Chianese, Giovanni (2023) Process modelling, monitoring and control for laser welded copper-to-steel battery tab connectors. [Tesi di dottorato]

Anteprima

Testo Chianese_Giovanni_35.pdf Download (5MB) | Anteprima

Abstract

The transition toward electric mobility is influencing the industrial strategies of many productors in the automotive field that are now planning to increase the share of electric vehicles (EVs) in their offer. Production of a finished battery pack can account for the 40% of the added value of a battery EV, therefore, large scale sustainable manufacturing of battery packs is emerging as a topic of strategic importance, and players in the automotive industry are focussing their efforts on research and development of technologies and methods to achieve Near Zero Defects (NZD) production. Structure of a battery packs for electric vehicles follows a pack-module-cell layout with up to several thousand of joints within a pack. For this reason, manufacturing connections between cells and modules is a task of critical importance in the entire production process, and repeatability is a strong requirement as, joints with different electrical resistance result in inhomogeneous current loads that can lead to detrimental effects on the performances and durability of the entire system. As it offers relevant technological advantages, such as high production rate, one side accessibility, narrow heat affected zone (HAZ), and possibility to reprocess defective seams, remote laser welding (RLW) enables good flexibility, automatic manufacturing processing and cost-effective mass production. Therefore, it is establishing itself as a key-enabler technology for sustainable manufacturing of connections within battery packs. Furthermore, connections between battery cells consist of joints between dissimilar metallic thin sheets, and RLW is potentially applicable to any cell type configuration and metals combination. Uncontrollable variations involved in the process pose significant challenge, as they can affect repeatability with detrimental effects on the quality of the weld joint and of the battery system. Variations from the manufacturing and clamping tolerances can cause geometric variations of the parts and, ultimately, result in lack of connection. Incorrect thermal management during welding can lead to damage to battery cells due to overpenetration with the unwanted risks of piercing and leaks. Additionally, welding of dissimilar metals with laser technology involves significant mixing, resulting in additional challenge in terms of control of cracking mechanisms and brittle Inter-Metallic Compounds (IMC). All these challenges urgently call for innovative solutions and models to control RLW of dissimilar metallic battery tab connectors. Deployment of control systems can have significant impact toward automatization of the process and achievement of NZD production target. However, its development and implementation consist of intermediate objectives. They are: (i) understanding of complex phenomena involved in the process, (ii) in-process monitoring of targeted nuisance factors, (iii) classification of the actual status of the RLW process, and (iv) development of an architecture for autonomous decision of corrective actions. This dissertation aimed to contribute to achievement of objectives (i), (ii), and (iii) and focused on variations of part-to-part gap and weld penetration depth during RLW of copper-to-steel thin sheets, by addressing the following research topics: 1. Development of a multi-physics CFD model for the simulation of RWL of copper-to-steel thin sheets with variable part-to-part gap and weld penetration depth, 2. Characterization of a photodiode-based sensor to variations of part-to-part gap and weld penetration depth during RLW of dissimilar metallic battery tab connectors, 3. Implementation of photodiodes and supervised Machine Learning algorithms for automatic isolation and diagnosis of weld defects during welding of copper-to-steel thin sheets.

Downloads

Actions (login required)

Modifica documento