Importance of the nozzle-exit boundary-layer state in subsonic turbulent jets

Abstract

To investigate the effects of the nozzle-exit conditions on jet flow and sound fields, large-eddy simulations of an isothermal Mach 0.9 jet issued from a convergent-straight nozzle are performed at a diameter-based Reynolds number of $1\times 10^{6}$. The simulations feature near-wall adaptive mesh refinement, synthetic turbulence and wall modelling inside the nozzle. This leads to fully turbulent nozzle-exit boundary layers and results in significant improvements for the flow field and sound predictions compared with those obtained from the typical approach based on laminar flow in the nozzle. The far-field pressure spectra for the turbulent jet match companion experimental measurements, which use a boundary-layer trip to ensure a turbulent nozzle-exit boundary layer to within 0.5 dB for all relevant angles and frequencies. By contrast, the initially laminar jet results in greater high-frequency noise. For both initially laminar and turbulent jets, decomposition of the radiated noise into azimuthal Fourier modes is performed, and the results show similar azimuthal characteristics for the two jets. The axisymmetric mode is the dominant source of sound at the peak radiation angles and frequencies. The first three azimuthal modes recover more than 97 % of the total acoustic energy at these angles and more than 65 % (i.e. error less than 2 dB) for all angles. For the main azimuthal modes, linear stability analysis of the near-nozzle mean-velocity profiles is conducted in both jets. The analysis suggests that the differences in radiated noise between the initially laminar and turbulent jets are related to the differences in growth rate of the Kelvin–Helmholtz mode in the near-nozzle region.

Resolvent-based modeling of coherent wave packets in a turbulent jet

“…In turn, changes in the optimal mode will be accompanied by changes in the orthogonal sub-optimals. The localisation of optimal forcing in the present results may furthermore be linked to the observed sensitivity of LES statistics with respect to flow details in the nozzle boundary layer [32]. Numerous previous studies addressing the linear modelling of wavepackets in turbulent jets, when only boundary forcing at the inflow was considered, observed discrepancies in the initial amplitude growth at low Strouhal numbers, typically St ≤ 0.3 [5,13,59], which play an important role in the generation of jet noise.…”

Spectral analysis of jet turbulence

“…The reader is referred to Brès et al (2017b) for further details on the numerical method, meshing strategy and subgrid-scale model. A detailed validation of the M j = 0.9 jet including the nozzle-interior turbulence modeling (i.e., synthetic turbulence, wall model) can be found in Brès et al (2017a). The subscripts j and ∞ refer to jet and free-stream conditions, a is the speed of sound, ρ density, D nozzle diameter, µ dynamic viscosity, T temperature, and U j to the axial jet velocity on the centerline of the nozzle exit, respectively.…”

An efficient streaming algorithm for spectral proper orthogonal decomposition

“…[22] for an early experimental and [23] for a recent numerical example, use SPOD to analyze jet turbulence. Our first is example is an LES of a Mach 0.9 jet at a jet diameter-based Reynolds number of 1.01 · 10 6 [19].…”