMRES/BiCGSTAB for nonsymmetric systems. Run record |
| TET10 | Linear static | gmres | FAILED | gmres did not converge (iteration limit reached, info=27690). Run record |
| TET10 | Linear static | minres | FAILED | minres residual is abnormal: relative=7.846963e-04, allowed=1.000000e-07. Run record |
| TET10 | Modal analysis | eigh | FAILED | Dense modal solve refused for 2769 free dofs; limit=2000. Use method='eigsh' and request fewer modes, or raise dense_modal_max_dofs explicitly. Run record |
| TET10 | Modal analysis | lobpcg | FAILED | Modal eigenpair residual is abnormal: relative=8.615588e-02, allowed=1.000000e-07. Run record |
| TET10 | Material nonlinear static | modified_newton | FAILED | Nonlinear step 1 did not converge in 40 iterations; relative residual=2.269845e+00. Run record |
| TET10 | Material nonlinear static | newton_line_search | TIMED_OUT | Case exceeded its frozen wall-time limit of 60 s. Run record |
| TET10 | Material nonlinear static | arc_length | TIMED_OUT | Case exceeded its frozen wall-time limit of 60 s. Run record |
| MITC3 | Linear static | cg | FAILED | CG requires a verified real SPD reduced matrix. For matrices larger than spd_dense_check_max_dofs, set assume_spd=true only when the model derivation establishes positive definiteness. Use MINRES for symmetric indefinite systems or GMRES/BiCGSTAB for nonsymmetric systems. Run record |
| MITC3 | Linear static | conjugate_gradient | FAILED | CG requires a verified real SPD reduced matrix. For matrices larger than spd_dense_check_max_dofs, set assume_spd=true only when the model derivation establishes positive definiteness. Use MINRES for symmetric indefinite systems or GMRES/BiCGSTAB for nonsymmetric systems. Run record |
| MITC3 | Linear static | gmres | FAILED | gmres did not converge (iteration limit reached, info=2880). Run record |
| MITC3 | Linear static | bicgstab | FAILED | bicgstab did not converge (iteration limit reached, info=2880). Run record |
| MITC3 | Linear static | minres | FAILED | minres residual is abnormal: relative=1.942813e-03, allowed=1.000000e-07. Run record |
| MITC3 | Modal analysis | lobpcg | FAILED | Modal eigenpair residual is abnormal: relative=1.156978e-04, allowed=1.000000e-07. Run record |
| MITC4 | Linear static | cg | FAILED | CG requires a verified real SPD reduced matrix. For matrices larger than spd_dense_check_max_dofs, set assume_spd=true only when the model derivation establishes positive definiteness. Use MINRES for symmetric indefinite systems or GMRES/BiCGSTAB for nonsymmetric systems. Run record |
| MITC4 | Linear static | conjugate_gradient | FAILED | CG requires a verified real SPD reduced matrix. For matrices larger than spd_dense_check_max_dofs, set assume_spd=true only when the model derivation establishes positive definiteness. Use MINRES for symmetric indefinite systems or GMRES/BiCGSTAB for nonsymmetric systems. Run record |
| MITC4 | Linear static | gmres | FAILED | gmres did not converge (iteration limit reached, info=2880). Run record |
| MITC4 | Linear static | bicgstab | FAILED | bicgstab did not converge (iteration limit reached, info=2880). Run record |
| MITC4 | Linear static | minres | FAILED | minres did not converge (iteration limit reached, info=1440). Run record |
| MITC4 | Modal analysis | lobpcg | FAILED | Modal eigenpair residual is abnormal: relative=2.897947e-02, allowed=1.000000e-07. Run record |
The figures and run records come from QF Solver 0.2.0a0 and Gmsh 4.15.2, started October 18, 2026. The current site release is 0.2.11; these calculations were not rerun for that release. The source bundle marks the campaign as COMPLETED_WITH_LIMITATIONS and records blocked source-baseline readiness with owner review at 0/168.
The dataset retains the archived run maturity and verdict, numerical checks, input and mesh hashes, Gmsh geometry, mesh, import report and legend. Each card looks up the recorded scope in the v0.2.8 registry. The public subset omits the vendored engine and unabridged solver-result arrays; their provenance hashes remain in the manifest.