Reza Vaziri
The University of British Columbia
Presenting in Track 2: Advanced Design and Information Technologies
Monday November 9, 8:00am - 8:45am
Presentation Title: A Systems-Based Computational Framework for Continuous Resistance Welding of Thermoplastic Composites
Abstract: Integrated design can extend the light-weighting potential of structural composite components in the aerospace industry. However, with increased component complexity and size, aspects such as manufacturing complexity and costs predominate, which is why differential construction is still the state of the art today. Welding technologies provide an approach for producing integrated, high-performance fibre-reinforced thermoplastic composite structures. A promising technique is continuous resistance welding (CRW), in which a localized heating zone is translated along the joint using a gantry or robotic system. This process uses Joule heating of a conductive insert to generate heat, locally melting the joint along the mating surfaces of the two consolidated components to be joined.
Complexity in CRW processing increases markedly when scaling to production environments. As with other thermoplastic welding technologies, non-invasive measurement of the temperature at the weld interface (i.e. without the use of embedded thermocouples), remains infeasible, despite being critical for controlling the process in order to ensure joint strength and quality. However, physics-based process simulation can help mitigate this limitation and reduce uncertainty in the welding process.
To this end, a rigorous, systems-based computational framework was developed and validated to simulate the CRW process, predict thermal histories, and qualitatively capture interfacial melting. This framework consists of a detailed three-dimensional finite element (FE) model to evaluate the thermal response of the welded components, coupled with an electrical model that generates heat maps accounting for contact resistances and current leakage.
The detailed simulation was used to establish the causal relationship between process parameters and weld interface temperature, thereby guiding the selection of experimental conditions. With a validated model, useful investigations into typical processing deviations were shown to be well captured with predictive control. The full model was further used to determine nominal process parameters, such as applied electric voltage and welding speed, to ensure operation within the desired processing window.
The efficiency limitations of the full FE model, however, hinders process design optimization, real-time control, uncertainty quantification, and large-scale parametric studies that are required for industrial settings. To address these challenges, a surrogate model was developed by learning the input-output relationships of the high-fidelity FE simulations. The surrogate model’s efficient inverse analysis capability enabled its deployment in a robotic CRW demonstration, where electrical current feedback was used to estimate weld temperatures in real time, predict peak temperatures, and compensate for process fluctuations through adaptive adjustments to the applied voltage and welding speed.
Biography: Dr. Vaziri is a Professor in the Department of Civil Engineering and Scientific Director of the Composites Research Network at the University of British Columbia (UBC). He is currently an Associate Dean of the Faculty of Applied Science and previously served as the Head of the Department of Civil Engineering at UBC. Dr. Vaziri obtained his B.Sc. in Civil Engineering from University College London, and his Ph.D. from UBC specializing in mechanics of composite structures. Dr. Vaziri’s contribution to engineering research has been the development of computational modelling techniques to solve a variety of problems in materials and structures and the transferring of the technology to the user community (e.g. industries, government research agencies, and private companies). His research has been focused on the development of analytical and numerical modelling techniques to simulate the manufacturing process of composite structures as well as their fracture and damage behaviour under high intensity impact and crash loadings. Dr. Vaziri’s research has attracted numerous research grants and contracts over the years from various government and industrial sources (e.g. aerospace, automotive and defence) both in Canada and in the US. The body of research has led to the training of several graduate students with over 20 doctoral students graduated. Dr. Vaziri serves on the technical advisory board of several international scientific and engineering organizations, including the International Journal of Impact Engineering. He served as the Co-Chair of the 16th World Congress on Computational Mechanics (WCCM 2024).