PSI - Issue 79

Umberto De Maio et al. / Procedia Structural Integrity 79 (2026) 386–393

391

boundaries. As the mesh density increases (Mesh 2 through Mesh 4), the path becomes progressively smoother and converges towards a more physically realistic trajectory. The paths for Mesh 3 and Mesh 4 are nearly identical, suggesting that solution convergence for the crack path has been achieved. In summary, these results validate the overall effectiveness and robustness of the proposed ALE-driven framework. The model successfully provides a consistent global response that is largely mesh-independent. However, they also clearly demonstrate that obtaining a physically accurate and geometrically correct crack path, a suitable mesh refinement within the critical propagation zone should be performed.

Figure 5. Predicted crack patterns.

Following the mesh sensitivity analysis, Mesh 3 was selected for further simulations. This discretization was chosen as it provides results that are nearly identical to those obtained with the finest mesh (Mesh 4), while significantly reducing the computational time required for the analysis. Figure 6 presents the comparison between the Load versus CMOD curve predicted by the present model using Mesh 3 and the reference results obtained by (Choi et al., 2025), which include both their experimental band (shaded area) and their own numerical simulation (red line). As can be observed, the numerical prediction obtained with the proposed ALE-driven cohesive model exhibits excellent agreement with both the experimental data and the numerical benchmark throughout the entire loading process, including the initial elastic response, the peak load, and the post-peak softening behavior. This good comparison effectively validates the accuracy and predictive capability of the proposed numerical framework for simulating complex fracture phenomena in heterogeneous materials.

Figure 6. Comparison between the numerical and experimental results.