Hybrid models for optimization of deep-drawn cups deploying scaling laws: a case study for sustainable development
S. P. Sundar Singh Sivam,N. Harshavardhana,D. Kumaran
DOI: https://doi.org/10.1007/s40430-024-04771-8
IF: 2.361
2024-03-16
Journal of the Brazilian Society of Mechanical Sciences and Engineering
Abstract:The deep drawing process is considered environmentally friendly as it generates minimal waste. This paper introduces the multi-draw deep drawing method and identifies the optimal number and parameters for scaled cup quality. This approach extends from macro to micro-parts, making it suitable for miniaturized devices in high-quality electronic applications. It involves unidirectional rolling (UDR) and bidirectional rolling, reducing the ultimate thickness to 50% of the original. Six-millimeter-thick copper sheets undergo five scaling factor, culminating in the ninth incremental rolling step to create the final scaled sheet. Formability characteristics of the rolled sheet are measured using the L 9 orthogonal Array. In the Altair Inspire Form, three random trials are performed to mimic deep-drawn cups. An integrated model involving AHP (Weights Determination), FEA (Trails Runs), ANN (Back Propagation Neural Network Predicted response), and TARAS (Ranking) is employed to forecast the MCDM response and validate parameters. This integrated model allows for the development of assessment and performance. It can predict and optimize process parameters for multi-draw deep drawing, covering macro to micro-sized cups. The study assesses various quality attributes, including maximum thinning rate, formability, minimum forming force in KN, and Springback in mm. The integrated approach provides optimal parameter values for scaled cup-drawn pieces, and FEA–ANN–AHP–TARAS helps identify the best values. These integrated tools and models yield robust results. For UDR rolling, trail run is the optimally balanced process parameter, with Trail 9 at 93.82% being the best and Trail 1 at 61.6% being the worst. The proposed sensitivity analysis is validated.
engineering, mechanical