Otimização MULTIescala de Revestimentos em baseada em Inteligência Artificial

The project aims to develop a comprehensive intelligence framework for WC-Co-MMCs coatings through a hybrid AM process, integrating both Wire Directed Energy Deposition (W-DED) and Powder Directed Energy Deposition (P-DED) technologies. Leveraging the powder material production expertise of Palbit S.A. in the Portuguese industry, this framework will optimize P-DED parameters independently of machine and powder suppliers, ensuring adaptability to various compositions. The primary objective is to minimize material waste and production time while addressing challenges related to melt pool stability and correlating desired properties with process parameters effectively. The project seeks to establish an efficient framework and roadmap for monitoring the deposition process, implementing intelligent control mechanisms for quality assurance, and enhancing the durability, sustainability, and longevity of coatings in demanding environments. General Objectives: 1.Enhance Energy Efficiency and Environmental Friendliness in Coating Production: Develop methods to make the coating production process more energy-efficient and environmentally friendly, leveraging laser-based additive manufacturing technologies. This includes minimizing material wastage and utilizing supplementary machining techniques and multi-layer materials for stronger, thinner coatings. 2. Improve Coating Properties: Optimize the hardness and toughness of WC-Co-MMCs coatings to enhance the performance and longevity of tools used in industries such as rock drilling and mining. Specific Objectives: i) Automate In-Situ Monitoring and Control: Develop an automated method for in-situ monitoring and control of the coating process, using a model trained with both experimental and simulated data to ensure optimal process parameters and coating quality. II) Integrate Multi-Physics Insights: Utilize multi-physics insights to enhance multimodal in-situ data representation, addressing the complex multiscale and multiphysics nature of the DED process, especially when working with WC-Co MMCs. This includes controlling melt pool behavior to achieve desired material properties and mitigate issues like part cracking and layer delamination. iii) Leverage Existing Infrastructure and Knowledge: Utilize the existing infrastructure (equipped machines) and the knowledge of the team to optimize the process, reduce the need for user input, and democratize DED technology. iV) Enhance Quality and Minimize User Input in DED: Aim to minimize user inputs and enhance the quality of DED processes by integrating simulation with experimental data, particularly for WC-Co MMCs. This involves using AI for swift fine-tuning of monitoring and control models based on changing material compositions. v) Optimize Coating Properties Through Multi-Objective Bayesian Optimization: Employ multi-objective Bayesian optimization to fine-tune controller setpoints based on AI monitoring predictions, targeting specific properties like hardness, toughness, and microstructure. This approach aims to achieve optimal coating properties with fewer samples and in scenarios where model evaluation is complex.