Shell models provide a simplified mathematical framework that captures essential features of incompressible fluid turbulence, such as the energy cascade and scaling of the fluid observables. We perform a precision analysis of the direct and inverse cascades in shell models of turbulence, where the velocity field is a real-valued function. We calculate the leading hundred anomalous scaling exponents, the marginal probability distribution functions of the velocity field at different shells, as well as the correlations between different shells. We find that the structure functions in both cascades exhibit a linear Kolomogorov scaling in the inertial range. We argue that the underlying reason for having no intermittency, is the strong correlations between the velocity fields at different shells. We analyze the tails of velocity distribution functions, which offer new insights to the structure of fluid turbulence.
{"title":"Direct and inverse cascades scaling in real shell models of turbulence","authors":"James Creswell, Viatcheslav Mukhanov, Yaron Oz","doi":"arxiv-2409.11898","DOIUrl":"https://doi.org/arxiv-2409.11898","url":null,"abstract":"Shell models provide a simplified mathematical framework that captures\u0000essential features of incompressible fluid turbulence, such as the energy\u0000cascade and scaling of the fluid observables. We perform a precision analysis\u0000of the direct and inverse cascades in shell models of turbulence, where the\u0000velocity field is a real-valued function. We calculate the leading hundred\u0000anomalous scaling exponents, the marginal probability distribution functions of\u0000the velocity field at different shells, as well as the correlations between\u0000different shells. We find that the structure functions in both cascades exhibit\u0000a linear Kolomogorov scaling in the inertial range. We argue that the\u0000underlying reason for having no intermittency, is the strong correlations\u0000between the velocity fields at different shells. We analyze the tails of\u0000velocity distribution functions, which offer new insights to the structure of\u0000fluid turbulence.","PeriodicalId":501125,"journal":{"name":"arXiv - PHYS - Fluid Dynamics","volume":"7 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142267235","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Thomas P. John, Jack R. C. King, Steven J. Lind, Cláudio P. Fonte
A liquid jet plunging into a quiescent bath of the same liquid is a fundamental fluid mechanical problem underpinning a range of processes in industry and the natural world. Significant attention has been given to the study of plunging laminar Newtonian jets and the associated air entrainment that can occur. However, there have been very few (if any) studies devoted to the equivalent case for non-Newtonian viscoelastic liquids. Here we consider the laminar plunging and associated air entrainment of a shear thinning viscoelastic jet into a still bath of the same liquid. We describe a previously unreported phenomenon, that we call ``bubbles-on-a-string'' (BUoaS), consisting of multiple stable toroidal bubbles rising co-axially around the submerged jet. In a qualitative sense, this new observation is akin to an inverse version of the well-known rheological phenomenon ``beads-on-a-string''. The BUoaS phenomenon is stable and repeatable and can be reproduced to a lesser extent in Newtonian surfactant solutions, indicating that low surface tension is key, but non-Newtonian rheology seems likely to provide the most favourable conditions for the onset of the phenomenon. A full characterisation and detailed study of this behaviour with accompanying numerical simulation is to follow in an upcoming publication.
{"title":"A new complex fluid flow phenomenon: Bubbles-on-a-String","authors":"Thomas P. John, Jack R. C. King, Steven J. Lind, Cláudio P. Fonte","doi":"arxiv-2409.11879","DOIUrl":"https://doi.org/arxiv-2409.11879","url":null,"abstract":"A liquid jet plunging into a quiescent bath of the same liquid is a\u0000fundamental fluid mechanical problem underpinning a range of processes in\u0000industry and the natural world. Significant attention has been given to the\u0000study of plunging laminar Newtonian jets and the associated air entrainment\u0000that can occur. However, there have been very few (if any) studies devoted to\u0000the equivalent case for non-Newtonian viscoelastic liquids. Here we consider\u0000the laminar plunging and associated air entrainment of a shear thinning\u0000viscoelastic jet into a still bath of the same liquid. We describe a previously\u0000unreported phenomenon, that we call ``bubbles-on-a-string'' (BUoaS), consisting\u0000of multiple stable toroidal bubbles rising co-axially around the submerged jet.\u0000In a qualitative sense, this new observation is akin to an inverse version of\u0000the well-known rheological phenomenon ``beads-on-a-string''. The BUoaS\u0000phenomenon is stable and repeatable and can be reproduced to a lesser extent in\u0000Newtonian surfactant solutions, indicating that low surface tension is key, but\u0000non-Newtonian rheology seems likely to provide the most favourable conditions\u0000for the onset of the phenomenon. A full characterisation and detailed study of\u0000this behaviour with accompanying numerical simulation is to follow in an\u0000upcoming publication.","PeriodicalId":501125,"journal":{"name":"arXiv - PHYS - Fluid Dynamics","volume":"49 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142267236","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The use of data-driven methods in fluid mechanics has surged dramatically in recent years due to their capacity to adapt to the complex and multi-scale nature of turbulent flows, as well as to detect patterns in large-scale simulations or experimental tests. In order to interpret the relationships generated in the models during the training process, numerical attributions need to be assigned to the input features. One important example are the additive-feature-attribution methods. These explainability methods link the input features with the model prediction, providing an interpretation based on a linear formulation of the models. The SHapley Additive exPlanations (SHAP values) are formulated as the only possible interpretation that offers a unique solution for understanding the model. In this manuscript, the additive-feature-attribution methods are presented, showing four common implementations in the literature: kernel SHAP, tree SHAP, gradient SHAP, and deep SHAP. Then, the main applications of the additive-feature-attribution methods are introduced, dividing them into three main groups: turbulence modeling, fluid-mechanics fundamentals, and applied problems in fluid dynamics and heat transfer. This review shows thatexplainability techniques, and in particular additive-feature-attribution methods, are crucial for implementing interpretable and physics-compliant deep-learning models in the fluid-mechanics field.
{"title":"Additive-feature-attribution methods: a review on explainable artificial intelligence for fluid dynamics and heat transfer","authors":"Andrés Cremades, Sergio Hoyas, Ricardo Vinuesa","doi":"arxiv-2409.11992","DOIUrl":"https://doi.org/arxiv-2409.11992","url":null,"abstract":"The use of data-driven methods in fluid mechanics has surged dramatically in\u0000recent years due to their capacity to adapt to the complex and multi-scale\u0000nature of turbulent flows, as well as to detect patterns in large-scale\u0000simulations or experimental tests. In order to interpret the relationships\u0000generated in the models during the training process, numerical attributions\u0000need to be assigned to the input features. One important example are the\u0000additive-feature-attribution methods. These explainability methods link the\u0000input features with the model prediction, providing an interpretation based on\u0000a linear formulation of the models. The SHapley Additive exPlanations (SHAP\u0000values) are formulated as the only possible interpretation that offers a unique\u0000solution for understanding the model. In this manuscript, the\u0000additive-feature-attribution methods are presented, showing four common\u0000implementations in the literature: kernel SHAP, tree SHAP, gradient SHAP, and\u0000deep SHAP. Then, the main applications of the additive-feature-attribution\u0000methods are introduced, dividing them into three main groups: turbulence\u0000modeling, fluid-mechanics fundamentals, and applied problems in fluid dynamics\u0000and heat transfer. This review shows thatexplainability techniques, and in\u0000particular additive-feature-attribution methods, are crucial for implementing\u0000interpretable and physics-compliant deep-learning models in the fluid-mechanics\u0000field.","PeriodicalId":501125,"journal":{"name":"arXiv - PHYS - Fluid Dynamics","volume":"33 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142267234","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Dean's approximation for curved pipe flow, valid under loose coiling and high Reynolds numbers, is extended to study three-dimensional travelling waves. Two distinct types of solutions bifurcate from the Dean's classic two-vortex solution. The first type arises through a supercritical bifurcation from inviscid linear instability, and the corresponding self-consistent asymptotic structure aligns with the vortex-wave interaction theory. The second type emerges from a subcritical bifurcation by curvature-induced instabilities and satisfies the boundary region equations. Despite the subcritical nature of the second type of solutions, it is not possible to connect their solution branches to the zero-curvature limit of the pipe. However, by continuing from known self-sustained exact coherent structures in the straight pipe flow problem, another family of three-dimensional travelling waves can be shown to exist across all Dean numbers. The self-sustained solutions also possess the two high-Reynolds-number limits. While the vortex-wave interaction type of solutions can be computed at large Dean numbers, their branch remains unconnected to the Dean vortex solution branch.
迪安曲线管道流近似法在松散卷曲和高雷诺数条件下有效,被扩展用于研究三维行波。从 Dean 的经典双涡解中分叉出两种不同类型的解。第一类是从粘性线性不稳定性的超临界分岔产生的,相应的自洽渐近结构与涡-波相互作用理论一致。第二种类型是由曲率诱导的不稳定性引起的亚临界分岔,并满足边界区域方程。尽管这些第二类解具有亚临界性质,但不可能将其解支与管道的零曲率极限连接起来。然而,通过延续直管流问题中已知的自持精确相干结构,可以证明存在跨越所有迪安数的另一个三维行波族。自持解也具有两个高雷诺数极限。虽然在大迪恩数下可以计算涡-波相互作用类型的解,但它们的分支仍然与迪恩涡解分支无关。
{"title":"Three-dimensional coherent structures in a curved pipe flow","authors":"Runjie Song, Kengo Deguchi","doi":"arxiv-2409.11105","DOIUrl":"https://doi.org/arxiv-2409.11105","url":null,"abstract":"Dean's approximation for curved pipe flow, valid under loose coiling and high\u0000Reynolds numbers, is extended to study three-dimensional travelling waves. Two\u0000distinct types of solutions bifurcate from the Dean's classic two-vortex\u0000solution. The first type arises through a supercritical bifurcation from\u0000inviscid linear instability, and the corresponding self-consistent asymptotic\u0000structure aligns with the vortex-wave interaction theory. The second type\u0000emerges from a subcritical bifurcation by curvature-induced instabilities and\u0000satisfies the boundary region equations. Despite the subcritical nature of the\u0000second type of solutions, it is not possible to connect their solution branches\u0000to the zero-curvature limit of the pipe. However, by continuing from known\u0000self-sustained exact coherent structures in the straight pipe flow problem,\u0000another family of three-dimensional travelling waves can be shown to exist\u0000across all Dean numbers. The self-sustained solutions also possess the two\u0000high-Reynolds-number limits. While the vortex-wave interaction type of\u0000solutions can be computed at large Dean numbers, their branch remains\u0000unconnected to the Dean vortex solution branch.","PeriodicalId":501125,"journal":{"name":"arXiv - PHYS - Fluid Dynamics","volume":"33 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142267529","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Robin A. Heinonen, Luca Biferale, Antonio Celani, Massimo Vergassola
In turbulent flows, tracking the source of a passive scalar cue requires exploiting the limited information that can be gleaned from rare, randomized encounters with the cue. When crafting a search policy, the most challenging and important decision is what to do in the absence of an encounter. In this work, we perform high-fidelity direct numerical simulations of a turbulent flow with a stationary source of tracer particles, and obtain quasi-optimal policies (in the sense of minimal average search time) with respect to the empirical encounter statistics. We study the trajectories under such policies and compare the results to those of the infotaxis heuristic. In the presence of a strong mean wind, the optimal motion in the absence of an encounter is zigzagging (akin to the well-known insect behavior ``casting'') followed by a return to the starting location. The zigzag motion generates characteristic $t^{1/2}$ scaling of the rms displacement envelope. By passing to the limit where the probability of detection vanishes, we connect these results to the classical linear search problem and derive an estimate of the tail of the arrival time pdf as a stretched exponential $p(T)sim exp(-ksqrt{T})$ for some $k>0,$ in agreement with Monte Carlo results. We also discuss what happens as the wind speed becomes smaller.
{"title":"Optimal trajectories for Bayesian olfactory search in the low information limit and beyond","authors":"Robin A. Heinonen, Luca Biferale, Antonio Celani, Massimo Vergassola","doi":"arxiv-2409.11343","DOIUrl":"https://doi.org/arxiv-2409.11343","url":null,"abstract":"In turbulent flows, tracking the source of a passive scalar cue requires\u0000exploiting the limited information that can be gleaned from rare, randomized\u0000encounters with the cue. When crafting a search policy, the most challenging\u0000and important decision is what to do in the absence of an encounter. In this\u0000work, we perform high-fidelity direct numerical simulations of a turbulent flow\u0000with a stationary source of tracer particles, and obtain quasi-optimal policies\u0000(in the sense of minimal average search time) with respect to the empirical\u0000encounter statistics. We study the trajectories under such policies and compare\u0000the results to those of the infotaxis heuristic. In the presence of a strong\u0000mean wind, the optimal motion in the absence of an encounter is zigzagging\u0000(akin to the well-known insect behavior ``casting'') followed by a return to\u0000the starting location. The zigzag motion generates characteristic $t^{1/2}$\u0000scaling of the rms displacement envelope. By passing to the limit where the\u0000probability of detection vanishes, we connect these results to the classical\u0000linear search problem and derive an estimate of the tail of the arrival time\u0000pdf as a stretched exponential $p(T)sim exp(-ksqrt{T})$ for some $k>0,$ in\u0000agreement with Monte Carlo results. We also discuss what happens as the wind\u0000speed becomes smaller.","PeriodicalId":501125,"journal":{"name":"arXiv - PHYS - Fluid Dynamics","volume":"1 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142267528","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Equilibrium, traveling-wave, and periodic-orbit solutions of the Navier-Stokes equations provide a promising avenue for investigating the structure, dynamics, and statistics of transitional flows. Many such invariant solutions have been computed for wall-bounded shear flows, including plane Couette, plane Poiseuille, and pipe flow. However, the organization of invariant solutions is not well understood. In this paper we focus on the role of symmetries in the organization and computation of invariant solutions of plane Poiseuille flow. We show that enforcing symmetries while computing invariant solutions increases the efficiency of the numerical methods, and that redundancies between search spaces can be eliminated by consideration of equivalence relations between symmetry subgroups. We determine all symmetry subgroups of plane Poiseuille flow in a doubly-periodic domain up to translations by half the periodic lengths and classify the subgroups into equivalence classes, each of which represents a physically distinct set of symmetries and an associated set of physically distinct invariant solutions. We calculate fifteen new traveling waves of plane Poiseuille flow in seven distinct symmetry groups and discuss their relevance to the dynamics of transitional turbulence. We present a few examples of subgroups with fractional shifts other than half the periodic lengths and one traveling wave solution whose symmetry involves shifts by one-third of the periodic lengths. We conclude with a discussion and some open questions about the role of symmetry in the behavior of shear flows.
{"title":"Symmetry groups and invariant solutions of plane Poiseuille flow","authors":"Pratik P. Aghor, John F. Gibson","doi":"arxiv-2409.11517","DOIUrl":"https://doi.org/arxiv-2409.11517","url":null,"abstract":"Equilibrium, traveling-wave, and periodic-orbit solutions of the\u0000Navier-Stokes equations provide a promising avenue for investigating the\u0000structure, dynamics, and statistics of transitional flows. Many such invariant\u0000solutions have been computed for wall-bounded shear flows, including plane\u0000Couette, plane Poiseuille, and pipe flow. However, the organization of\u0000invariant solutions is not well understood. In this paper we focus on the role\u0000of symmetries in the organization and computation of invariant solutions of\u0000plane Poiseuille flow. We show that enforcing symmetries while computing\u0000invariant solutions increases the efficiency of the numerical methods, and that\u0000redundancies between search spaces can be eliminated by consideration of\u0000equivalence relations between symmetry subgroups. We determine all symmetry\u0000subgroups of plane Poiseuille flow in a doubly-periodic domain up to\u0000translations by half the periodic lengths and classify the subgroups into\u0000equivalence classes, each of which represents a physically distinct set of\u0000symmetries and an associated set of physically distinct invariant solutions. We\u0000calculate fifteen new traveling waves of plane Poiseuille flow in seven\u0000distinct symmetry groups and discuss their relevance to the dynamics of\u0000transitional turbulence. We present a few examples of subgroups with fractional\u0000shifts other than half the periodic lengths and one traveling wave solution\u0000whose symmetry involves shifts by one-third of the periodic lengths. We\u0000conclude with a discussion and some open questions about the role of symmetry\u0000in the behavior of shear flows.","PeriodicalId":501125,"journal":{"name":"arXiv - PHYS - Fluid Dynamics","volume":"33 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142267527","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Liquid mobility on fibers is critical to the effectiveness of fiber matrices in face masks, water harvesting and aerosol filtration, but is typically affected by Plateau-Rayleigh instability. However, the spontaneous flow within precursor films arising from this instability has been largely overlooked, particularly regarding its fundamental flow pattern and the potential for liquid mobilization. This study reveals the pivotal role of spontaneous flow on ribbon-like fibers in enhancing liquid transport. The non-axisymmetric curvature of these fibers induces long-wave instabilities, generating a sustained flow that enables film-wise transport over centimeter-scale distances at velocities of several millimeters per second. Using particle-image velocimetry, we uncover intricate hydrodynamics, including opposing flows within the film and organized vortices in the shear layer, driven by capillary effects at the liquid-vapor interfaces. Building on these insights, we demonstrate a network structure capable of achieving planar liquid transport over a 10 cm2 area. The ribbon-like fibers investigated exhibit the longest transport distances relative to biomimetic structures and aerodynamic propulsion. The unique transport dynamics and planar configuration of the fiber matrix offer substantial potential for advanced fiber-based liquid transport systems, with enhanced mass/heat transfer, laminar mixing and aerodynamic characteristics.
{"title":"Long-distance Liquid Transport Along Fibers Arising From Plateau-Rayleigh Instability","authors":"Yunqiao Huang, Xianguo Li, Zhongchao Tan","doi":"arxiv-2409.11607","DOIUrl":"https://doi.org/arxiv-2409.11607","url":null,"abstract":"Liquid mobility on fibers is critical to the effectiveness of fiber matrices\u0000in face masks, water harvesting and aerosol filtration, but is typically\u0000affected by Plateau-Rayleigh instability. However, the spontaneous flow within\u0000precursor films arising from this instability has been largely overlooked,\u0000particularly regarding its fundamental flow pattern and the potential for\u0000liquid mobilization. This study reveals the pivotal role of spontaneous flow on\u0000ribbon-like fibers in enhancing liquid transport. The non-axisymmetric\u0000curvature of these fibers induces long-wave instabilities, generating a\u0000sustained flow that enables film-wise transport over centimeter-scale distances\u0000at velocities of several millimeters per second. Using particle-image\u0000velocimetry, we uncover intricate hydrodynamics, including opposing flows\u0000within the film and organized vortices in the shear layer, driven by capillary\u0000effects at the liquid-vapor interfaces. Building on these insights, we\u0000demonstrate a network structure capable of achieving planar liquid transport\u0000over a 10 cm2 area. The ribbon-like fibers investigated exhibit the longest\u0000transport distances relative to biomimetic structures and aerodynamic\u0000propulsion. The unique transport dynamics and planar configuration of the fiber\u0000matrix offer substantial potential for advanced fiber-based liquid transport\u0000systems, with enhanced mass/heat transfer, laminar mixing and aerodynamic\u0000characteristics.","PeriodicalId":501125,"journal":{"name":"arXiv - PHYS - Fluid Dynamics","volume":"119 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142267526","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pranay P. Nagrani, Amy M. Marconnet, Ivan C. Christov
Understanding transport phenomena through porous media is essential for applications ranging from water treatment systems to heat pipes. In many of these systems, packed-bed reactors (PBRs) are crucial components, and understanding and quantifying the pressure drop due to flow through the PBR is critical to effective operation. Recent experiments conducted by NASA measured the pressure drop due to gas-liquid flow through a PBR under microgravity conditions. Based on these experiments, we develop correlations for the interphase drag in a two-fluid model (TFM). Specifically, two closure relations are needed for the TFM: the liquid-solid $f_{ls}$ and gas-liquid $f_{gl}$ interphase force. We use an Ergun-type closure for $f_{ls}$. Then, under a 1D flow assumption, the TFM equations are rewritten with $f_{gl}$ as the only unknown. We employ data-driven calculations to determine $f_{gl}$, which we correlate (via composite fits) as a function of the liquid and gas Reynolds numbers, $Re_{l}$ and $Re_{g}$, respectively, and the Suratman number $Su_{l}$. To validate the proposed $f_{gl}(Re_{l},Re_{g},Su_{l})$ closure, we perform two-dimensional (2D) transient, multiphase computational fluid dynamics (CFD) simulations at low $Re_{l}$ and $Re_{g}$ (laminar flow) in ANSYS Fluent employing an Euler-Euler formulation. We find good agreement between the CFD simulations based on the proposed $f_{gl}$ closure and the experimental data.
{"title":"New correlations for the interphase drag in the two-fluid model of gas-liquid flows through packed-bed reactors","authors":"Pranay P. Nagrani, Amy M. Marconnet, Ivan C. Christov","doi":"arxiv-2409.10674","DOIUrl":"https://doi.org/arxiv-2409.10674","url":null,"abstract":"Understanding transport phenomena through porous media is essential for\u0000applications ranging from water treatment systems to heat pipes. In many of\u0000these systems, packed-bed reactors (PBRs) are crucial components, and\u0000understanding and quantifying the pressure drop due to flow through the PBR is\u0000critical to effective operation. Recent experiments conducted by NASA measured\u0000the pressure drop due to gas-liquid flow through a PBR under microgravity\u0000conditions. Based on these experiments, we develop correlations for the\u0000interphase drag in a two-fluid model (TFM). Specifically, two closure relations\u0000are needed for the TFM: the liquid-solid $f_{ls}$ and gas-liquid $f_{gl}$\u0000interphase force. We use an Ergun-type closure for $f_{ls}$. Then, under a 1D\u0000flow assumption, the TFM equations are rewritten with $f_{gl}$ as the only\u0000unknown. We employ data-driven calculations to determine $f_{gl}$, which we\u0000correlate (via composite fits) as a function of the liquid and gas Reynolds\u0000numbers, $Re_{l}$ and $Re_{g}$, respectively, and the Suratman number $Su_{l}$.\u0000To validate the proposed $f_{gl}(Re_{l},Re_{g},Su_{l})$ closure, we perform\u0000two-dimensional (2D) transient, multiphase computational fluid dynamics (CFD)\u0000simulations at low $Re_{l}$ and $Re_{g}$ (laminar flow) in ANSYS Fluent\u0000employing an Euler-Euler formulation. We find good agreement between the CFD\u0000simulations based on the proposed $f_{gl}$ closure and the experimental data.","PeriodicalId":501125,"journal":{"name":"arXiv - PHYS - Fluid Dynamics","volume":"5 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142267533","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Semi-analytical multi-front solutions of water table response due to periodic forcing in a partially saturated vertical porous column are developed, tested and compared to finite volume solutions of the Richards equation. The multi-front solutions are useful for capturing parametrically the frequency response of the vertical column to tidal oscillations while taking into account both capillary and gravitational effects. Vertical oscillations are examined, accounting for unsaturated flow above the oscillating water table as well as saturated flow below it. The multi-front models are conceived as successive generalizations of the Green-Ampt piston flow approach. The single front model is an "inverted" Green Ampt model, with an abrupt front separating the saturated and dry regions. It is adapted to the case of an oscillatory pressure imposed at the bottom of the column (rather than a fixed pressure imposed at the top). The N front models (N>=2) further generalize this concept, using a (Theta(h), K(h)) parametrization to take into account the capillary properties of the unsaturated medium. The resulting systems of ODE's are non linear with time variable coefficients. The solutions obtained for N = 10 fronts are satisfactory both in terms of water table fluctuations and moisture profiles, even for fine grained soils (Guelph Loam). They are computed much faster than space-time discretized solutions of the non linear Richards PDE. For sandy soils, even the 2-front solution (N=2) is satisfactory in terms of water table response Zs(t). The 2-front model itself is a significant improvement on the single front Green-Ampt model, and it appears potentially useful for analyzing the response of unsaturated flow systems under various types of oscillatory and transient forcing. Overall, the N-front method is useful for exploring the frequency response of the water table to tidal forcing.
{"title":"Generalized Green-Ampt Approach to 1D Oscillatory Flows in Partially Saturated/Unsaturated Media: Capillary Effects in Beach Hydrodynamics (Semi-analytical & Numerical studies)","authors":"Khalil Alastal, Rachid Ababou, Dominique Astruc","doi":"arxiv-2409.10718","DOIUrl":"https://doi.org/arxiv-2409.10718","url":null,"abstract":"Semi-analytical multi-front solutions of water table response due to periodic\u0000forcing in a partially saturated vertical porous column are developed, tested\u0000and compared to finite volume solutions of the Richards equation. The\u0000multi-front solutions are useful for capturing parametrically the frequency\u0000response of the vertical column to tidal oscillations while taking into account\u0000both capillary and gravitational effects. Vertical oscillations are examined,\u0000accounting for unsaturated flow above the oscillating water table as well as\u0000saturated flow below it. The multi-front models are conceived as successive\u0000generalizations of the Green-Ampt piston flow approach. The single front model\u0000is an \"inverted\" Green Ampt model, with an abrupt front separating the\u0000saturated and dry regions. It is adapted to the case of an oscillatory pressure\u0000imposed at the bottom of the column (rather than a fixed pressure imposed at\u0000the top). The N front models (N>=2) further generalize this concept, using a\u0000(Theta(h), K(h)) parametrization to take into account the capillary properties\u0000of the unsaturated medium. The resulting systems of ODE's are non linear with\u0000time variable coefficients. The solutions obtained for N = 10 fronts are\u0000satisfactory both in terms of water table fluctuations and moisture profiles,\u0000even for fine grained soils (Guelph Loam). They are computed much faster than\u0000space-time discretized solutions of the non linear Richards PDE. For sandy\u0000soils, even the 2-front solution (N=2) is satisfactory in terms of water table\u0000response Zs(t). The 2-front model itself is a significant improvement on the\u0000single front Green-Ampt model, and it appears potentially useful for analyzing\u0000the response of unsaturated flow systems under various types of oscillatory and\u0000transient forcing. Overall, the N-front method is useful for exploring the\u0000frequency response of the water table to tidal forcing.","PeriodicalId":501125,"journal":{"name":"arXiv - PHYS - Fluid Dynamics","volume":"52 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142267532","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Benjamin Wilfong, Anand Radhakrishnan, Henry A. Le Berre, Steve Abbott, Reuben D. Budiardja, Spencer H. Bryngelson
GPUs are the heart of the latest generations of supercomputers. We efficiently accelerate a compressible multiphase flow solver via OpenACC on NVIDIA and AMD Instinct GPUs. Optimization is accomplished by specifying the directive clauses 'gang vector' and 'collapse'. Further speedups of six and ten times are achieved by packing user-defined types into coalesced multidimensional arrays and manual inlining via metaprogramming. Additional optimizations yield seven-times speedup in array packing and thirty-times speedup of select kernels on Frontier. Weak scaling efficiencies of 97% and 95% are observed when scaling to 50% of Summit and 95% of Frontier. Strong scaling efficiencies of 84% and 81% are observed when increasing the device count by a factor of 8 and 16 on V100 and MI250X hardware. The strong scaling efficiency of AMD's MI250X increases to 92% when increasing the device count by a factor of 16 when GPU-aware MPI is used for communication.
{"title":"OpenACC offloading of the MFC compressible multiphase flow solver on AMD and NVIDIA GPUs","authors":"Benjamin Wilfong, Anand Radhakrishnan, Henry A. Le Berre, Steve Abbott, Reuben D. Budiardja, Spencer H. Bryngelson","doi":"arxiv-2409.10729","DOIUrl":"https://doi.org/arxiv-2409.10729","url":null,"abstract":"GPUs are the heart of the latest generations of supercomputers. We\u0000efficiently accelerate a compressible multiphase flow solver via OpenACC on\u0000NVIDIA and AMD Instinct GPUs. Optimization is accomplished by specifying the\u0000directive clauses 'gang vector' and 'collapse'. Further speedups of six and ten\u0000times are achieved by packing user-defined types into coalesced\u0000multidimensional arrays and manual inlining via metaprogramming. Additional\u0000optimizations yield seven-times speedup in array packing and thirty-times\u0000speedup of select kernels on Frontier. Weak scaling efficiencies of 97% and 95%\u0000are observed when scaling to 50% of Summit and 95% of Frontier. Strong scaling\u0000efficiencies of 84% and 81% are observed when increasing the device count by a\u0000factor of 8 and 16 on V100 and MI250X hardware. The strong scaling efficiency\u0000of AMD's MI250X increases to 92% when increasing the device count by a factor\u0000of 16 when GPU-aware MPI is used for communication.","PeriodicalId":501125,"journal":{"name":"arXiv - PHYS - Fluid Dynamics","volume":"5 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142267531","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}