FILAMIN
The cylindrical lamination process (“flow forming”) of tubes, an incremental deformation process that requires high precision in tolerances and finishes. Currently, there is limited knowledge about the interaction between the elastic deformation of the tooling and the elastoplastic deformation of the piece, which causes defects such as scales and folds, as well as inaccuracies in diameters and final thicknesses. In addition, the elastic recovery of the tube after deformation is difficult to predict and correct, generating rejection problems due to deviations in the thicknesses.
The objective of the project is to provide the machine with the necessary intelligence to anticipate and correct defects in real time, achieving pieces with dimensional, mechanical and microstructural reliability. Currently, the system is being modeled, characterizing the materials and rigidity of the machine, and simulations are carried out to predict the formation of defects. The key variables of the process, the requirements of the tools and the elastic recovery of the materials are also being analyzed.
The sensors needed to obtain data from the process in real time (dimensions, temperature, vibrations, etc.) are being identified and a machine learning model based on these data and simulations is being developed. An artificial intelligence algorithm is indicating to the machine how to correct the process parameters (speed of rotation and advance, strength, etc.) in order to obtain the desired dimensions and eliminate defects.
The project is allowing to incorporate a system of pioneering “zero defects” worldwide in its laminating shifts. In addition, the knowledge that is being acquired is key for CIM and PROCOMAME to extend the benefits of AI and advanced sensors to other processes and sectors, driving the transition towards Industry 4.0.
Project reference: CPP2022-009710
Program and call: Project Public Private Collaboration, State Subprogramme for Knowledge Transfer, State Program to promote scientific-technical research and its transfer
Participants: CIEFMA-PROCOMAME, DigiFact, Grouping of Centers for the Advanced Digital Factory (Technium Network/CIT UPC)
CPP2022-009710 grant from the project funded by MICIU/AEI/10.13039/501100011033 and by the European Union NextGenerationEU/PRTR.
