PSI - Issue 79

Ibrahim T. Teke et al. / Procedia Structural Integrity 79 (2026) 17–25

19

Thus, this study closes a critical gap in the structural simulation of dissimilar metal joints by coupling thermomechanical effects with established mechanical load modeling. The proposed approach has direct implications for durability prediction, weld qualification, and design of lightweight, multi-material assemblies in automotive and aerospace sectors. 2. Material and Methods 2.1. Numerical Framework and Fatigue Modeling To simulate the fatigue behavior of spot-welded joints under cyclic loading, a nonlinear finite element (FE) framework was developed using ANSYS Mechanical APDL. The fatigue life prediction was based on the strain-life approach, specifically utilizing the Morrow mean stress correction model, which is well-suited for welded joints where mean stress effects due to residual stresses and asymmetric loading are significant. The Morrow model modifies the elastic strain component of the total strain-life equation to account for the mean stress, and is formulated as: 1 ( ′ − )(2 ) + ′ (2 ) (1) where: • σ max is the maximum principal stress during a load cycle, extracted as ABS (Max Principal Stress) from ANSYS simulation results. • ε a is the strain amplitude. • ′ = 920 (fatigue strength coefficient) • b=−0.106 (fatigue strength exponent) • ′ =0.213 (fatigue ductility coefficient) • c=−0.47 (fatigue ductility exponent) • E is the elastic modulus of the base material. For cyclic plasticity, the cyclic stress – strain behavior was also considered: 2 = ′ ( 2 ) ′ (2) where: • K′=1000 MPa (cyclic strength coefficient) • n′=0.2 (cyclic strain hardening exponent) • Stress and strain data were extracted from each simulation node at critical regions (typically near the weld nugget). • Thermal prestress fields were included via a pre-loading step based on differential thermal expansion modeling, as previously described. Stress results were post- processed to extract the maximum principal stress, from which the mean stress (σ m ) was computed over the load cycle and used in the Morrow model. The integration of this strain – life model with thermal strain-based prestress fields, described in the following section, enabled fatigue life prediction across multiple joint types and load configurations with a unified numerical setup.