Research and development of design tools and methods for Additive Manufacturing: virtual prototyping and artificial intelligence

The Industry 4.0 revolution promotes the integration of computers and control systems into manufacturing processes. Additive Manufacturing technologies are a key part of Industry 4.0 because they align with the industrial revolution's essential principles of connectivity, flexibility, and integration with Artificial Intelligence. Additive methods enable high personalisation and flexibility, allowing on-demand creation of customised items without the lengthy and costly retooling required by traditional methods. Additive Manufacturing promotes innovation by enabling rapid prototyping and shortening the time to market for new products. Its digital nature permits direct compatibility with other Industry 4.0 technologies, such as the Internet of Things and Digital Twins. Design for Additive Manufacturing is a field of study that focuses on reducing errors and costs in additive projects from the early design phases. This approach is a crucial practice for avoiding failures and enhancing 3D part quality by optimising the part geometry in function of the manufacturing process and boundary conditions. One of the main advantages of Additive Manufacturing, and particularly the Laser-Powder Bed Fusion process for metals, is its ability to produce highly complex geometries. These intricate shapes can justify the higher cost of metal Laser-Powder Bed Fusion components, compared to traditional manufacturing. Although research continues to improve part quality, reducing porosity, surface roughness, and geometric or dimensional deviations, further studies are needed to fully understand the mechanical performance of such complex geometries under loads. Using a Design for Additive Manufacturing technique, the designer can minimise wasting raw materials and time, resulting in higher performance and printability. The additive process diffusion is slowed down, especially for metals, by the high cost and low repeatability of results. The application of Machine Learning methods in the Additive Manufacturing field is already well-documented in the literature, showing various application levels. Among the most studied are geometric optimisation (Topological Optimisation, Generative Design, Lattice Structures, etc.), process parameter optimisation, and production process control practices. Industry 4.0 delivers vast volumes of data that can be used as training material for a Machine Learning system, creating a direct connection between these two worlds. This thesis proposes a method that combines Artificial Intelligence techniques, advanced CAD modelling, and virtual prototyping into Design for Additive Manufacturing workflows to enhance the integration of these innovative technologies with classical design techniques.