Quality assurance of battery laser welding: A data-driven approach
Panagiotis Stavropoulos, Harry Bikas, Kyriakos Sabatakakis, Christos Theoharatos, Stefano Grossi
Abstract
Battery packs manufactured for electromobility application consist of battery cells/modules connected with joints. While their quality has been significantly improved with the utilization of Laser welding in terms of automation, minimizing the heat-affected zone, and precision, challenges have arisen in the case of joining dissimilar materials. Ranging from the low absorptivity of non-ferrous materials such as Copper (Cu) and Aluminum (Al) when welded using infrared lasers to the formation of brittle intermetallic connections these challenges are increasing the probability of a joint being defective in terms of low electrical conductivity and/or pure mechanical strength. Within the context of a battery pack production scenario, this study introduces a novel online data-driven approach for assessing the resistance and maximum tensile shear strength of Tab-to-Tab Al-Cu laser joints. Basic statistical features were extracted from the data captured using a high-speed infrared camera and used to optimize two different Machine Learning models.