This conference paper was submitted for presentation at the NAFEMS World Congress 2025, held in Salzburg, Austria from May 19–22, 2025.

Abstract

This work presents an innovative collaborative research initiative between Nissan Technical Centre Europe, RBF Morph, and the University of Rome Tor Vergata aimed at advancing the design of automotive road wheels through a multi-physics optimization approach. Road wheels are a critical component where aesthetics, performance, and efficiency converge, making their optimization a multifaceted challenge. The proposed methodology addresses three core requirements. The first is styling, as the wheel'™s design must align with consumer appeal and brand identity, making it a key selling point. The second is structural performance, achieved through rigorous finite element analysis (FEA) and experimental validation to ensure compliance with impact and fatigue durability requirements as well as guarantee optimal NVH and handling characteristics. The third is aerodynamics, which is particularly vital for electric vehicles due to the significant impact of spoke design on drag and vehicle range. Computational fluid dynamics (CFD) simulations and wind tunnel tests are integral to this analysis. An integrated computer-aided engineering (CAE) workflow has been developed, allowing design iterations generated from styling considerations to be evaluated against structural and aerodynamic key performance indicators (KPIs). The key enabler of this approach is advanced mesh morphing, which facilitates rapid geometry updates while maintaining high-fidelity simulations. To further accelerate the process, reduced-order models and artificial intelligence are employed to refine designs and achieve superior performance outcomes efficiently. The software platform rbfCAE is adopted to orchestrate shape control across CAE solvers as it allows to adapt CAD defined variations onto computational meshes both for FEA, in this specific case solid models for the NX Nastran solver, and CFD in this specific case the volume mesh for the HELYX solver. The automation can be upfront driven or evolutive. In both cases a set of key shape parameters is defined so that the high fidelity models can be updated. In the upfront drive case a dataset of high fidelity simulations is computed by means of intense HPC automation and snapshots are created. Principal orthogonal decomposition combined with AI is then adopted to compress the multi-physics results and an interactive inspection of the obtained reduced order models can be performed in the rbfCAE UI and/or in its companion rbfVR tool; in the evolutive approach the shapes are computed in sequence and the optimisation process converges toward the optimal design. This study highlights the synergy between advanced simulation tools and cross-disciplinary expertise, offering a robust solution to enhance both the performance and aesthetics of wheels, with a particular emphasis on the unique demands of electric vehicles. The two approaches proposed are compared showing how the innovative AI/ROM, at an higher upfront cost, allows to look for optimal performances whilst controlling the aesthetic results. The more traditional optimisation tool requires less iterations to get better performances but the resulting shape has to be accepted regardless the style requirement.