Elizaveta Ivanova

A Numerical Study on the Turbulent Schmidt Numbers in a Jet in Crossflow

This work presents a numerical study on the turbulent Schmidt numbers in jets in crossflow. This study contains two main parts. In the first part, the problem of the proper choice of the turbulent Schmidt number in the Reynolds-averaged Navier-Stokes (RANS) jet in crossflow mixing simulations is outlined. The results of RANS employing the shear-stress transport (SST) model of Menter and its curvature correction modification and different turbulent Schmidt number values are validated against experimental data. The dependence of the optimal value of the turbulent Schmidt number on the dynamic RANS model is studied. Furthermore, a comparison is made with the large-eddy simulation (LES) results obtained using the wall-adapted local eddy viscosity (WALE) model. The accuracy given by LES is superior in comparison to RANS results. This leads to the second part of the current study, in which the time-averaged mean and fluctuating velocity and scalar fields from LES are used for the evaluation of the turbulent viscosities, turbulent scalar diffusivities, and the turbulent Schmidt numbers in a jet in crossflow configuration. The values obtained from the LES data are compared with those given by the RANS modeling. The deviations are discussed, and the possible ways for the RANS model improvements are outlined.

Impact of Fischer-Tropsch fuels on aero-engine combustion performance

In this paper a detailed numerical investigation of an aero-engine single sector combustor model is carried out. The main goal of this study is to compare its combustion characteristics when fueled with a single component surrogate Jet A-1 fuel to the ones when fueled with a multicomponent Jet A-1 as well as with multicomponent Fischer-Tropsch (F-T) fuels under the same operating conditions. The combustor geometry under investigation mirrors at smaller scale many salient features of modern aero-engine combustors such as high swirling flow, eusion cooling at the walls, head cooling flow and, dilution flow. The numerical simulations are performed by means of an Eulerian-Lagrangian method for the gas and liquid phase, respectively. A finite-rate chemistry combustion model, which explicitly takes all species into account and an assumed-PDF approach for the turbulence-chemistry interaction are used to model the turbulent reacting flow. A multicomponent-fuel droplet evaporation model is implemented and enables to study real-fuel eects on the overall combustion process. A former numerical study had shown good results when comparing the reacting velocity and temperature fields and spray pattern for Jet A-1 combustion with LDA, CARS and Mie scattering measurements respectively. In the present study, a fully synthetic jet fuel which contains 50 % Coal-to-Liquid (CtL) fuel and a 100 % Gas-to-Liquid (GtL) fuel, both synthesized from a Fischer-Tropsch process are substituted to Jet A-1. The simulations are repeated under the same nominal operating conditions. A comparison in terms of temperature distribution, OH radical concentration and, main reactants concentration is carried out in order to outline the influence of the fuel composition on the combustor performance. Additionally, an analysis of certain soot precursors concentration and spray evaporation characteritics is delivered.

Numerical Simulations of Turbulent Mixing and Autoignition of Hydrogen Fuel at Reheat Combustor Operating Conditions

Turbulent mixing and autoignition of H2 -rich fuels at relevant reheat combustor operating conditions are investigated in the present numerical study. The flow configuration under consideration is a fuel jet perpendicularly injected into a crossflow of hot flue gas (T > 1000 K, p = 15 bar). Based on the results of the experimental study for the same flow configuration and operating conditions two different fuel blends are chosen for the numerical simulations. The first fuel blend is a H2 /natural gas/N2 mixture at which no autoignition events were observed in the experiments. The second fuel blend is a H2 /N2 mixture at which autoignition in the mixing section occurred. First, the non-reacting flow simulations are performed for the H2 /natural gas/N2 mixture in order to compare the accuracy of different turbulence modeling methods. Here the steady-state Reynolds-averaged Navier-Stokes (RANS) as well as the unsteady scale-adaptive simulation (SAS) turbulence modeling methods are applied. The velocity fields obtained in both simulations are directly validated against experimental data. The SAS method shows better agreement with the experimental results. In the second part of the present work the autoignition of the H2 /N2 mixture is numerically studied using the 9-species 21-steps reaction mechanism of O’Conaire et al. [1]. As in the reference experiments, autoignition can be observed in the simulations. Influences of the turbulence modeling as well as of the hot flue gas temperature are investigated. The onset and the propagation of the ignition kernels are studied based on the SAS modeling results. The obtained numerical results are discussed and compared with data from experimental autoignition studies.Copyright

Unsteady Simulations of Turbulent Mixing in Jet in Crossflow

This paper presents numerical simulations of turbulent mixing of a jet in crossflow. The testcase is chosen to resemble the scalar mixing processes in the premixing zones of gas turbine combustion chambers. Unsteady simulations employing two different computational approaches are presented: unsteady Reynolds-Averaged Navier-Stokes (URANS) and Large-Eddy Simulations (LES). Presented results comprise the (time-averaged) profiles of flow velocities, the turbulent kinetic energy of the flow, Reynolds stresses, passive scalar distribution, turbulent scalar fluxes, and the turbulent scalar variance. All presented results are directly validated against experimental data. A parametric study related to the dependence of the scalar mixing predictions on the turbulent Schmidt number value for both URANS and LES is presented. The effective values of the turbulent scalar diffusivity predicted by both simulation methods are evaluated.

Unsteady Simulations of Flow Field and Scalar Mixing in Transverse Jets

This paper presents numerical simulations of flow and scalar mixing in two different jet in crossflow configurations. The testcases are chosen to resemble the dilution mixing processes in gas turbine combustion chambers. Unsteady simulations employing two different computational approaches are presented: unsteady Reynolds-Averaged Navier-Stokes (URANS) and Scale-Adaptive Simulations (SAS). The results obtained by each method are compared, analyzed, and validated against experimental data. The importance of the reproduction of the large-scale unsteady coherent vortical structures in the numerical simulation is demonstrated. Both URANS and SAS revealed the typical jet in crossflow vortical structures. The SAS method was able to resolve smaller structures than URANS on the same computational grid. The quantitative prediction accuracy of time-averaged velocities and temperatures is satisfactory for both methods. In contrast, the steady-state Reynolds-Averaged Navier-Stokes (RANS) computations failed for the present testcases.Copyright

Improvement and Assessment of RANS Scalar Transport Models for Jets in Crossflow

The present work proposes a new form of two-equation model for turbulent scalar transport. The aim is a better prediction of turbulent scalar mixing by the Reynolds-averaged Navier-Stokes (RANS) methods. The presented model allows a direct determination of the local turbulent scalar diffusivity avoiding the assumption of a constant global turbulent Prandtl/Schmidt number. The main difference between the present model and the models of this type published earlier is an improved form of the scalar dissipation equation. Comparisons are made with three other two-equation turbulent scalar diffusivity models in different jet in crossflow test cases. A sensitivity study with respect to model parameter variations is also given. The results show significant improvements in scalar field predictions for jet in crossflow test cases.

LES-based evaluation of the turbulent Schmidt numbers for confined coaxial jets

This paper is devoted to the challenges arising in Reynolds-Averaged Navier-Stokes (RANS) modeling of confined coaxial jet mixing. The main drawbacks of the gradient diffusion hypothesis and of the constant turbulent Schmidt number approach for the considered flow type are elucidated. In the first part of the present work, the mean and fluctuating velocity and scalar fields obtained in RANS simulations are directly validated against experimental data and compared with large-eddy simulation (LES) results. Three different turbulence models and three different turbulent Schmidt numbers values are selected for the present RANS computations. The accuracy given by LES is superior in comparison to RANS results especially for scalar mixing. The inability of the chosen RANS approach to represent counter-gradient diffusion in the axial direction characteristic for the considered test case is discussed. In the second part of the present study, the time-averaged mean and fluctuating velocity and scalar fields from LES are used for the evaluation of turbulent viscosities, turbulent scalar diffusivities, and turbulent Schmidt numbers as they appear in the RANS equations. Specific problems of the evaluation of these parameters from LES data are addressed. The importance of accounting for the turbulent scalar fluxes in all coordinate directions for evaluation of turbulent scalar diffusivity is revealed. The values obtained from LES results are compared with those given by RANS modeling. The deviations and their consequences for RANS predictions are discussed.

Computational Modelling of Turbulent Mixing of a Transverse Jet

This paper presents numerical simulations of turbulent mixing of a jet in crossflow. The test case is chosen to resemble scalar mixing processes in the premixing zones of gas turbine combustion chambers. Steady and unsteady simulations employing three different computational approaches are presented: steady Reynolds-Averaged Navier-Stokes (RANS), unsteady Reynolds-Averaged Navier-Stokes (URANS), and Scale-Adaptive Simulations (SAS). Presented results comprise the (time-averaged) profiles of flow velocities, turbulent kinetic energy of the flow, Reynolds stresses, passive scalar distribution, turbulent scalar fluxes, and the turbulent variance of the passive scalar. All presented results are directly validated against experimental data. Additionally two parameter studies are presented. Both studies are related to the accuracy of the turbulent scalar mixing predictions for all used simulation methods. In the first study the dependence of the scalar mixing predictions on the value of the turbulent Schmidt number is considered. In the second study the dependence of the predicted turbulent scalar variance on the used modelling approach is analysed.Copyright