Data of "Bayesian inference of high-dimensional finite-strain visco-elastic-visco-plastic model parameters for additive manufactured polymers and neural network based material parameters generator"
DOI10.5281/zenodo.7792804Zenodo7792804MaRDI QIDQ6703053FDOQ6703053
Dataset published at Zenodo repository.
Miguel Monclus, Javier Segurado, Fatma Karayagiz, Ubiratan Santos Freitas, Ludovic Noels, Ling Wu, Lucia Cobian, Cyrielle Anglade
Publication date: 2 April 2023
Copyright license: Creative Commons Attribution 4.0 International
General Data of https://doi.org/10.1016/j.ijsolstr.2023.112470 related to MOAMMM project. Data related to the publication (we would be grateful if you could cite the paper in the case in which you are using the data): title = Bayesian inference of high-dimensional finite-strain visco-elastic-visco-plastic model parameters for additive manufactured polymers and neural network based material parameters generator., journal = International Journal of Solids and Structures, year = 2023, volume = 283, pages = 112470, doi = 10.1016/j.ijsolstr.2023.112470, author = Ling Wu, Cyrielle Anglade, Lucia Cobian, Miguel Monclus, Javier Segurado, Fatma Karayagiz, Ubiratan Freitas, and Ludovic Noels This project has received funding from the European Unions Horizon 2020 research and innovation programme under grant agreement No 862015. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union. Neither the European Union nor the granting authority can be held responsible for them. Description BI code and results of the inference of a pressure-dependent visco-elastic visco-plastic model developed in [NGU16] with a umat implementation in https://gitlab.uliege.be/moammm/moammmPublic/code/-/tree/main/MaterialModels/FiniteStrain/Finite_VEVP. The BI is described in [WU23] .The experimental results used in the BI are reported in [COB22,COB22b]. To run the BI you need the open source code https://gitlab.onelab.info/cm3/cm3Libraries If you use these data or model, we would be grateful if you could cite the related papers. Bibliography [WU23] L. Wu, C. Anglade, L. Cobian, M. Monclus, J. Segurado, F. Karayagiz, U. Santos Freitas, L. Noels, Bayesian inference of high-dimensional finite-strain visco-elastic-visco-plastic model parameters for additive manufactured polymers and neural network based material parameters generator, International Journal of Solids and Structures (2023) 112470: https://doi.org/10.1016/j.ijsolstr.2023.112470 [COB22] L. Cobian, M. Rueda-Ruiz, J.P. Fernandez-Blazquez, V. Martinez, F. Galvez, F. Karayagiz, T. Lck, J. Segurado, M.A. Monclus, Micromechanical characterization of the material response in a PA12-SLS fabricated lattice structure and its correlation with bulk behaviour, Polymer Testing 110 (2022) 107556: https://doi.org/10.1016/j.polymertesting.2022.107556 (in Open access) [COB22b] Data of . Cobian, M. Rueda-Ruiz, J.P. Fernandez-Blazquez, V. Martinez, F. Galvez, F. Karayagiz, T. Lck, J. Segurado, M.A. Monclus, Micromechanical characterization of the material response in a PA12-SLS fabricated lattice structure and its correlation with bulk behaviour, Polymer Testing 110 (2022) 107556 http://dx.doi.org/10.5281/zenodo.6136935 (in Open access) [NGU16] V. D. Nguyen, F. Lani, T. Pardoen, X. Morelle, L. Noels, A large strain hyperelastic viscoelastic-viscoplastic-damage constitutive model based on a multi-mechanism non-local damage continuum for amorphous glassy polymers. International Journal of Solids and Structures 96 (2016): 192-216; https://dx.doi.org/10.1016/j.ijsolstr.2016.06.008, Open access: https://orbi.uliege.be/handle/2268/197898 Directories All the codes and experimental results are in five directories: experimentalTests: experimental data, see the README.txt in each subdirectory for details BayesianVE: BI of the visco-elastic parameters PlotExperimentalCurves: to vizualize the experimental curves and prepare the observations for the BI in the VE range Loadcase_H.py and Loadcase_V.py read experimental results and create Load_ExpVE_H.dat and Load_ExpVE_V.dat, which keep the experimental observations and loading conditions to perform the BI. PrintDir_H PrintDir_V subdirectories with the functions called by Loadcase_H.py and Loadcase_V.py Load_ExpVE_H.dat and Load_ExpVE_V.dat created files with the observations and loading conditions to perform the BI VE_V2Step and VE_H: BI for viscoelastic properties of V specimen (VE_V2Step) and H specimen (VE_H) BI_allpos_sequence.py runs the BI using Predict_VETest.py and creates the MCMC_VE_....dat WarmStart = True is used to restart an inference MCMC_VE_....dat in the VE_V2Step and VE_H directories are the BI results When proceeding in two steps in VE_V2Step, a first step generates MCMC_VE_VN8_1st.dat whose posterior is used as prior in the second step to generate MCMC_VE_VN8_2nd.dat CheckBayRes: to visualize predictions of a BI sample and experimental curves MCMCRes.py is used to check the numerical predictions of a BI parameter sample (read last sample by default, V or H direction can be selected at line ResKGEmu.py plots the evolution of elastic properties with time uses as input VE_V2Step/MCMC_VE_....dat or VE_H/MCMC_VE_....dat uses local ViscoElasticTest.py, line.geo, line. msh as interface with https://gitlab.onelab.info/cm3/cm3Libraries code uses local functions plotExpLoad_Unload.py, plotExp.py ViscoElasticTest.py, line.geo, line.msh: interface with https://gitlab.onelab.info/cm3/cm3Libraries code used by VE_V2Step and VE_H to call the VEVP model BayesianVEVP: BI of the visco-elastic and visco-plastic parameters PlotExperimentalCurves: to vizualize the experimental curves and prepare the observations for the BI in the VE-VP ranges Loadcase_H.py and Loadcase_V.py read experimental results and create Load_ExpVEVP_H.dat and Load_ExpVEVP_V.dat, which keep the experimental observations and loading conditions to perform BI at the viscoplastic stage. PrintDir_H PrintDir_V subdirectories with the functions called by Loadcase_H.py and Loadcase_V.py Load_ExpVEVP_H.dat and Load_ExpVEVP_V.dat created files with the observations and loading conditions to perform the BI VP_V2step and VP_H2step: BI for viscoelastic-viscoplastic properties of V specimen (VP_V2Step) and H specimen (VP_H2Step) BI_allpos_sequence.py runs the BI using Predict_VETest.py and creates the MCMC_VP_....dat WarmStart = True is used to restart an inference It starts from the VE prosterior as prior, see point 2, and generates a MCMC_VP_?_1of2Steps.dat (? being H or V) Then using MCMC_VP_?_1of2Steps.dat posterior to get a new prior, it generates MCMC_VP_?_2of2Steps.dat (? being H or V) CheckBayRes: visualize predictions of a BI sample and experimental curves MCMCRes.py is used to check the numerical predictions with 3 BI parameter samples ([28000, 45000,70000] by default, V or H direction can be selected at line 12) using the samples of BayesianVEVP/VP_?2step/MCMC_VP_?_2of2Steps.dat (? being H or V) plot_hist.py is used to plot histograms of all the inferred parameters using the samples of BayesianVEVP/VP_?2step/MCMC_VP_?_2of2Steps.dat (? being H or V) Plot_Prop.py plots joints histograms of the inferred parameters using the samples of BayesianVEVP/VP_?2step/MCMC_VP_?_2of2Steps.dat (? being H or V) ResKGEmu.py plots the evolution of elastic properties with time VEVPTest.py: interface with https://gitlab.onelab.info/cm3/cm3Libraries code used by VP_V2Step and VP_H2Step to call the VEVP model RandomParametersGenerator: used to generate the parameters from the BI samples, with the same statistical content Generator DataProcess.py: creates normalized data for training from final inferred parameters in ../MCMC_ResData and creates ?_dirNormData (? being H or V) KmeanDataProcess.py: performs clustering for the data of H_dirNormDat and creates H_dirNormData_2cluster (no need for V direction because not bimodal) Gan_V.py and Gan_H.py are used to train the random material parameter generators and create the VDir_Gan or HDir_Gan200_0/HDir_Gan200_1 GenerateParameters.py generates random parameters using the Gan files VDir_Gan or HDir_Gan200_0/HDir_Gan200_1 and checks the joint histograms of generated parameters, generated parameters are in V_GenData and H_GenData Ganlib.py is used by the generator CheckRes GenDataRes.py is used to check the numerical predictions with the generated parameter samples, see point 4) (using V_GenData and H_GenData). Plot_PropGen.py plots joints histograms of the generated parameters using the samples of V_GenData or H_GenData MCMC_ResData:All final data used in the paper (they can substitute the ones used here above) H_direction and V_direction keep the MCMC random walk results of BI. RandomParameterGenerator keeps results of the generator Paper Figures of [WU23] Fig. 5: From directory BayesianVE/PlotExperimentalCurves, run python3 ./PrintDir_V/plotExp_T.py or ./PrintDir_V/plotExp_C.py or ./PrintDir_V/plotExp_R.py Fig. 7: BayesianVEVP/CheckBayRes/Plot_Prop.py with direct = V and then with direct = H and with Var = [0,1,20,24,28,29,30,31] Fig. 8: BayesianVEVP/CheckBayRes/MCMCRes.py with direct = V (requires https://gitlab.onelab.info/cm3/cm3Libraries code) Fig. 9: BayesianVEVP/CheckBayRes/MCMCRes.py with direct = H (requires https://gitlab.onelab.info/cm3/cm3Libraries code) Fig. 11: RandomParametersGenerator/CheckRes/Plot_PropGen.py with direct = V and then with direct = H and with Var = [0,1,20,24,28,29,30,31] Fig. 12: RandomParametersGenerator/CheckRes/GenDataRes.py with direct = V (requires https://gitlab.onelab.info/cm3/cm3Libraries code) Fig. 13: RandomParametersGenerator/CheckRes/GenDataRes.py with direct = H (requires https://gitlab.onelab.info/cm3/cm3Libraries code) Fig. 14A: From directory BayesianVE/PlotExperimentalCurves, run python3 ./PrintDir_H/plotExp_T.py or ./PrintDir_H/plotExp_C.py or ./PrintDir_H/plotExp_R.py Fig. 15C: BayesianVEVP/CheckBayRes/Plot_Prop.py with direct = V, Var = [2,3,8,9,14,15,18,19] and [20,21,22,23,24,25,26,27] Fig. 16C: BayesainVEVP/CheckBayRes/plot_hist.py with direct = V Fig. 17C: BayesianVEVP/CheckBayRes/plot_hist.py with direct = V Fig. 18C: BayesianVEVP/CheckBayRes/Plot_Prop.py with direct = H, Var = [2,3,8,9,14,15,18,19] and [20,21,22,23,24,25,26,27] Fig. 19C: BayesianVEVP/CheckBayRes/plot_hist.py with direct = H Fig. 20C: BayesianVEVP/CheckBayRes/plot_hist.py with direct = H Fig. 21D: RandomParametersGenerator/CheckRes/Plot_PropGen.py with direct = V , Var = [2,3,8,9,14,15,18,19] and [20,21,22,23,24,25,26,27] Fig. 22D: RandomParametersGenerator/CheckRes/Plot_PropGen.py with direct = H , Var = [2,3,8,9,14,15,18,19] and [20,21,22,23,24,25,26,27]
This page was built for dataset: Data of "Bayesian inference of high-dimensional finite-strain visco-elastic-visco-plastic model parameters for additive manufactured polymers and neural network based material parameters generator"
