Data of "Discrete non-local interactions formulation of deep material networks for damage-enhanced constitutive material models"

Discrete Non-Local Interaction-Based Deep Material Network This directory contains the data and algorithms generated in publication1 Table of Contents Dependencies and Prerequisites Structure of Repository Generating the data-base Training the IB-DMN Run simulations and visualisation Reproduce paper[^1] figures Dependencies and Prerequisites Python, pandas, matplotlib, texttabble and latextable are pre requisites for visualizing and navigating the data. For generating mesh and for vizualization, gmsh (www.gmsh.info) is required. For running simulations, cm3Libraries (http://www.ltas-cm3.ulg.ac.be/openSource.htm) is required. Instructions using apt & pip3 package manager Instructions for Debian/Ubuntu based workstations are as follows. python, pandas and dependencies sudo apt install python3 python3-scipy libpython3-dev python3-numpy python3-pandas matplotlib, texttabble and latextable pip3 install matplotlib texttable latextable Structure of Repository ElasticTrainingData: Generate the training data base, see Generating the data-base DMN_Training: Train the IB-DMN, see Training the IB-DMN ELP: Run simulations (DNS and IB-DMN), keep the predictions, and vizualize the graphs for elliptical inclusions, see Run simulations and visualisation H12: Run simulations (DNS and IB-DMN), keep the predictions, and vizualize the graphs for circular inclusions with SVE size of 12x12 micrometres2, see Run simulations and visualisation H20: Run simulations (DNS and IB-DMN), keep the predictions, and vizualize the graphs for circular inclusions with SVE size of 20x20 micrometres2, see Run simulations and visualisation Generating the data-base All the files are in ElasticTrainingData: MicroSample_LocalInfo.py: Generate homogenised elasticity tensors for linear elastic training. As an option, the local strain and stress can be collected. Elastic_Tensor_LocalInfo.py and utils.py: The functions will be called in MicroSample_LocalInfo.py. SVE_Data/: Directory that keeps the generated data. Training the IB-DMN All the files are in DMN_Training with names in H12, H20 for circular fibres and SVEs of size 12x12 or 20x20 micrometres2, and name in ELP for elliptical fibres: SingleH12SVEPara.py, SingleH20SVEPara.py and SingleELPSVEPara.py: Training DMN for each SVE of each kinf and keeping the training results according to required epochs. WriteH12Para.py, WriteH20Para.py and WriteELPPara.py: Compute the weight matrix using the parameters of DMN and the defined non-local length. Write the input data files to perform the nonlinear simulation. DMN_Tree.py: The functions will be called in the previously listed python files. DMNPara: Save the training results of SingleH12SVEPara.py, SingleH20SVEPara.py and SingleELPSVEPara.py. Run simulations and visualisation All the files are either in ELP for elliptical fibres, H12 for circular fibres and SVEs of size 12x12 micrometres2 or H20 for circular fibres and SVEs of size 20x20 micrometres2: H12/SVE_Data: Keeps the geometry information of the 6 SVEs for direct simulations. H12/DMN_1Dam: Script to run and visualise IB-DMN results for the UD fibre-reinforced single damage-enhanced J2-elasto-plastic matrix: DMNPara: Archive the output files of WriteH12Para.py, see Training the IB-DMN, which are the input data files for non-linear IB-DMN simulations. SimulationDMN_H12.py: Scripts to run the IB-DMN non-linear simulations; The simulation results are saved in directory DMN_simulation. Plot_Energy.py: Plot DNS and IB-DMN simulation irreversible energy results with parameters from different non-local lengths for IB-DMN and DNS. Plot_Length.py: Plot DNS and IB-DMN stress-strain curve results with parameters from different non-local lengths for IB-DMN and DNS. Plot_Max.py: Plot maximum stresses of DNS and DMN simulation results with parameters from different non-local lengths. H12/DNS_1Dam: Direct simulations scripts (requires cm3Libraries) and save results in Path_Res on SVEs for the UD fibre-reinforced single damage-enhanced J2-elasto-plastic matrix l03: scripts for DNS with a non-local length of 3 micrometres. l06: scripts for DNS with a non-local length of 6 micrometres. l10: scripts for DNS with a non-local length of 10 micrometres. H12/DMN_2Dam: Script to run and visualise IB-DMN results for the UD fibre-reinforced multi-damage-enhanced non-associated pressure-dependent elasto-plastic matrix: DMNPara: Archive the output files of WriteH12Para.py, see Training the IB-DMN, which are the input data files for non-linear IB-DMN simulations. SimulationDMN_2Dam.py: Scipts to run the IB-DMN non-linear simulations; The simualtion results are saved in directory DMN_simulation. Plot_Max.py: Plot maximum stresses of DNS and DMN simulation results with parameters from different non-local lengths for IB-DMN only. Plot_rve_x.py: Plot DNS and IB-DMN simulation results with parameters from different non-local lengths for IB-DMN only. H12/DNS_2Dam: Direct simulations scripts (requires cm3Libraries) and save results in Path_Res on SVEs for the UD fibre-reinforced multi-damage-enhanced non-associated pressure-dependent elasto-plastic . H20/SVE_Data: Keeps the geometry information of the 6 SVEs for direct simulations. H20/DMN_1Dam: Script to run and visualise IB-DMN results for the UD fibre-reinforced single damage-enhanced J2-elasto-plastic matrix: DMNPara: Archive the output files of WriteH20Para.py, see Training the IB-DMN, which are the input data files for non-linear IB-DMN simulations. SimulationDMN_H20.py: Scripts to run the IB-DMN non-linear simulations; The simulation results are saved in directory DMN_simulation. LocalDamDMN.py: Scripts to run the local damage IB-DMN non-linear simulations; The simulation results are saved in directory DMN_simulation. Plot_Dam.py: Plot leaf nodes damage distribution with parameters from different non-local lengths for IB-DMN and for local IB-DMN. Plot_Energy.py: Plot DNS and IB-DMN simulation irreversible energy results with parameters from different non-local lengths for IB-DMN and DNS. Plot_Level.py: Plot DNS and IB-DMN stress-strain curve results for different IB-DMN levels. Plot_Length.py: Plot DNS and IB-DMN stress-strain curve results with parameters from different non-local lengths for IB-DMN and DNS. Plot_Length3to4.py: Plot DNS and IB-DMN stress-strain curve results with parameters from different non-local lengths for IB-DMN only. Plot_Max.py: Plot maximum stresses of DNS and DMN simulation results with parameters from different non-local lengths for IB-DMN and DNS. Plot_Max3to4.py: Plot maximum stresses of DNS and DMN simulation results with parameters from different non-local lengths for IB-DMN only. Plot_LocalDam.py: Plot DNS and local IB-DMN simulation results. H20/DNS_1Dam: Direct simulations scripts (requires cm3Libraries) and save results in Path_Res on SVEs for the UD fibre-reinforced single damage-enhanced J2-elasto-plastic matrix l03: scripts for DNS with a non-local length of 3 micrometres. l06: scripts for DNS with a non-local length of 6 micrometres. l10: scripts for DNS with a non-local length of 10 micrometres. H20/DMN_2Dam: Script to run and visualise IB-DMN results for the UD fibre-reinforced multi-damage-enhanced non-associated pressure-dependent elasto-plastic matrix: DMNPara: Archive the output files of WriteH20Para.py, see Training the IB-DMN, which are the input data files for non-linear IB-DMN simulations. SimulationDMN_2Dam.py: Scipts to run the IB-DMN non-linear simulations; The simulation results are saved in directory DMN_simulation. Plot_rve_x.py: Plot DNS and IB-DMN simulation results with parameters from different non-local lengths for the IB-DMN only. H20/DNS_2Dam: Direct simulations scripts (requires cm3Libraries) and save results in Path_Res on SVEs for the UD fibre-reinforced multi-damage-enhanced non-associated pressure-dependent elasto-plastic matrix. ELP/ELP_sve.geo: Keeps the geometry information of the elliptical fibre SVE for direct simulations. ELP/DMN: Script to run and visualise IB-DMN results for the elliptical fibre-reinforced single damage-enhanced J2-elasto-plastic matrix: DMNPara: Archive the output files of WriteELPPara.py, see Training the IB-DMN, which are the input data files for non-linear IB-DMN simulations. SimulationDMN_ELP.py: Scripts to run the IB-DMN non-linear simulations; The simulation results are saved in directory DMN_simulation. Plot_Length.py: Plot DNS and IB-DMN stress-strain curve results with parameters from different non-local lengths for IB-DMN only. ELP/DNS: Direct simulations scripts (requires cm3Libraries) and save results in Path_Res on SVEs for the elliptical fibre-reinforced single damage-enhanced J2-elasto-plastic matrix Reproduce paper[^1] figures Figures 5a and 6a: H12/DMN_1Dam/Plot_Length.py Figures 5b and 6b: H12/DMN_1Dam/Plot_Max.py Figures 7a and 8a: H20/DMN_1Dam/Plot_Length.py Figures 7b and 8b: H20/DMN_1Dam/Plot_Max.py Figures 9, 10 and 11: H20/DMN_1Dam/Plot_Dam.py Figures 12a and 13a: H20/DMN_1Dam/Plot_Length3to4.py Figures 12b and 13b: H20/DMN_1Dam/Plot_Max3to4.py Figure 14: H20/DMN_1Dam/Plot_Energy.py Figures 16 and 17: ELP/DMN/Plot_Length.py Figure 18: H12/DMN_2Dam/Plot_rve_x.py Figure 19: H20/DMN_2Dam/Plot_rve_x.py Figure 20: DMN_Training/Plot_LossH20 Figure 21: H20/DMN_1Dam/Plot_Level.py Disclaimer This project has received funding from the European Union’s Horizon Europe Framework Programme under grant agreement No. 101056682 for the project “DIgital DEsign strategies to certify and mAnufacture Robust cOmposite sTructures (DIDEAROT)”. The contents of this publication are the sole responsibility of ULiege and do not necessarily reflect the opinion of the European Union. Neither the European Union nor the granting authority can be held responsible for them. The work is described in:"Wu, L. and Noels, L. (2026). Discrete non-local interactions formulation of deep material networks for damage-enhanced constitutive material models ??: ??, doi: ???" which can be downloaded on repository. We would be grateful if you could cite this publication in case you use the files. ↩