Leveraging Cyclic Bending Under Tension Data and an Artificial Neural Network to Predict Extrapolated Strain Hardening Behavior of Dual Phase Steels