Pub Date : 2024-10-10DOI: 10.1016/j.compfluid.2024.106451
Chunyuan Xu, Zhijun Shen, Qinghong Zeng
In this work, a hybrid nodal solver is constructed to incorporate the directional effect of wave propagation in a cell-centered Lagrangian scheme. First, the direction of wave is determined via an assumption based on the Rankine–Hugoniot condition. Next, two different approximation methods are used to calculate the velocity jump and numerical fluxes. Correspondingly, a hybridization strategy is proposed to formulate a hybrid nodal solver. It is shown that the developed nodal solver can replicate two well-known ones, and its effectiveness is shown in numerical experiments. Finally, an adaptive hybridization method based on vorticity is proposed. The accuracy and robustness of the adaptive method is assessed in various tests.
{"title":"Formulation of hybrid nodal solver based on directional effect of wave propagation in a cell-centered Lagrangian scheme","authors":"Chunyuan Xu, Zhijun Shen, Qinghong Zeng","doi":"10.1016/j.compfluid.2024.106451","DOIUrl":"10.1016/j.compfluid.2024.106451","url":null,"abstract":"<div><div>In this work, a hybrid nodal solver is constructed to incorporate the directional effect of wave propagation in a cell-centered Lagrangian scheme. First, the direction of wave is determined via an assumption based on the Rankine–Hugoniot condition. Next, two different approximation methods are used to calculate the velocity jump and numerical fluxes. Correspondingly, a hybridization strategy is proposed to formulate a hybrid nodal solver. It is shown that the developed nodal solver can replicate two well-known ones, and its effectiveness is shown in numerical experiments. Finally, an adaptive hybridization method based on vorticity is proposed. The accuracy and robustness of the adaptive method is assessed in various tests.</div></div>","PeriodicalId":287,"journal":{"name":"Computers & Fluids","volume":"285 ","pages":"Article 106451"},"PeriodicalIF":2.5,"publicationDate":"2024-10-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142528815","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-10-05DOI: 10.1016/j.compfluid.2024.106447
Andrei I. Tolstykh, Dmitrii A. Shirobokov
Numerical solutions of the Navier–Stokes equations describing the onset and the development of the boundary layer instability of a subsonic flow about a finite width flat plate are described. They were obtained via exciters-free calculations with the 16th-order multioperators-based scheme optimized to resolve very small solutions scales. Some relevant details of the scheme are briefly outlined. The occurrence of the Tollmien–Schlichting waves near the leading edge, their downstream propagation and breakdowns some distance away are displayed. The role of the oblique waves seen in the solutions is emphasized. The nonlinear stage in the form of structured vortices followed by the solutions randomization are described.
{"title":"Breakdown of laminar regime in flat plate boundary layer seen in numerical solutions of the Navier–Stokes equations obtained with 16th-order scheme","authors":"Andrei I. Tolstykh, Dmitrii A. Shirobokov","doi":"10.1016/j.compfluid.2024.106447","DOIUrl":"10.1016/j.compfluid.2024.106447","url":null,"abstract":"<div><div>Numerical solutions of the Navier–Stokes equations describing the onset and the development of the boundary layer instability of a subsonic flow about a finite width flat plate are described. They were obtained via exciters-free calculations with the 16th-order multioperators-based scheme optimized to resolve very small solutions scales. Some relevant details of the scheme are briefly outlined. The occurrence of the Tollmien–Schlichting waves near the leading edge, their downstream propagation and breakdowns some distance away are displayed. The role of the oblique waves seen in the solutions is emphasized. The nonlinear stage in the form of structured vortices followed by the solutions randomization are described.</div></div>","PeriodicalId":287,"journal":{"name":"Computers & Fluids","volume":"285 ","pages":"Article 106447"},"PeriodicalIF":2.5,"publicationDate":"2024-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142433288","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-28DOI: 10.1016/j.compfluid.2024.106446
Swapnil Tupkari, Hrishikesh Gadgil, Vineeth Nair
Three-dimensional large-eddy simulations (LES) of supercritical nitrogen injection are carried out over a range of swirl numbers () to investigate the characteristics of the flow field. The swirl number is varied by varying the tangential velocity, keeping mass flow rate unchanged. Based on the observed flow dynamics, the swirl numbers are classified as (i) low, (ii) medium, or (iii) high. For low swirl numbers, recirculation zones are not observed and the flow pattern consists primarily of free shear layer instabilities arising out of forced convection that move helically around an intact jet potential core. For medium swirl numbers, a sudden drop in axial centerline velocity is observed due to large adverse pressure gradients. These gradients lead to recirculation regions in front of the injector forming a bubble-type aerodynamic bluff body outside the injector. As the swirl number is increased in this range, the bubble disappears forming a venturi-type flow with local acceleration of axial centerline velocity. For high swirl numbers, the recirculation region enters the injector, and a precessing vortex core (PVC) is observed in the flow field. Mean jet length and initial jet cone angle were found to have non-monotonic variations for medium and high swirl numbers. Finally, spectral analysis reveals the presence of a hydrodynamic frequency corresponding to shear layer instabilities and an acoustic mode corresponding to the injector for low and medium swirl numbers. For high swirl numbers, both these frequencies are suppressed noticeably and the dynamics is completely described by the PVC frequency.
{"title":"A numerical study of dynamic flow patterns in supercritical jet flows for various swirl numbers","authors":"Swapnil Tupkari, Hrishikesh Gadgil, Vineeth Nair","doi":"10.1016/j.compfluid.2024.106446","DOIUrl":"10.1016/j.compfluid.2024.106446","url":null,"abstract":"<div><div>Three-dimensional large-eddy simulations (LES) of supercritical nitrogen injection are carried out over a range of swirl numbers (<span><math><mrow><mn>0</mn><mo>≤</mo><mi>S</mi><mo><</mo><mn>2</mn></mrow></math></span>) to investigate the characteristics of the flow field. The swirl number is varied by varying the tangential velocity, keeping mass flow rate unchanged. Based on the observed flow dynamics, the swirl numbers are classified as (i) low, (ii) medium, or (iii) high. For low swirl numbers, recirculation zones are not observed and the flow pattern consists primarily of free shear layer instabilities arising out of forced convection that move helically around an intact jet potential core. For medium swirl numbers, a sudden drop in axial centerline velocity is observed due to large adverse pressure gradients. These gradients lead to recirculation regions in front of the injector forming a bubble-type aerodynamic bluff body outside the injector. As the swirl number is increased in this range, the bubble disappears forming a venturi-type flow with local acceleration of axial centerline velocity. For high swirl numbers, the recirculation region enters the injector, and a precessing vortex core (PVC) is observed in the flow field. Mean jet length and initial jet cone angle were found to have non-monotonic variations for medium and high swirl numbers. Finally, spectral analysis reveals the presence of a hydrodynamic frequency corresponding to shear layer instabilities and an acoustic mode corresponding to the injector for low and medium swirl numbers. For high swirl numbers, both these frequencies are suppressed noticeably and the dynamics is completely described by the PVC frequency.</div></div>","PeriodicalId":287,"journal":{"name":"Computers & Fluids","volume":"285 ","pages":"Article 106446"},"PeriodicalIF":2.5,"publicationDate":"2024-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142417419","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-27DOI: 10.1016/j.compfluid.2024.106445
Chengzhi Zhang, Supei Zheng, Jianhu Feng, Shasha Liu
An entropy stable scheme based on adaptive unstructured meshes for solving ideal magnetohydrodynamic (MHD) equations is proposed. Firstly, a semi-discrete finite volume scheme is constructed on unstructured meshes, which includes entropy conservative flux and Roe-type dissipation operator. Particularly, a special discrete Godunov source term is added to control magnetic field divergence, and it is proved that the new scheme is entropy stable. Secondly, the accuracy of the basic entropy stable scheme is enhanced through reconstruction of the entropy dissipation operator using the minmod slope limiter. Finally, based on the adaptive moving meshes, a new monitor function is designed for the properties of the ideal MHD equation solution, which can effectively identify the large gradient areas of the solution and optimize the mesh distribution. Several numerical examples illustrate that the novel scheme exhibits high accuracy and proficiently captures shock waves.
{"title":"Entropy stable scheme for ideal MHD equations on adaptive unstructured meshes","authors":"Chengzhi Zhang, Supei Zheng, Jianhu Feng, Shasha Liu","doi":"10.1016/j.compfluid.2024.106445","DOIUrl":"10.1016/j.compfluid.2024.106445","url":null,"abstract":"<div><div>An entropy stable scheme based on adaptive unstructured meshes for solving ideal magnetohydrodynamic (MHD) equations is proposed. Firstly, a semi-discrete finite volume scheme is constructed on unstructured meshes, which includes entropy conservative flux and Roe-type dissipation operator. Particularly, a special discrete Godunov source term is added to control magnetic field divergence, and it is proved that the new scheme is entropy stable. Secondly, the accuracy of the basic entropy stable scheme is enhanced through reconstruction of the entropy dissipation operator using the minmod slope limiter. Finally, based on the adaptive moving meshes, a new monitor function is designed for the properties of the ideal MHD equation solution, which can effectively identify the large gradient areas of the solution and optimize the mesh distribution. Several numerical examples illustrate that the novel scheme exhibits high accuracy and proficiently captures shock waves.</div></div>","PeriodicalId":287,"journal":{"name":"Computers & Fluids","volume":"285 ","pages":"Article 106445"},"PeriodicalIF":2.5,"publicationDate":"2024-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142417418","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-27DOI: 10.1016/j.compfluid.2024.106441
Xin Li 李鑫 , Zhiwen Deng 邓志文 , Rui Feng 冯瑞 , Ziyang Liu 刘子扬 , Renkun Han 韩仁坤 , Hongsheng Liu 刘红升 , Gang Chen 陈刚
Artificial intelligence-based three-dimensional fluid modeling has gained significant attention recently. However, the accuracy of such models is often limited by the preprocessing of irregular flow data. In order to bolster the credibility of near-wall flow prediction, we present a deep learning-based reduced order model for three-dimensional unsteady flow using the transformation and stitching techniques for multi-block structured meshes. To begin with, full-order flow data is provided by numerical simulations that rely on multi-block structured meshes. The mesh transformation technique is applied to convert each structured grid with data into a corresponding uniform and orthogonal grid, which is subsequently stitched and filled. The resulting snapshots in the transformed domain contain accurate flow information for multiple meshes and can be directly fed into a structured neural network without requiring any interpolation operation. Subsequently, a network model based on a fully convolutional neural network is constructed to predict flow dynamics accurately. To validate the strategy's feasibility, the flow around a sphere with Re=300 was investigated, and the results obtained using traditional Cartesian interpolation were used as the baseline for comparison. All the results demonstrate the preservation and accurate prediction of flow details near the wall, with the pressure correlation coefficient on the wall achieving a remarkable value of 0.9997. The stitching scheme that follows the proposed standard can effectively reduce the accumulation of inferring errors. Moreover, the periodic behavior of flow fields can be faithfully predicted during long-term inference. This work demonstrates that the proposed strategy has the advantage of high efficiency while providing accuracy assurance for downstream tasks.
{"title":"Deep learning-based reduced order model for three-dimensional unsteady flow using mesh transformation and stitching","authors":"Xin Li 李鑫 , Zhiwen Deng 邓志文 , Rui Feng 冯瑞 , Ziyang Liu 刘子扬 , Renkun Han 韩仁坤 , Hongsheng Liu 刘红升 , Gang Chen 陈刚","doi":"10.1016/j.compfluid.2024.106441","DOIUrl":"10.1016/j.compfluid.2024.106441","url":null,"abstract":"<div><div>Artificial intelligence-based three-dimensional fluid modeling has gained significant attention recently. However, the accuracy of such models is often limited by the preprocessing of irregular flow data. In order to bolster the credibility of near-wall flow prediction, we present a deep learning-based reduced order model for three-dimensional unsteady flow using the transformation and stitching techniques for multi-block structured meshes. To begin with, full-order flow data is provided by numerical simulations that rely on multi-block structured meshes. The mesh transformation technique is applied to convert each structured grid with data into a corresponding uniform and orthogonal grid, which is subsequently stitched and filled. The resulting snapshots in the transformed domain contain accurate flow information for multiple meshes and can be directly fed into a structured neural network without requiring any interpolation operation. Subsequently, a network model based on a fully convolutional neural network is constructed to predict flow dynamics accurately. To validate the strategy's feasibility, the flow around a sphere with <em>Re</em>=300 was investigated, and the results obtained using traditional Cartesian interpolation were used as the baseline for comparison. All the results demonstrate the preservation and accurate prediction of flow details near the wall, with the pressure correlation coefficient on the wall achieving a remarkable value of 0.9997. The stitching scheme that follows the proposed standard can effectively reduce the accumulation of inferring errors. Moreover, the periodic behavior of flow fields can be faithfully predicted during long-term inference. This work demonstrates that the proposed strategy has the advantage of high efficiency while providing accuracy assurance for downstream tasks.</div></div>","PeriodicalId":287,"journal":{"name":"Computers & Fluids","volume":"285 ","pages":"Article 106441"},"PeriodicalIF":2.5,"publicationDate":"2024-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142417420","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-26DOI: 10.1016/j.compfluid.2024.106437
Joris Labarbe, Alexandre Vieira, Didier Clamond
We consider the problem of recovering the surface wave profile from noisy bottom pressure measurements with (a priori unknown) arbitrary pressure at the surface. Without noise, the direct approach developed in Clamond and Labarbe (2023) provides an effective way to recover the sea surface. However, the assumption of analyticity for the measured pressure renders this method inefficient in the presence of noise. In order to address this issue, we introduce here an optimisation procedure based on the minimisation of a distance between a recovered bottom pressure and its measurement. Such method proves to be well-designed to handle perturbed signals. We illustrate the effectiveness of this approach in the recovery of gravity-capillary waves from unfiltered noisy data.
{"title":"Optimal reconstruction of water-waves from noisy pressure measurements at the seabed","authors":"Joris Labarbe, Alexandre Vieira, Didier Clamond","doi":"10.1016/j.compfluid.2024.106437","DOIUrl":"10.1016/j.compfluid.2024.106437","url":null,"abstract":"<div><div>We consider the problem of recovering the surface wave profile from noisy bottom pressure measurements with (<em>a priori</em> unknown) arbitrary pressure at the surface. Without noise, the direct approach developed in Clamond and Labarbe (2023) provides an effective way to recover the sea surface. However, the assumption of analyticity for the measured pressure renders this method inefficient in the presence of noise. In order to address this issue, we introduce here an optimisation procedure based on the minimisation of a distance between a recovered bottom pressure and its measurement. Such method proves to be well-designed to handle perturbed signals. We illustrate the effectiveness of this approach in the recovery of gravity-capillary waves from unfiltered noisy data.</div></div>","PeriodicalId":287,"journal":{"name":"Computers & Fluids","volume":"284 ","pages":"Article 106437"},"PeriodicalIF":2.5,"publicationDate":"2024-09-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142326739","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The study presents a comprehensive numerical investigation of the Kelvin–Helmholtz Rayleigh–Taylor Instability (KHRTI) onset using highly resolved peta-scale direct numerical simulations by solving the compressible Navier–Stokes equations (NSE). The numerical framework incorporates a three-dimensional (3D) cuboidal domain with differential heating applied to two air streams, fostering the development of the KHRTI. A novel numerical methodology with selective mesh refinement near critical regions is employed with the help of a non-uniform compact scheme to capture small-scale phenomena accurately. Analysis of pressure disturbances during early KHRTI stages reveal distinct wave propagation patterns influenced by Rayleigh–Taylor (RT) and Kelvin–Helmholtz (KH) mechanisms. Enstrophy dynamics are quantified through the compressible enstrophy transport equation (CETE), highlighting dominant contributions from viscous stresses during early receptivity stages. The study provides insights into KHRTI evolution, shedding light on shear-buoyancy-driven instabilities and their implications for transition to turbulence.
{"title":"Highly resolved peta-scale direct numerical simulations: Onset of Kelvin–Helmholtz Rayleigh–Taylor instability via pressure pulses","authors":"Bhavna Joshi , Tapan K. Sengupta , Prasannabalaji Sundaram , Aditi Sengupta","doi":"10.1016/j.compfluid.2024.106442","DOIUrl":"10.1016/j.compfluid.2024.106442","url":null,"abstract":"<div><div>The study presents a comprehensive numerical investigation of the Kelvin–Helmholtz Rayleigh–Taylor Instability (KHRTI) onset using highly resolved peta-scale direct numerical simulations by solving the compressible Navier–Stokes equations (NSE). The numerical framework incorporates a three-dimensional (3D) cuboidal domain with differential heating applied to two air streams, fostering the development of the KHRTI. A novel numerical methodology with selective mesh refinement near critical regions is employed with the help of a non-uniform compact scheme to capture small-scale phenomena accurately. Analysis of pressure disturbances during early KHRTI stages reveal distinct wave propagation patterns influenced by Rayleigh–Taylor (RT) and Kelvin–Helmholtz (KH) mechanisms. Enstrophy dynamics are quantified through the compressible enstrophy transport equation (CETE), highlighting dominant contributions from viscous stresses during early receptivity stages. The study provides insights into KHRTI evolution, shedding light on shear-buoyancy-driven instabilities and their implications for transition to turbulence.</div></div>","PeriodicalId":287,"journal":{"name":"Computers & Fluids","volume":"284 ","pages":"Article 106442"},"PeriodicalIF":2.5,"publicationDate":"2024-09-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142356814","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-24DOI: 10.1016/j.compfluid.2024.106444
Fulin Tong , Xiangxin Ji , Siwei Dong , Xianxu Yuan , Xinliang Li
The direct numerical simulation of an impinging oblique shock wave interacting with a spatially evolving turbulent boundary layer on a flat plat at Mach number 2.25 is used to investigate the characteristics of turbulent heat flux and temperature variance. Downstream of the interaction, the turbulent heat flux and temperature variance attain very large values in the outer region. The observed amplification of the turbulent heat flux is independent of the pressure–velocity correlation and is mainly characterized by mass flux. The coupling between temperature variance and turbulent kinetic energy is analyzed by examining the turbulent time-scale ratio. Across the interaction, the nearly constant time-scale ratio found in most parts of the boundary layer is generally smaller than the commonly accepted value of 0.5. The near-wall asymptotic behavior of the temperature variance is verified. Bidimensional empirical mode decomposition is adopted to analyze the contributions of different scale structures to the turbulent heat flux and temperature variance. This scale-decomposed analysis reveals that, compared with the upstream boundary layer, the shock interaction leads to increasingly pronounced contributions of the greatly enhanced outer large-scale structures and decreased contributions associated with the near-wall small-scale structures. In addition, an analysis of the primary budget terms in the transport of turbulent heat flux and temperature variance was performed. Unlike the upstream budget, the balance of production, destruction, and redistribution changes significantly in the downstream region.
{"title":"On the turbulent heat flux and temperature variance in supersonic shock-wave boundary-layer interaction","authors":"Fulin Tong , Xiangxin Ji , Siwei Dong , Xianxu Yuan , Xinliang Li","doi":"10.1016/j.compfluid.2024.106444","DOIUrl":"10.1016/j.compfluid.2024.106444","url":null,"abstract":"<div><div>The direct numerical simulation of an impinging oblique shock wave interacting with a spatially evolving turbulent boundary layer on a flat plat at Mach number 2.25 is used to investigate the characteristics of turbulent heat flux and temperature variance. Downstream of the interaction, the turbulent heat flux and temperature variance attain very large values in the outer region. The observed amplification of the turbulent heat flux is independent of the pressure–velocity correlation and is mainly characterized by mass flux. The coupling between temperature variance and turbulent kinetic energy is analyzed by examining the turbulent time-scale ratio. Across the interaction, the nearly constant time-scale ratio found in most parts of the boundary layer is generally smaller than the commonly accepted value of 0.5. The near-wall asymptotic behavior of the temperature variance is verified. Bidimensional empirical mode decomposition is adopted to analyze the contributions of different scale structures to the turbulent heat flux and temperature variance. This scale-decomposed analysis reveals that, compared with the upstream boundary layer, the shock interaction leads to increasingly pronounced contributions of the greatly enhanced outer large-scale structures and decreased contributions associated with the near-wall small-scale structures. In addition, an analysis of the primary budget terms in the transport of turbulent heat flux and temperature variance was performed. Unlike the upstream budget, the balance of production, destruction, and redistribution changes significantly in the downstream region.</div></div>","PeriodicalId":287,"journal":{"name":"Computers & Fluids","volume":"284 ","pages":"Article 106444"},"PeriodicalIF":2.5,"publicationDate":"2024-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142423691","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-24DOI: 10.1016/j.compfluid.2024.106443
Joel Ho , Nick Pepper , Tim Dodwell
A probabilistic machine learning model is introduced to augment the turbulence model in order to improve the modelling of separated flows and the generalisability of learnt corrections. Increasingly, machine learning methods have been used to leverage experimental and high-fidelity simulation data, improving the accuracy of the Reynolds Averaged Navier–Stokes (RANS) turbulence models widely used in industry. A significant challenge for such methods is their ability to generalise to unseen geometries and flow conditions. Furthermore, heterogeneous datasets containing a mix of experimental and simulation data must be efficiently handled. In this work, field inversion and an ensemble of Gaussian Process Emulators (GPEs) is employed to address both of these challenges. The ensemble model is applied to a range of benchmark test cases, demonstrating improved turbulence modelling for cases involving separated flows with adverse pressure gradients, where RANS simulations are understood to be unreliable. Perhaps more significantly, the simulation reverted to the uncorrected model in regions of the flow exhibiting physics outside of the training data.
{"title":"Probabilistic machine learning to improve generalisation of data-driven turbulence modelling","authors":"Joel Ho , Nick Pepper , Tim Dodwell","doi":"10.1016/j.compfluid.2024.106443","DOIUrl":"10.1016/j.compfluid.2024.106443","url":null,"abstract":"<div><div>A probabilistic machine learning model is introduced to augment the <span><math><mrow><mi>k</mi><mo>−</mo><mi>ω</mi><mspace></mspace><mi>S</mi><mi>S</mi><mi>T</mi></mrow></math></span> turbulence model in order to improve the modelling of separated flows and the generalisability of learnt corrections. Increasingly, machine learning methods have been used to leverage experimental and high-fidelity simulation data, improving the accuracy of the Reynolds Averaged Navier–Stokes (RANS) turbulence models widely used in industry. A significant challenge for such methods is their ability to generalise to unseen geometries and flow conditions. Furthermore, heterogeneous datasets containing a mix of experimental and simulation data must be efficiently handled. In this work, field inversion and an ensemble of Gaussian Process Emulators (GPEs) is employed to address both of these challenges. The ensemble model is applied to a range of benchmark test cases, demonstrating improved turbulence modelling for cases involving separated flows with adverse pressure gradients, where RANS simulations are understood to be unreliable. Perhaps more significantly, the simulation reverted to the uncorrected model in regions of the flow exhibiting physics outside of the training data.</div></div>","PeriodicalId":287,"journal":{"name":"Computers & Fluids","volume":"284 ","pages":"Article 106443"},"PeriodicalIF":2.5,"publicationDate":"2024-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142326740","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-21DOI: 10.1016/j.compfluid.2024.106436
Junjie Gao , Daiying Deng , Xiaoguang Luo , Haitao Han , Jijun Yu
In this paper, the moving particle semi-implicit method (MPS) is extended from calculating free mobility to simulating the extremely viscous and temperature-dependent molten layer flow of SiO2f/SiO2 composite material under aerodynamic heating conditions, which includes strong heating and shear of incoming flow. A method for applying heat flux and airflow shear, based on the conceptual particle approach, has been established. Heat transfer, melting, solidification, and evaporation behaviors are considered, with temperature-dependent viscosity variations also accounted for. The ablative regression of the plate model is verified using experimental results of the SiO2f/SiO2 composite material, and results from convergence analysis demonstrate the accuracy of the space step size selection. Surface morphology analysis through three-dimensional computation indicates that the extended particle method also accurately describes the surface morphology of SiO2f/SiO2 composite material under aerodynamic heating conditions. Thus, the extended particle method accurately simulates both the ablation process and the surface morphology of the SiO2f/SiO2 composite material. The influences of acceleration and surface tension are discussed. Ablative recession, when subject to acceleration, is smaller than that observed in its absence. When exposed to surface tension, the liquid layer tends to form a spherical shape, and the particles behave as a cohesive unit, resulting in smaller ablative recession than in the absence of surface tension.
{"title":"Ablation and molten layer flow simulation for plate model of SiO2f/SiO2 composite material using particle method","authors":"Junjie Gao , Daiying Deng , Xiaoguang Luo , Haitao Han , Jijun Yu","doi":"10.1016/j.compfluid.2024.106436","DOIUrl":"10.1016/j.compfluid.2024.106436","url":null,"abstract":"<div><div>In this paper, the moving particle semi-implicit method (MPS) is extended from calculating free mobility to simulating the extremely viscous and temperature-dependent molten layer flow of SiO<sub>2f</sub>/SiO<sub>2</sub> composite material under aerodynamic heating conditions, which includes strong heating and shear of incoming flow. A method for applying heat flux and airflow shear, based on the conceptual particle approach, has been established. Heat transfer, melting, solidification, and evaporation behaviors are considered, with temperature-dependent viscosity variations also accounted for. The ablative regression of the plate model is verified using experimental results of the SiO<sub>2f</sub>/SiO<sub>2</sub> composite material, and results from convergence analysis demonstrate the accuracy of the space step size selection. Surface morphology analysis through three-dimensional computation indicates that the extended particle method also accurately describes the surface morphology of SiO<sub>2f</sub>/SiO<sub>2</sub> composite material under aerodynamic heating conditions. Thus, the extended particle method accurately simulates both the ablation process and the surface morphology of the SiO<sub>2f</sub>/SiO<sub>2</sub> composite material. The influences of acceleration and surface tension are discussed. Ablative recession, when subject to acceleration, is smaller than that observed in its absence. When exposed to surface tension, the liquid layer tends to form a spherical shape, and the particles behave as a cohesive unit, resulting in smaller ablative recession than in the absence of surface tension.</div></div>","PeriodicalId":287,"journal":{"name":"Computers & Fluids","volume":"284 ","pages":"Article 106436"},"PeriodicalIF":2.5,"publicationDate":"2024-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142314636","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号