Artificial Neural Network-Based Springback Prediction in Sheet Metal Bending for Industry 4.0 Applications

Springback is a major challenge in sheet metal forming, affecting the dimensional accuracy of bent components. This study proposes an Artificial Neural Network (ANN) to predict springback angles from material properties and bending parameters, trained on an industrial dataset including sheet thickness, internal radius, yield strength, elastic modulus, and applied force. The model achieved high predictive accuracy, generalized across materials and conditions, and reduced output variability while slightly underestimating extreme cases. Validation tests confirmed its ability to capture nonlinear elastic recovery, highlighting its potential for integration into Industry 4.0 manufacturing environments.