arXiv - PHYS - Fluid Dynamics最新文献_第5页

Surface roughness effects in a transonic axial flow compressor operating at near-stall conditions 在近失速条件下运行的跨音速轴流压缩机中的表面粗糙度效应

arXiv - PHYS - Fluid Dynamics Pub Date : 2024-09-11 DOI: arxiv-2409.07344

Prashant B. Godse, Harshal D. Akolekar, A. M. Pradeep

{"title":"Surface roughness effects in a transonic axial flow compressor operating at near-stall conditions","authors":"Prashant B. Godse, Harshal D. Akolekar, A. M. Pradeep","doi":"arxiv-2409.07344","DOIUrl":"https://doi.org/arxiv-2409.07344","url":null,"abstract":"Surface roughness is a major contributor to performance degradation in gas\u0000turbine engines. The fan and the compressor, as the first components in the\u0000engine's air path, are especially vulnerable to the effects of surface\u0000roughness. Debris ingestion, accumulation of grime, dust, or insect remnants,\u0000typically at low atmospheric conditions, over several cycles of operation are\u0000some major causes of surface roughness over the blade surfaces. The flow in\u0000compressor rotors is inherently highly complex. From the perspective of the\u0000component designers, it is thus important to study the effect of surface\u0000roughness on the performance and flow physics, especially at near-stall\u0000conditions. In this study, we examine the effect of surface roughness on flow\u0000physics such as shock-boundary layer interactions, tip and hub flow\u0000separations, the formation and changes in the critical points, and tip leakage\u0000vortices amongst other phenomena. Steady and unsteady Reynolds Averaged Navier\u0000Stokes (RANS) calculations are conducted at near-stall conditions for smooth\u0000and rough NASA (National Aeronautics and Space Administration) rotor 67 blades.\u0000Surface streamlines, Q-criterion, and entropy contours aid in analyzing the\u0000flow physics qualitatively and quantitatively. It is observed that from the\u0000onset of stall, to fully stalled conditions, the blockage varies from 21.7% to\u000059.6% from 90% span to the tip in the smooth case, and from 40.5% to 75.2%\u0000in the rough case. This significant blockage, caused by vortex breakdown and\u0000chaotic flow structures, leads to the onset of full rotor stall.","PeriodicalId":501125,"journal":{"name":"arXiv - PHYS - Fluid Dynamics","volume":"12 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142212690","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}

引用次数: 0

Hammering at the entropy: A GENERIC-guided approach to learning polymeric rheological constitutive equations using PINNs 锤击熵：利用 PINNs 学习聚合物流变构造方程的 GENERIC 引导方法

arXiv - PHYS - Fluid Dynamics Pub Date : 2024-09-11 DOI: arxiv-2409.07545

David Nieto Simavilla, Andrea Bonfanti, Imanol García de Beristain, Pep Español, Marco Ellero

{"title":"Hammering at the entropy: A GENERIC-guided approach to learning polymeric rheological constitutive equations using PINNs","authors":"David Nieto Simavilla, Andrea Bonfanti, Imanol García de Beristain, Pep Español, Marco Ellero","doi":"arxiv-2409.07545","DOIUrl":"https://doi.org/arxiv-2409.07545","url":null,"abstract":"We present a versatile framework that employs Physics-Informed Neural\u0000Networks (PINNs) to discover the entropic contribution that leads to the\u0000constitutive equation for the extra-stress in rheological models of polymer\u0000solutions. In this framework the training of the Neural Network is guided by an\u0000evolution equation for the conformation tensor which is GENERIC-compliant. We\u0000compare two training methodologies for the data-driven PINN constitutive\u0000models: one trained on data from the analytical solution of the Oldroyd-B model\u0000under steady-state rheometric flows (PINN-rheometric), and another trained on\u0000in-silico data generated from complex flow CFD simulations around a cylinder\u0000that use the Oldroyd-B model (PINN-complex). The capacity of the PINN models to\u0000provide good predictions are evaluated by comparison with CFD simulations using\u0000the underlying Oldroyd-B model as a reference. Both models are capable of\u0000predicting flow behavior in transient and complex conditions; however, the\u0000PINN-complex model, trained on a broader range of mixed flow data, outperforms\u0000the PINN-rheometric model in complex flow scenarios. The geometry agnostic\u0000character of our methodology allows us to apply the learned PINN models to\u0000flows with different topologies than the ones used for training.","PeriodicalId":501125,"journal":{"name":"arXiv - PHYS - Fluid Dynamics","volume":"15 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142212688","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}

引用次数: 0

Some effects of limited wall-sensor availability on flow estimation with 3D-GANs 墙壁传感器可用性有限对 3D-GAN 流量估算的一些影响

arXiv - PHYS - Fluid Dynamics Pub Date : 2024-09-11 DOI: arxiv-2409.07348

Antonio Cuéllar, Andrea Ianiro, Stefano Discetti

{"title":"Some effects of limited wall-sensor availability on flow estimation with 3D-GANs","authors":"Antonio Cuéllar, Andrea Ianiro, Stefano Discetti","doi":"arxiv-2409.07348","DOIUrl":"https://doi.org/arxiv-2409.07348","url":null,"abstract":"In this work we assess the impact of the limited availability of\u0000wall-embedded sensors on the full 3D estimation of the flow field in a\u0000turbulent channel with Re{tau} = 200. The estimation technique is based on a\u00003D generative adversarial network (3D-GAN). We recently demonstrated that\u00003D-GANs are capable of estimating fields with good accuracy by employing\u0000fully-resolved wall quantities (pressure and streamwise/spanwise wall shear\u0000stress on a grid with DNS resolution). However, the practical implementation in\u0000an experimental setting is challenging due to the large number of sensors\u0000required. In this work, we aim to estimate the flow fields with substantially\u0000fewer sensors. The impact of the reduction of the number of sensors on the\u0000quality of the flow reconstruction is assessed in terms of accuracy degradation\u0000and spectral length-scales involved. It is found that the accuracy degradation\u0000is mainly due to the spatial undersampling of scales, rather than the reduction\u0000of the number of sensors per se. We explore the performance of the estimator in\u0000case only one wall quantity is available. When a large number of sensors is\u0000available, pressure measurements provide more accurate flow field estimations.\u0000Conversely, the elongated patterns of the streamwise wall shear stress make\u0000this quantity the most suitable when only few sensors are available. As a\u0000further step towards a real application, the effect of sensor noise is also\u0000quantified. It is shown that configurations with fewer sensors are less\u0000sensitive to measurement noise.","PeriodicalId":501125,"journal":{"name":"arXiv - PHYS - Fluid Dynamics","volume":"13 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142212689","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}

引用次数: 0

Coupling Machine Learning Local Predictions with a Computational Fluid Dynamics Solver to Accelerate Transient Buoyant Plume Simulations 将机器学习本地预测与计算流体力学求解器相结合，加速瞬态浮力烟羽模拟

arXiv - PHYS - Fluid Dynamics Pub Date : 2024-09-11 DOI: arxiv-2409.07175

Clément Caron, Philippe Lauret, Alain Bastide

{"title":"Coupling Machine Learning Local Predictions with a Computational Fluid Dynamics Solver to Accelerate Transient Buoyant Plume Simulations","authors":"Clément Caron, Philippe Lauret, Alain Bastide","doi":"arxiv-2409.07175","DOIUrl":"https://doi.org/arxiv-2409.07175","url":null,"abstract":"Data-driven methods demonstrate considerable potential for accelerating the\u0000inherently expensive computational fluid dynamics (CFD) solvers. Nevertheless,\u0000pure machine-learning surrogate models face challenges in ensuring physical\u0000consistency and scaling up to address real-world problems. This study presents\u0000a versatile and scalable hybrid methodology, combining CFD and machine\u0000learning, to accelerate long-term incompressible fluid flow simulations without\u0000compromising accuracy. A neural network was trained offline using simulated\u0000data of various two-dimensional transient buoyant plume flows. The objective\u0000was to leverage local features to predict the temporal changes in the pressure\u0000field in comparable scenarios. Due to cell-level predictions, the methodology\u0000was successfully applied to diverse geometries without additional training.\u0000Pressure estimates were employed as initial values to accelerate the\u0000pressure-velocity coupling procedure. The results demonstrated an average\u0000improvement of 94% in the initial guess for solving the Poisson equation. The\u0000first pressure corrector acceleration reached a mean factor of 3, depending on\u0000the iterative solver employed. Our work reveals that machine learning estimates\u0000at the cell level can enhance the efficiency of CFD iterative linear solvers\u0000while maintaining accuracy. Although the scalability of the methodology to more\u0000complex cases has yet to be demonstrated, this study underscores the\u0000prospective value of domain-specific hybrid solvers for CFD.","PeriodicalId":501125,"journal":{"name":"arXiv - PHYS - Fluid Dynamics","volume":"33 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142212697","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}

引用次数: 0

Three-dimensional generative adversarial networks for turbulent flow estimation from wall measurements 根据壁面测量结果估算湍流的三维生成对抗网络

arXiv - PHYS - Fluid Dynamics Pub Date : 2024-09-10 DOI: arxiv-2409.06548

Antonio Cuéllar, Alejandro Güemes, Andrea Ianiro, Óscar Flores, Ricardo Vinuesa, Stefano Discetti

{"title":"Three-dimensional generative adversarial networks for turbulent flow estimation from wall measurements","authors":"Antonio Cuéllar, Alejandro Güemes, Andrea Ianiro, Óscar Flores, Ricardo Vinuesa, Stefano Discetti","doi":"arxiv-2409.06548","DOIUrl":"https://doi.org/arxiv-2409.06548","url":null,"abstract":"Different types of neural networks have been used to solve the flow sensing\u0000problem in turbulent flows, namely to estimate velocity in wall-parallel planes\u0000from wall measurements. Generative adversarial networks (GANs) are among the\u0000most promising methodologies, due to their more accurate estimations and better\u0000perceptual quality. This work tackles this flow sensing problem in the vicinity\u0000of the wall, addressing for the first time the reconstruction of the entire\u0000three-dimensional (3-D) field with a single network, i.e. a 3-D GAN. With this\u0000methodology, a single training and prediction process overcomes the limitation\u0000presented by the former approaches based on the independent estimation of\u0000wall-parallel planes. The network is capable of estimating the 3-D flow field\u0000with a level of error at each wall-normal distance comparable to that reported\u0000from wall-parallel plane estimations and at a lower training cost in terms of\u0000computational resources. The direct full 3-D reconstruction also unveils a\u0000direct interpretation in terms of coherent structures. It is shown that the\u0000accuracy of the network depends directly on the wall footprint of each\u0000individual turbulent structure. It is observed that wall-attached structures\u0000are predicted more accurately than wall-detached ones, especially at larger\u0000distances from the wall. Among wall-attached structures, smaller sweeps are\u0000reconstructed better than small ejections, while large ejections are\u0000reconstructed better than large sweeps as a consequence of their more intense\u0000footprint.","PeriodicalId":501125,"journal":{"name":"arXiv - PHYS - Fluid Dynamics","volume":"1 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142212703","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}

引用次数: 0

nekCRF: A next generation high-order reactive low Mach flow solver for direct numerical simulations nekCRF：用于直接数值模拟的下一代高阶反应性低马赫数流动求解器

arXiv - PHYS - Fluid Dynamics Pub Date : 2024-09-10 DOI: arxiv-2409.06404

Stefan Kerkemeier, Christos E. Frouzakis, Ananias G. Tomboulides, Paul Fischer, Mathis Bode

引用次数: 0

Mixing in viscoelastic fluids using elastic turbulence 利用弹性湍流在粘弹性流体中进行混合

arXiv - PHYS - Fluid Dynamics Pub Date : 2024-09-10 DOI: arxiv-2409.06391

Reinier van Buel, Holger Stark

引用次数: 0

Synchronization of wave-propelled capillary spinners 波浪推进毛细管旋流器的同步运行

arXiv - PHYS - Fluid Dynamics Pub Date : 2024-09-10 DOI: arxiv-2409.06652

Jack-William Barotta, Giuseppe Pucci, Eli Silver, Alireza Hooshanginejad, Daniel M. Harris

{"title":"Synchronization of wave-propelled capillary spinners","authors":"Jack-William Barotta, Giuseppe Pucci, Eli Silver, Alireza Hooshanginejad, Daniel M. Harris","doi":"arxiv-2409.06652","DOIUrl":"https://doi.org/arxiv-2409.06652","url":null,"abstract":"When a millimetric body is placed atop a vibrating liquid bath, the relative\u0000motion between the object and interface generates outward propagating waves\u0000with an associated momentum flux. Prior work has shown that isolated chiral\u0000objects, referred to as spinners, can thus rotate steadily in response to their\u0000self-generated wavefield. Here, we consider the case of two co-chiral spinners\u0000held at a fixed spacing from one another but otherwise free to interact\u0000hydrodynamically through their shared fluid substrate. Two identical spinners\u0000are able to synchronize their rotation, with their equilibrium phase difference\u0000sensitive to their spacing and initial conditions, and even cease to rotate\u0000when the coupling becomes sufficiently strong. Non-identical spinners can also\u0000find synchrony provided their intrinsic differences are not too disparate. A\u0000hydrodynamic wave model of the spinner interaction is proposed, recovering all\u0000salient features of the experiment. In all cases, the spatially periodic nature\u0000of the capillary wave coupling is directly reflected in the emergent\u0000equilibrium behaviors.","PeriodicalId":501125,"journal":{"name":"arXiv - PHYS - Fluid Dynamics","volume":"21 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142212701","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}

引用次数: 0

Particle-Laden Fluid on Flow Maps 流动图上的含颗粒流体

arXiv - PHYS - Fluid Dynamics Pub Date : 2024-09-10 DOI: arxiv-2409.06246

Zhiqi Li, Duowen Chen, Candong Lin, Jinyuan Liu, Bo Zhu

引用次数: 0

Mapping Driven Oscillations in the Size of a Bubble to the Dynamics of a Newtonian Particle in a Potential 将气泡大小的驱动振荡映射到势能中的牛顿粒子动力学

arXiv - PHYS - Fluid Dynamics Pub Date : 2024-09-09 DOI: arxiv-2409.05961

Uri Shimon, Ady Stern

引用次数: 0