ISSN 1068-3356, Bulletin of the Lebedev Physics Institute, 2018, Vol. 45, No. 10, pp. 308–310. c  Allerton Press, Inc., 2018. Original Russian Text c  I.A. Zubrilin, S.G. Matveev, A. Marrone, D.G. Pastrone, 2018, published in Kratkie Soobshcheniya po Fizike, 2018, Vol. 45, No. 10, pp. 28–32. Simulation of Pollutant Emissions in a Small-Size GTE Based on the Reactor Network Model I. A. Zubrilin a* , S. G. Matveev a , A. Marrone b , and D. G. Pastrone b a Samara National Research University, Moskovskoye shosse 34, Samara, 443086 Russia; * e-mail: zubrilin416@mail.ru b Department of Mechanical and Aerospace Engineering (DIMEAS), Politecnico of Turin 24 Corso Duca degli Abruzzi, 10129 Torino, Italy Received September 19, 2018 Abstract—The results of simulation of the NO x and CO formation in a small-sized gas turbine en- gine are presented. A comparison of the results of CFD calculations with experimental data showed that calculations in standard three-dimensional formulation in combination with partitioning into a network of reactors can be used to simulate NO x and CO formation. When using the JetSurf 2.0 chemical reaction mechanism, satisfactory agreement between calculated and experimental data is achieved with a number of elementary reactors above several hundreds. DOI: 10.3103/S1068335618100056 Keywords: reactors network, GTE, pollutant emission, CFD. Introduction. An increase in computational power and progress in the development of mathematical models of combustion in the last few decades allowed the formulation of approaches to simulation of toxic discharges such as CO and NO x in GTD combustion chambers, which significantly reduces expenses for experimental studies. However, currently these approaches strategy still require further correction. The objective of this study is to estimate the effect of the model parameters on the results of calculations of NO x and CO emission indices during Jet-A kerosene combustion in a small-sized turbojet engine. Simulation was performed in the ANSYS FLUENT software package followed by a comparison with experimental data. The results of calculations are presented within computational fluid dynamics (CFD) for both the steady-state Reynolds–Averaged Navier–Stokes (RANS) approach and the unsteady Large Eddy Simulation (LES) approach. Combustion chamber and experimental data. The combustion chamber of the engine under study is ring-shaped with counterflow organization of the working process and contains 6 fuel nozzles. In experimental studies at a rotation frequency from 40000 to 80000 rpm, pollutant concentrations at the engine output were measured. This is described in more detail in [1]. Simulation conditions. Due to axial symmetry of the combustion chamber, only one sixth part of the entire chamber was studied with periodic boundary conditions in the simulation. To simulate Jet-A kerosene, its substitute consisting of 9.1% of hexane, 18.2% of benzene, and 72.7% of decane (mass fractions) was used [2]. Chemical kinetics in the combustion chamber was described by the JetSurf 2.0 kinetic mechanism of chemical reactions, including 348 components and 2163 reactions [3]. Simulation results: cases under study. Turbulent transport was simulated in the steady-state formulation within the Reynolds stress transport model. The combustion process was described by the Flamelet-Generated Manifold (FGM) model in combination with the probability density function (PDF) approach for stochastic description of the turbulent-chemical interaction model [4]. The NO x formation was described by the extended Zeldovich mechanism [5]. The effect of such parameters as the number of network cells (1.2 or 3.3 millions), the approach to the O and CH concentration calculation in the Zeldovich mechanism (partially equilibrium or unsteady transient), the PDF approach (for the temperature or mixture fraction), and the fuel drop size on the results of calculations of adverse substance emission was studied. For the latter parameter, two different values for each mode were considered: those presented in [1] (larger diameter) or [6] (smaller diameter). The simulation plan is given in Table 1. 308