Dataset for Machine Learning-Based Identification of Material Parameters from Cruciform Biaxial Tensile Test Simulations

# Dataset for Machine Learning-Based Identification of Material Parameters from Cruciform Biaxial Tensile Test Simulations ## Description This dataset contains numerical simulation results of cruciform biaxial tensile tests generated using finite element analysis. The data are intended for the development and evaluation of machine learning models for inverse identification of material parameters. Two anisotropic yield criteria were considered:- Hill'48- CPB2006 Both datasets assume Swift isotropic hardening. The files are organized into input features (`X`) describing the simulated mechanical response and target variables (`y`) corresponding to the constitutive model parameters. ## Files | File | Description ||-----|-------------|| X_cruciform_CPB06_Swift.csv | Simulation features using the CPB2006 yield criterion || y_cruciform_CPB06_Swift.csv | Corresponding CPB2006 material parameters || X_cruciform_Hill48_Swift.csv | Simulation features using the Hill'48 yield criterion || y_cruciform_Hill48_Swift.csv | Corresponding Hill'48 material parameters | Total dataset size: ~7.3 GB. ## Project Context This dataset was generated within the research project COMPETE2030-FEDER-00778700 (LSD-TRIP - Learning with Small Data: Towards Robust and Interpretable Predictions with Efficient Data-Driven Material Modelling in Forming Simulations), with the reference 2023.16367.ICDT (doi: https://doi.org/10.54499/2023.16367.ICDT), supported by the Innovation and Digital Transition Programme –Compete 2030, through the ERDF. ## Related Publications Details on the numerical modelling used to generate this dataset are available in: Prates et al., Materials Research Proceedings 41 (2024). https://doi.org/10.21741/9781644903131-200 Prates et al., Materials Research Proceedings 54 (2025). https://doi.org/10.21741/9781644903599-172 ## Author Pedro Prates University of Aveiro