Integrated Computational Materials, Process and Product Engineering – IC-MPPE 2022

The manufacturing industries have already started to incorporate data driven and hybrid techniques into their associated material processes where utilization of physical and data-driven technologies along with Machine Learning (ML) and smart data handling have recently gained a tremendous momentum. Optimizing process controls through digital twining and improving operations and maintenances along with minimization of overheads and deficiencies are among main goals of the process digitalization drive. Hybrid physical-data driven modelling with its associated ML schemes along with Reduced Order Modelling (ROM) techniques are among the new trends in design and monitoring of light weight material processes. In this research work, an effective use of hybrid ROM scheme for predictive modelling of material processes have been investigated and its agility of dealing with optimization of process parameters and part performance has briefly been reviewed. Further attempts have also been made to employ the hybrid physical-data driven modelling for the real-time optimization of the process parameters using fast predictive-corrective models. With the introduction of these hybrid models, the combination of experimental, numerical simulation and mined data sources have been considered to setup proper semantic databases for data processing and filtering. Furthermore, the combination of hybrid models and ROM techniques along with ML modules have been employed for selected cases of material processes as pilot studies.