Base metal
S-N and ε-N evaluations run in the same analysis, on shells and solids, with any of the solvers: inertia-relief, inertia-relief via normal modes and modal-transient.
Goodman, Gerber, Walker or FKM for S-N; Morrow or SWT for ε-N. Neuber plasticity correction. Materials mapped by property id or material id.
Absolute maximum principal, signed von Mises, Tresca and critical plane are evaluated together. No separate run per criterion.
Damage is evaluated on every candidate plane through a point. The worst plane, its normal vector and its life are reported. One algorithm covers shells and solids, which matters in thick cast, forged and 3D-printed sections.
The open material database holds S-N, ε-N, brittle, crack-growth and weld cards. Every card carries its source.
Welds
Weld life from grid-point force histories, for time-history and PSD loads, with every solver. One block in the deck defines the welds.
*Weld_Definition Spot_1D_Set beams ! CBAR CBEAM CBUSH CWELD (element set) Spot_3D_Set solid ! ACM hexas, 2T / 3T stacks Spot_Method RUPP NAKAHARA DONG Weld_PIDs 5 ! PSHELL id(s) of the seam-weld elements Seam_Method DONG VOLVO
77,100 cyclespredicted at the test crack site of a welded T-tube. The twelve tested specimens averaged 75,000 (29,000 to 190,000).
4000 lbf at R = −1. The toe stress is within 7 % of the published value, and the result holds on 8 mm and 16 mm meshes.
In a three-sheet stack the inner sheet carries the load transfer of both interfaces, and FatDyn-X reports it separately. Applied to an EV body with more than 5,000 spot welds and over a million time steps.
Events solved by different solvers are summed into one damage map. The damage is summed per weld and per angle before the maximum is taken, so the crack direction belongs to the whole duty cycle.
Frequency domain
Normal modes plus a PSD give fatigue life. For load PSDs no frequency-response run in a commercial solver is needed, because FatDyn-X forms the transfer functions itself.
Spectral moments, RMS, irregularity factor and peak rate of the stress response, then Dirlik damage per element for shells and solids.
Load (force) PSDs go through the internal frequency response. Acceleration PSDs at the base use a solver deck that FatDyn-X writes for you.
Spot and seam weld fatigue runs under PSD loads, for load PSDs and for base-excitation PSDs.
On the same model and loading, PSD fatigue finds the same hotspots as full time-history fatigue. To compress a duty cycle into one shaker test, see accelerated PSD tests.
Beyond ductile metals
One material card carries two mechanisms: cyclic S-N with no endurance limit and tension-only damage, and slow crack growth from time at stress. Every constant is temperature-dependent.
Stress and temperature histories straight from OptiStruct, Nastran or LS-DYNA thermal transients. Solids and shells, with S-N, ε-N and brittle materials mixed in one run.
Verification
| Case | FatDyn-X | Reference |
|---|---|---|
| Perforated plate, CQUAD4, S-N Gerber, life | 1,026 | 1,026 (benchmark) |
| Perforated plate, CTRIA3, S-N Gerber, life | 604 | 604 (benchmark) |
| Seam weld, welded T-tube, cycles at R = −1 | 77,100 | 75,000 (mean of 12 tests) |
| Spot weld, inertia-relief via modes, damage per repeat | 1.08E-06 | 1.08E-06 (static inertia-relief) |
| Brittle: SiC block cycles, damage | 0.200270 | 0.200267 (hand) |
| Brittle: alumina ±180 MPa at 900 °C, damage | 0.13915 | 0.1392 (hand) |