Critical analysis of classical turbulent combustion experiments on the basis of RANS simulations

Abstract

The purpose of the work is to analyze the consistency of experimental data concerning classical turbulent combustion tests and their applicability for CFD programs validation. Efforts are directed to development and verification of a computationally efficient method for numerical simulation of turbulent combustion in a high-speed flow, based on RANS equations and taking into account turbulence-combustion interaction (TCI). Computations are fulfilled with zFlare module from application package EWT-TsAGI [1] which is based on in-house code ZEUS (TsAGI) [2] and intended to describe the flows of ideal compressible gases mixtures with variable heat capacities and nonequilibrium chemical reactions. Several kinetics models for H2 and hydrocarbon fuel combustion in air together with two-parameter turbulence models based on the Boussinesq hypothesis are implemented into the program. The paper describes some results of zFlare verification, as well as detailed analyses of experimental data. Five classical turbulent combustion experiments are regarded: Evans et al. experiments [3-4], Burrows & Kurkov [5], A.D.Cutler [6-7], Cheng et al. [8] and Moreau et al. backward-facing step combustor [9]. In addition, several approaches for TCI based on methods of Partially Stirred Reactor model (PaSR, [10]) have been studied and compared.