Measurement of biocompatible polymers. Assessing technical properties and modelling their environmental impact relationships for new innovative Decathlon products

Prototyping is a fundamental process in product development, particularly in the field of innovative sports equipment, where the rapid and efficient introduction of new products is essential. By val- idating designs early, prototyping helps identify and correct potential issues, thereby reducing the risk of post-production modifications. Investing in advanced technologies is crucial for maintain- ing leadership in the sector. To avoid producing an excessive number of prototypes, finite element analysis (FEA) is utilized. Virtual simulation is significantly less costly compared to building and testing physical prototypes and allows for the identification and analysis of potential structural and behavioral issues in the design before constructing physical prototypes. However, FEA can be very time-consuming.This research aims to develop a machine learning (ML) model capable of replacing FEA calculations, providing quick and accurate results while minimizing measure- ment uncertainty. The goal is to achieve both qualitative and quantitative analysis of displacement responses within seconds. Partial Least Squares Regression (PLSR) is particularly suitable for pre- dicting FEA strain distributions, as it uses an algorithm that reduces dimensionality and maximizes the correlation between independent and dependent variables. In the literature, the integration of FEA and PLSR has been used for biomedical purposes, such as the rapid prediction of stresses and strains in the retina after a soccer ball impact, or quickly predicting brain injury patterns in mild traumatic brain injury (mTBI), with an average error of 3% between the model and the simulation. In this study, before developing a PLSR model, a large dataset containing FEA results of simple 3D geometries with a consistent design was trained, varying only the force application points and materials associated with the 3D geometry, while keeping the mesh unchanged in each test. The database acquisition process was automated using ANSYS APDL software. Subsequently, this study will focus on Decathlon bicycle helmets, where efficient prototyping is crucial not only to meet rigorous safety standards, such as those required for crash tests that assess impact absorption at various points, but also to incorporate innovative sustainable materials and achieve ecological transition goals by 2026. This approach will not only significantly reduce the time required for design validation but also allow for rapid iteration and optimization of designs. Consequently, the company will be able to integrate innovative materials and features more quickly, ensuring high-performance and environmentally sustainable products.