A finite element model updating method for silicone foam spacer using adaptive surrogate model

To meet the growing demand for cushioning and vibration damping in large-scale equipment and precision structures, preloaded spacer structures are required to apply preload forces for component assembly. Silicone foam spacers, as rubber-like materials with excellent compressive properties, are commonly used as soft connection components between parts. Their static mechanical characteristics directly affect load transfer accuracy and equipment safety. Conducting high-precision finite element analysis before use is a critical means of controlling their performance. Traditional finite element model updating methods face challenges such as numerous parameters due to strong material nonlinearity, high experimental complexity and cost, and low computational efficiency. This study integrates low-discrepancy sequence sampling, the Minimizing Surrogate Prediction (MSP) method, and the Particle Swarm Optimization (PSO) algorithm into an adaptive Kriging surrogate model, proposing a finite element model updating approach. The robustness of the proposed method is investigated through multiple numerical examples. Finally, to better apply this method to the finite element model updating of silicone foam spacers, a key-point polling strategy is introduced to achieve efficient inversion of silicone foam material parameters. As a result, the static response of the finite element model is updated to achieve a correlation coefficient R 2 of 0.9990 with experimental data.