Abstract

Anti-vibration rubber components in automobiles require reliable finite element (FE) analysis in order to control vehicle NVH performance. However, the Ogden hyperelastic model is highly sensitive to data preprocessing and fitting algorithms, so that different analysis programs or procedures can produce inconsistent parameters and even loss of convergence at the product level. In industry, there is a practical need for a simple but robust procedure that links specimen tests to bush stiffness analysis without trial-and-error. In this study, uniaxial and equibiaxial tensile tests were conducted on a natural rubber-based compound (NR/BR, natural rubber-butadiene rubber blend) used in a 4-bridge bush, and input data for the evaluation process were prepared in three different datasets: raw data, zero-setting data and simply preprocessed data. For each dataset, third-order Ogden parameters were obtained using the Evaluate function in Abaqus and the Experimental Data Fit function in Marc, and their stability was assessed by the stability limit and stress–stretch response at the specimen level. The parameter sets were then applied to a 4-bridge bush FE model and compared with Y and Z-direction load–displacement tests, while data fit analysis with Microsoft Excel (Microsoft Corp, USA) was additionally implemented as a supplementary tool. The results show that raw and zero-setting data frequently cause stability warnings and early loss of convergence in bush analysis, whereas the simply preprocessed dataset provides monotonic behavior and the widest stable strain range. Among all combinations, the Marc program with simply preprocessed data exhibits the most robust convergence in both directions while maintaining reasonable agreement with bush stiffness, and is therefore recommended as a practical single reference material model. The proposed procedure demonstrates that appropriate preprocessing and solver selection can significantly improve the robustness and reproducibility of industrial anti-vibration rubber analysis. In addition, data fitting using Microsoft Excel suggests a potential low-cost extension for environments without commercial fitting modules.

Keywords: Anti-vibration rubber;Ogden 3rd-order model;data preprocessing;numerical stability;rubber bush;finite element analysis