Conclusion

Additive manufacturing of metal parts is of considerable importance because of its flexibility, both in terms of design (complex shapes with no cost issues, optimized topology, etc.) and production (low waste, no tooling, reactivity, etc.). While additive manufacturing of metal parts is experiencing sustained growth and a real craze in the aerospace industry, its scale in production is limited by the difficulty of controlling the properties and performance of functional parts. Current approaches to finding the most favorable manufacturing parameters involve systematic exploration of the parametric space, necessitating costly and time-consuming test campaigns. Physical models can partly replace experiments, but they remain limited by the immensity of the space to be investigated. To meet these challenges, it will be necessary to make use of artificial intelligence by coupling learning algorithms...