Structure of turbulent sonic underexpanded free jets

Abstract

Turbulent mixing in adapted and weakly underexpanded (underexpansion ratios less than 1.4) round jets, involving fully developed pipe flows injected into still air, was studied experimentally. Measurements included mean and fluctuating concentrations and mean static pressures using laser-induced iodine flourescence and mean and fluctuating streamwise velocities using laser Doppler anemometry. Predictions were used to help interpret the measurements and to initiate evaluation of methods for analyzing these processes. The predictions were based on k-e turbulence models, including a proposed extension to treat compressibility effects at high convection Mach numbers. In conjunction with other measurements, the results show that the near-field region of underexpanded jets is influenced by compressibility, which tends to reduce turbulent mixing rates at high convective Mach numbers, and high turbulence levels at the jet exit, which tends to increase turbulent mixing rates. Predictions based on effective-adapted-jet exit conditions yielded reasonably good estimates of mixing levels near the exit of underexpanded jets for both fully developed and slug flow jet exit conditions; however, such methods provide no information concerning the near-field region containing the shock waves. Predictions based on solution of parabolized Navier-Stokes governing equations, using the SCIPVIS algorithm, were encouraging for slug flow exit conditions, but this approach must be extended to treat fully developed flow at the jet exit. Nomenclature a = acceleration of gravity C M = constant in turbulence model d =jet exit diameter / = mixture fraction Fr = Froude number k -turbulence kinetic energy K = compressibility correction M = Mach number M c = convective Mach number p = static pressure r = radial distance u = streamwise velocity x = axial distance e = rate of dissipation of turbulence kinetic energy H t = turbulent viscosity v = kinematic viscosity Subscripts c = centerline quantity e =jet exit condition oo = ambient condition Superscripts(")(")' = time-averaged mean and root-mean-squared fluctuating quantity

Large eddy simulation of highly turbulent under-expanded hydrogen and methane jets for gaseous-fuelled internal combustion engines

“…Fluorescence (PLIF) in order to measure the mixing parameters and visualise the near-field shock structure of under-expanded air/nitrogen jets [23,25,26,29,[34][35][36][37]. Computational studies of under-expanded jets have been performed using compressible Euler equations [38][39][40][41], Reynolds Averaged Navier-Stokes (RANS) and Large Eddy Simulation (LES) methodologies [41][42][43][44][45][46][47][48][49][50][51][52], mainly for air jets or by assuming the injection of a passive scalar.…”

Numerical predictions of turbulent underexpanded sonic jets using a pressure-based methodology

“…The earliest work [9] used the Euler equations with varying success. Several studies have employed a parabolized Navier-Stokes method [8,10,11]. The method was found to overpredict the shock structure and to underpredict the mixing rate of sonic jets.…”

“…Measurements of the flow properties of underexpanded jets were initially limited to pressure probe measurements [5,6] or else the measurements were of the far field flow only [7]. More recently laser Doppler anemometry and laser-induced fluorescence have been used [8], although the measurements were limited to moderately underexpanded jets. Reliable measurements of velocity and turbulence quantities are not at present available but are clearly important for the validation of modelling techniques and for jets undergoing combustion.…”

“…Several works have therefore considered compressibility corrections. Experimental data were used in an empirical manner to attenuate the turbulent viscosity as the convective Mach number increased [8]. Reference [11] applied a model based on two length scales obtained from the k-å model and experimental data.…”

Reducing Computational Time in Turbulent Jet Modelling for Gas Dispersion Simulation

“…Thus, new boundary conditions namely, the jet nozzle becomes and the exit velocity, are considered in the numerical simulation. As the potential core extends up to ⁄ = 0.6 for sonic jets (Chuech, 1989) monitoring points were adjusted accordingly to consider the very first recording velocity immediately after the release.…”