PSI - Issue 57

Ewelina Czerlunczakiewicz et al. / Procedia Structural Integrity 57 (2024) 743–753 / Structural Integrity Procedia 00 (2019) 000 – 000

750 8

Higher displacements in linear simulation

Higher displacements in nonlinear simulation

Fig. 7. Displacement results, in chosen time step, comparison

In the next step, the stress curves and RMS von Mises stress results were compared to examine the differences in behavior predicted by the nonlinear modeling approach. Depending on the specific region being analyzed, different trends were observed (Fig. 8). For the high temperature radiator (HTR), which contains numerous clippin g connections with nonlinear definitions, and where stress hot spots were identified near these connections, the nonlinear model predicted significantly higher damage compared to the linear model. In the areas where stress hot spots were observed at a distance from the clipping connections, an opposite trend was observed. The nonlinear model predicted significantly lower damage compared to the linear model. This suggests that in this particular case, the nonlinear effects have a decreasing effect on the structural response and resulting damage. These findings highlight the importance of considering nonlinear behavior in the analysis of clipping connections and stress hot spots, as it can lead to more accurate predictions of damage and failure in such regions.

Fig. 8. Damage comparison for HTR and LTR components (original PSD)

To further investigate the accuracy and necessity of the nonlinear model, additional analyses were conducted with a different excitation. The initial excitation used in the analysis was highly accelerated (ALT). In order to assess the model's performance under different loading conditions, a reduced gRMS PSD signal was chosen as the new excitation for the analysis. The less severe PSD has a spectral value of 0.75 gRMS, since the spectral amplitude has been reduced on purpose. The frequency range of the signals is the same as before. This allowed for a comparison of the model's response and damage predictions under different loading scenarios. The reduced PSD (Fig. 9) revealed differences between the linear and nonlinear models, although the