PSI - Issue 57

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

751

9

differences were smaller compared to the ALT test. This indicates that even for less accelerated random load the nonlinear model still captures certain nonlinear effects and behaviors that are not accounted for in the linear model.

Fig. 9. Damage comparison for HTR and LTR components (reduced PSD)

3.3. Problem linearization using parameters optimization Non-linear results obtained from FEA calculations emphasize the pressing need to modify the current approach in order to accurately capture real damages in the model. Consequently, the author decided to explore a simplified solution for capturing non-linearities by linearizing the solution within a defined range to be able to use this method as a standard approach in future projects without increasing calculation and preparation time. By employing a well-optimized workflow, a comprehensive parametric optimization study was conducted, utilizing stiffness and damping values as parameters for the connector elements. In the non-linear scenario, the FEA connectors were modeled as non-linear elements with stop boundary conditions. Conversely, in the proposed solution, these elements were replaced with linear elements with predefined stiffness and damping properties. The optimization objectives were established to achieve similar levels of damage at a few defined points and to obtain a comparable stress spectrum. The optimization procedure employed the Steady State Dynamic FEA method, while a multi-start point algorithm served as the optimization algorithm.

BEST PARAMETERS

Fig. 10. Designs tested to find optimal parameters for linearization