Research on Advanced Materials and Design Optimization of Lightweight Automotive Structures under Uncertain Conditions

This paper presents a novel reliability-based design optimization (RBDO) methodology for lightweight automotive structures that explicitly addresses uncertainties stemming from unknown manufacturing precision.Key design variables are modeled as interval variables, while other parameters are treated as probabilistic variables.The proposed approach integrates interval model transformation techniques and probabilistic model decoupling strategies to formulate a comprehensive reliability design framework.To efficiently handle the mixed uncertainties inherent in structural topology optimization variables, a multi-stage surrogate modeling strategy is developed.This strategy sequentially constructs: a deterministic global surrogate model for system response, surrogate models for the upper and lower bounds of the interval response using an adaptive update strategy, and computes the probabilistic output response via Polynomial Chaos Expansion (PCE) based on the interval response bounds.Comparative studies with established methods, such as the Sequential Optimization and Reliability Assessment (SORA) approach, demonstrate the superior efficiency of the proposed method, reducing the required number of function evaluations by up to 32% (e.g., from 234 to 159 calls in a test case) and significantly decreasing CPU computation time.Case studies on automotive components (e.g., hood, anti-collision beam) yielded optimal design solutions and identified the maximum permissible deviation ranges for varying reliability requirements, confirming the method's efficacy.The optimization results indicate that both the optimal nominal design variables and their manufacturable tolerance ranges can be achieved under practical conditions.Crucially, the findings reveal that by strategically increasing the permissible deviation range of design variables, the required manufacturing precision specifications can be relaxed by approximately 20%, leading to substantial reductions in production costs while maintaining structural reliability and achieving a 15% mass reduction.The method offers a practical pathway for designing cost-effective, reliable lightweight automotive structures under uncertainty.