Dataset: A dual-stage constitutive modeling framework based on finite strain data-driven identification and physics-augmented neural networks

############################################################## README########################################################### Dataset for benchmark test of a perforated plate under uniaxial tension(based on: *A dual-stage constitutive modeling framework based on finite strain data-driven identification and physics-augmented neural networks*) ############################################################## REFERENCE########################################################### If you use this dataset, please cite the associated paper: Lennart Linden, Karl Kalina, Jörg Brummund, Brain Riemer, Markus Kästner "A dual-stage constitutive modeling framework based on finite strain data-driven identification and physics-augmented neural networks" Computer Methods in Applied Mechanics and Engineering (CMAME), 2025. This dataset corresponds to Section 4.2: *Benchmarking the DDI formulations with ideal data*. ############################################################## AUTHORS & CONTACT########################################################### Dataset prepared by: Lennart Linden, Karl Kalina, Jörg Brummund, Brain Riemer, Markus KästnerAffiliation: Chair of Computational and Experimental Solid Mechanics, TU Dresden, Dresden, 01069, GermanyContact: [e-mail] Version: v1.0 (August 2025) ############################################################## LICENSE########################################################### © 2025 Lennart Linden, Karl Kalina, Jörg Brummund, Brain Riemer, Markus Kästner. All rights reserved. This dataset is licensed under Creative Commons Attribution-NoDerivatives (CC BY-ND 4.0).More information: https://creativecommons.org/licenses/by-nd/4.0/ ############################################################## DESCRIPTION########################################################### - Geometry: Thin perforated disk (plane stress assumption), domain of interest: 100 mm × 100 mm × 5 mm- Loading: Uniaxial tensile test, displacement at the boundary increased linearly in all increments - number of increments: n_increments = 10- Material model: Neo-Hookean, - material parameter: - Young’s modulus E = 1000 kPa - Poisson’s ratio ν = 0.3- Discretization: Linear triangular finite elements - number of nodes: n_nodes = 4403 - number of elements: n_elements = 8444 - number of quadrature points: n_quadrature_points = 8444- Software: in-house code was used ############################################################## DIRECTORY STRUCTURE########################################################### benchmark_perforated_disk_uniaxial_v1.0/├── README.txt└── Data/ ├── mesh.msh # Finite element mesh ├── displacement.npy # Nodal displacements ├── force.npy # External nodal forces ├── thickness_initial.npy # Initial plate thickness ├── thickness.npy # Plate thickness at quadrature points per increment ├── deformation_gradient.npy # In-plane deformation gradient ├── strain.npy # Euler-Almansi strain field ├── stress.npy # Cauchy stress field ############################################################## DATA FORMAT########################################################### All files (except mesh.msh) are stored as NumPy binary files (.npy). ############################################################## FILE DETAILS########################################################### - mesh.msh Finite element mesh in Gmsh format. - displacement.npy Array shape: (n_increments, 2*n_nodes) Displacements (ux, uy) per node and increment. - force.npy Array shape: (n_increments, 2*n_nodes) External forces (fx, fy) per node and increment. - thickness_initial.npy Scalar: initial plate thickness. - thickness.npy Array shape: (n_increments, n_quadrature_points) Plate thickness at quadrature points for each increment. - deformation_gradient.npy Array shape: (n_increments, n_quadrature_points, 4) In-plane deformation gradient components F11, F12, F21, F22. - strain.npy Array shape: (n_increments, n_quadrature_points, 4) Strain components e11, e22, e12, e33. - stress.npy Array shape: (n_increments, n_quadrature_points, 3) Stress components sigma11, sigma22, sigma12. ############################################################## USAGE NOTES########################################################### Data can be loaded in Python using NumPy: import numpy as np displacements = np.load("Data/displacements.npy") All quantities are provided in SI units:- Length in millimeter (mm)- Forces / 1000 in newton (N)- Stresses in kilopascal (kPa)