Manufacturing Analytics for Cold Flow Forming of H30 Aluminum Tubes

Flow forming is a complex manufacturing process and accuracy of the flow formed component largely depends on its selection of process parameters. Very limited work is available in the literature in this regard. Mostly because of non-availability of experimental data and due to the involvement of costly equipment. A large number of experimental data is made available in this book. Initially, regression analysis using full factorial DOE was used for modeling the process. However, it is a multi-input-multi-output process and regression analysis can only model multi-input-single-output systems. Therefore, three soft computing-based approaches such as Back-propagation neural network (BPNN), genetic-neural network (GANN) and Adaptive neuro-fuzzy system (ANFIS) were applied for modeling of the process. We also wanted to obtain the best input combination for a pre-determined output combination. It is called reverse modeling and regression analysis is not possible to be applied for such a case. Therefore, BPNN and GANN were employed for this purpose.

Autorentext

Dr. Bikramjit Podder is presently working as a Scientist at the Defence Research and Development Laboratory(DRDL), Hyderabad, India. He received his MTech and PhD from IIT, Kharagpur and NIT, Durgapur respectively. His research interests include flow forming, machining and modeling & simulation of manufacturing processes.