Data-Driven Multi-Objective Topology Optimization of Electric Vehicle Body-in-White under Real Crashworthiness and Battery Protection Constraints Using Euro NCAP and NHTSA Datasets

Abstract

The increasing structural complexity of modern electric vehicles has introduced substantial challenges in achieving simultaneous lightweight performance, crashworthiness enhancement, and battery safety protection. Unlike conventional internal combustion vehicles, electric vehicle body-in-white (BIW) structures must withstand severe multi-directional crash loads while minimizing deformation transfer to battery enclosures and preserving occupant survivability. Existing optimization approaches often focus on either lightweighting or crash energy absorption independently and rarely integrate real-world crash datasets with multi-objective topology optimization frameworks. This limitation reduces the practical applicability of many current structural optimization methodologies in industrial electric vehicle development. This study proposes a data-driven multi-objective topology optimization framework for electric vehicle BIW structures using real crashworthiness constraints extracted from Euro NCAP and NHTSA datasets. The proposed methodology integrates finite element crash simulations, machine learning-based surrogate modeling, and non-dominated sorting genetic algorithm II (NSGA-II) optimization to simultaneously minimize structural mass and peak battery intrusion while maximizing crash energy absorption performance. Frontal offset impact, side pole impact, and progressive deformable barrier crash scenarios were incorporated into the optimization process to represent realistic multi-load operating conditions. The topology optimization domain was defined around critical BIW load paths, including rocker panels, side members, cross beams, and battery-supporting substructures. A hybrid surrogate-assisted optimization architecture was developed to reduce computational cost while maintaining high prediction accuracy for crash responses. Multiple crash indicators, including intrusion displacement, specific energy absorption, peak deceleration, and structural stiffness, were considered as objective functions and constraints. The optimized BIW configuration demonstrated substantial improvements in crashworthiness behavior compared with the baseline structure. Numerical analyses indicated reductions in battery intrusion and structural mass while preserving compliance with Euro NCAP side-impact safety thresholds. The obtained Pareto-optimal solutions revealed strong nonlinear relationships between lightweight design strategies and crash energy management mechanisms. The results further demonstrated that integrating data-driven surrogate models with topology optimization significantly improves optimization efficiency under multi-load crash conditions. The proposed framework provides a scalable methodology for next-generation electric vehicle structural design and offers practical implications for safety-oriented lightweight vehicle development under increasingly stringent international crash regulations.