Experience of defect tolerant design for Additively Manufactured components in high performance cars

Additive Manufacturing (AM) has rapidly advanced in the motorsport field, enabling the production of custom lightweight components with high-performance alloys. Rapid design iterations and short production lead times make AM ideal to improve component performance. However, predicting fatigue resistance remains challenging due to the inherent presence of manufacturing defects. This work presents the application of a defect-tolerant methodology to predict the impact of manufacturing defects on component performance under operating conditions. First, the manufacturing defects of standard fatigue specimens and a selected component were revealed by X-ray Computed Tomography. Then, machine learning-assisted Extreme Value Statistics was adopted to estimate the occurrence of different defect types in critical regions of the component. Finally, a probabilistic fracture-based design model was applied to quantify the influence of defect size on fatigue performance.