Machine Learning-Based Models for the Estimation of the Energy Consumption in Metal Forming Processes

Abstract

This research provides an insight on the performances of machine learning (ML)-based algorithms for the estimation of the energy consumption in metal forming processes and is applied to the radial-axial ring rolling process. To define the mutual influence between ring geometry, process settings, and ring rolling mill geometries with the resulting energy consumption, measured in terms of the force integral over the processing time (FIOT), FEM simulations have been implemented in the commercial SW Simufact Forming 15. A total of 380 finite element simulations with rings ranging from 650 mm < DF < 2000 mm have been implemented and constitute the bulk of the training and validation datasets. Both finite element simulation settings (input), as well as the FI (output), have been utilized for the training of eight machine learning models, implemented with Python scripts. The results allow defining that the Gradient Boosting (GB) method is the most reliable for the FIOT prediction in forming processes, being its maximum and average errors equal to 9.03% and 3.18%, respectively. The trained ML models have been also applied to own and literature experimental cases, showing a maximum and average error equal to 8.00% and 5.70%, respectively, thus proving once again its reliability.

Modeling Energy Consumption Using Machine Learning

“…Predictive intelligence of the smart manufacturing approach has been extended to predict energy consumption based on historical data (Essien et al, 2020;Kumar et al, 2021). Machine learning models were used to determine energy consumption in the metal forming process and are applied to the radial-axial ring rolling process (Mirandola et al, 2021). For complex additive manufacturing systems, a hybrid machine learning approach that integrates extreme gradient boosting (XGBoost) decision tree and density-based spatial clustering of applications with noise (DBSCAN) technique are applied (Li et al, 2021).…”

On Improving Sheet Metal Formability in Hydrodynamic Deep Drawing Process through a Solid‐Oil‐Type Mineral Grease Pressure Medium

“…However, the increase in viscosity of the SMG in comparison to the water medium also results in a 21% (pure copper) and 10% (DIN 1.0338 steel) increase in the experimental maximum forming force. Although the minimization of the maximum forming force is normally considered one of the most important optimization criteria in all metal forming processes, [32][33][34] the strong improvement in the thickness uniformity proved by the employment of the SMG medium can be considered as justified when compared with the fairly limited increase in the forming force.…”

Finite Element Simplifications and Simulation Reliability in Single Point Incremental Forming

“…In 1998, Forcellese et al [ 34 ] evaluated the effect of the training set size of ANN on the reliability of the prediction of the springback in the free-bending process, which was also presented in the overview work by Pattanaik in 2013 [ 35 ]. In the following years, the ANN methods were further developed and applied to several forming technologies including deep drawing [ 36 ], ring rolling [ 37 ], electrohydraulic forming [ 38 ], bending [ 39 , 40 ], incremental forming [ 41 ], and several other application areas [ 42 , 43 , 44 , 45 , 46 ]. Hamouche et al [ 47 ] have developed a novel approach to select and classify a sheet metal process by machine-learning method from the final part geometry and achieved an accuracy of 89%.…”