Experimental and Computational Investigations of Flow and Mixing in a Single-Annular Combustor Configuration

A complementary experimental and computational study of the flow and mixing in a single annular gas turbine combustor has been carried out. The object of the investigation is a generic mixing chamber model, representing an unfolded segment of a simplified Rich-Quick-Lean (RQL) combustion chamber operating under isothermal, non-reacting conditions at ambient pressure. Two configurations without and with secondary air injection were considered. To provide an appropriate reference database several planar optical measurement techniques (time-resolved flow visualisation, PIV, QLS) were used. The PIV measurements have been performed providing profiles of all velocity and Reynolds-stress components at selected locations within the combustor. Application of a two-layer hybrid LES/RANS (HLR) method coupling a near-wall k − ε RANS model with conventional LES in the core flow was the focus of the computational work. In addition to the direct comparison with the experimental results, the HLR performance is comparatively assessed with the results obtained by using conventional LES using the same (coarser) grid as HLR and two eddy-viscosity-based RANS models. The HLR model reproduced all important flow features, in particular with regard to the penetrating behaviour of the secondary air jets, their interaction with the swirled main flow, swirl-induced free recirculation zone evolution and associated precessing-vortex core phenomenon in good agreement with experimental findings.     

NUMERICAL MODELLING AND PARTICLE IMAGE VELOCIMETRY MEASUREMENT OF THE LAMINAR FLOW FIELD INDUCED BY AN ENCLOSED ROTATING DISC

The fluid flow field within an enclosed cylindrical chamber with a rotating flat disc was calculated using a finite volume computational fluid dynamics (CFD) model and compared with particle image velocimetry (PIV) measurements. Two particular laminar cases near the Transitional flow regime were investigated: Reynolds number Re=2.5×1 ⁴, chamber aspect ratio G (h/R_d)=0.2 and Re=4.2×10⁴, G (h/R_d)=0.217. This enabled direct comparison with the numerical and experimental results reported by other researchers. The computational details and some major factors that affect the computed accuracy and convergence speed are also discussed in detail. PIV results containing some 4300 velocity vector points in each of seven planes for each case were obtained from the flow field parallel to the rotating disc. It was found that PIV results could be obtained in planes within the boundary layers as well as the core flow by careful use of a thin laser illumination sheet and correct choice of laser pulse separation. There was close agreement between numerical results, the present PIV measurements and other reported experimental and numerical results.     

Stereo PIV measurements in an enclosed rotor–stator system with pre-swirled cooling air

In order to validate computational fluid dynamics (CFD) calculations, accurate velocity measurements were performed in a so called pre-swirl system. The objective was to determine the three-dimensional velocity field in the enclosed rotor–stator gap by using an adapted stereo particle image velocimetry (stereo PIV) setup. Particular attention was invested in the design of the optical access, thus, offering interesting possibilities to investigate various geometrical configurations of the pre-swirl system. The measurements impressively showed the spreading of the jet inside the wheelspace and the unsteady aspect of the flow, confirming that stereo PIV can successfully be applied in an enclosed rotor–stator system.     

Computational and experimental study of annular photo-reactor hydrodynamics

The performance of ultraviolet (UV) reactors used for water treatment is greatly influenced by the reactor hydrodynamics due to the non-homogeneity of the radiation field. Reliable modeling of the reactor flow structures is therefore crucial for the design process. In this study, the turbulent flow through two characteristic annular UV-reactor configurations, with inlets concentric (L-shape) and normal (U-shape) to the reactor axis, was investigated through computational fluid dynamic (CFD). The modeling results were evaluated with the velocity profiles from particle image velocimetry (PIV) experiments. The influence of mesh structure and density, as well as three turbulence models: Standard κ–, Realizable κ–, and Reynolds stress model (RSM), on the simulation results were evaluated. Mesh-independent solutions were achieved at mean cell volumes of 5 × 10⁻⁹ m³. The Realizable κ– displayed the best overall match to the experimental PIV measurements. In general, the CFD models showed a close agreement with the experimental data for the majority of the reactor domain and captured the influences of reactor configuration and internal reactor structures on the flow distribution. The validated CFD hydrodynamic models could be integrated with kinetic and radiation distribution models for UV-reactor performance simulation.     

Experimental evaluation of numerical simulation of cavitating flow around hydrofoil

Cavitation in hydraulic machines causes different problems that can be related to its unsteady nature. An experimental and numerical study of developed cavitating flow was performed. Until now simulations of cavitating flow were limited to the self developed “in house” CFD codes. The goal of the work was to experimentally evaluate the capabilities of a commercial CFD code (Fluent) for simulation of a developed cavitating flow. Two simple hydrofoils that feature some 3D effects of cavitation were used for the experiments. A relatively new technique where PIV method combined with LIF technique was used to experimentally determine the instantaneous and average velocity and void ratio fields (cavity shapes) around the hydrofoils. Distribution of static pressure on the hydrofoil surface was determined. For the numerical simulation of cavitating flow a bubble dynamics cavitation model was used to describe the generation and evaporation of vapour phase. An unsteady RANS 3D simulation was performed. Comparison between numerical and experimental results shows good correlation. The distribution and size of vapour structures and the velocity fields agree well. The distribution of pressure on the hydrofoil surface is correctly predicted. The numerically predicted shedding frequencies are in fair agreement with the experimental data.     

Planar flow field measurements in atmospheric and pressurized combustion chambers

The investigation of velocity fields in complex combustor flows is an important and necessary subject of propulsion technology. A persisting problem in computational fluid dynamics (CFD) modeling is that current numerical design tools have a number of deficiencies in accurately predicting the complex combustor flow. Using planar techniques such as planar Doppler velocimetry (PDV) or particle image velocimetry (PIV) it is possible to provide detailed information of the flow field inside the combustor. This paper reports on the applicability of PIV in combustor flows at realistic operating conditions.     

Application of an Explicit Algebraic Reynolds Stress Model within a Hybrid LES–RANS Method

Large-eddy simulations (LES) still suffer from extremely large resources required for the resolution of the near-wall region, especially for high-Re flows. That is the main motivation for setting up hybrid LES–RANS methods. Meanwhile a variety of different hybrid concepts were proposed mostly relying on linear eddy-viscosity models. In the present study a hybrid approach based on an explicit algebraic Reynolds stress model (EARSM) is suggested. The model is applied in the RANS mode with the aim of accounting for the Reynolds stress anisotropy emerging especially in the near-wall region. For the implementation into a CFD code this anisotropy-resolving closure can be formally expressed in terms of a non-linear eddy-viscosity model (NLEVM). Its extra computational effort is small, still requiring solely the solution of one additional transport equation for the turbulent kinetic energy. In addition to this EARSM approach, a linear eddy-viscosity model (LEVM) is used in order to verify and emphasize the advantages of the non-linear model. In the present formulation the predefinition of RANS and LES regions is avoided and a gradual transition between both methods is assured. A dynamic interface criterion is suggested which relies on the modeled turbulent kinetic energy and the wall distance and thus automatically accounts for the characteristic properties of the flow. Furthermore, an enhanced version guaranteeing a sharp interface is proposed. The interface behavior is thoroughly investigated and it is shown how the method reacts on dynamic variations of the flow field. Both model variants, i.e. LEVM and EARSM, have been tested on the basis of the standard plane channel flow and even more detailed on the flow over a periodic arrangement of hills using fine and coarse grids.     

The development of algebraic stress models using a novel evolutionary algorithm

This work presents developments to a novel evolutionary framework that symbolically regresses algebraic forms of the Reynolds stress anisotropy tensor. This work contributes to the growing trend in machine-learning for modelling physical phenomena. Our framework is shown to be computational inexpensive and produce accurate and robust models that are tangible mathematical expressions. This transparency in the result allows us to diagnose issues with the regressed formulae and appropriately make amendments, as we further understand the regression tools. Such models are created using hybrid RANS/LES flow field data and a passive solving of the RANS transport equations to obtain the modelled time scale. This process shows that models can be regressed from a qualitatively correct flow field and fully resolved DNS is not necessarily required. Models are trained and tested using rectangular ducts, an example flow genus that linear RANS models even qualitatively fail to predict correctly. A priori and a posteriori testing of the new models show that the framework is a viable methodology for RANS closure development. This a posteriori agenda includes testing on an asymmetric diffuser, for which the new models vastly outperform the baseline linear model. Therefore this study presents one of the most rigorous and complete CFD validation of machine learnt turbulent stress models to date.     

Prediction of losses in the flow through the last stage of low‐pressure steam turbine

In this paper, the numerical modelling of the flow through the low‐pressure steam turbine last stage was presented. On the basis of predicted wet steam flow‐field, the aerodynamic as well as thermodynamic losses were estimated. For calculations of the wet steam steady flow‐field three numerical methods were employed. The first method was a streamline curvature method (SCM). The commercial CFD code (CFX‐TACflow) and an in‐house code, both solving the 3‐D RANS equations, were the next two methods. In the wet steam region, by means of all three methods, the equilibrium flow was modelled. Additionally, the in‐house CFD code was used for modelling of the non‐equilibrium steam condensing flow. In this work, the comparison of the cascade loss coefficient for stator and rotor and selected flow parameters for the stage were presented, compared and discussed. Copyright © 2006 John Wiley & Sons, Ltd.     

Non-intrusive load characterization of an airfoil using PIV

An assessment is made of the feasibility of using PIV velocity data for the non-intrusive aerodynamic force characterization (lift, drag and pitching moment) of an airfoil. The method relies upon the application of control-volume approaches in combination with the deduction of the pressure from the PIV experimental data, by making use of the momentum equation. First, the consistency of the method is verified by means of synthetic data obtained from CFD. Subsequently, the procedure was applied in an experimental investigation, in which the PIV approach is validated against standard pressure-based methods (surface pressure distribution and wake rake).     

Large Eddy Simulation of Scalar Mixing in a Coaxial Confined Jet

The highly turbulent flow occurring inside gas-turbine combustors requires accurate simulation of scalar mixing if CFD methods are to be used with confidence in design. This has motivated the present paper, which describes the implementation of a passive scalar transport equation into an LES code, including assessment/testing of alternative discretisation schemes to avoid over/undershoots and excessive smoothing. Both second order accurate TVD and higher order accurate DRP schemes are assessed. The best performance is displayed by a DRP method, but this is only true on fine meshes; it produces similar (or larger) errors to a TVD scheme on coarser meshes, and the TVD approach has been retained for LES applications. The unsteady scalar mixing performance of the LES code is validated against published DNS data for a slightly heated channel flow. Excellent agreement between the current LES predictions and DNS data is obtained, for both velocity and scalar statistics. Finally, the developed methodology is applied to scalar transport in a confined co-axial jet mixing flow, for which experimental data are available. Agreement with statistically averaged fields for both velocity and scalar, is demonstrated to be very good, and a considerable improvement over the standard eddy viscosity RANS approach. Illustrations are presented of predicted time-resolved information e.g. time histories, and scalar pdf predictions. The LES results are shown, even using a simple Smagorinsky SGS model, to predict (correctly) lower values of the turbulent Prandtl number in the free shear regions of the flow, compared to higher values in the wall-affected regions. The ability to predict turbulent Prandtl number variations (rather than input these as in combustor RANS CFD models) is an important and promising feature of the LES approach for combustor flow simulation since it is known to be important in determining combustor exit temperature traverse.     

Assessment of 2D/3D numerical modeling for deep dynamic stall experiments

The results of computational fluid dynamics (CFD) simulations in two and three spatial dimensions are compared to pressure measurements and particle image velocimetry (PIV) flow surveys to assess the suitability of numerical models for the simulation of deep dynamic stall experiments carried out on a pitching NACA 23012 airfoil. A sinusoidal pitching motion with a 10° amplitude and a reduced frequency of 0.1 is imposed around two different mean angles of attack of 10° and 15°. The comparison of the airloads curves and of the pressure distribution over the airfoil surface shows that a three-dimensional numerical model can better reproduce the flow structures and the airfoil performance for the deep dynamic stall regime. Also, the vortical structures observed by PIV in the flow field are better captured by the three-dimensional model. This feature highlighted the relevance of three-dimensional effects on the flow field in deep dynamic stall.     

Cellular-level near-wall unsteadiness of high-hematocrit erythrocyte flow using confocal μPIV

In hemodynamics, the inherent intermittency of two-phase cellular-level flow has received little attention. Unsteadiness is reported and quantified for the first time in the literature using a combination of fluorescent dye labeling, time-resolved scanning confocal microscopy, and micro-particle image velocimetry (μPIV). The near-wall red blood cell (RBC) motion of physiologic high-hematocrit blood in a rectangular microchannel was investigated under pressure-driven flow. Intermittent flow was associated with (1) the stretching of RBCs as they passed through RBC clusters with twisting motions; (2) external flow through local obstacles; and (3) transitionary rouleaux formations. Velocity profiles are presented for these cases. Unsteady flow clustered in local regions. Extra-cellular fluid flow generated by individual RBCs was examined using submicron fluorescent microspheres. The capabilities of confocal μPIV post-processing were verified using synthetic raw PIV data for validation. Cellular interactions and oscillating velocity profiles are presented, and 3D data are made available for computational model validation.     

Prediction of unsteady,separated boundary layer over a blunt body for laminar,turbulent, and transitional flow

The focus of this paper is to study the ability of unsteady RANS‐based CFD to predict separation over a blunt body for a wide range of Reynolds numbers particularly the ability to capture laminar‐to‐turbulent transition. A perfect test case to demonstrate this point is the cylinder‐in‐crossflow for which a comparison between experimental results from the open literature and a series of unsteady simulations is made. Reynolds number based on cylinder diameter is varied from 10⁴ to 10⁷ (subcritical through supercritical flow). Two methods are used to account for the turbulence in the simulations: currently available eddy–viscosity models, including standard and realizable forms of the k–ε model; and a newly developed eddy–viscosity model capable of resolving boundary layer transition, which is absolutely necessary for the type and range of flow under consideration. The new model does not require user input or ‘empirical’ fixes to force transition. For the first time in the open literature, three distinct flow regimes and the drag crisis due to the downstream shift of the separation point are predicted using an eddy–viscosity based model with transition effects. Discrepancies between experimental and computational results are discussed, and difficulties for CFD prediction are highlighted. Copyright © 2004 John Wiley & Sons, Ltd.     

Modeling of Internal Reacting Flows and External Static Stall Flows Using RANS and PRNS

The reacting flow in a research lean direct injection (LDI) hydrogen combustor and the static stall of NACA0012 airfoil were simulated using both Reynolds averaged Navier–Stokes (RANS) and partially resolved numerical simulation (PRNS) approaches. The concept and the main features of the PRNS approach are briefly described. The PRNS basic equations are grid independent or grid invariant; the subscale models are a dynamic equation system. We consider PRNS as an engineering tool for the very large eddy simulation of complex turbulent flows. Two CFD codes, NCC and Wind-US, with two different subscale models (i.e. two- and one-transport equation models, respectively) are used in the presented PRNS simulations. Based on the comparisons with available experimental data, the numerical results indicate that the PRNS subscale models seem to be able to capture important large scale turbulent structures and to improve the quality of numerical simulations while keeping a relatively low cost comparable to the unsteady RANS simulations.     

An embedded boundary framework for compressible turbulent flow and fluid–structure computations on structured and unstructured grids

The finite volume method with exact two‐phase Riemann problems (FIVER) is a two‐faceted computational method for compressible multi‐material (fluid–fluid, fluid–structure, and multi‐fluid–structure) problems characterized by large density jumps, and/or highly nonlinear structural motions and deformations. For compressible multi‐phase flow problems, FIVER is a Godunov‐type discretization scheme characterized by the construction and solution at the material interfaces of local, exact, two‐phase Riemann problems. For compressible fluid–structure interaction (FSI) problems, it is an embedded boundary method for computational fluid dynamics (CFD) capable of handling large structural deformations and topological changes. Originally developed for inviscid multi‐material computations on nonbody‐fitted structured and unstructured grids, FIVER is extended in this paper to laminar and turbulent viscous flow and FSI problems. To this effect, it is equipped with carefully designed extrapolation schemes for populating the ghost fluid values needed for the construction, in the vicinity of the fluid–structure interface, of second‐order spatial approximations of the viscous fluxes and source terms associated with Reynolds averaged Navier–Stokes (RANS)‐based turbulence models and large eddy simulation (LES). Two support algorithms, which pertain to the application of any embedded boundary method for CFD to the robust, accurate, and fast solution of FSI problems, are also presented in this paper. The first one focuses on the fast computation of the time‐dependent distance to the wall because it is required by many RANS‐based turbulence models. The second algorithm addresses the robust and accurate computation of the flow‐induced forces and moments on embedded discrete surfaces, and their finite element representations when these surfaces are flexible. Equipped with these two auxiliary algorithms, the extension of FIVER to viscous flow and FSI problems is first verified with the LES of a turbulent flow past an immobile prolate spheroid, and the computation of a series of unsteady laminar flows past two counter‐rotating cylinders. Then, its potential for the solution of complex, turbulent, and flexible FSI problems is also demonstrated with the simulation, using the Spalart–Allmaras turbulence model, of the vertical tail buffeting of an F/A‐18 aircraft configuration and the comparison of the obtained numerical results with flight test data. Copyright © 2014 John Wiley & Sons, Ltd.     

A study of the flow field characteristics around star-shaped artificial reefs

In this paper, a three-dimensional numerical model is devised to calculate the unsteady flow field around star-shaped artificial reefs. The model is based on Reynolds-averaged Navier–Stokes (RANS) equations embedded within a renormalization group (RNG) k–ε turbulence model. The RANS equations are solved using the finite volume method (FVM) with an unstructured tetrahedral mesh. The pressure and velocity coupling is solved at each time step with the SIMPLEC algorithm. Non-invasive particle image velocimetry (PIV) laboratory measurements are employed to verify the simulation results. Good agreement is found between the simulation and experimental results with respect to the major flow fields. Based on the flow-field verification, the influence of arrangement and spacing on the flow field of one and two artificial reefs are discussed in light of the numerical method. A large-scale slow flow region is obtained when the reef is arranged in the second form. In the parallel combination, a slight mutual effect exists between the two reefs when the spacing is larger than 3.0L. In the streamwise combination, the interaction of two reefs is at its strongest at spacings of 3.0L to 4.0L.     

Dissipative particle dynamics for complex geometries using non‐orthogonal transformation

Dissipative particle dynamics (DPD) was applied to fluid flow in irregular geometries using non‐orthogonal transformation, where an irregular domain is transformed into a simple rectangular domain. Transformation for position and velocity was used to relate the physical and computational domains. This approach was described by simulating fluid flow inside a two‐dimensional convergent–divergent nozzle. The nozzle geometry is controlled by the contraction ratio (CR) in the middle of the channel. The range of Reynolds number and CR, in this paper, was Re = 10hbox??200 and CR = 0.8 and 0.6, respectively. The DPD results were validated against in‐house computational fluid dynamic (CFD) finite volume code based on the stream function vorticity approach. The results revealed an excellent agreement between DPD and CFD. The maximum deviation between the DPD and CFD results was within 2%. Local and average coefficients of friction was calculated and it compared well with the CFD results. Copyright © 2011 John Wiley & Sons, Ltd.