Experimentally Supported Nonlinear Finite Element and Topology Optimization Framework for Lightweight Design of a GH60 Cast Iron Steering Knuckle
摘要
This study presents an experimentally supported nonlinear finite element and topology optimization framework for the lightweight design of an automotive steering knuckle manufactured from GH60 grey cast iron. The baseline component was produced via sand casting and evaluated under three critical loading scenarios: braking, vertical impact, and lateral rim impact through physical bench tests. Nonlinear finite element analyses were performed using Simcenter software to replicate the experimental conditions, incorporating elastic–plastic material behaviour. An experimental–numerical correlation was established based on the observed failure load under lateral rim impact, resulting in an approximate deviation of 5.8% between the predicted stress and the material strength limit. This correlation was used as an engineering-level calibration of the numerical model at the governing loading condition. Following this calibration, topology optimization was carried out in Siemens NX under identical loading and boundary conditions. Three optimized knuckle geometries with different mass reduction levels were reconstructed and subsequently evaluated using nonlinear finite element analyses under braking, vertical, and lateral loading conditions. Among the investigated alternatives, Model 2 achieved approximately 14% mass reduction relative to the baseline configuration while maintaining acceptable stress levels across the considered loading scenarios. The results indicate that integrating experimental observations with nonlinear numerical analysis can provide a consistent basis for strength-based lightweight design of safety-critical components. The additional analyses demonstrated that the experimentally derived calibration approach remained applicable to the investigated topology-optimized geometries, as the governing load case and critical stress concentration regions were preserved after optimization. Nevertheless, the experimental calibration was established using the baseline geometry, and further experimental validation of optimized configurations would be beneficial to assess the applicability of the proposed approach to a broader range of geometries and loading conditions.