Differential diffusion and pressure effects on heavily sooting 2D Kerosene/Air shear flames

Abstract

Purpose: The current study aims to test the limits of the Unity Lewis number simplification coupled with both the ideal gas (IGL) and a real gas (RGL) equation of state (EOS) for predicting flame and soot characteristics of heavily sooting Kerosene/Air shear flames at 4 different operating pressures. Theory and Methods: Fully compressible Navier-Stokes equations are adopted in DNS environment. For the computational mesh, equivalently spaced grid points from 0 < x1 < L1 are used, while in the cross-stream direction the mesh has been stretched in the x2 direction. An 8th order central explicit finite difference method and a 4th order Runge-Kutta are employed to solve for the spatial and time derivatives, respectively. The temperature contour plot created using 2D DNS data are provided in Figure A. Boundary conditions of the problem: I -in x2 direction, far from the flame kernel, flow is set to have free stream characteristics and II -in x1 direction, periodic boundary conditions are embedded to solve the turbulent flow. Results: Comparisons to the earlier studies have been revealed that the mean peak flame temperature predictions done by the unity-Le number model are lessened; implying that the soot production/oxidation rates calculations will be debilitated comparably to the low mean flame temperature of the unity-Le number cases regardless of the operating pressure. The unity-Le number model's under-prediction of the flame temperatures results in intensifying incomplete combustion which not only emanates the soot load but also weaken the soot oxidation rate for the RGL EOS cases, while causes under-prediction of mass fraction of soot for the IGL EOS cases in comparison to the non-unity Le number results. As the operating pressure increases to 35 atm, the mean flame temperature increases hence, the soot load in the flame kernels increases even though the soot oxidation process is enhanced by the increasing flame temperature. An increase in the mean flame temperature has been noted for the IGL EOS model predictions at 35 atm case, which will cause an inevitable increment in soot load throughout the flame kernel. Conclusion: The analyses have revealed the that the unity-Le number simplification estimates the average flame temperature lower than expected for each operating pressure increasing the possibility of an incomplete combustion or even local flame weakening and extinction which stems 'abnormally' high soot load throughout the flame kernel essentially for 10 and 35 atm operating pressure cases of the IGL EOS model.