Design and Optimization of Radiator End Tanks using FEA Stress Data for enhanced performance in F1 Car

Formula One radiator end tanks are lightweight thin-walled structures operating under severe internal pressure and thermal loads while also needing to satisfy aerodynamic, packaging, and structural requirements. In such components, even a slight reduction in wall thickness can significantly reduce mass but can also increase stress at the tube-to-header interface. The conventional tank wall thickness optimization procedure is highly based on manual CAD modifications guided by the tank stress contours, but these methods are time consuming and do not adequately account for the non-linear response at the tube-header interface in the optimization process. As a result, weight reduction efforts may unintentionally compromise the structural integrity of the radiator. This thesis investigates a field driven optimization framework for structural light-weighting of Formula One radiator end tanks by combining finite element analysis (FEA), analytical structural mechanics, Python-based data processing, Physics-informed scaling factor, and Implicit modeling in the nTop environment to establish a continuous variable-thickness optimization workflow. A baseline structural model of the radiator assembly was first developed and validated under internal pressure loading. A parametric study was then performed to quantify the influence of tank wall thickness and flow channel cavity height on tube-interface stress, strengthening the argument of radiator structural behavior analogy to classical plate and Euler- Bernoulli beam theory. Based on these findings, a linear field-driven optimization workflow was implemented in nTop as a preliminary analysis, and its results were compared against the scaled factor optimization workflow developed with the help of the corrected stress field using a physics informed coupling factor derived from the structural response of the flow tubes. Furthermore, an independent analytical model based on beam-bending theory and Castigliano’s theorem was also developed to validate the observed non-linear behavior. The observed results show the effectiveness of the scaled factor approach in reducing the tube-header interface stress penalty observed in the linear workflow while preserving mass savings of nearly 17.2% compared to the baseline model. The proposed method improves the reliability of weight optimization by accounting for the coupled structural response of the end tanks and the tube interface, while preserving compatibility with existing industrial design workflows. Overall, the thesis demonstrates the effectiveness of field driven implicit optimization, when corrected for structural coupling, which offers a more effective route to achieve lightweight while preserving structural integrity of radiator end tank design.