CHARACTERIZATION OF PHOTODIODES FOR DETECTION OF VARIATIONS IN PART-TO-PART GAP AND WELD PENETRATION DEPTH DURING REMOTE LASER WELDING OF COPPER-TO-STEEL BATTERY TAB CONNECTORS

This paper addresses sensor characterization to detect variations in part-to-part gap and weld penetration depth using photodiode-based signals during Remote Laser Welding (RLW) of battery tab connectors. Photodiode-based monitoring has been implemented largely for structural welds due to its relatively low cost and ease of automation. However, research in sensor characterization, monitoring and diagnosis of weld defects during joining of battery tab connectors is at an infancy and results are inconclusive. Motivated by the high variability during the welding process of dissimilar metallic thin foils, this paper aims to characterize the signals generated by a photodiode-based sensor to determine whether variations in weld quality can be isolated and diagnosed. Photodiode-based signals were collected during RLW of copper-to-steel thin-foil lap joint (Ni-plated copper 300 μm to Ni-plated steel 300 μm). The presented methodology is based on the evaluation of the energy intensity and scatter level of the signals. The energy intensity gives information about the amount of radiation emitted during the welding process, and the scatter level is associated with the accumulated and un-controlled variations. Findings indicated that part-to-part gap variations can be diagnosed by observing the step-change in the plasma signal, with no significant contribution given by the back-reflection. Results further suggested that over-penetration corresponds to significant increment of the scatter level in the sensor signals. Opportunities for automatic isolation and diagnosis of defective welds based on supervised machine learning are discussed.

Photodiode-Based In-Process Monitoring of Part-to-Part Gap and Weld Penetration Depth in Remote Laser Welding of Copper-to-Steel Battery Tab Connectors

This paper addresses in-process monitoring of part-to-part gap and weld penetration depth using photodiode-based signals during Remote Laser Welding (RLW) of battery tab connectors. Photodiode-based monitoring has been largely implemented for structural welds due to its relatively low cost and ease of automation. However, the application of photodiode-based monitoring to RLW of thin foils of dissimilar metals for battery tab connectors remains an unexplored area of research and will be addressed in this paper. Motivated by the high variability during the welding process of thin foils of dissimilar metals, this paper aims to evaluate the photodiode-based signals to determine if variations in weld quality can be isolated and diagnosed. The main focus is in diagnosing defective weld conditions caused by part-to-part gap variations and/or excessive weld penetration depth. Photodiode-based signals have been collected during RLW of copper-to-steel thin foils lap joint (Ni-plated copper 300 μm to Ni-plated steel 300 μm). The methodology is based on the evaluation of the energy intensity and scatter level of the signals. The energy intensity gives information about the amount of radiation emitted during the welding process, and the scatter level is associated to the accumulated and un-controlled variations. Findings indicated that part-to-part gap variations can be diagnosed by observing the step-change in the plasma signal, with no significant contribution given by the back-reflection. Results further suggested that over-penetration corresponds to significant increment of the scatter level in the sensor signals. Opportunities for automatic isolation and diagnosis of defective welds based on supervised machine learning will be discussed throughout the paper.

Effect of Micro Solidification Crack on Mechanical Performance of Remote Laser Welded AA6063 Fillet Lap Joint in Automotive Battery Tray Construction

Remote laser welding (RLW) has shown a number of benefits of joining 6xxx aluminium alloys such as high processing speed and process flexibility. However, the crack susceptibility of 6xxx aluminium alloys during RLW process is still an open problem. This paper experimentally assesses the impact of transverse micro cracks on joint strength and fatigue durability in remote laser welding of AA6063-T6 fillet lap joints. Distribution and morphology of transverse micro cracks were acquired by scanning electron microscope (SEM) on cross-sections. Grain morphology in the weld zone was determined by electron backscatter diffraction (EBSD) while static tensile and dynamic fatigue tests were carried out to evaluate weld mechanical performance. Results revealed that increasing welding speed from 2 m/min to 6 m/min did not introduce additional transverse micro cracks. Additionally, welding at 2 m/min resulted in tensile strength improvement by 30% compared to 6 m/min due to the expansion of fusion zone, measured by the throat thickness, and refinement of columnar grains near fusion lines. Furthermore, the weld fatigue durability is significantly higher when fracture occurs in weld root instead of fusion zone. This can be achieved by increasing weld root angle with optimum weld fatigue durability at around 55°.

Power Control during Remote Laser Welding Using a Convolutional Neural Network

The increase in complex workpieces with changing geometries demands advanced control algorithms in order to achieve stable welding regimes. Usually, many experiments are required to identify and confirm the correct welding parameters. We present a method for controlling laser power in a remote laser welding system with a convolutional neural network (CNN) via a PID controller, based on optical triangulation feedback. AISI 304 metal sheets with a cumulative thickness of 1.5 mm were used. A total accuracy of 94% was achieved for CNN models on the test datasets. The rise time of the controller to achieve full penetration was less than 1.0 s from the start of welding. The Gradient-weighted Class Activation Mapping (Grad-CAM) method was used to further understand the decision making of the model. It was determined that the CNN focuses mainly on the area of the interaction zone and can act accordingly if this interaction zone changes in size. Based on additional testing, we proposed improvements to increase overall controller performance and response time by implementing a feed-forward approach at the beginning of welding.