Session G29: CFD: LES and RANS

Not Participating

One of the methods used for the production of multiphase flow is the Flash NanoPrecipitation achievable by the usage of macro-size Multi-Inlet Vortex Reactor (MIVR). Previous studies showed that steady-state solvers such as the standard Reynolds Averaged Navier-Stokes (RANS) weren’t the most appropriate solvers for the simulation of a MIVR flow as it wasn’t capturing many highly temporal-fluctuating quantities. For example, the time-fluctuating formation of small vortices through the reactors flow [1] or even the wandering motion of the center of the core vortex in the mixing chamber [2] indicates that transient solvers can be promising for a more accurate simulation of the swirling flow and turbulent mixing taking place in a macro-MIVR. The study presented here take a comparative look on the solving of two different inlet Re flows in a macro-MIVR using a transient solver. This transient solver will essentially solve the cases using 4 different turbulence models: the SSG (Speziale Sarkar and Gatski), standard k-ε, LRR (Launder, Reece and Rodi) and SST k-ω models. The comparison of the results will be made by the analysis of the statistical information concerning the distribution of the passive scalar and the turbulent mixing provided by the probability density function (PDF) [3]. The comparison will also be made on the basis of experimental unconditional mean velocity extracted by using stereo-PIV and PLIF.

[1]E. Hitimana, R. O. Fox and J. C. Hill, Coherent structure characteristics of the swirling flow during turbulent mixing in a multi-inlet vortex reactor, Physics of Fluid 33, 065119 , 2021.

[2]Z. Liu, M. Ramezani, R. O. Fox, J. C. Hill and M. G. Olsen, Flow Characteristics in a Scaled-up Multi-inlet Vortex Nanoprecipitation Reactor, I&EC Research 54, pp. 4512-4525, 2015.

[3]E. Madadi-Kandjani, R. O. Fox and A. Passalacqua, Application of the Fokker-Planck molecular mixing model to turbulent scalar mixing using moment methods, Physics of fluids 29, 065109, 2017.   

Performance of Reynolds Averaged Navier Stokes Models for Unsteady Separated flows

Unsteady separation is a phenomenon that occurs in many flows due to time-varying adverse pressure gradients, and results in increased drag, decreased lift, and loss of efficiency or failure in flow devices. Therefore, it is important to predict and analyze unsteady separation. RANS models are commonly used in the industrial design process due to their low computational cost; however, their performance in predicting separations is unsatisfactory. Our goal is to use high-fidelity Large-Eddy Simulation results carried out by Ambrogi et al. [J. Fluid Mech. 245, A10, 2022] to validate RANS models for predicting unsteady separation. By using an identical grid, numerical scheme and consistent boundary conditions to the LES calculations we are able to isolate modelling errors and modify terms accordingly. The Spalart-Allmaras eddy viscosity model predicts the unsteady behaviour of the flow with fair accuracy; the location of the separation region agrees with the LES results, hysteresis is captured and the mean velocity profiles match that of the LES results well. Calculations using the k-ε and k-Ω models will also be discussed.   

Improving RANS Modeling of Heat Transfer in Hypersonic Boundary Layers Using Learning and Inference Assisted by Feature-space Engineering

To address the stagnation in the accuracy of RANS models, data-driven model augmentation frameworks have been developed over the past few years with varying levels of success. This work presents an augmentation procedure which enforces consistency between the learning and prediction environments by simultaneously inferring and learning the model discrepancy during the training process. The approach, termed Learning and Inference assisted by Feature-space Engineering (LIFE), emphasizes a careful introduction of the augmentation, a meticulous design of the features and feature space, and the novel notion of localized learning, improving the generalizability and robustness of the augmentation. This work applies the LIFE framework to augment the Wilcox-2006 k-ω turbulence model to improve heat transfer predictions for hypersonic boundary layers. The structural form of a transport equation related to the Turbulent Prandtl number is inferred and learned. The generalization capabilities of the present approach are evaluated, and the impact of several modeling choices is examined.   

Comparison of RANS, LES, and wind tunnel experiments for the calculation of wind loads on a low-rise building in its urban environment.

The majority of buildings are low-rise and their wind-resistant design plays an important role to reduce losses due to extreme wind events. Computational fluid dynamics is an attractive approach to study wind loads and offers high-resolution data output, but validation is required to assess its performance. The low-fidelity Reynolds-Averaged Navier-Stokes simulations offer fast calculations of the mean quantities and require additional empirical relationships to calculate turbulent statistics, such as the root-mean-square (rms) pressure coefficients (C_p). In contrast, the high-fidelity Large-eddy Simulations (LES) solve for the instantaneous fields allowing direct estimation of turbulent statistics. This study compares wind tunnel tests, RANS, and LES of a low-rise building in terms of mean, fluctuating, and peak C_p. The analysis considers 1) the isolated building and 2) the building in its urban environment, to evaluate the interference effects for different wind directions. Preliminary results for the most dominant wind direction show good agreement for the mean C_p, but RANS fails to predict rms C_p in regions of flow separation and recirculation of vortices at the building wakes. Future work will extend the comparison to turbulent statistics for all wind directions.   

Computational study of Flow Past an Axially Aligned Spinning Cylinder

Computational investigations are performed for flow past a spinning cylinder aligned along its axis. In this study computational findings from both RANS and LES investigations are reported and compared with available experimental results. The RANS model is based on modified energy-spectrum to incorporate the effects of swirl and rotation using a parametric characterization of the model coefficients. An efficient finite-volume algorithm is used to solve the time-averaged equations of motion along with the modeled form of transport equations for the turbulence kinetic energy and the scalar form of turbulence dissipation. The LES approach is based on a continuous turbulence model that is suitable for representing boththe subgrid scale stresses in large eddy simulation and the Reynolds stresses in theReynolds averaged Navier-Stokes formulation. Experimental data for a range of spin rates and free stream flow conditions have been obtained from a subsonic wind tunnel for flow past an axially aligned spinning cylinder.