Abstract
Mathematical simulation of non-Newtonian fluid flow is an enduring problem with imperative influence on numerous industrial processes such as oil and gas drilling, food processing, etc. The relation between shear rate and shear stress is nonlinear for non-Newtonian fluids, which results in a highly nonlinear governing equation for fluid flow in an irregular geometry. Analytical solution does not exist for such governing equations and is generally solved by algebraic and iterative methods, which is computationally intensive. Convolutional Networks can learn a complex and high-dimensional functional space solution and may have high accuracy but depend significantly on the quality of training data. One of the prominent challenges in using a Convolutional network is the limiting performance, and the proposed solution may become inconclusive in a small data regime over an irregular geometry. A novel algorithm, Boundary Fitted Convolutional Network, is proposed in this research, which can proficiently solve a partial differential equation for an irregular geometry. This research aims to simulate a Power-Law non-Newtonian fluid in an eccentric annulus with a convolutional network without using training data. The governing equations are transformed from physical domain to computational plane using boundary-fitted coordinate system and then solved by minimizing physics-based residuals. Thus, establishing a benchmark investigation in non-Newtonian fluid flow. The Dirichlet and Neumann boundary conditions are applied in a ‘hard’ manner. The simulated results and parametric analysis conclude that the proposed algorithm can decipher the non-Newtonian fluid mechanics from the governing equations. The algorithms also explain the effect of geometric and rheological parameters on the fluid flow attributes. The performance of the algorithm is validated with experimental data available from published studies. The statistical error estimation exhibits an average root mean squared error of 0.228 and mean absolute percentage error of 8.21% for four different samples of Power-Law fluid, with varying eccentricities. A comprehensive discussion to train the unsupervised convolutional network, and the spectrum of hyperparameters considered to expedite convergence is also highlighted.
Similar content being viewed by others
Availability of data and materials
The research uses the published experimental data, and the data could be availed from the cited sources in the manuscript.
Code availability
The computer codes developed for this research will be available online once the manuscript is published.
Change history
12 August 2022
A Correction to this paper has been published: https://doi.org/10.1007/s00521-022-07629-z
Abbreviations
- AD:
-
Automatic differentiation
- BFC-NET:
-
Boundary fitted convolutional network
- CFD:
-
Computational fluid dynamics
- CNN:
-
Convolutional neural network
- DNN:
-
Deep neural network
- FD:
-
Finite difference
- FE:
-
Finite element
- FV:
-
Finite volume
- HB:
-
Herschel Bulkley
- MAE:
-
Mean absolute error
- MAPE:
-
Mean absolute percentage error
- ML:
-
Machine learning
- NN:
-
Neural network
- PDE:
-
Partial differential equation
- PINN:
-
Physics informed machine learning
- PL:
-
Power law
- QoI:
-
Quantity of interest
- RMSE:
-
Root mean squared error
- RF:
-
Random forest
- \(R^{2}\) :
-
Coefficient of determination
- \(\dot{\gamma }\) :
-
Shear rate
- \(\mu_{a}\) :
-
Apparent viscosity
- \(\frac{\partial P}{{\partial z}}\) :
-
Pressure loss gradient
- \(\Delta\) :
-
Laplace operator
- \(J\) :
-
Jacobian matrix
- \(K\) :
-
Fluid consistency index
- \({\text{Re}}\) :
-
Generalized reynolds number
- \(e\) :
-
Eccentricity
- \(f\) :
-
Friction factor
- \(n\) :
-
Fluid behavior index
- \(u,v,w\) :
-
Velocity along \(x,y\) and \(z\)
- \(x,y,z\) :
-
Axial coordinates (physical plane)
- \(\mu\) :
-
Viscosity
- \(\xi ,\eta\) :
-
Computational plane
- \(\tau\) :
-
Stress tensor
- \(\lambda\) :
-
Convolutional filter
References
Peng Y, Lv BH, Yuan JL, et al (2014) Application and prospect of the non-Newtonian fluid in industrial field. In: Materials science forum. Trans Tech Publications Ltd, pp 396–401
Sun L, Gao H, Pan S, Wang JX (2020) Surrogate modeling for fluid flows based on physics-constrained deep learning without simulation data. Comput Methods Appl Mech Eng 361:112732. https://doi.org/10.1016/j.cma.2019.112732
Adariani YH (2005) Numerical simulation of laminar flow of non-Newtonian fluids in eccentric annuli. MSc thesis. The University of Tulsa
Fang P, Manglik RM, Jog MA (1999) Characteristics of laminar viscous shear-thinning fluid flows in eccentric annular channels. J Nonnewton Fluid Mech 84:1–17. https://doi.org/10.1016/S0377-0257(98)00145-1
Raissi M, Perdikaris P, Karniadakis GE (2019) Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. J Comput Phys 378:686–707. https://doi.org/10.1016/j.jcp.2018.10.045
Beck C, Weinan E, Jentzen A (2019) Machine learning approximation algorithms for high-dimensional fully nonlinear partial differential equations and second-order backward stochastic differential equations. J Nonlinear Sci 29:1563–1619. https://doi.org/10.1007/s00332-018-9525-3
Han J, Jentzen A, Weinan E (2018) Solving high-dimensional partial differential equations using deep learning. Proc Natl Acad Sci USA 115:8505–8510. https://doi.org/10.1073/pnas.1718942115
González Cervera JA (2019) Solution of the Black-Scholes equation using artificial neural networks. In: J. Phys. Conf. Ser. https://doi.org/10.1088/1742-6596/1221/1/012044. Accessed 13 Mar 2020
Fang Z, Zhan J (2020) Deep physical informed neural networks for metamaterial design. IEEE Access 8:24506–24513. https://doi.org/10.1109/ACCESS.2019.2963375
Wang N, Zhang D, Chang H, Li H (2020) Deep learning of subsurface flow via theory-guided neural network. J Hydrol 584:124700. https://doi.org/10.1016/j.jhydrol.2020.124700
Perdikaris P, Grinberg L, Karniadakis GE (2016) Multiscale modeling and simulation of brain blood flow. Phys Fluids 28:021304. https://doi.org/10.1063/1.4941315
Winovich N, Ramani K, Lin G (2019) ConvPDE-UQ: convolutional neural networks with quantified uncertainty for heterogeneous elliptic partial differential equations on varied domains. J Comput Phys 394:263–279. https://doi.org/10.1016/j.jcp.2019.05.026
Yang L, Meng X, Karniadakis GE (2021) B-PINNs: Bayesian physics-informed neural networks for forward and inverse PDE problems with noisy data. J Comput Phys 425:109913. https://doi.org/10.1016/j.jcp.2020.109913
Berg J, Nyström K (2019) Data-driven discovery of PDEs in complex datasets. J Comput Phys 384:239–252. https://doi.org/10.1016/j.jcp.2019.01.036
Gao H, Sun L, Wang JX (2021) PhyGeoNet: physics-informed geometry-adaptive convolutional neural networks for solving parameterized steady-state PDEs on irregular domain. J Comput Phys 428:110079. https://doi.org/10.1016/j.jcp.2020.110079
Geneva N, Zabaras N (2020) Modeling the dynamics of PDE systems with physics-constrained deep auto-regressive networks. J Comput Phys 403:109056. https://doi.org/10.1016/j.jcp.2019.109056
Zhu Y, Zabaras N, Koutsourelakis PS, Perdikaris P (2019) Physics-constrained deep learning for high-dimensional surrogate modeling and uncertainty quantification without labeled data. J Comput Phys 394:56–81. https://doi.org/10.1016/j.jcp.2019.05.024
He J, Xu J (2019) MgNet: a unified framework of multigrid and convolutional neural network. Sci China Math 62:1331–1354. https://doi.org/10.1007/s11425-019-9547-2
Joshi A, Shah V, Ghosal S et al (2019) Generative models for solving nonlinear partial differential equations. In: Proc. of NeurIPS Workshop on ML for Physics
Ranade R, Hill C, Pathak J (2021) DiscretizationNet: a machine-learning based solver for Navier–Stokes equations using finite volume discretization. Comput Methods Appl Mech Eng 378:113722. https://doi.org/10.1016/j.cma.2021.113722
Subramaniam A, Wong ML, Borker RD, Nimmagadda S, Lele SK (2018) Turbulence enrichment using physics-informed generative adversarial networks. https://arxiv.org/abs/2003.01907
Mohan AT, Lubbers N, Livescu D, Chertkov M (2020) Embedding hard physical constraints in neural network coarse-graining of 3D turbulence. https://arxiv.org/abs/2002.00021
Kim B, Azevedo VC, Thuerey N et al (2019) Deep fluids: a generative network for parameterized fluid simulations. Comput Graph Forum 38:59–70. https://doi.org/10.1111/cgf.13619
Erge O, Vajargah K, Ozbayoglu A et al (2015) Frictional pressure loss of drilling fluids in a fully eccentric annulus. J Nat Gas Sci Eng 26:1119–1129. https://doi.org/10.1016/j.jngse.2015.07.030
Fredrickson A, Bird RB (1958) Non-Newtonian flow in annuli. Ind Eng Chem 50:347–352. https://doi.org/10.1021/ie50579a035
Herschel WH, Bulkley R (1926) Konsistenzmessungen von Gummi-Benzollösungen. Kolloid-Z 39:291–300. https://doi.org/10.1007/BF01432034
Yao L-S, Mamun Molla M, Ghosh Moulic S (2013) Fully-developed circular-pipe flow of a non-Newtonian pseudoplastic fluid. Univ J Mech Eng 1:23–31. https://doi.org/10.13189/ujme.2013.010201
Hanks RW (1979) The axial laminar flow of yield-pseudoplastic fluids in a concentric annulus. Ind Eng Chem Process Des Dev 18:488–493. https://doi.org/10.1021/i260071a024
Buchtelová M (1988) Comments on “The axial laminar flow of yield-pseudoplastic fluids in a concentric annulus.” Ind Eng Chem Res 27:1557–1558
Haciislamoglu M, Langlinais J (1990) Non-Newtonian flow in eccentric annuli. J Energy Resour Technol Trans ASME 112:163–169. https://doi.org/10.1115/1.2905753
Luo Y, Peden JM (1987) Flow of drilling fluids through eccentric annuli. SPE Annu Tech Conf Exhib Dallas, Texas, USA. https://doi.org/10.2118/16692-ms
Iyoho AW, Azar JJ (1981) Accurate slot-flow model for non-Newtonian fluid flow through eccentric annuli. Soc Pet Eng J 21:565–572. https://doi.org/10.2118/9447-PA
Haciislamoglu M (1989) Non-Newtonian fluid flow in eccentric annuli and its application to petroleum engineering problems, PhD thesis
Kozicki W, Chou CH, Tiu C (1966) Non-Newtonian flow in ducts of arbitrary cross-sectional shape. Chem Eng Sci 21:665–679. https://doi.org/10.1016/0009-2509(66)80016-7
Bertola V, Cafaro E (2003) Analogy between pipe flow of non-Newtonian fluids and 2-D compressible flow. J Nonnewton Fluid Mech 109:1–12. https://doi.org/10.1016/S0377-0257(02)00146-5
Azouz I, Shirazi SA, Pilehvari A, Azar JJ (1993) Numerical simulation of laminar flow of yield-power-law fluids in conduits of arbitrary cross-section. J Fluids Eng Trans ASME 115:710–716. https://doi.org/10.1115/1.2910203
Hussain QE, Sharif MAR (1998) Analysis of yield-power-law fluid flow in irregular eccentric annuli. J Energy Resour Technol Trans ASME 120:201–207. https://doi.org/10.1115/1.2795036
Hashemian Y, Yu M, Miska S et al (2014) Accurate predictions of velocity profiles and frictional pressure losses in annular YPL-fluid flow. J Can Pet Technol 53:355–363. https://doi.org/10.2118/173181-PA
Escudier MP, Oliveira PJ, Pinho FT (2002) Fully developed laminar flow of purely viscous non-Newtonian liquids through annuli, including the effects of eccentricity and inner-cylinder rotation. Int J Heat Fluid Flow 23:52–73. https://doi.org/10.1016/S0142-727X(01)00135-7
Tong TA, Yu M, Ozbayoglu E, Takach N (2020) Numerical simulation of non-Newtonian fluid flow in partially blocked eccentric annuli. J Pet Sci Eng 193:107368. https://doi.org/10.1016/j.petrol.2020.107368
Fluent Thoery Guide (2013) Ansys fluent theory guide. ANSYS Inc, USA 15317:724–746
Ozbayoglu ME, Omurlu C (2006) Analysis of the effect of eccentricity on flow characteristics of annular flow of non-newtonian fluids using finite element method. In: 2006 SPE-ICoTA coiled tubing and well intervention conference and exhibition, proceedings, pp 293–298
Eesa M (2009) CFD studies of complex fluid. PhD thesis
Singh AP, Samuel R (2009) Effect of eccentricity and rotation on annular frictional pressure losses with standoff devices. In: Proceedings—SPE annual technical conference and exhibition, pp 1244–1255
Dokhani V, Shahri MP, Karimi M, Salehi S (2013) Evaluation of annular pressure losses while casing drilling. In: Proceedings—SPE annual technical conference and exhibition, pp 388–402
Rojas S, Ahmed R, Elgaddafi R, George M (2017) Flow of power-law fluid in a partially blocked eccentric annulus. J Pet Sci Eng 157:617–630. https://doi.org/10.1016/j.petrol.2017.07.060
Karimi Vajargah A, van Oort E (2015) Determination of drilling fluid rheology under downhole conditions by using real-time distributed pressure data. J Nat Gas Sci Eng 24:400–411. https://doi.org/10.1016/j.jngse.2015.04.004
Mokhtari M, Ermila M, Tutuncu AN (2012) Accurate bottomhole pressure for fracture gradient prediction and drilling fluid pressure program—Part I. American Rock Mechanics Association
Bralts VF, Kelly SF, Shayya WH, Segerlind LJ (1993) Finite element analysis of microirrigation hydraulics using a virtual emitter system. Trans Am Soc Agric Eng 36:717–725. https://doi.org/10.13031/2013.28390
Sablani SS, Shayya WH, Kacimov A (2003) Explicit calculation of the friction factor in pipeline flow of Bingham plastic fluids: a neural network approach. Chem Eng Sci 58:99–106. https://doi.org/10.1016/S0009-2509(02)00440-2
Swamee PK, Aggarwal N (2011) Explicit equations for laminar flow of Bingham plastic fluids. J Pet Sci Eng 76:178–184. https://doi.org/10.1016/j.petrol.2011.01.015
Lei H, Wu L, Weinan W (2020) Machine-learning-based non-Newtonian fluid model with molecular fidelity. Phys Rev E 102:043309. https://doi.org/10.1103/PhysRevE.102.043309
Kumar A, Ridha S, Ganet T et al (2020) Machine learning methods for herschel-bulkley fluids in annulus: pressure drop predictions and algorithm performance evaluation. Appl Sci. https://doi.org/10.3390/app10072588
Muravleva E, Oseledets I, Koroteev D (2018) Application of machine learning to viscoplastic flow modeling. Phys Fluids 30:103102. https://doi.org/10.1063/1.5058127
Reyes B, Howard AA, Perdikaris P, Tartakovsky AM (2021) Learning unknown physics of non-Newtonian fluids. Phys Rev Fluids. https://doi.org/10.1103/physrevfluids.6.073301
Mahmoudabadbozchelou M, Caggioni M, Shahsavari S et al (2021) Data-driven physics-informed constitutive metamodeling of complex fluids: a multifidelity neural network (MFNN) framework. J Rheol (NY) 65:179–198. https://doi.org/10.1122/8.0000138
Goodfellow I, Bengio Y, Courville A (2017) Deep learning. MIT Press
Karpatne A, Watkins W, Read J, Kumar V (2017) Physics-guided Neural Networks (PGNN): an application in lake temperature modeling. https://arxiv.org/abs/1710.11431
Maglione R, Ferrario G (1996) Equations determine flow states for yield-pseudoplastic drilling fluids. PennWell Publ, Co
Hartnett JP, Kostic M (1990) Turbulent friction factor correlations for power law fluids in circular and non-circular channels. Int Commun Heat Mass Transf 17:59–65. https://doi.org/10.1016/0735-1933(90)90079-Y
Shah SN (1984) Correlations predict friction pressures of fracturing gels. Oil Gas J 82:92–98
Han J, Tao J, Wang C (2020) FlowNet: a deep learning framework for clustering and selection of streamlines and stream surfaces. IEEE Trans Vis Comput Graph 26:1732–1744. https://doi.org/10.1109/TVCG.2018.2880207
Han J, Tao J, Zheng H et al (2019) Flow field reduction via reconstructing vector data from 3-D streamlines using deep learning. IEEE Comput Graph Appl 39:54–67. https://doi.org/10.1109/MCG.2018.2881523
Guo X, Li W, Iorio F (2016) Convolutional neural networks for steady flow approximation. In: dl.acm.org 13–17-August-2016, pp 481–490. https://doi.org/10.1145/2939672.2939738
Bhatnagar S, Afshar Y, Pan S et al (2019) Prediction of aerodynamic flow fields using convolutional neural networks. Comput Mech 64:525–545. https://doi.org/10.1007/s00466-019-01740-0
Thompson JF, Thames FC, Mastin CW (1974) Automatic numerical generation of body-fitted curvilinear coordinate system for field containing any number of arbitrary two-dimensional bodies. J Comput Phys 15:299–319. https://doi.org/10.1016/0021-9991(74)90114-4
Anderson (1995) Computational fluid dynamics: the basics with applications. McGraw-Hill
Ahmed R (2005) Experimental study and modeling of yield power-law fluid flow in pipes and annuli. Oklahoma. https://edx.netl.doe.gov/dataset/experimental-study-and-modeling-of-yield-power-law-fluid-flow-in-pipes-and-annuli-effects-of-drill/resource/da965317-3ac8-4f9e-8d86-99fb0155a3d5/download/NT40637%20Flow%20in%20Pipes%202006%20Nov.pdf
Kumar A, Ridha S, Ganet T et al (2020) Machine learning methods for herschel-bulkley fluids in annulus: pressure drop predictions and algorithm performance evaluation. Appl Sci 10:2588. https://doi.org/10.3390/app10072588
Belimane Z, Hadjadj A, Ferroudji H et al (2021) Modeling surge pressures during tripping operations in eccentric annuli. J Nat Gas Sci Eng 96:104233. https://doi.org/10.1016/J.JNGSE.2021.104233
Smith LN, Topin N (2017) Super-convergence: very fast training of neural networks using large learning rates. https://arxiv.org/abs/1708.07120
Glorot X, Bengio Y (2010) Understanding the difficulty of training deep feedforward neural networks. https://proceedings.mlr.press/v9/glorot10a/glorot10a.pdf
Acknowledgements
The authors acknowledge the support from the Petroleum Engineering Department and Institute of Hydrocarbon Recovery at Universiti Teknologi PETRONAS, Malaysia. This research is funded by YUTP, Grant Number 015lC0-231.
Funding
This research is funded by YUTP, Grant Number 015lC0-231.
Author information
Authors and Affiliations
Contributions
Abhishek Kumar: Conceptualization, Methodology, Software, Investigation, Validation, Formal analysis, Writing—Original Draft. Syahrir Ridha: Writing—Review & Editing, Supervision, Project administration, Funding acquisition. Suhaib Umer Ilyas: Resources, Data Curation, Writing—Review & Editing, Visualization. Iskandar Dzulkarnain: Data Curation, Visualization. Agus Pratama: Data Curation, Visualization.
Corresponding author
Ethics declarations
Conflicts of interest
The authors declare that they have no known competing financial interests or conflict of interest that could have appeared to influence the work reported in this paper.
Ethics approval
The research presented in this manuscript has not been submitted to any other simultaneous journal or any part of the work published elsewhere.
Consent for publication
The author and co-authors of this manuscript consent for the publication in the journal.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendix A
The governing equation used in this study is given by Eq. (A1)
The mathematical relation used to transform the cartesian-based governing equation into boundary fitted coordinate system is given by Eqs. (A2)–(A3).
where the values of \(y_{\eta } , y_{\xi } ,x_{\xi } ,x_{\eta }\) are determined from Eqs. (A4.1)–(A.4.4).
Substituting Eq. (A2) in \(\frac{\partial }{\partial x}\left( {\mu \frac{\partial w}{{\partial x}}} \right)\) yields the expression given by Eq. (A5.1).
Substituting Eq. (A3) in \(\frac{\partial }{\partial y}\left( {\mu \frac{\partial w}{{\partial y}}} \right)\) yields the expression given by Eq. (A5.2).
Combining Eqs. (A5.1) and (A5.2) into Eq. (A6)
Appendix B
The central nodes are discretized for the first derivative using fourth-order accuracy given by Eqs. (B1.1) and (B1.2).
The convolutional filter for the central discretization defined in Eqs. (B1.1) and (B1.2) is given by Eqs. (B2.1) and (B2.2)
Similarly, the discretization for forward and backward schemes used on the boundary is approximated by third-order accuracy. The forward scheme is given by Eqs. (B3.1) and (B3.2)
The backward scheme is given by Eqs. (B4.1) and (B4.2)
Rights and permissions
Springer Nature or its licensor holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Kumar, A., Ridha, S., Ilyas, S.U. et al. Application of boundary-fitted convolutional neural network to simulate non-Newtonian fluid flow behavior in eccentric annulus. Neural Comput & Applic 34, 12043–12061 (2022). https://doi.org/10.1007/s00521-022-07092-w
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00521-022-07092-w