Session: 12-08-02: Computational Heat Transfer and Applications II

Paper Number: 166359

Large-Eddy Simulation of Non-Premixed Turbulent Diffusion Flame Using High-Order Finite Difference Schemes 

The fundamentals of the physicochemical process in turbulent combustion may be investigated through direct numerical simulations (DNS); the applicability of which, however, can be limited by the combustor geometry, a high Damkohler number (Da), and high Reynolds numbers (Re). Additionally, to capture the complete nonlinear effects of turbulence in combustion, the governing equations must resolve the smallest turbulence scales, known as Kolmogorov scales. These challenges are often addressed through the implementation of large eddy simulations (LES). At the expense of slightly reduced computational accuracy, i.e., instead of resolving all the scales of the flow that requires a dense computational grid; the LES models solve for larger eddies that contain the majority of the turbulent kinetic energy of the flow, providing reasonable accuracy with lower computational resources. This study presents a FORTRAN-based compressible computational fluid dynamics (CFD) code for high-Re non-premixed diffusion flames that utilizes the Smagorinsky-Lilly sub-grid scale model with a higher-order compact finite difference scheme. The first derivatives of the NS equations along with the mixture fraction equation in the wall-normal direction were discretized through fifth-order-accurate upwind and downwind-biased compact finite difference schemes. For the viscous terms, the discretization were performed using a fourth-order-accurate compact finite difference scheme. The tridiagonal equations from discretization were solved using the Thomas algorithm, while streamwise and wall-normal derivatives were computed via Fast Fourier Transforms (FFTs) in Fourier space. A fourth-order Runge-Kutta scheme was used for time-dependent components. The compact finite difference discretization schemes provide lower numerical dissipation in non-linear problems and are also able to achieve higher accuracy with lower computational grids compared to the conventional counterparts. Moreover, a high-order finite difference scheme can be mor effective than the finite volume method due to its advantages in structured grids, spectral-like resolution, and lower memory requirements. The turbulence chemistry interaction model decouples the flow field and the combustion chemistry through the assumption of steady flamelets. The model predicts the flow temperature and the species concentrations using the mixture fraction, mixture fraction variance, and scalar dissipation rate obtained from the model and coupled with the probability density function (PDF) model using the beta distribution function. The specific heat capacity (Cp) of the individual species is calculated using the NASA polynomials coefficient.  A counterflow configuration of fuel and oxidizer was studied (for methane-air diffusion flame), with the flamelet library generated using the Cantera 0-D chemical kinetics solver. The validity of the model will be demonstrated through the comparison of numerical data to the experimentally obtained data for the Sandia Flame D configuration.

Presenting Author: Rajib Mahamud Idaho State Unversity

Presenting Author Biography: Dr. Rajib Mahamud received his PhD in Mechanical Engineering from the University of South Carolina and afterwards obtained postdoctoral research experience in Aerospace Engineering at Texas A&M University and in the T-3 Division at Los Alamos National Laboratory. He holds BSc and MSc degrees in Mechanical Engineering from Bangladesh University of Engineering & Technology and the University of Nevada Reno, respectively. His research experience and interests broadly cover plasma, combustion, turbulence-chemistry interactions, aerothermodynamics, and nonequilibrium plasma processes and sources (e.g., microplasma, dual-pulse laser, etc.). He makes extensive use of Multiphysics FEM simulation code in his research.

Authors:

Rajib Mahamud Idaho State Unversity
Md Azazul Haque Idaho State University
Md. Kamrul Hasan Virginia Military Institute