Session G41: Turbulence: Jets

Turbulence budgets in time-resolved variable density round jets

Variable density turbulence occurs in flows with large density gradients and plays an important role in natural and engineered systems from the astrophysical (stellar evolution, supernovae) to the very small (inertial confinement fusion). The Turbulent Mixing Tunnel at Los Alamos National Laboratory was built to study this problem using simultaneous time-resolved planar density and stereoscopic velocity measurements (Particle Image Velocimetry and Laser-Induced Fluorescence). Because of the density field’s active role simultaneous measurements are critical for testing turbulence models such as LANL’s BHR. We made measurements of two momentum-matched jets in air and SF₆ (At = 0.16 and 0.6; Re = 7,000 and 12,000) and the flow was sampled at locations ranging from 2-16 initial jet diameters with 10,000 snapshots taken at each. Applying Taylor’s frozen turbulence hypothesis the full velocity gradient tensor was estimated and the pressure fields calculated using omni-directional integration, allowing the computation of the complete turbulence budgets. The effect of variable density conditions will be explored and compared to previous experiments on the same facility at Re = 20,000, and the validity of Lumley’s model for pressure-velocity correlations will be examined.   

Lagrangian properties in turbulent round jet flow from Large-Eddy Simulation

Turbulent round jet flows with Taylor Reynolds numbers of 70 and 210 were obtained using Large-Eddy Simulation (LES). Eulerian flow results are in accordance with theory, numerical and laboratory experiments from the literature. These results include the self similar radial profiles of velocity statistics and turbulence kinetic energy (TKE) budget, in addition to the power-law axial evolution along the centerline for several flow properties (velocity variance, TKE dissipation rate and integral timescale, for example). Furthermore, the Lagrangian tracking of point particles seeded at different axial locations is performed. The Lagrangian tracking uses the LES resolved flow field, and the inclusion of a subgrid scale velocity obtained from a Langevin equation is tested. Furthermore, we test the stationarization technique for inhomogeneous, self-similar flows known as the compensation method, which comprises the rescale of the Lagrangian velocity by the Eulerian velocity and integral timescale at the particle position. The success of the method is observed by the collapse of the Lagrangian second-order structure functions from different axial positions. Additional Lagrangian properties such as acceleration autocorrelation functions and their zero-crossing times are investigated as a function of Reynolds number and axial position. Possible impacts of the LES unresolved scales on the results are discussed. Research funded by the Fapesp-Université de Lyon Cooperation Grant 2020/06273-2.   

Modal analysis of forced turbulent jets

A variety of actuation methods have been employed to force turbulent jets, where in addition to the strong deterministic tonal response, the underlying stochastic turbulence is modulated by the applied forcing. However, the impact of the forcing on the turbulence and coherent structures is lacking. Since the forcing modulates the turbulence, the flow can not be analyzed using tools such as spectral proper orthogonal decomposition (SPOD) and resolvent analysis that have been successful in studying the mechanisms and coherent structures present in the unforced flow since they rely on a statistically-stationary assumption. Instead, the turbulence is cyclostationary, meaning that the statistics vary periodically. We investigate the impact of the forcing using recent extensions to SPOD and resolvent analysis, called cyclostationary SPOD (CS-SPOD) and harmonic resolvent analysis, which are suitable for analyzing flows with periodically varying statistics. We perform large-eddy simulations of forced turbulent jets using low-frequency St_f=0.3 and high-frequency St_f=1.5 forcing at various amplitudes. We find that the coherent structures are modulated by the applied forcing at both forcing frequencies. Good agreement between the dominant coherent structures determined via CS-SPOD and those predicted by the linearized harmonic resolvent analysis are seen.   

Can the turbulent-nonturbulent interface be associated with a transport barrier?

In this study, we explore the interface between turbulent and non-turbulent regions in jet flows. We employ simultaneous Particle Image Velocimetry (PIV) and Laser-Induced Fluorescence (LIF) measurements to capture the interface dynamics. Our primary objective is to track the mechanisms of large-scale engulfments and small-scale nibblings that shape the interface. This is achieved using a traversing camera system that moves along with the jet's streamwise evolution, covering a range from 10D to 60D downstream of the nozzle. We quantify the diffusive transport of linear momentum and the advective transport of passive particles. The small-scale nibbling structures at the interface, which were previously only visible in scalar LIF images, are now brought to light by diffusive transport barriers. Our findings reveal that the interface does not act as a barrier to passive particles. However, it shows a strong correlation with the diffusive transport of momentum. These significant insights contribute to our understanding of the mixing process between turbulent and non-turbulent flows.   

Turbulent/Non-turbulent Interface properties of a temporally developing turbulent jet

In many free shear flows such as turbulent boundary layers, wakes, or jets, a turbulent/non-turbulent interface (TNTI) separates the turbulent and the irrotational flow regions. This sharp and highly non-homogeneous region of the flow raises new modeling problems in the frame of large-eddy simulation for instance. Important exchanges of mass, momentum, and scalar quantities take place across this interface and lead to the propagation of the turbulent region known as turbulent entrainment. In this study, we investigate the geometrical properties and interscale energy transfer mechanisms at the TNTI in a temporally developing turbulent jet using data from highly resolved direct numerical simulations. We calculate the mean velocities and turbulence statistics in the vicinity of the TNTI thin layer conditional on characteristics such as TNTI position, orientation, curvature and local propagation speed relative to the fluid.  A very special attention has been paid to the influence of the overall resolution and quality of the simulation on the TNTI properties.