Bulletin of the American Physical Society
72nd Annual Meeting of the APS Division of Fluid Dynamics
Volume 64, Number 13
Saturday–Tuesday, November 23–26, 2019; Seattle, Washington
Session H19: CFD: Advanced Turbulence Simulations |
Hide Abstracts |
Chair: Krishnan Mahesh, University of Minnesota, Twin Cities Room: 401 |
Monday, November 25, 2019 8:00AM - 8:13AM |
H19.00001: Turbulent channel flow at $Re_\tau=10000$ Sergio Hoyas, Martin Oberlack, Stefanie Kraheberger, Francisco Alcantara-Avila A new simulation of a turbulent channel flow was conducted up to the limit of $Re_\tau=10.000$. The domain size is $2\pi h \times2h\times \pi h$. This domain is thought to be large enough to accurately compute the one point statistics of the flow. The simulation has been carried out on 2048 SuperMUC phase II cores, at a mesh of $(6144, 2101, 6144)\approx 8\times10e10$ grid points. A database with approximately 100 TB has already been created, which will be analyzed further at a later stage. As it was expected, a long logarithmic layer exists with $\kappa \approx 0.40$ and extending from $y^+\approx 70$ to $y^+\approx 2000$. The first maximum of the indicator function is not growing anymore and remains constant. A first analysis of the intensities shows that the near wall peaks of $u'$, $w'$ and $p'$ are still growing with Reynolds number. The possible secondary maximum of $u'$ is barely present. New scaling laws of $U$ and $u'$ based on symmetry theory will be also shown [Preview Abstract] |
Monday, November 25, 2019 8:13AM - 8:26AM |
H19.00002: Wall-resolved LES and RANS/LES hybrid analyses in turbulent heat exchanger with modified oblique wavy walls Kenichi Morimoto, Shu-Qun Jin, Junyu Chen High-performance turbulent heat exchangers play a key role in diversified energy systems/devices. In our recent work, double-pipe turbulent heat exchangers with V-shaped oblique wavy walls, with which the ratio of the heat transfer to the pressure loss is much larger than that for conventional techniques, have been proposed. The present study aims to clarify the detailed mechanism of the turbulent heat transfer enhancement, and to explore a practical numerical approach to deal with turbulence anisotropy and unsteady flow separation at high Reynolds number condition. Here we perform wall-resolved large eddy simulations based on dynamic sigma model in which a dynamic procedure is originally applied to both the velocity and thermal fields. Also we perform improved delayed detached eddy simulation (IDDES) with k-$\omega$ SST model as a near-wall model. It is shown that, with the present wall undulations, large-scale vortical structures are embedded in a steady manner inside the turbulent boundary layer, leading to remarkable enhancement of the heat transfer performance. The global and local quantities as well as turbulent statistics are compared between RANS, LES and hybrid approaches. The feasibility of RANS/LES hybrid approach is explored for further shape-optimization study. [Preview Abstract] |
Monday, November 25, 2019 8:26AM - 8:39AM |
H19.00003: Learning about High Schmidt scalar mixing in turbulent round jets from 3D periodic box simulations Guillaume Blanquart, Kyupaeck Jeff Rah Jet-centerline (JC) forcing technique was developed to create velocity and scalar fields of turbulent round jets in triply periodic box. This forcing method is here utilized to simulate high Schmidt number passive scalars mixing. A series of DNS have been performed to compare scalar statistics over a range of Schmidt and Reynolds numbers. For a given Schmidt number, the turbulent scalar flux increases with Reynolds number. This increase is in contrast with the values for unity Schmidt number cases, whose increase with Reynolds number is either small or unapparent. For a given Reynolds number, the flux values decrease with the Schmidt number. Turbulent scalar flux values for infinity Schmidt numbers are extrapolated from the simulations at finite Schmidt numbers. These estimates are in good agreement with jet experiments. Scalar energy spectra have been computed as well, and their scaling exponents, n, have been estimated. At a fix Reynolds number, the value of n decreases as the Schmidt number increases. At a fix Schmidt number, the n value increases with the Reynolds number. Once again, values extrapolated for infinite Schmidt numbers compare well with experimental observations. [Preview Abstract] |
Monday, November 25, 2019 8:39AM - 8:52AM |
H19.00004: Data-driven predictions of root mean square pressure fluctuations from RANS simulations Giacomo Lamberti, Catherine Gorle Wind engineering applications often employ computationally efficient Reynolds-averaged Navier-Stokes (RANS) simulations, even if the turbulence models can compromise the accuracy of the prediction. Furthermore, the estimation of turbulent quantities such as fluctuating pressure loads requires additional, potentially inaccurate, models. For example, empirical models to estimate the root mean square (rms) pressure fluctuations from of the mean pressure, turbulent kinetic energy and velocity, are known to produce incorrect results on the lateral facades of tall buildings. To address this problem, we propose a data-driven approach to determine the best functional form that relates high-fidelity data of rms pressure fluctuations to time-averaged variables predicted by RANS. The high-fidelity data is obtained from large-eddy simulations or wind tunnel experiments of a tall building at different wind directions. We perform RANS simulations for the same conditions and construct features from the resulting time-averaged quantities. Then, we employ machine learning to relate these features to the available training data. We investigate the ability of the data-driven model to provide predictions for wind directions or regions of the building that were not included in the training data. [Preview Abstract] |
Monday, November 25, 2019 8:52AM - 9:05AM |
H19.00005: Area of Scalar Isosurfaces in Homogeneous Isotropic Turbulence Kedar Prashant Shete, Stephen De Bruyn Kops A fundamental effect of fluid turbulence is turbulent mixing, which results in the stretching and wrinkling of scalar isosurfaces. Thus, the area of isosurfaces is of interest in understanding turbulence in general with specific applications in, e.g., combustion and the identification of turbulent/non-turbulent interfaces. We report measurements of isosurface areas in 28 direct numerical simulations (DNSs) of homogeneous isotropic turbulence with a mean scalar gradient resolved on up to $14256^3$ grid points with Taylor Reynolds number $Re$ ranging from 24 to 633 and Schmidt number $Sc$ ranging from 0.1 to 7. The continuous equation we evaluate converges exactly to the area in the limit of zero layer thickness. We demonstrate a method for numerically integrating this equation that, for a test case with an analytical solution, converges linearly towards the exact solution with decreasing layer width. By applying the technique to DNS data and testing for convergence with resolution of the simulations, we verify the resolution requirements for DNS that have been recently published. We conclude that isosurface areas scale with the square root of the Taylor P\'eclet number $Pe$ between approximately 50 and 4429. No independent effect of either $Re$ or $Sc$ were observed. [Preview Abstract] |
Monday, November 25, 2019 9:05AM - 9:18AM |
H19.00006: A study of locking phenomenon of elliptical particle in shear flow with DNS Zhizhong Ding, Chenguang Zhang, Shashank Tiwari, Jyeshtharaj Joshi, Krishnaswamy Nandakumar The understanding of the dynamics of arbitrary shaped particle(s) under shear flow is of great importance in design and scale-up Chemical Engineering equipment. In this study, we have carried out a series of numerical experiments on an elliptical cylinder particle, subjected to a 2D shear flow for moderate shear-based Reynolds number. The simulations have been performed using an in-house code Signed Distance Field - Immersed Boundary Method (SDFIBM) (Zhang et al., 2018). The degrees of freedom of the particle is limited and thus allows the particle to rotate only about a fixed position in space. Any translational and vertical movements are strictly limited to phase out the noises. This work demonstrates that there are two stages for elliptical cylinder particle in shear flow: the periodical rotating stage and stationary stage, divided by a critical Reynolds number. We further extend our investigation to include the effects of changing particle aspect ratio, flow confinement size and distance of particle from the wall. Frequency domain analysis have been carried out on the data to develop a better understanding of the locking phenomenon. The effect of a second ellipse in the vicinity of primary particle and its impact on the locking phenomenon has also been included. [Preview Abstract] |
Monday, November 25, 2019 9:18AM - 9:31AM |
H19.00007: Direct Numerical Simulation in the Sub-critical and Super-critical regimes for Flow past a Stationary Sphere Shashank S Tiwari, Shivkumar Bale, Zhizhong Ding, Ashwin W. Patwardhan, Krishnaswamy Nandakumar, Jyeshtharaj B. Joshi The physical understanding of separating flows which exhibit critical phenomena under various flow conditions, have helped in designing various drag-reduction devices, turbulence generators, controllers, etc. Direct Numerical Simulation (DNS) of such separating flows helps to identify the flow structures and decipher the corresponding effects they have on the resulting forces. DNS being computationally intensive, the investigations for flow past a sphere has been limited to Re $=$10,000. In this study, we test the capability of OpenFOAM in performing fully resolved DNS for 1000 \textless Re \textless 10$^{\mathrm{5}}_{\mathrm{.\thinspace }}$Appropriate time and length scales have been used to adequately resolve the boundary layer. Extensive simulations were performed to test, optimize and set a benchmarking case for the domain size, mesh size, time-step and discretization schemes required for performing such computationally intensive simulations on a scalable parallel platform. Simulations were run till a statistically converged solution was obtained. The drag coefficients and pressure coefficients from the simulations were compared against experimental results available in literature and were found to be in good agreement. [Preview Abstract] |
Monday, November 25, 2019 9:31AM - 9:44AM |
H19.00008: Large-Eddy Simulation of Propeller Crashback Using an Unstructured Overset Grid Method Thomas Kroll, Wyatt Horne, Krishnan Mahesh We discuss the application of a novel unstructured overset grid methodology to compute the flow around a marine propeller. We consider the crashback mode of operation where the propeller rotates in the reverse direction while the vessel moves in the forward direction yielding highly unsteady loads. The numerical algorithm is that developed by Horne and Mahesh [J. Comput. Phys (2019) 376:585-596] and addresses two significant challenges posed by the overset methodology - discrete conservation and scaling. The simulations consider open and ducted propellers. The results are compared to available experimental data and previous LES studies. Details of the flow field are discussed. [Preview Abstract] |
Monday, November 25, 2019 9:44AM - 9:57AM |
H19.00009: Large-eddy simulation of a submarine propeller downstream of a rudder Antonio Posa, Riccardo Broglia, Elias Balaras The influence of an upstream rudder on the wake properties of a submarine propeller is investigated using Large-Eddy Simulation and an Immersed-Boundary method. The flow problem is simulated using a cylindrical grid composed of about 1.7 billion nodes. Earlier computations of the isolated propeller (open-water condition) demonstrated the accuracy of the overall methodology, via comparisons with both dynamometric measurements and Particle Imaging Velocimetry (PIV) visualizations. Three incidence conditions of the rudder are considered, corresponding to $0^o$, $10^o$ and $20^o$. Comparisons with the results in open-water conditions demonstrate that in the near wake the topology of the typical coherent structures shed by submarine propellers (tip and hub vortices) is not modified by the presence of the upstream hydrofoil at small incidence angles. In contrast, levels of turbulence within the wake are dramatically increased in the last configuration, featuring the rudder at $20^o$ of incidence. This substantial change is triggered by the separation occurring on the suction side of the hydrofoil, leading to a significant perturbation of the inflow conditions of the propeller. [Preview Abstract] |
Monday, November 25, 2019 9:57AM - 10:10AM |
H19.00010: DNS and LES of Transitional Flow around The T106C Turbine blade Using the High-Order FR/CPR Method Mohammad Alhawwary, Z. J. Wang A high-order Navier-Stokes solver based on the flux reconstruction (FR) or the correction procedure via reconstruction (CPR) formulation is employed to perform direct numerical simulations (DNS) and large eddy simulations (LES) of a well-known benchmark problem -- transitional flow over the low-pressure turbine T106C cascade. Using both h- and p-refinements to achieve a DNS resolution we were able to establish a "converged" solution, including the mean pressure and skin-friction distribution, and the wake loss. The identified DNS levels were achieved using a 4th order solution on a coarse mesh and a 3rd order one on a medium mesh, both agreed very well. Then LES on the coarse mesh with 3rd order schemes and varying the (X$+$,Y$+$,Z$+)$ resolutions was conducted to assess the mesh dependence of the solution. In particular, we study the error in the transition location and the mean skin-friction due to the different coarse (X$+$,Y$+$,Z$+)$ resolutions of these LES results. These h and p-refinement (coarse and medium meshes, at 2nd, 3rd and 4th order accuracy) as well as coarse (X$+$,Y$+$,Z$+)$ studies will provide much needed guideline in mesh resolution to achieve a certain level of accuracy for flow parameters of interest to designers. [Preview Abstract] |
Monday, November 25, 2019 10:10AM - 10:23AM |
H19.00011: On the effect of wake passing on a low pressure turbine cascade using spectral/\textit{hp} element methods Andrea Cassinelli, Paolo Adami, Francesco Montomoli, Spencer J Sherwin The developing and future interaction of HPC, high fidelity CFD and high order unstructured grid algorithms has the potential to allow for simulation-based research, analysis and design capability. Previous work focused on developing guidelines to leverage the use of high order spectral/\textit{hp} element methods as a virtual cascade for turbomachinery applications. Building on the knowledge previously reported, we analyze a representative industrial low pressure turbine cascade subject to wake passing interactions at moderate Reynolds number, adopting the incompressible Navier-Stokes solver implemented in the Nektar$++$ software framework. The rotor-stator interaction is modelled by imposing appropriate Dirichlet boundary conditions. The impact of flow coefficient and reduced frequency is studied in conjunction with the Reynolds sensitivity. The analysis focuses in detail on the dynamics of the separation bubble on the suction surface looking at mean flow properties and turbulence kinetic energy budgets, comparing the main findings with experimental data. [Preview Abstract] |
Monday, November 25, 2019 10:23AM - 10:36AM |
H19.00012: Numerical investigation of the flow in gas turbine blade trailing edge internal cooling passages Jaehyun Ryu, Wontae Hwang Gas turbine blades operate at very high temperatures, often beyond material limits. Internal and external cooling enable the blade to survive these extreme temperatures. The trailing edge of the blade is designed to be sharp for high aerodynamic efficiency, but this makes internal convective cooling poor at the corner. The effect of ribs was assessed in a right triangle channel containing a sharp corner, representing a simplified trailing edge. First, 45° angled ribs on the pressure and suction side walls were investigated. The Reynolds-averaged Navier-Stokes (RANS) results show fairly good agreement with previous results from magnetic resonance velocimetry (MRV) and large eddy simulation (LES). Different turbulence models were assessed, and the baseline explicit algebraic Reynolds stress model (BSL EARSM) was adequate in capturing the flow structure. Next, we optimized rib geometry via design of experiments (DOE), by changing the rib height to channel height ratio and rib angle. 3-level full factorial design (FFD) was used to determine the DOE points. Streamwise flow velocity at the sharp corner and friction factor were set as objective functions. The corner flow velocity increases as the rib is angled less toward the flow, at the penalty of an increase in friction factor. [Preview Abstract] |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2020 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
1 Research Road, Ridge, NY 11961-2701
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700