3D-CNN; CNN-LSTM; Convolutional Neural Networks; Image-to-image regression; Natural convection in porous media; Surrogate and inverse modeling; 3d-convolutional neural network; Convection in porous medias; Convolutional neural network; Convolutional neural network-long short-term memory; Image regression; Inverse modelling; Natural convection in porous medium; Surrogate modeling; Condensed Matter Physics; Mechanical Engineering; Fluid Flow and Transfer Processes
Abstract :
[en] Convolutional Neural Networks (CNNs) are gaining significant attention in applications related to coupled flow and transfer processes in porous media, especially when dealing with image-like data. In this context, the most important applications are related to surrogate modeling, where data obtained from simulators is used to train a CNN model. CNNs are also used as optimizers for inverse modeling or parameter estimation. For natural convection in porous media, applications of CNNs are scarce and limited to steady-state data. The main goal of this paper is to extend the applications of CNNs to transient data, by developing new CNN models that allow for integrating time-variant images. Thus, we suggest using an Encoder-Decoder CNN (ED-CNN) for surrogate modeling and a 3D-CNN for inverse modeling. Besides surrogate and inverse modeling, we suggest using CNN for time prediction by coupling it with long short-term memory (LSTM). The performances of these suggested approaches are investigated by applying them to the benchmark of natural convection in porous cavity with heterogeneous property fields, and by comparing the suggested approaches to other alternatives such as standard deep neural network (DNN) and 2D-CNN trained on steady-state data. The results show that, for surrogate modeling, with the same amount of data and equivalent training times, ED-CNN is more practical than DNN because it provides spatially distributed prediction while DNN is limited to local data. The transient data allows for improving the performance of CNN in inverse modeling because it provides more information about heat transfer across the different material zones and thus heterogeneity. The 3D-CNN approach is more efficient than 2D-CNN as it allows for considering the time progress in the training. For instance, the error with 3D-CNN and transient data is about 11 %, while it is about 18 % with 2D-CNN. Coupling CNN with LSTM allows for improving the performance of CNN in time series prediction.
Disciplines :
Mechanical engineering
Author, co-author :
Virupaksha, Adhish Guli; Université de Strasbourg, ENGEES, CNRS, ITES UMR 7063, Strasbourg, France ; Geotechnical Institute, Technische Universität Bergakademie Freiberg, Freiberg, Germany
Nagel, Thomas ; Geotechnical Institute, Technische Universität Bergakademie Freiberg, Freiberg, Germany
Lehmann, François ; Université de Strasbourg, ENGEES, CNRS, ITES UMR 7063, Strasbourg, France
RAJABI, Mohammadmahdi ; University of Luxembourg > Faculty of Science, Technology and Medicine (FSTM) > Department of Engineering (DoE)
Hoteit, Hussein; Physical Science and Engineering Division, King Abdullah University of Science and Technology (KAUST), Saudi Arabia
Fahs, Marwan ; Université de Strasbourg, ENGEES, CNRS, ITES UMR 7063, Strasbourg, France
Le Ber, Florence ; Université de Strasbourg, ENGEES, CNRS, ICube UMR 7357, Strasbourg, France
External co-authors :
yes
Language :
English
Title :
Modeling transient natural convection in heterogeneous porous media with Convolutional Neural Networks
Deutsche Forschungsgemeinschaft Technische Universität Bergakademie Freiberg
Funding text :
We thank Aqeel Afzal Chaudhry from the Geotechnical Institute at TU Bergakademie Freiberg for help in setting up the OGS simulations. We further acknowledge funding by the German Research Foundation (DFG, project INFRA NA1528/2–1 and MA4450/5–1). This work was motivated by results from a joint MSc thesis of the first author between Technische Universität Bergakademie Freiberg and the University of Strasbourg which was supported by the PROCOPE program.We thank Aqeel Afzal Chaudhry from the Geotechnical Institute at TU Bergakademie Freiberg for help in setting up the OGS simulations. We further acknowledge funding by the German Research Foundation (DFG, project INFRA NA1528/2–1 and MA4450/5–1). This work was motivated by results from a joint MSc thesis of the first author between Technische Universität Bergakademie Freiberg and the University of Strasbourg which was supported by the PROCOPE program.
Ai, Y., Han, S., Lei, C., Cheng, J., The characteristics extraction of weld seam in the laser welding of dissimilar materials by different image segmentation methods. Opt. Laser Technol., 167, 2023, 109740, 10.1016/j.optlastec.2023.109740 a.
Ai, Y., Yan, Y., Dong, G., Han, S., Investigation of microstructure evolution process in circular shaped oscillating laser welding of Inconel 718 superalloy. Int. J. Heat Mass Transf., 216, 2023, 124522, 10.1016/j.ijheatmasstransfer.2023.124522 b.
Alsabery, A.I., Sheremet, M.A., Chamkha, A.J., Hashim, I., MHD convective heat transfer in a discretely heated square cavity with conductive inner block using two-phase nanofluid model. Sci. Rep., 8, 2018, 7410, 10.1038/s41598-018-25749-2.
Alsenafi, A., Ferdows, M., Dual solution for double-diffusive mixed convection opposing flow through a vertical cylinder saturated in a Darcy porous media containing gyrotactic microorganisms. Sci. Rep., 11, 2021, 19918, 10.1038/s41598-021-99277-x.
Bai, T., Tahmasebi, P., Efficient and data-driven prediction of water breakthrough in subsurface systems using deep long short-term memory machine learning. Comput. Geosci. 25 (2021), 285–297, 10.1007/s10596-020-10005-2.
Bidarmaghz, A., Narsilio, G.A., Is natural convection within an aquifer a critical phenomenon in deep borehole heat exchangers’ efficiency?. Appl. Therm. Eng., 212, 2022, 118450, 10.1016/j.applthermaleng.2022.118450.
Bilke, L., Flemisch, B., Kalbacher, T., Kolditz, O., Helmig, R., Nagel, T., Development of open-source porous media simulators: principles and experiences. Transp. Porous Media 130 (2019), 337–361, 10.1007/s11242-019-01310-1.
Cao, K., Kim, H., Hwang, C., Jung, H., CNN-LSTM coupled model for prediction of waterworks operation data. J. Inf. Process. Syst. 14 (2018), 1508–1520.
Ekechukwu, G.K., de Loubens, R., Araya-Polo, M., Long short-term memory-driven forecast of CO2 injection in porous media. Phys. Fluids, 34, 2022, 056606, 10.1063/5.0092697.
Fahs, M., Graf, T., Tran, T.V., Ataie-Ashtiani, B., Simmons, C.T., Younes, A., Study of the effect of thermal dispersion on internal natural convection in porous media using fourier series. Transp. Porous Media. 131 (2020), 537–568, 10.1007/s11242-019-01356-1.
Fahs, M., Younes, A., Makradi, A., A reference benchmark solution for free convection in a square cavity filled with a heterogeneous porous medium. Numer. Heat Transf. Part B Fundam. 67 (2015), 437–462, 10.1080/10407790.2014.977183.
Fajraoui, N., Fahs, M., Younes, A., Sudret, B., Analyzing natural convection in porous enclosure with polynomial chaos expansions: effect of thermal dispersion, anisotropic permeability and heterogeneity. Int. J. Heat Mass Transf. 115 (2017), 205–224, 10.1016/j.ijheatmasstransfer.2017.07.003.
Filippi, J.B., Durand, J., Tulet, P., Bielli, S., Multiscale modeling of convection and pollutant transport associated with volcanic eruption and lava flow: application to the April 2007 eruption of the Piton de la Fournaise (Reunion Island). Atmosphere, 12, 2021, 507, 10.3390/atmos12040507.
Ghazvinian, M., Zhang, Y., Seo, D.J., He, M., Fernando, N., A novel hybrid artificial neural network - Parametric scheme for postprocessing medium-range precipitation forecasts. Adv. Water Resour., 151, 2021, 103907, 10.1016/j.advwatres.2021.103907.
Graczyk, K.M., Matyka, M., Predicting porosity, permeability, and tortuosity of porous media from images by deep learning. Sci. Rep., 10, 2020, 21488, 10.1038/s41598-020-78415-x.
Habib, A.B., Akhter, M.E., Sultaan, R., Zahir, Z.B., Arfin, R., Haque, F., Amir, S.A.B., Hussain, M.S., Palit, R., Performance analysis of different 2D and 3D CNN model for liver semantic segmentation: a review. R. Su, H. Liu (Eds.) Medical Imaging and Computer-Aided Diagnosis, 2020, Springer Singapore, Singapore, 166–174, 10.1007/978-981-15-5199-4_17.
Hirthe, E.M., Graf, T., Non-iterative adaptive time-stepping scheme with temporal truncation error control for simulating variable-density flow. Adv. Water Resour. 49 (2012), 46–55, 10.1016/j.advwatres.2012.07.021.
Islam, A.W., Lashgari, H.R., Sephernoori, K., Double diffusive natural convection of CO2 in a brine saturated geothermal reservoir: study of non-modal growth of perturbations and heterogeneity effects. Geothermics 51 (2014), 325–336, 10.1016/j.geothermics.2014.03.001.
Islam, M.D.Z., Islam, M.D.M., Asraf, A., A combined deep CNN-LSTM network for the detection of novel coronavirus (COVID-19) using X-ray images. Inform. Med. Unlocked, 20, 2020, 100412, 10.1016/j.imu.2020.100412.
Javaheri, M., Abedi, J., Natural convection in CO2-based VAPEX. All Days, 2013, SPE, Monterey, California, USA, 10.2118/165352-MS SPE-165352-MS.
Ji, X., Yan, Q., Huang, D., Wu, B., Xu, X., Zhang, A., Liao, G., Zhou, J., Wu, M., Filtered selective search and evenly distributed Convolutional Neural Networks for casting defects recognition. J. Mater. Process. Technol., 292, 2021, 117064, 10.1016/j.jmatprotec.2021.117064.
Jiang, F., Guo, Y., Tsuji, T., Kato, Y., Shimokawara, M., Esteban, L., Seyyedi, M., Pervukhina, M., Lebedev, M., Kitamura, R., Upscaling permeability using multiscale X-ray-CT images with digital rock modeling and deep learning techniques. Water. Resour Res., 2023, 10.1029/2022WR033267.
Jiang, Z., Tahmasebi, P., Mao, Z., Deep residual U-net convolution neural networks with autoregressive strategy for fluid flow predictions in large-scale geosystems. Adv. Water Resour., 150, 2021, 103878, 10.1016/j.advwatres.2021.103878.
Keyes, D.E., McInnes, L.C., Woodward, C., Gropp, W., Myra, E., Pernice, M., Bell, J., Brown, J., Clo, A., Connors, J., Constantinescu, E., Estep, D., Evans, K., Farhat, C., Hakim, A., Hammond, G., Hansen, G., Hill, J., Isaac, T., Jiao, X., Jordan, K., Kaushik, D., Kaxiras, E., Koniges, A., Lee, K., Lott, A., Lu, Q., Magerlein, J., Maxwell, R., McCourt, M., Mehl, M., Pawlowski, R., Randles, A.P., Reynolds, D., Rivière, B., Rüde, U., Scheibe, T., Shadid, J., Sheehan, B., Shephard, M., Siegel, A., Smith, B., Tang, X., Wilson, C., Wohlmuth, B., Multiphysics simulations: challenges and opportunities. Int. J. High Perform. Comput. Appl. 27 (2013), 4–83, 10.1177/1094342012468181.
Kolditz, O., Nagel, T., Shao, H., Wang, W., Bauer, S., Thermo-Hydro-Mechanical-Chemical Processes in Fractured Porous Media: Modelling and Benchmarking: from Benchmarking to Tutoring. 2018, Springer International Publishing, Cham, 10.1007/978-3-319-68225-9.
Kong, Z., Zhang, C., Lv, H., Xiong, F., Fu, Z., Multimodal feature extraction and fusion deep neural networks for short-term load forecasting. IEEE Access 8 (2020), 185373–185383, 10.1109/ACCESS.2020.3029828.
Kreyenberg, P.J., Bauser, H.H., Roth, K., Velocity field estimation on density-driven solute transport with a convolutional neural network. Water Resour. Res. 55 (2019), 7275–7293, 10.1029/2019WR024833.
A. Latrach, M.L. Malki, M. Morales, M. Mehana, M. Rabiei, A critical review of physics-informed machine learning applications in subsurface energy systems, (2023). 10.48550/ARXIV.2308.04457.
Liu, G., Zhou, Z., The natural convection of aquifers with constant heat sources and its influence on temperature fields. J. Hydrodyn. 24 (2012), 617–627, 10.1016/S1001-6058(11)60285-0.
Neto, H.L, Quaresma, J.N.N., Cotta, R.M., Natural convection in three-dimensional porous cavities: integral transform method. Int. J. Heat Mass Transf. 45 (2002), 3013–3032, 10.1016/S0017-9310(02)00015-7.
S. Mo, Y. Zhu, N. Zabaras, X. Shi, J. Wu, Deep convolutional Encoder-Decoder networks for uncertainty quantification of dynamic multiphase flow in heterogeneous media, (2018). 10.48550/ARXIV.1807.00882.
Mohd Razak, S., Jahandideh, A., Djuraev, U., Jafarpour, B., Deep learning for latent space data assimilation in subsurface flow systems. SPE J. 27 (2022), 2820–2840, 10.2118/203997-PA.
Papagiannopoulos, A., Clausen, P., Avellan, F., How to teach neural networks to mesh: application on 2-D simplicial contours. Neural Netw. 136 (2021), 152–179, 10.1016/j.neunet.2020.12.019.
Plúa, C., Vu, M.N., Armand, G., Rutqvist, J., Birkholzer, J., Xu, H., Guo, R., Thatcher, K.E., Bond, A.E., Wang, W., Nagel, T., Shao, H., Kolditz, O., A reliable numerical analysis for large-scale modelling of a high-level radioactive waste repository in the Callovo-Oxfordian claystone. Int. J. Rock Mech. Min. Sci., 140, 2021, 104574, 10.1016/j.ijrmms.2020.104574.
Pyle, R.J., Bevan, R.L.T., Hughes, R.R., Rachev, R.K., Ali, A.A.S., Wilcox, P.D., Deep learning for ultrasonic crack characterization in NDE. IEEE Trans. Ultrason. Ferroelectr. Freq. Control 68 (2021), 1854–1865, 10.1109/TUFFC.2020.3045847.
Rajabi, M.M., Fahs, M., Panjehfouladgaran, A., Ataie-Ashtiani, B., Simmons, C.T., Belfort, B., Uncertainty quantification and global sensitivity analysis of double-diffusive natural convection in a porous enclosure. Int. J. Heat Mass Transf., 162, 2020, 120291, 10.1016/j.ijheatmasstransfer.2020.120291.
Rajabi, M.M., Hajizadeh Javaran, M.R., Bah, A., Frey, G., Ber, F.L., Lehmann, F., Fahs, M., Analyzing the efficiency and robustness of deep Convolutional Neural Networks for modeling natural convection in heterogeneous porous media. Int. J. Heat Mass Transf., 183, 2022, 122131, 10.1016/j.ijheatmasstransfer.2021.122131.
Rao, P.S., Barman, P., Natural convection in a wavy porous cavity subjected to a partial heat source. Int. Commun. Heat Mass Transf., 120, 2021, 105007, 10.1016/j.icheatmasstransfer.2020.105007.
Riley, M.F., Firoozabadi, A., Compositional variation in hydrocarbon reservoirs with natural convection and diffusion. AIChE J. 44 (1998), 452–464, 10.1002/aic.690440221.
Ronneberger, O., Fischer, P., Brox, T., U-Net: convolutional networks for biomedical image segmentation. N. Navab, J. Hornegger, W.M. Wells, A.F. Frangi (Eds.) Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015, 2015, Springer International Publishing, Cham, 234–241, 10.1007/978-3-319-24574-4_28.
Santos, J.E., Xu, D., Jo, H., Landry, C.J., Prodanović, M., Pyrcz, M.J., PoreFlow-Net: a 3D convolutional neural network to predict fluid flow through porous media. Adv. Water Resour., 138, 2020, 103539, 10.1016/j.advwatres.2020.103539.
Shahane, S., Aluru, N.R., Vanka, S.P., Uncertainty quantification in three dimensional natural convection using polynomial chaos expansion and deep neural networks. Int. J. Heat Mass Transf. 139 (2019), 613–631, 10.1016/j.ijheatmasstransfer.2019.05.014.
Shahane, S., Guleryuz, E., Abueidda, D.W., Lee, A., Liu, J., Yu, X., Chiu, R., Koric, S., Aluru, N.R., Ferreira, P.M., Surrogate neural network model for sensitivity analysis and uncertainty quantification of the mechanical behavior in the optical lens-barrel assembly. Comput. Struct., 270, 2022, 106843, 10.1016/j.compstruc.2022.106843.
Shahriari, A., Ashorynejad, H.R., Pop, I., Entropy generation of MHD nanofluid inside an inclined wavy cavity by lattice Boltzmann method. J. Therm. Anal. Calorim. 135 (2019), 283–303, 10.1007/s10973-018-7061-x.
Shahzadi, G., Soulaïmani, A., Deep neural network and polynomial chaos expansion-based surrogate models for sensitivity and uncertainty propagation: an application to a rockfill dam. Water, 13, 2021, 1830, 10.3390/w13131830.
Sheremet, M.A., Pop, I., Natural convection in a square porous cavity with sinusoidal temperature distributions on both side walls filled with a nanofluid: buongiorno's Mathematical Model. Transp. Porous Media 105 (2014), 411–429, 10.1007/s11242-014-0375-7.
Shiranthika, C., Premakumara, N., Chiu, H.L., Samani, H., Shyalika, C., Yang, C.Y., Human activity recognition using CNN & LSTM. Proceedings of the 2020 5th International Conference on Information Technology Research (ICITR), Moratuwa, Sri Lanka, 2020, IEEE, 1–6, 10.1109/ICITR51448.2020.9310792.
Singh, S.P., Wang, L., Gupta, S., Goli, H., Padmanabhan, P., Gulyás, B., 3D deep learning on medical images: a review. Sensors, 20, 2020, 5097, 10.3390/s20185097.
Su, Y., Davidson, J.H., Modeling Approaches to Natural Convection in Porous Media. 1st ed., 2015, Springer International Publishing: Imprint: Springer, Cham, 10.1007/978-3-319-14237-1 2015.
Sun, G., Wang, S., A review of the artificial neural network surrogate modeling in aerodynamic design. Proc. Inst. Mech. Eng. Part G J. Aerosp. Eng. 233 (2019), 5863–5872, 10.1177/0954410019864485.
Tabrizinejadas, S., Fahs, M., Hoteit, H., Younes, A., Ataie-Ashtiani, B., Simmons, C.T., Carrayrou, J., Effect of temperature on convective-reactive transport of CO2 in geological formations. Int. J. Greenh. Gas Control, 128, 2023, 103944, 10.1016/j.ijggc.2023.103944.
Tahmasebi, P., Kamrava, S., Bai, T., Sahimi, M., Machine learning in geo- and environmental sciences: from small to large scale. Adv. Water Resour., 142, 2020, 103619, 10.1016/j.advwatres.2020.103619.
Tamer, A.A., Jalal, N.A., Möller, K., A convolutional neural network with a two-stage LSTM model for tool presence detection in laparoscopic videos. Curr. Dir. Biomed. Eng., 6, 2020, 20200002, 10.1515/cdbme-2020-0002.
Tang, P., Zhang, D., Li, H., Predicting permeability from 3D rock images based on CNN with physical information. J. Hydrol., 606, 2022, 127473, 10.1016/j.jhydrol.2022.127473.
Tartakovsky, A.M., Marrero, C.O., Perdikaris, P., Tartakovsky, G.D., Barajas-Solano, D., Physics-informed deep neural networks for learning parameters and constitutive relationships in subsurface flow problems. Water Resour. Res., 56, 2020, 10.1029/2019WR026731.
Thavarajah, R., Zhai, X., Ma, Z., Castineira, D., Fast modeling and understanding fluid dynamics systems with encoder–decoder networks. Mach. Learn. Sci. Technol., 2, 2021, 025022, 10.1088/2632-2153/abd1cf.
Wang, Z., Simoncelli, E.P., Bovik, A.C., Multiscale structural similarity for image quality assessment. Proceedings of the Thrity-Seventh Asilomar Conference on Signals, Systems & Computers, 2003, Pacific Grove, CA, USA, 2003, IEEE, 1398–1402, 10.1109/ACSSC.2003.1292216.
Wei, H., Zhao, S., Rong, Q., Bao, H., Predicting the effective thermal conductivities of composite materials and porous media by machine learning methods. Int. J. Heat Mass Transf. 127 (2018), 908–916, 10.1016/j.ijheatmasstransfer.2018.08.082.
Wong, J.C., Ooi, C.C., Chattoraj, J., Lestandi, L., Dong, G., Kizhakkinan, U., Rosen, D.W., Jhon, M.H., Dao, M.H., Graph neural network based surrogate model of physics simulations for geometry design. Proceedings of the 2022 IEEE Symposium Series on Computational Intelligence (SSCI), Singapore, Singapore, 2022, IEEE, 1469–1475, 10.1109/SSCI51031.2022.10022022.
Xu, L., Huang, J., Controlling thermal waves with transformation complex thermotics. Int. J. Heat Mass Transf., 159, 2020, 120133, 10.1016/j.ijheatmasstransfer.2020.120133.
Yang, S., Wang, J., Dai, G., Yang, F., Huang, J., Controlling macroscopic heat transfer with thermal metamaterials: theory, experiment and application. Phys. Rep. 908 (2021), 1–65, 10.1016/j.physrep.2020.12.006 a.
Yang, X., Shao, Q., Hoteit, H., Carrera, J., Younes, A., Fahs, M., Three-dimensional natural convection, entropy generation and mixing in heterogeneous porous medium. Adv. Water Resour., 155, 2021, 103992, 10.1016/j.advwatres.2021.103992 b.
Yasrab, R., Gu, N., Zhang, X., An Encoder-Decoder based convolution neural network (CNN) for future advanced driver assistance system (ADAS). Appl. Sci., 7, 2017, 312, 10.3390/app7040312.
Yoon, H., Melander, D.J., Verzi, S.J., Permeability prediction of porous media using Convolutional Neural Networks with physical properties. Sandia National Laboratories Report, 2019, SAND2019–14739C.
Yu, H., Taleghani, A.D., Al Balushi, F., Wang, H., Machine learning for rock mechanics problems; an insight. Front. Mech. Eng., 8, 2022, 1003170, 10.3389/fmech.2022.1003170.
