Armenio / Geurts / Fröhlich Direct and Large-Eddy Simulation VII

Proceedings of the Seventh International ERCOFTAC Workshop on Direct and Large-Eddy Simulation, held at the University of Trieste, September 8-10, 2008
1. Auflage 2010
ISBN: 978-90-481-3652-0
Verlag: Springer Netherlands
Format: PDF
Kopierschutz: 1 - PDF Watermark

E-Book, Englisch, Band 13, 646 Seiten

Reihe: ERCOFTAC Series

After Surrey in 1994, Grenoble in 1996, Cambridge in 1999, Enschede in 2001, Munich in 2003 and Poiters in 2005, the 7th Workshop, DLES7, will be held in Trieste, again under the auspices of ERCOFTAC. Following the spirit of the series, the goal of this latest workshop is to establish a state-of-the-art of DNS and LES techniques for the computation and modeling of transitional/turbulent flows covering a broad scope of topics such as aerodynamics, acoustics, combustion, multiphase flows, environment, geophysics and bio-medical applications. This gathering of specialists in the field should once again be a unique opportunity for discussions about the more recent advances in the prediction, understanding and control of turbulent flows in academic or industrial situations.

Autoren/Hrsg.

Armenio, Vincenzo

Geurts, Bernard

Fröhlich, Jochen

Inhaltsverzeichnis

1;Direct and Large-Eddy Simulation VII;2
1.1;Preface;6
1.2;Part I Fundamentals;18
1.2.1;Wall-Modeled Large-Eddy Simulations: Present Status and Prospects;19
1.2.1.1;1 Statement of the problem;19
1.2.1.2;2 Near-wall modelling;20
1.2.1.3;3 Sources of error in WMLES;22
1.2.1.4;4 Conclusions;26
1.2.1.5;References;27
1.2.2;A Study of the Influence of the Reynolds Number on Jet Self-Similarity Using Large-Eddy Simulation;29
1.2.2.1;1 Introduction;29
1.2.2.2;2 Simulation parameters;30
1.2.2.3;3 Results;31
1.2.2.4;4 Conclusion;34
1.2.2.5;References;34
1.2.3;Direct Numerical Simulation of Fractal-Generated Turbulence;35
1.2.3.1;1 Introduction;35
1.2.3.2;2 Numerical methods;35
1.2.3.3;3 The fractal square grid;36
1.2.3.4;4 Results;37
1.2.3.5;5 Conclusion;40
1.2.3.6;References;41
1.2.4;Turbulent Oscillating Channel Flow Subjected to Wind Stress;42
1.2.4.1;1 Introduction;42
1.2.4.2;2 Problem description;43
1.2.4.3;3 Mean velocity;44
1.2.4.4;4 Structures of the turbulent flow;45
1.2.4.5;5 Conclusion;46
1.2.4.6;References;48
1.2.5;DNS of a Periodic Channel Flow with IsothermalAblative Wall;49
1.2.5.1;1 Introduction;49
1.2.5.2;2 Results;51
1.2.5.3;3 Conclusion;54
1.2.5.4;References;55
1.2.6;Diagnostic Properties of Structure Tensors in Turbulent Flows;56
1.2.6.1;1 Introduction;56
1.2.6.1.1;1.1 Definitions;56
1.2.6.2;2 Results;57
1.2.6.3;3 Conclusions;61
1.2.6.4;References;61
1.2.7;Development of Brown–Roshko Structures in the Mixing Layer Behind a Splitter Plate;63
1.2.7.1;1 Introduction;63
1.2.7.2;2 Methodology;64
1.2.7.3;3 Results;65
1.2.7.4;4 Conclusion;67
1.2.7.5;References;68
1.2.8;DNS of Spatially-Developing Three-Dimensional Turbulent Boundary Layers;69
1.2.8.1;1 Introduction and flow configuration;69
1.2.8.2;2 Results;70
1.2.8.3;3 Conclusions;74
1.2.8.4;References;74
1.2.9;Direct Numerical Simulation and Experimental Results of a Turbulent Channel Flow with Pin Fins Array;76
1.2.9.1;1 Introduction;76
1.2.9.2;2 Experimental set up;77
1.2.9.3;3 Numerical procedure;77
1.2.9.4;4 Results;78
1.2.9.5;5 Conclusions;81
1.2.9.6;References;81
1.2.10;New Experimental Results for a LES Benchmark Case;82
1.2.10.1;1 Introduction;82
1.2.10.2;2 Experimental setup;83
1.2.10.3;3 Velocity measurements;84
1.2.10.3.1;3.1 PIV measurements;84
1.2.10.3.2;3.2 LDA measurements;84
1.2.10.4;4 Conducted experiments;85
1.2.10.5;5 Conclusions;87
1.2.10.6;References;87
1.2.11;Direct Computation of the Sound Radiated by Shear Layers Using Upwind Compact Schemes;88
1.2.11.1;1 Introduction;88
1.2.11.2;2 Numerical techniques;89
1.2.11.3;3 Results and comments;90
1.2.11.4;4 Discussion and conclusion;91
1.2.11.5;References;93
1.2.12;DNS of Orifice Flow with Turbulent Inflow Conditions;94
1.2.12.1;1 Introduction;94
1.2.12.2;2 Computational approach;94
1.2.12.3;3 Results;96
1.2.12.4;4 Conclusions;97
1.2.12.5;References;97
1.2.13;The Mean Flow Profile of Wall-Bounded Turbulence and Its Relation to Turbulent Flow Topology;98
1.2.13.1;1 Introduction;98
1.2.13.2;2 Theory;99
1.2.13.3;3 Phenomenology;100
1.2.13.4;4 Conclusion;100
1.2.13.5;References;101
1.2.14;Large Eddy Simulation of a Rectangular Turbulent Jet in Crossflow;102
1.2.14.1;1 Numerical details;102
1.2.14.2;2 Boundary conditions;103
1.2.14.2.1;2.1 Jet flow;103
1.2.14.2.2;2.2 Crossflow;103
1.2.14.3;3 Results and discussion;104
1.2.14.3.1;3.1 Instantaneous flow field and vortex structures;104
1.2.14.3.2;3.2 Jet inflow conditions effect;104
1.2.14.4;4 Conclusion;105
1.2.14.5;References;105
1.2.15;Numerical Simulation of a 2D Starting-Plume Cloud-Flow;107
1.2.15.1;1 Introduction;107
1.2.15.2;2 Simulation details;107
1.2.15.3;3 Code validation;108
1.2.15.4;4 Results;109
1.2.15.5;5 Conclusions;110
1.2.15.6;References;110
1.3;Part II Methodologies and Modelling Techniques;111
1.3.1;Variational Multiscale Theory of LES Turbulence Modeling;112
1.3.1.1;1 Variational multiscale formulation of the incompressible Navier–Stokes equations ;112
1.3.1.1.1;1.1 Incompressible Navier–Stokes equations;112
1.3.1.1.2;1.2 Scale separation;115
1.3.1.2;2 Turbulent channel flow;118
1.3.1.3;3 Conclusions;119
1.3.1.4;References;121
1.3.2;An Immersed Interface Method in the Framework of Implicit Large-Eddy Simulation;122
1.3.2.1;1 Introduction;122
1.3.2.2;2 Conservative immersed interface method (CIIM);123
1.3.2.2.1;2.1 Cut-cell volume balance;123
1.3.2.2.2;2.2 Friction force;124
1.3.2.2.3;2.3 Homogeneous Neumann condition for pressure;124
1.3.2.3;3 Numerical examples;124
1.3.2.3.1;3.1 Square cylinder at Re = 100 ;125
1.3.2.3.2;3.2 Round cylinder at Re = 100 ;125
1.3.2.3.3;3.3 Round cylinder at Re = 3,900;126
1.3.2.4;4 Concluding remarks;127
1.3.2.5;References;127
1.3.3;Simulation of Gravity-Driven Flows Using an Iterative High-Order Accurate Navier–Stokes Solver;129
1.3.3.1;1 Introduction and governing equations;129
1.3.3.2;2 Numerical approach;130
1.3.3.3;3 Flow configuration;131
1.3.3.4;4 Results;132
1.3.3.4.1;4.1 Spatial growth of 2D and weak 3D disturbances;132
1.3.3.4.2;4.2 3D flow configuration and influence of suspended particles;133
1.3.3.5;5 Conclusions;134
1.3.3.6;References;135
1.3.4;Compact Fourth-Order Finite-Volume Method for Numerical Solutions of Navier–Stokes Equations on Staggered Grids;136
1.3.4.1;1 Introduction;136
1.3.4.2;2 Cartesian grid system;137
1.3.4.3;3 Numerical approximations;137
1.3.4.3.1;3.1 Cell-centered interpolation for the computation of mass fluxes;137
1.3.4.3.2;3.2 Discretisation of Poisson equation;138
1.3.4.3.3;3.3 Divergence-free convective fluxes;138
1.3.4.3.3.1;Divergence-free interpolation for convective fluxes;138
1.3.4.3.4;3.4 Nonlinear correction;139
1.3.4.4;4 Validation;139
1.3.4.5;5 Conclusion;141
1.3.4.6;References;141
1.3.5;An Accurate Numerical Method for DNS of Turbulent Pipe Flow;142
1.3.5.1;1 Introduction;142
1.3.5.2;2 Governing equations and numerical method;142
1.3.5.3;3 Results;144
1.3.5.4;References;147
1.3.6;Local Large Scale Forcing of Unsheared Turbulence;148
1.3.6.1;1 Introduction;148
1.3.6.2;2 Random force construction;149
1.3.6.3;3 Results;151
1.3.6.4;4 Conclusion;153
1.3.6.5;References;153
1.3.7;Large-Eddy Simulations of a Turbulent Magnetohydrodynamic Channel Flow;154
1.3.7.1;1 Introduction;154
1.3.7.2;2 Equations of motion and subgrid modeling;155
1.3.7.3;3 Numerical methods;155
1.3.7.4;4 Results;156
1.3.7.5;5 Conclusions;158
1.3.7.6;References;159
1.3.8;Development of a DNS-FDF Approach to Inhomogeneous Non-Equilibrium Mixing for High Schmidt Number Flows;160
1.3.8.1;1 Introduction;160
1.3.8.2;2 The DNS-FDF approach;161
1.3.8.3;3 The LMSE model for micro mixing;162
1.3.8.4;4 Modelling the subgrid scalar dissipation rate;163
1.3.8.5;5 Conclusions;164
1.3.8.6;References;165
1.3.9;Multi-Scale Simulation of Near-Wall Turbulent Flows;167
1.3.9.1;1 Introduction;167
1.3.9.2;2 Results;170
1.3.9.3;3 Conclusion;172
1.3.9.4;References;172
1.3.10;Explicit Algebraic Subgrid Stress Models for Large Eddy Simulation;173
1.3.10.1;1 Introduction;173
1.3.10.2;2 Model;174
1.3.10.3;3 Results;175
1.3.10.4;4 Conclusion;178
1.3.10.5;References;178
1.3.11;Scrutinizing the Leray-Alpha Regularization for LES in Turbulent Axisymmetric Free Jets;179
1.3.11.1;1 Introduction;179
1.3.11.2;2 Results and discussions;180
1.3.11.3;References;184
1.3.12;Localization of Unresolved Regions in the Selective Large-Eddy Simulation of Hypersonic Jets;186
1.3.12.1;1 Small scale detection criterion;186
1.3.12.2;2 Results;188
1.3.12.3;3 Concluding remarks;191
1.3.12.4;References;191
1.3.13;An ADM-Based Subgrid Scale Reconstruction Procedure for Large Eddy Simulation;193
1.3.13.1;1 Introduction;193
1.3.13.2;2 Triple-scale decomposition;193
1.3.13.3;3 Use of ADM to reconstruct the sub-filter field;194
1.3.13.4;4 Evaluation of the subgrid kinetic energy;196
1.3.13.4.1;4.1 The double deconvolution approach;196
1.3.13.4.2;4.2 A priori tests on synthetic turbulent fields;197
1.3.13.4.3;4.3 Tests on DNS fields;198
1.3.13.5;5 Conclusions and future works;198
1.3.13.6;References;199
1.3.14;Large-Eddy Simulation of Turbulent Flow in a Plane Asymmetric Diffuser by the Spectral-Element Method;200
1.3.14.1;1 Introduction;200
1.3.14.2;2 Numerical method and simulation setup;201
1.3.14.3;3 Validation by turbulent channel flow;201
1.3.14.4;4 Diffuser;202
1.3.14.4.1;4.1 Geometry and parameter settings;202
1.3.14.4.2;4.2 Results;202
1.3.14.5;5 Conclusion and outlook;205
1.3.14.6;References;205
1.3.15;h and p Refinement with Wall Modelling in Spectral-Element LES;207
1.3.15.1;1 Introduction;207
1.3.15.2;2 Method;207
1.3.15.3;3 Error determination;208
1.3.15.4;4 Results;209
1.3.15.5;5 Summary;210
1.3.15.6;References;210
1.3.16;Error-Landscape Assessment of LES Accuracy Using Experimental Data;211
1.3.16.1;1 Introduction;211
1.3.16.2;2 Methodology;212
1.3.16.3;3 Results and discussion;213
1.3.16.4;4 Conclusions;215
1.3.16.5;References;216
1.3.17;The Role of Different Errors in Classical LES and in Variational Multiscale LES on Unstructured Grids;217
1.3.17.1;1 Introduction;217
1.3.17.2;2 Basic ingredients for numerics and modeling;218
1.3.17.3;3 Results and discussion;219
1.3.17.4;References;222
1.4;Part III LES Modelling Errors;223
1.4.1;Practical Quality Measures for Large-Eddy Simulation;224
1.4.1.1;1 Introduction;224
1.4.1.2;2 Grid resolution measures;225
1.4.1.3;3 Application;226
1.4.1.4;4 Discussion and conclusions;228
1.4.1.5;References;229
1.4.2;The Simplest LES;230
1.4.2.1;1 Introduction;230
1.4.2.2;2 The two-point sum and difference operators;231
1.4.2.3;3 The two-point LES of a passive scalar;233
1.4.2.4;4 The simplest LES of a passive scalar in the case of a homogeneous turbulent field;234
1.4.2.5;5 Conclusions;235
1.4.2.6;References;235
1.4.3;Application of an Anisotropy Resolving Algebraic Reynolds Stress Model within a Hybrid LES-RANS Method;237
1.4.3.1;1 Introduction;237
1.4.3.2;2 Hybrid LES-URANS methodology;238
1.4.3.3;3 Numerical method and test case;240
1.4.3.4;4 Results and conclusions;240
1.4.3.5;References;243
1.4.4;LES Meets FSI – Important Numerical and Modeling Aspects;244
1.4.4.1;1 Introduction;244
1.4.4.2;2 Important steps for joining LES and FSI;245
1.4.4.2.1;2.1 LES on moving grids;245
1.4.4.2.2;2.2 Partitioned coupled predictor–corrector scheme;247
1.4.4.3;References;249
1.4.5;A New Multiscale Model with Proper Behaviour in Both Vortex Flows and Wall Bounded Flows;251
1.4.5.1;1 Introduction;251
1.4.5.2;2 Presentation of the model;251
1.4.5.3;3 LES of the turbulent channel flow;252
1.4.5.4;4 LES of a counter-rotating four-vortex system;254
1.4.5.5;5 LES of a two-vortex system in ground effect;254
1.4.5.6;6 Conclusion;256
1.4.5.7;References;256
1.4.6;LES Based POD Analysis of Jet in Cross Flow;257
1.4.6.1;1 Introduction;257
1.4.6.2;2 Large Eddy simulation (LES) details;257
1.4.6.3;3 Results;259
1.4.6.4;4 Conclusions;262
1.4.6.5;References;263
1.4.7;A Dissipative Scale-Similarity Model;264
1.4.7.1;1 The dissipative scale-similarity model;264
1.4.7.2;2 Results;267
1.4.7.2.1;2.1 Decaying grid turbulence;267
1.4.7.2.2;2.2 Fully developed channel flow;267
1.4.7.3;3 Concluding comments;269
1.4.7.4;References;270
1.4.8;Optimization of Turbulent Mixing Restricted by Linear and Nonlinear Constraints;271
1.4.8.1;1 Introduction;271
1.4.8.2;2 Cost functionals;272
1.4.8.3;3 Constrained optimization method;272
1.4.8.4;4 Computational setup and discretization;273
1.4.8.5;5 Results;273
1.4.8.5.1;5.1 Comparison of the augmented Lagrangian and gradient projection method;273
1.4.8.5.2;5.2 Optimization with different cost functionals;274
1.4.8.6;6 Conclusion;276
1.4.8.7;References;276
1.4.9;Stochastic Coherent Adaptive LES of Forced Isotropic Turbulence;277
1.4.9.1;1 Introduction;277
1.4.9.2;2 Adaptive LES;278
1.4.9.3;3 Numerical experiments;279
1.4.9.4;References;282
1.4.10;An Improvement of Increment Model by Using Kolmogorov Equation of Filtered Velocity;283
1.4.10.1;1 Introduction;283
1.4.10.2;2 Improved increment model;284
1.4.10.3;3 A priori numerical verifications;285
1.4.10.4;4 Conclusion;288
1.4.10.5;References;288
1.4.11;Symmetry-Preserving Regularization Modelling of a Turbulent Plane Impinging Jet;290
1.4.11.1;1 Introduction;290
1.4.11.2;2 C4 regularization modelling;291
1.4.11.3;3 Numerical method: symmetry preserving discretization;292
1.4.11.3.1;3.1 Kinetic energy conservation;292
1.4.11.3.2;3.2 Solving the pressure–velocity coupling – Checkerboard problem;293
1.4.11.4;4 Numerical results;294
1.4.11.5;5 Conclusions;295
1.4.11.6;References;295
1.4.12;Progress in the Development of Stochastic Coherent Adaptive LES Methodology;297
1.4.12.1;1 Introduction;297
1.4.12.2;2 SCALES methodology;297
1.4.12.3;References;301
1.5;Part IV Scalars;302
1.5.1;LES of Heat Transfer in a Channel with a Staggered Pin Matrix;303
1.5.1.1;1 Introduction;303
1.5.1.2;2 Flow specification and computational details;304
1.5.1.3;3 Discussion of results;305
1.5.1.4;4 Conclusions;308
1.5.1.5;References;308
1.5.2;Turbulent Channel Flow with -Shape Turbulators on One Wall;309
1.5.2.1;1 Introduction;309
1.5.2.2;2 Numerical procedure;311
1.5.2.3;3 Results;311
1.5.2.4;4 Conclusions;314
1.5.2.5;References;314
1.5.3;Implicit Large-Eddy Simulation of Passive-Scalar Mixing in a Confined Rectangular-Jet Reactor;315
1.5.3.1;1 Introduction;315
1.5.3.2;2 Experimental configuration;315
1.5.3.3;3 Numerical method;316
1.5.3.4;4 Implicit subgrid-scale modeling for passive-scalar transport;317
1.5.3.5;5 Computational details and numerical results;317
1.5.3.6;References;320
1.5.4;Direct Numerical Simulation of a Turbulent Boundary Layer with Passive Scalar Transport;321
1.5.4.1;1 Introduction;321
1.5.4.2;2 Numerical approach;322
1.5.4.3;3 Results;322
1.5.4.4;4 Conclusion;325
1.5.4.5;References;326
1.6;Part V Active Scalars;328
1.6.1;Numerical Experiments on Turbulent Thermal Convection;329
1.6.1.1;1 Introduction;329
1.6.1.2;2 The problem;331
1.6.1.3;3 Experimental setups ;332
1.6.1.4;4 Numerical simulations ;333
1.6.1.5;5 Results ;334
1.6.1.6;References;336
1.6.2;Direct Numerical Simulation of Turbulent Reacting and Inert Mixing Layers Laden with Evaporating Droplets;337
1.6.2.1;1 Introduction;337
1.6.2.2;2 Direct numerical simulation;338
1.6.2.3;3 Results and discussion;340
1.6.2.4;4 Conclusions;342
1.6.2.5;References;342
1.6.3;Large Eddy Simulation of a Two-Phase Reacting Flow in an Experimental Burner;344
1.6.3.1;1 Introduction;344
1.6.3.2;2 Numerical configuration and simulations;345
1.6.3.2.1;2.1 Calculation domain and mesh;345
1.6.3.2.2;2.2 Description of the Euler–Euler (EE) solver;345
1.6.3.2.3;2.3 Description of the evaporation and combustion models;346
1.6.3.2.4;2.4 Description of the kerosene injection;346
1.6.3.3;3 Results;347
1.6.3.3.1;3.1 Gas flow without droplets (case I);347
1.6.3.3.2;3.2 Gas flow with evaporating droplets (case II);347
1.6.3.3.3;3.3 Reacting two-phase flow (case III);348
1.6.3.4;4 Conclusions;349
1.6.3.5;References;349
1.6.4;Hybrid LES/CAA Simulation of a Turbulent Non-Premixed Jet Flame;351
1.6.4.1;1 Introduction;351
1.6.4.2;2 LES/CAA methodology;352
1.6.4.3;3 Experimental test case and numerical setup;353
1.6.4.4;4 Results;353
1.6.4.5;5 Conclusions and outlook;356
1.6.4.6;References;356
1.6.5;LES/CMC of Forced Ignition of a Bluff-Body Stabilised Non-Premixed Methane Flame;358
1.6.5.1;1 Introduction;358
1.6.5.2;2 Modelling;359
1.6.5.3;3 Results and discussion;360
1.6.5.4;4 Conclusions;363
1.6.5.5;References;363
1.6.6;Large Eddy Simulation of a High Reynolds Number Swirling Flow in a Conical Diffuser;364
1.6.6.1;1 Introduction;364
1.6.6.2;2 Physical and numerical modelling;364
1.6.6.2.1;2.1 Axisymetric diffuser;365
1.6.6.2.2;2.2 Wall modelling;366
1.6.6.2.3;2.3 Generation of realistic inlow;367
1.6.6.3;3 Results;368
1.6.6.3.1;3.1 Profiles of mean velocity;368
1.6.6.3.2;3.2 Instantaneous flow;368
1.6.6.4;4 Conclusion;370
1.6.6.5;References;370
1.6.7;Direct Numerical Simulation of Hot and Highly Pulsated Turbulent Jet Flows;372
1.6.7.1;1 Introduction;372
1.6.7.2;2 Formulation of the problem;373
1.6.7.2.1;2.1 Modeling assumptions and governing equations;373
1.6.7.2.2;2.2 Numerical procedure;374
1.6.7.3;3 Numerical results;375
1.6.7.4;References;377
1.6.8;DNS of Convective Heat Transfer in a Rotating Cylinder;379
1.6.8.1;1 Introduction;379
1.6.8.2;2 Numerical procedure;380
1.6.8.3;3 Coherent structures in the flow;381
1.6.8.4;4 Heat transfer;383
1.6.8.5;5 Concluding remarks;384
1.6.8.6;References;384
1.6.9;Numerical Simulations of Thermal Convection at High Prandtl Numbers;385
1.6.9.1;1 Introduction;385
1.6.9.2;2 Physical and numerical setup;386
1.6.9.3;3 Results;387
1.6.9.3.1;3.1 Nusselt number;387
1.6.9.3.2;3.2 Characteristic velocity and Reynolds number;388
1.6.9.4;4 Conclusions;390
1.6.9.5;References;390
1.6.10;Influence of the Lateral Walls on the Thermal Plumes in Turbulent Rayleigh–Bénard Convection in Rectangular Containers;391
1.6.10.1;1 Introduction;391
1.6.10.2;2 Computational setup;392
1.6.10.3;3 Results;393
1.6.10.4;4 Conclusions;395
1.6.10.5;References;396
1.6.11;DNS of Mixed Convection in Enclosed 3D-Domains with Interior Boundaries;397
1.6.11.1;1 Introduction;397
1.6.11.2;2 Problem definition;398
1.6.11.3;3 Numerical method;399
1.6.11.4;4 Poisson solver;400
1.6.11.5;5 Results;402
1.6.11.6;6 Conclusion;402
1.6.11.7;References;403
1.6.12;LES and Hybrid RANS/LES of Turbulent Flow in Fuel Rod Bundle Arranged with a Triangular Array;404
1.6.12.1;1 Introduction;404
1.6.12.2;2 Hybrid RANS/LES;405
1.6.12.3;3 Case description;406
1.6.12.4;4 Results;407
1.6.12.5;5 Conclusion;408
1.6.12.6;References;409
1.6.13;Large-Scale Patterns in a Rectangular Rayleigh–Bénard Cell;410
1.6.13.1;1 Introduction;410
1.6.13.2;2 Numerical set-up;410
1.6.13.3;3 Results;412
1.6.13.4;References;413
1.6.14;LES and Laser Measurements of Dynamic Flame/Vortex Interactions;414
1.6.14.1;1 Introduction;414
1.6.14.2;2 Experimental work;414
1.6.14.3;3 Large Eddy simulation (LES) model;415
1.6.14.4;4 Results and discussion;416
1.6.14.5;References;417
1.6.15;3D Direct Simulation of a Nonpremixed Hydrogen Flame with Detailed Models;418
1.6.15.1;1 Introduction;418
1.6.15.2;2 Physical and computational model;419
1.6.15.3;3 Results;420
1.6.15.3.1;3.1 Turbulent flame structure;420
1.6.15.3.2;3.2 Reconstructing the PDF of mixture fraction;421
1.6.15.4;4 Conclusions;421
1.6.15.5;References;422
1.7;Part VI Environmental and Multiphase Flows;423
1.7.1;Large Eddy Simulation of Pollen Dispersion in the Atmosphere;424
1.7.1.1;1 Introduction;424
1.7.1.2;2 Model description;424
1.7.1.3;3 Numerical discretization;426
1.7.1.4;4 Validation;426
1.7.1.5;5 Pollen dispersion from a ragweed field;428
1.7.1.6;6 Conclusions;430
1.7.1.7;References;431
1.7.2;Internal Wave Breaking in Stratified Flows Past Obstacles;432
1.7.2.1;1 Introduction;432
1.7.2.2;2 Results;432
1.7.2.3;3 Conclusions;437
1.7.2.4;References;438
1.7.3;DNS of a Gravity Current Propagating over a Free-Slip Boundary;439
1.7.3.1;1 Introduction;439
1.7.3.2;2 Approach;440
1.7.3.2.1;2.1 Implementation;441
1.7.3.3;3 Results;441
1.7.3.4;References;444
1.7.4;Large Eddy Simulation of Turbulent Mixing in an Estuary Region;445
1.7.4.1;1 Introduction;445
1.7.4.2;2 The mathematical model;446
1.7.4.3;3 Application to an estuarine flow: results and discussion;448
1.7.4.4;References;450
1.7.5;Dispersion of (Light) Inertial Particles in Stratified Turbulence;451
1.7.5.1;1 Introduction;451
1.7.5.2;2 Numerical approach;451
1.7.5.3;3 Results;453
1.7.5.3.1;3.1 Single-particle dispersion;453
1.7.5.3.2;3.2 Preferential concentration;454
1.7.5.3.3;3.3 Forces acting on the particles;455
1.7.5.4;4 Concluding remarks;456
1.7.5.5;References;456
1.7.6;The Influence of Magnetic Fields on the Rise of Gas Bubbles in Electrically Conductive Liquids;458
1.7.6.1;1 Introduction;458
1.7.6.2;2 Numerical method;459
1.7.6.3;3 Results;460
1.7.6.3.1;3.1 Gas bubbles rising on linear paths;460
1.7.6.3.2;3.2 Gas bubbles on unsteady paths;462
1.7.6.4;4 Conclusions;463
1.7.6.5;References;463
1.7.7;Large Eddy Simulation of a Turbulent Droplet LadenMixing Layer;465
1.7.7.1;1 Introduction;465
1.7.7.2;2 Mathematical modelling;466
1.7.7.2.1;2.1 Filtered Navier–Stokes equations;466
1.7.7.2.2;2.2 PDF modelling of fuel sprays;466
1.7.7.3;3 Results and discussion;467
1.7.7.4;4 Conclusions;470
1.7.7.5;References;470
1.7.8;The Diffuse Interface Method with Korteweg Approach for Isothermal, Two-Phase Flow of a Van der Waals Fluid;471
1.7.8.1;1 Introduction;471
1.7.8.2;2 The numerical method;472
1.7.8.3;3 Benchmark simulations;474
1.7.8.3.1;3.1 Drop retraction;474
1.7.8.3.2;3.2 Two-drop collision;474
1.7.8.3.3;3.3 Determination of surface tension;475
1.7.8.4;4 Concluding remarks;476
1.7.8.5;References;476
1.7.9;Numerical Simulation of Air Flows in Street Canyons Using Mesh-Adaptive LES;477
1.7.9.1;1 Methodology;477
1.7.9.1.1;1.1 Inlet boundary conditions;478
1.7.9.1.2;1.2 Traffic induced turbulence;478
1.7.9.2;2 Results and conclusions;479
1.7.9.3;References;480
1.8;Part VII Aerodynamics and Wakes;482
1.8.1;LES of the Flow Around a Two-Dimensional Vehicle Model with Active Flow Control;483
1.8.1.1;1 Introduction;483
1.8.1.2;2 Description of the model and numerical set-up;484
1.8.1.2.1;2.1 Boundary conditions and the actuation;485
1.8.1.2.2;2.2 Numerical simulations;485
1.8.1.3;3 Results;485
1.8.1.4;References;489
1.8.2;Wake-Vortex Decay in External Turbulence;490
1.8.2.1;1 Introduction;490
1.8.2.2;2 Computational setting and regularization modeling;491
1.8.2.3;3 Direct numerical simulation;492
1.8.2.4;4 Regularization modeling of vortex decay;494
1.8.2.5;5 Concluding remarks;494
1.8.2.6;References;495
1.8.3;DNS of Aircraft Wake Vortices: The Effect of Stable Stratification on the Development of the Crow Instability;496
1.8.3.1;1 Introduction;496
1.8.3.2;2 Approach;497
1.8.3.3;3 Results;498
1.8.3.4;References;501
1.8.4;On the Download Alleviation for the XV-15 Wing by Active Flow Control Using Large-Eddy Simulation;503
1.8.4.1;1 Introduction;503
1.8.4.2;2 Numerical method;503
1.8.4.3;3 Results;505
1.8.4.3.1;3.1 Drag and lift;505
1.8.4.3.2;3.2 Pressure coefficient distribution and its RMS;505
1.8.4.3.3;3.3 Mean velocity, pressure and resolved kinetic energy;507
1.8.4.4;4 Conclusion;508
1.8.4.5;References;508
1.8.5;Turbulent Flow Simulations Around an Airfoil At High Incidences Using URANS, DES and ILES Approaches;509
1.8.5.1;1 Introduction;509
1.8.5.2;2 Methodologies;509
1.8.5.3;3 Simulation conditions and setup;510
1.8.5.4;4 Numerical results and discussions;511
1.8.5.5;5 Conclusions;515
1.8.5.6;References;516
1.8.6;Large Eddy Simulation of Flow Around an Airfoil Near Stall;517
1.8.6.1;1 Introduction;517
1.8.6.2;2 Results;518
1.8.6.3;3 Conclusions;520
1.8.6.4;References;520
1.8.7;Large Eddy Simulation of Turbulent Flows Around a Rotor Blade Segment Using a Spectral Element Method;522
1.8.7.1;1 Numerical method and computational parameters;522
1.8.7.2;2 Results of large eddy simulations;523
1.8.7.3;3 Outlook and future work;525
1.8.7.4;References;525
1.9;Part VIII Compressible Flows;526
1.9.1;DNS of Compressible Turbulent Flows;527
1.9.1.1;1 Introduction;527
1.9.1.2;2 Comparison of supersonic channel and pipe flow;528
1.9.1.2.1;2.1 Mean flow variables;529
1.9.1.2.2;2.2 Reynolds stresses and budgets;531
1.9.1.3;3 Supersonic nozzle and diffuser flow;533
1.9.1.4;4 Conclusions;536
1.9.1.5;References;537
1.9.2;Large-Eddy Simulation of Transonic Buffet over a Supercritical Airfoil;538
1.9.2.1;1 Introduction;538
1.9.2.2;2 Description of the computation;539
1.9.2.3;3 Mean field analysis;539
1.9.2.4;4 Spectral analysis;541
1.9.2.5;5 Space and time scales;541
1.9.2.6;6 Discussion;543
1.9.2.7;References;543
1.9.3;Detached-Eddy and Delayed Detached-Eddy Simulation of Supersonic Flow over a Three-Dimensional Cavity;544
1.9.3.1;1 Introduction;544
1.9.3.2;2 Numerical methodology;545
1.9.3.3;3 Details of the test case;545
1.9.3.4;4 Results;545
1.9.3.4.1;4.1 Comparison of DES, DES-MB and DDES;545
1.9.3.4.2;4.2 Influence of grid resolution;547
1.9.3.4.3;4.3 Influence of momentum thickness of the boundary layer at the cavity leading edge;548
1.9.3.5;5 Summary;549
1.9.3.6;References;549
1.9.4;A WALE-Similarity Mixed Model for Large-Eddy Simulation of Wall Bounded Compressible Turbulent Flows;551
1.9.4.1;1 Introduction;551
1.9.4.2;2 Mathematical formulation;552
1.9.4.2.1;2.1 The WALE-similarity model (WSM);553
1.9.4.2.2;2.2 Flow configuration and numerics;553
1.9.4.3;3 Results and discussion;554
1.9.4.4;References;556
1.9.5;Parametric Study of Compressible Turbulent Spots;558
1.9.5.1;1 Introduction;558
1.9.5.2;2 Method;559
1.9.5.3;3 Results;560
1.9.5.4;4 Conclusion;563
1.9.5.5;References;563
1.9.6;Azimuthal Resolution Effects in LES of Subsonic Jet Flow and Influence on Its Noise;564
1.9.6.1;1 Numerical methods;564
1.9.6.2;2 Resolution effect;566
1.9.6.2.1;2.1 Effect on the flow field;566
1.9.6.2.2;2.2 Effect on the acoustic near-field;568
1.9.6.3;3 Conclusions;569
1.9.6.4;References;570
1.9.7;Large Eddy Simulations of Compressible MHD Turbulence in Heat-Conducting Fluid;571
1.9.7.1;1 Introduction;571
1.9.7.2;2 LES formulation;571
1.9.7.3;3 Results;573
1.9.7.4;References;575

Leseproben

"LES Meets FSI – Important Numerical and Modeling Aspects (p. 245-246)

M. Breuer1,2 and M. Münsch2

1 Dept. of Fluid Mechanics, Institute of Mechanics, Helmut-Schmidt-University Hamburg, Holstenhofweg 85, D-22043 Hamburg, Germany, [e-mail] 2 Institute of Fluid Mechanics, University of Erlangen-N¨urnberg, Cauerstr. 4, D-1058 Erlangen, Germany, [e-mail]

Abstract The paper is concerned with two main aspects, which should be considered when large–eddy simulation (LES) is married to ?uid–structure interaction (FSI). First, the in?uence of moving grids leading to temporally varying ?lter widths and thus additional commutation errors on the quality of the predicted results is thoroughly investigated. Second, a new partitioned coupling method based on the predictor–corrector scheme often used for LES is evaluated. A strongly coupled but nevertheless still explicit time–stepping algorithm results, which is very e?cient in the LES–FSI context. This new scheme is evaluated in detail based simulations around elastically supported cylindrical structures and a swiveling ?at plate.

1 Introduction

Fluid–structure interaction (FSI) plays a dominant role in many technical applications such as suspension bridges, o?-shore platforms or even vocal folds. Therefore, a strong need for appropriate numerical simulation tools exists for such coupled problems. In previous studies, FSI applications in the regime of laminar ?ows as well as turbulent ?ows using the RANS approach [5, 6] were numerically investigated. For that purpose, a partitioned fully implicit scheme was applied which coupled a three-dimensional ?nite-volume based multi-block ?ow solver for incompressible ?uids with a ?nite-element code for the structural problem.

This coupling scheme works e?ciently for large time step sizes typically used for implicit time-stepping schemes within RANS predictions. However, ?ow problems involving large-scale ?ow structures such as vortex shedding or instantaneous separation and reattachment are often not reliably predicted by RANS and more advanced techniques such as largeeddy simulation (LES) are required. To resolve the turbulent ?ow ?eld in time, LES uses small time steps.

Thus, in general explicit time-marching schemes are favored, especially predictor–corrector schemes [1,2]. Furthermore, for FSI applications the solution domain changes in time due to the displacement of the boundaries linked to the structure. Thus moving grids have to be used which has a direct in?uence on the ?ltering approach in LES. Thus the paper addresses the aspects of additional errors introduced (e.g., commutation errors) and code coupling, which should be considered when LES is married to FSI.

2 Important steps for joining LES and FSI

2.1 LES on moving grids

Within an FSI application the ?uid forces acting on the structure lead to the displacement or deformation of the structure. Thus the computational domain is no longer ?xed but changes in time. Besides other numerical techniques to account for these variations, the most popular one is the so-called Arbitrary Lagrangian–Eulerian (ALE) formulation. Here the conservation equations for mass, momentum (and energy) are re-formulated for a temporally varying domain."