W. Halsey, D. Pokkalla, V. Paquit, R. Davies, L. Huang, A. Ilinich, W. Wu, K. Murali, L. Huang, K. Kannan, K. Li, S. Dev, S. Kim
Oak Ridge National Laboratory,
United States
Keywords: AI, manufacturing, optimization, stamping
Summary:
Sheet metal stamping processes are widely used in various industries, including automotive, aerospace, electronics casing, and home appliance owing to their cost-effectiveness, rapid cycle times, mass production capabilities, and relatively high precision. The quality and efficiency of these processes hinge on several critical factors, including material properties, blank dimension and uniformity, die geometry, and processing parameters such as binder force, lubricants, and spacer utilization. To enhance production efficiency, process simulation tools are commonly employed. A typical optimization procedure for manufacturing simulation follows a sequence of steps. Firstly, a set of material properties and stamping parameters is proposed. Subsequently, a computational simulation is executed using this initial parameter set. The third step involves analyzing the simulation results and making parameter adjustments as necessary. In the fourth step, a subsequent computational simulation is conducted using the updated parameter set. These third and fourth steps are iteratively performed to achieve desired forming feasibility and quality. This iterative optimization process is important and acceptable during the engineering of the stamping process and continuous improvements during production; however, it demands substantial computational time, making it impractical for real-time feedback towards rapid corrective actions required for in-line control for running production processes. To overcome this challenge, artificial intelligence (AI) can be leveraged to determine optimal manufacturing parameters within a single manufacturing cycle time. This research proposes an in-line optimization framework incorporating a trained AI model to predict kidney-shaped die forming using AA6114 material. Fifteen process variables are identified, while resultant blank draw-in values are used as the forming quality characteristic. Preliminary results indicate that the AI framework can accurately predict draw-in values based on a given parameter set, a process referred to as forward prediction. Furthermore, the AI framework can also predict the optimal parameter set that leads to the desired draw-in values, referred to as backward prediction. This research is performed in collaborations with AutoFORM and USCAR (United States Council for Automotive Research). The members of USCAR are Ford, GM, and Stellantis.