he primary objective of this study is to validate and/or identify issues for available numerical methods and turbulence models in OpenFOAM 2.0.0. Such a study will provide a guideline for users, will aid acceptance of OpenFOAM as one of the research solvers at institutions and also guide future multidisciplinary research using OpenFOAM. In addition, a problem of aerospace interest such as the flow features and vortex breakdown around a VFE-II model is obtained for SA, SST RANS and SA-DDES models and compared with DLR experiment. The available numerical methods such as time schemes, convection schemes, P-V couplings and turbulence models are tested as available for a fundamental case of a backward facing step for RANS and Hybrid RANSLES prediction of fully turbulent flow at a Reynolds number of 32000 and the OpenFOAM predictions are validated against experimental data by Driver et.al and compared with Fluent predictions.

In hybrid RANS/LES, Reynolds-averaged Navier-Stokes (RANS) equations method is used to treat the near-wall region and large-eddy simulation (LES) is applied to the core turbulent region. Owing to the incompatibility of these two numerical modelling approaches, an artificial (i.e., non-physical) buffer layer forms along the interface where the model switches from RANS to LES. In this thesis, a novel dynamic forcing scheme incorporating backscatter is proposed in order to remove the artificial buffer layer. In contrast to previous forcing techniques, the proposed forcing is determined dynamically from the flow field itself, and does not require any extraction of turbulent fields from reference direct numerical simulation (DNS) or high-resolution LES databases. The proposed forcing model has been tested on three types of wall-bounded turbulent flows, namely, turbulent flow in a plane channel; turbulent flow in a spanwise rotating channel; and turbulent flow in a spanwise rotating rib-roughened channel. In order to validate the present hybrid approach, turbulence statistics obtained from hybrid RANS/LES simulations are thoroughly compared with the available DNS results and laboratory measurement data. Based on the study of a plane channel flow, transport equations for the resolved turbulent stresses and kinetic energy are introduced to investigate the effects of dynamic forcing on reduction of the thickness and impact of the artificial buffer layer. As long as the dynamic forcing is in use, the artificial buffer layer have been successfully removed, indicating that the proposed hybrid approach is insensitive to the choices of the forcing region or interface location. The predictive performance of the dynamic forcing scheme has been further evaluated by considering turbulent flows subjected to a special type of body force, i.e., the non-inertial and non-conservative Coriolis force. Due to the effects of system rotation, turbulence level is enhanced on the pressure side and suppressed on the suction side of the rotating channel. Furthermore, it is reported in this thesis that the classification of the roughness type now relies not only on the pitch ratio, but also on the rotation number in the context of rotating rib-roughened flows. / February 2016

In order to increase results accuracy and to provide some time-dependency to CFD results, embedded RANS/LES simulations are getting more and more interesting. In order to run these simulations accurate LES boundary conditions are required, not to affect the downstream results with a poor quality synthetic turbulence generation. Considering the currently developped methodologies, it is not possible to generate a divergence free turbulent flow which satisfy a non isotropic state of turbulence. The author started from the Synthetic Eddy Method (SEM) defined by Jarrin (2009), and defined a new shape function with the ability to satisfy continuity. The new methodology, named Divergence Free SEM (DFSEM), is able to reproduce almost any kind of turbulence anisotropy by using a special shape function and adapting the eddies intensities in order to match the Reynolds stress tensor rather than using the Lund coefficients, as most of the precursor methodologies did. Results comparisons against SEM and some other very popular synthetic turbulence models in some academic cases, proved that a reduce influence on the downstream flow was achieved. In most of the cases the friction coefficient Cf , used as a performance parameter, benefit by reducing the downstream developping zone by almost 50% in most cases, when compared against SEM. Another issue that has been tackled regards the unphysical pressure fluctuations present because of the synthetic turbulence, due to non perfectly constant mass-flow rate imposed in stochastic methodologies. The new methodology also showed an increased flexibility as it has been tested in embedded DDES simulation, by using the blending function to activate/deactivate it, and again it showed improved performances when compared against standard SEM.

Computational investigations were carried out to evaluate the effectiveness and usability of hybrid RANS/LES techniques for predicting the unsteady separated flow over wings with ice accretion. RANS and hybrid RANS/LES computations were performed using the viscous flow solver CHEM with the SST turbulence model. Two configurations were considered during the study: an extruded wing with a glaze-ice shape and a swept wing with a simulated glaze-ice accretion. Hybrid RANS/LES results, in general, predict a less active shear layer ``roll up' than seen in the experimental data. Qualitative improvements are seen in the hybrid RANS/LES results over corresponding RANS results. The extruded wing results show that the CHEM hybrid RANS/LES results are similar to the AVUS DES results. The use of preconditioning and a different turbulent model in CHEM showed a slight improvement in results.

Ce travail porte sur l’évaluation de la méthode ZDES pour la simulation de jets propulsifs. L’analyse se concentre sur le positionnement des cellules de chocs et le développement des couches de mélange d’une tuyère double-flux avec plug externe, typique des moteurs d’avions modernes. Les champs statistiques sont comparés aux résultats expérimentaux et discutés en termes de grandeurs moyennes, fluctuantes et dans le domaine fréquentiel. L’intérêt d’utiliser un schéma spatial peu dissipatif ainsi qu’une échelle de longueur sous-maille basée sur la vorticité locale est mis en évidence, notamment pour le dévelopement de la couche de mélange interne, et le mode 2 ("automatique") de la ZDES a démontré un comportement similaire au mode 1 ("manuel") dans les couches de mélange. Par ailleurs, la technique Random Flow Generation (RFG) mise en oeuvre afin de reproduire la turbulence amont existant au coeur des jets primaire et secondaire a permis d’accélérer la transition RANS-LES dans les deux couches de mélanges, plus conformément à l’expérience. La transition est d’autant plus rapide que le taux de turbulence est élevé et l’échelle de la turbulence injectée est petite. Le positionnement des cellules de choc est également amélioré, soulignant l’importance de prendre en compte la turbulence amont dans les simulations de jets. / In this thesis, the ZDES method is assessed for the simulation of propulsive jets. This work focuses on the shock-cell positioning and the mixing layer development of a dual-stream nozzle configuration with an external plug, typical of modern aircraft engines. Reynolds averaged data are discussed in terms of mean and fluctuating quantities as well as in the frequency domain and compared with experimental data. First, the advantage of using a low dissipative spatial scheme as well as a subgrid length scale based on the local vorticity is demonstrated, especially for the development of the core mixing layer. Besides, the "automatic" mode of ZDES (mode 2) is found to provide similar mixing layers as the user defined mode.Then, the use of the Random Flow Generation (RFG) technique at the inlet boundaries of the core and fan channels in order to reproduce the turbulence rate at the center of the nozzle ducts is shown to accelerate the RANS-to-LES transition in both external and internal mixing layers, which is in better agreement with the experimental results. The transition length is further reduced when the injected turbulent ratio is higher, but also when the injected turbulent length scale is smaller. Of interest, the shock-cell positioning in the fan jet is also improved using RFG, which emphasizes the importance of accounting for upstream turbulence for this type of simulations.

No description available.

Computational fluid dynamics (CFD) has been widely applied to simulate turbulent flows in an urban environment. The two basic methodologies in CFD that have been applied here are a Reynolds-averaged Navier-Stokes (RANS) modeling and a large-eddy simulation (LES). The nature of the flow in a built-up urban area consisting of an arbitrary aggregation of buildings is dominated by unsteady large-scale turbulent structures. Recognizing that RANS is unable to correctly capture these turbulent structures while LES is associated with high computational costs, a hybrid RANS/LES methodology that combines the computational efficiency of RANS with the predictive accuracy of LES can be a promising simulation approach for the application to urban flows. In the non-zonal approach of hybrid RANS/LES methodology, a single generalized turbulence model is used in the entire computational domain. This model can function as a RANS turbulence closure model or as a LES subgrid scale model, depending on the local grid resolution or flow properties. A variant of non-zonal approaches, referred as partially resolved numerical simulation (PRNS) in this study, obtains the generalized turbulence model from the rescaling of a conventional RANS model through the incorporation of a resolution control function (F_R). The resolution control function F_R is used to characterize the degree of modeling required to represent the unresolved scales of motion. A new generalized functional form for F_R in PRNS is proposed in this thesis. The predictive performance of PRNS is compared with unsteady RANS (URANS) and LES computations, for a plane channel flow, and for fully-developed and developing flows over a matrix of cubes resembling a group of buildings. It is demonstrated that PRNS behaves similarly to LES, in terms of the predictions of the mean flow and turbulence, but outperforms URANS in general. This indicates PRNS is a promising approach for the simulation of complex turbulent flows in an urban environment.

Computational fluid dynamics (CFD) has been widely applied to simulate turbulent flows in an urban environment. The two basic methodologies in CFD that have been applied here are a Reynolds-averaged Navier-Stokes (RANS) modeling and a large-eddy simulation (LES). The nature of the flow in a built-up urban area consisting of an arbitrary aggregation of buildings is dominated by unsteady large-scale turbulent structures. Recognizing that RANS is unable to correctly capture these turbulent structures while LES is associated with high computational costs, a hybrid RANS/LES methodology that combines the computational efficiency of RANS with the predictive accuracy of LES can be a promising simulation approach for the application to urban flows. In the non-zonal approach of hybrid RANS/LES methodology, a single generalized turbulence model is used in the entire computational domain. This model can function as a RANS turbulence closure model or as a LES subgrid scale model, depending on the local grid resolution or flow properties. A variant of non-zonal approaches, referred as partially resolved numerical simulation (PRNS) in this study, obtains the generalized turbulence model from the rescaling of a conventional RANS model through the incorporation of a resolution control function (F_R). The resolution control function F_R is used to characterize the degree of modeling required to represent the unresolved scales of motion. A new generalized functional form for F_R in PRNS is proposed in this thesis. The predictive performance of PRNS is compared with unsteady RANS (URANS) and LES computations, for a plane channel flow, and for fully-developed and developing flows over a matrix of cubes resembling a group of buildings. It is demonstrated that PRNS behaves similarly to LES, in terms of the predictions of the mean flow and turbulence, but outperforms URANS in general. This indicates PRNS is a promising approach for the simulation of complex turbulent flows in an urban environment.

The work presented in this dissertation concerns continued development, validation and verification of the partially averaged Navier-Stokes (PANS) method - a variable resolution closure model for turbulence. Linear eddy viscosity models (LEVM), which are popular because of their simplicity and affordability in terms of computational cost have fundamental deficiencies and cannot be trusted to accurately represent turbulence in realistic complex flows. The more high fidelity approaches such as large eddy simulations (LES) and direct numerical simulations (DNS) are out of realm of engineering applicability because of their high requirements in computing power. PANS, a variable resolution approach considered in this study, lies between LEVM and LES in terms of computational cost and is designed to prudently utilize the available computing power to improve accuracy. This dissertation presents the various studies performed to characterize the PANS fluctuations and extend the model for use in various wall bounded flows. The road map towards our goal includes: (i) Comparing a-priori and a-posteriori eddy viscosity values to establish whether PANS is capable of producing the pre-specified level of reduction. (ii) Investigating the scaling of PANS fluctuations for different levels of prescribed resolution and establishing if the fluctuations abide by known turbulence scaling laws. (iii) Extending PANS to k-w formulation which is better suited for wall-bounded shear flows, and (iv) Modifying the present LEVM to yield reasonable behavior in the rapid distortion limit where the turbulence is elastic in nature which ultimately affects PANS performance. Results reported in this dissertation illustrate that the PANS closure yields reliable and predictable reduction in the modeled viscosity. The accuracy of the simulations improve as the effective damping is reduced by lowering the specified viscosity providing credibility to the PANS method as a bridging model that performs as intended.

Thermal stripping is a major safety challenge in nuclear power generation and propulsion systems. It arises as a consequence of the heat transfer from fluid to surrounding solid components varying in time and typically occurs in regions where the mixing of hot and cold fluids results in turbulent temperature fluctuations. It can occur in a range of components in reactors and thermal-hydraulics systems and may lead to structural failure by high-cycle thermal fatigue. Cases of cooling system pipes failing by this mechanism have been reported at the French Civaux and the Japanese Tsuruga-2 & Tomari-2 pressurised water reactor plants. CFD has great potential to provide predictions for flow fields in the pipe bends and junctions of nuclear plant thermal-hydraulics systems. The current project aims to use CFD to explore the physics of thermal mixing in plant components, and to develop \& validate CFD techniques for studying such problems in industry. Firstly, wall-resolved LES is used to demonstrate the importance of including nearby upstream pipe bends in CFD studies of thermal mixing in T-junctions. Swirl-switching of the Dean vortices generated at an upstream bend can give rise to an unsteady secondary flow about the pipe axis. This provides an additional mechanism for low-frequency near-wall temperature fluctuations downstream of the T-junction, over those that would be produced by mixing in the same T-junction with straight inlets. Wall-resolved LES is however currently computationally unaffordable for studying plant components in industry. Wall-functions offer a solution to this problem by imposing empirical results near walls, such that a coarser grid can be used. LES with blended wall-function predictions for flows in a 90 degree pipe bend and a simple T-junction with straight inlets are compared to experimental data. These studies highlight limitations in the predictive capabilities of the LES with wall-function approach. Predictions from a number of RANS models are also benchmarked. Finally, the consistent dual-mesh hybrid LES/RANS framework proposed by Xiao and Jenny (2012) is further developed as an alternative solution to the high computational cost of wall-resolved LES. Numerous modifications to the coupling between the two meshes are presented, which improve automation and accuracy. The approach is also extended to a passive temperature scalar field. Predictions for channel flows, a flow through periodic hills and thermal mixing in a T-junction between channel flows are all in excellent agreement with reference data.