EWI and materialsIN Launch Real-Time UMW Process Monitoring Based Battery Tab Quality Prediction Product at The Battery Show (Detroit, MI)

Traditional quality assurance (QA) approaches are being confronted with the ever-increasing speed of product development, a decreasing workforce size, and the always present need to manage costs. Today, process monitoring and AI/machine learning can provide fully automated quality assurance of manufactured components. The right combination of the optimal sensors, a robust analytical approach, and the use of a design of experiments (DOE) approach to develop balanced datasets can produce a successful solution. This paper demonstrates a case in which a solution developed by EWI and materialsIN provided 100% quality assurance for ultrasonic welding of lithium-ion battery tabs. 

Battery Tab Welding

Electric vehicles require thousands of successful battery tab joints. EWI partnered with materialsIN to develop a process monitoring solution for quality assurance of battery tab welds (Figure 1). In most cases, destructive statistical process-control-based testing is used for quality assurance. However, the high production rates 

Figure 1: Process monitoring solution for quality assurance developed by materialsIN and EWI

and volumes can lead to a high quantity of work-in progress, a high quantity of uninspected components, and significant scrap-out when failures occur.EWI partnered with materialsIN to develop a process monitoring solution for quality assurance of battery tab welds. Based on EWI’s substantial ultrasonic metal welding experience, this solution employed acoustic microphones to allow for an unintrusive retrofit of production ultrasonic welding systems. As shown in Figure 2, there was a clear qualitative difference between the sound of a good and bad weld. The materialsIN platform was used for real-time analysis of the measured waveforms. In addition to robust analytics, this platform provided straightforward and actionable insights as shown in Figure 3. The materialsIN program was initially taught using stack-up-specific dataset generated by EWI. DOE-based approaches were employed to generate datasets that encompass production relevant quality scenarios. This joint solution demonstrates significant benefits: 100% quality assessment that can increase customer confidence and reduce warranty claims, instant quality results that can reduce scrap costs and work-in progress, and the trends in part quality that can predict maintenance issues and increase uptime.

Figure 2. Comparison of microphone signals of successful (left) and failed (right) welds

Conclusions

The marriage of sensor selection, AI, and intelligent training data generation enables the development of fully automated, real-time quality assurance solutions. These solutions can be developed across the range of welding and manufacturing processes within EWI’s expertise. The materialsIN platform can be applied to 1D, 2D, and 3D datasets, while supporting visualization and AI. It allows inspection of a range of anomalies including micro cracks and pores.  

Figure 3. Example image shows a microphone waveform and the analysis results in the materialsIN platform

If you seek quality assurance solutions, EWI and materialsIN have the experience and engineering resources to help you achieve your goals. Contact Doug Myers at [email protected]. 

Note: Any reference to specific equipment and/or materials is for informational purposes only. Any reference made to a specific product does not constitute or imply an endorsement by EWI of the product or its producer or provider.

Reference

Lindamood, L., Mohr, L., Moghaddas, A., Kitt, A., & Frech, T. (2021, March). Investigation of monitoring methods for ultrasonic metal welding. In Sensors and Smart Structures Technologies for Civil, Mechanical, and Aerospace Systems 2021 (Vol. 11591, pp. 62-70). SPIE.

Alex Kitt, Director of Data Science, has more than a decade of experience performing, leading, and identifying needs for industry relevant research and development. His investigative research work has involved data science, metal additive manufacturing, x-ray CT and optical metrology. augmented reality, and graphene. In his current role, he leads strategic planning and R&D efforts for data science and works to integrate it with EWI’s technical disciplines.

Zachary Corey is a Post-doctoral Fellow with the EWI data science group. He received his undergraduate degree at SUNY – ESF (Syracuse, NY) and both his M.S. and Ph.D. in Materials Design and Innovation at University at Buffalo. Zach is based at EWI’s Buffalo Manufacturing Works facility in New York.

EWI helps industry leaders solve complex design and production challenges, integrate new processes, and bring products to market faster. We provide comprehensive engineering services to help companies identify, develop, and implement the best options for specific applications. With unmatched expertise, state-of-the-art lab facilities, and technology resources, we offer customized solutions that deliver game-changing results faster, more efficiently, and with less risk. For more information about EWI. visit ewi.org.

materialsIN harnesses advanced data analytic techniques and material science for the development, design, discovery and handling of materials. The highly scalable and automated approach empowers clients with real-time insights that allow rapid decision-making in material usage, selection and discovery and enables accelerated innovation in the use of materials in an economically viable manner. For more information about materialsIN, contact Frits Abell at [email protected].