Vagesh D Narasimhamurthy

We have considered the three-dimensional wake behind a cross formed by two intersecting flat plates using direct numerical simulations. The Reynolds number based on the uniform inflow velocity U0 and the plate width d was 1000. The vortex shedding in the wake was totally suppressed in a 4d wide intersection region and this gave rise to a massive zone of recirculating flow. Quasi two-dimensional vortex shedding with a primary frequency 0.165 U0/d occurred behind the outer branches more than 7d from the intersection. The wake behind the outer branches of the crossing plates closely resembled the wake behind a single flat plate. However, the wake flow in an intermediate region (located between the intersection region and the outer branches) was affected by persistent secondary flows. Further, shear-layer (K-H) instabilities have been observed in this region. The mean wake structure revealed the formation of four symmetrically positioned pairs of swirling vortices close to the intersection corner next to the plate's edges.

This study assesses different turbulence modeling approaches for simulation of two-phase coaxial annular swirling jet flows. The problem selected from literature involves an analytical inlet profile for an annular liquid sheet sandwiched between two coaxial annular gaseous jets. The liquid-gas interface is resolved using the volume-of-fluid (VOF) model with continuum surface force approximation. 3D unsteady Reynolds averaged Navier-Stokes simulations using up to 8.4 million grid cells and 64 HPC cores are conducted using the Fluent 17.2 software to obtain transient multiphase CFD data for this problem. Different turbulence models explored include the k-epsilon RNG with swirl modification, the Reynolds stress model (RSM), and RSM with scale adaptive simulations (RSM-SAS). Comparisons with the direct numerical results from literature suggest that the scale-adaptive simulation using RSM-SAS approach better predicts the onset of instability, liquid jet column collapse, jet mixing, vortex breakup, and the overall characteristics of this flow.

The effects of a pintle-shaped orifice on a planar turbulent jet flow at Reynolds number 4000, based on the inlet bulk mean velocity and the jet width, are studied using direct numerical simulations. Flapping of the jet along with a low-frequency modulation of the Kelvin-Helmholtz (KH) instability, in the presence of a pintle-shaped orifice, is observed. To compare the pintle-jet behavior, a free-jet is simulated as a reference case. The effects of the near-field region on the far-field flow characteristics have been investigated. In both the cases, the KH instability in the near-field influences the far-field jet, whereas the pintle-jet also exhibits a low-frequency flapping. In addition, oblique vortex pattern has been observed in the case of pintle-jet. The far-field flow statistics of the pintle-jet with a top-hat inlet interestingly agree with those of the free-jet with a hyperbolic tangent inlet. Temporal variation of the jet characteristics has been analyzed using spatiotemporal plots. In addition, the large- and small-scale turbulent motion have been studied using three anisotropic invariant maps (turbulence triangles, eigenvalue, and barycentric maps). Moreover, that the barycentric map gives a better visualization of the anisotropic behavior has been observed in the current study.

This article aims to make a detailed analysis of co-flowing plane Couette flows. Particularly, the variation of flow quantities from the turbulent to non-turbulent region is studied. While the enstrophy exhibits a sharp jump, the other quantities (e.g., mean velocity, Reynolds normal stress, and kinetic energy) show a continuous variation across the interface. The budget analysis of Reynolds normal stresses reveals that the terms playing a key role in turbulence transportation vary depending on the Reynolds normal stress under study. The terms production, diffusion, and redistribution play an important role in streamwise Reynolds stress (u ′ u ′ ¯). In the spanwise Reynolds stress (v ′ v ′ ¯), the diffusion terms play a significant role. In the wall-normal Reynolds stress (w ′ w ′ ¯), only the redistribution term is significant. The influence of one flow over another in the co-flow state was observed through the additional mean velocity and Reynolds normal stress found in the system compared to a standard plane Couette flow (pCf). Comparing the co-flow system with a conventional pCf system, the former exhibits greater vorticity, vortex stretching, and kinetic energy. A detailed analysis on the geometry and topology of flow structures was studied using flow invariants.

On roughening one of the walls in a planar Couette flow, it was reported that turbulence augments near the opposite wall [Javanappa and Narasimhamurthy, "Turbulent plane Couette flow with a roughened wall,"Phys. Rev. Fluids 6, 104609 (2021)]. The current direct numerical simulation work further explores this interesting phenomenon by investigating the flow dynamics and anisotropic nature of turbulence. For roughening, transverse square ribs are placed only on the bottom wall with streamwise pitch separations s = 5 r and 10r, where r = 0.2 h is the rib height and h is the channel half height. The time series of spanwise vorticity fluctuation in the case of s = 10 r shows the presence of coherent Kelvin-Helmholtz-like structures behind the ribs. Phase analysis using Hilbert transform reveals that the flow within the cavity for the s = 5 r case is in-phase, while a phase shift is observed for the s = 10 r case. The visualization of enstrophy production rate (ω i S i j ω j) reveals that regions of intense positive ones are observed to be topologically "sheet-like,"while the regions of negative ones are found to be "spotty."Anisotropy tensors and anisotropic invariant maps are used to explore turbulence anisotropy at both large and small scales of motion. It is observed that anisotropy is reduced in both the cases near the vicinity of roughness.

The wake behind T-shaped intersecting flat plates has been studied by direct numerical simulations and compared with the wake behind intersecting plates forming a cross. The Reynolds number based on the uniform inflow velocity and the plate width d was 1000. Similar to the cross-plate the vortex shedding was suppressed in a 4d wide intersection region with a substantial base suction pressure reduction. Shear-layer (K-H) instabilities have been observed and its characteristic frequency obtained. In contrast to the cross-plate, a main feature of the mean wake structure behind the T-plate is the formation of two symmetrically positioned swirling vortices close to the internal corners of the T. This was examined by considering pressure contours and the turbulent production terms of mean streamwise vorticity. In spite of some similarities, major features of the wake behind the T-plate turned out to be distinctly different from the wake behind a cross-plate configuration.

Direct numerical simulation (DNS) of rib-roughened turbulent channel flow rotating about its spanwise axis, by Narasimhamurthy and Andersson [Turbulence statistics in a rotating ribbed channel. International Journal of Heat and Fluid Flow 51, 29–41. (2015)], is revisited to seek complementary insights into the combined effects of roughness and Coriolis force on the turbulence. Flow in the channel was maintained at a friction Reynolds number, Reτ=400 and the non-inertial reference frame was rotated at different speeds quantified by the dimensionless rotation number, Ro=0, 2 and 6. Both the aforementioned parameters are based on the friction velocity uτ and half-height of the channel h. The channel walls were symmetrically mounted with transverse square ribs with cross-section of side k=0.1h and pitch λ=8k. Rotation causes preferential aligning of the near-wall vortical structures. The Taylor-Görtler–like roll-cells similar to those found in the rotating smooth-channel flows, survive the presence of the transverse ribs, but exhibit transient behavior. Increased transport of turbulent kinetic energy from the pressure side at higher Ro is evident from the variation of the vertical transport velocity Vk. The rotational production rates assume increasingly significant roles in distributing the kinetic energy in different directions. Anisotropy invariant maps and Taylor microscales show that the structure of turbulence is affected by rotation in a significant manner.

This paper investigates the dispersion of tiny inertial particles in the two-dimensional laminar wake from a pair of square cylinders placed side-by-side to each other. The flow Reynolds number (defined based on the cylinder size h and uniform inflow velocity Uc) in our study is fixed at 75. The wake pattern resulting from the interaction between vortices shed from individual cylinders depends on the cross-stream spacing between the cylinders, i.e. spacing ratio s/h. We examine the effect of varying the cross-stream spacing between the cylinders on the body impaction and dispersion of particles in the wake flow over a range of their responses. Three different spacing ratios, s/h = 0.3, 2, and 4 are considered. We find that the impaction efficiency of Stokes number St = 0.1 particles is not sensitive to the spacing ratio. However, the impaction efficiency of St = 1 and 10 particles is dependent on the spacing ratio and increases with the latter. We illustrate how the distribution and clustering of representative inertial particles (i.e., with Stokes number unity) evolve over a vortex shedding cycle uniquely due to the wake flow in each spacing ratio configuration and explain the physical mechanism of particle clustering using backward tracking. The non-uniform clustering of particles with different responses to the wake flow in each spacing ratio configuration is analyzed using the Voronoï diagrams. The spacing ratio effects on the dispersion of particles are quantified in terms of mean statistical quantities such as mean local concentration, clustering intensity, and velocity statistics. The local concentration of St = 1 and 10 particles at streamwise positions away from cylinders is significantly influenced by the wake flow behavior governed by the choice of the spacing ratio. Further, the combined effect of gravity and hydrodynamic drag forces on the impaction and clustering behavior of particles is reported, where the Froude number Fr=0.32. In all the cases, particles exhibit ballistic behavior, showing a significantly reduced tendency to form clusters and a drastic increase in the impaction efficiency.

The effect of coupled transverse and in-line motion of an elastically mounted rigid circular cylinder, subjected to vortex induced vibrations (VIV), is predicted using a reduced-order model. The model comprises of coupled wake and structural oscillators, where the nonlinearities in the fluid damping and forcing terms of the structural oscillator are retained. The classical van der Pol equation is used to model the wake oscillator. The unknown model constants are tuned to fit to experimental data. The influence of these tuning constants on the model performance are identified. The nonlinear contributions are shown to be insignificant in predicting the VIV characteristics associated with the transverse (y-only) oscillations of the cylinder at low Re. Surprisingly, the nonlinear terms were found to play a key role in predicting the two degree-of-freedom (2 DoF) motion of the cylinder. The model results for the cylinder with mass ratios in the low and moderate ranges are in good agreement with the experiments.

This paper reports occurrence of vortex shedding behind bluff-bodies in gas explosions, methods to suppress them using passive flow control techniques, and their overall impact on explosion overpressures. The pressure-time histories from a series of explosion tests, using an initially quiescent propane-air mixture in a vented channel of dimensions 1.5 m × 0.28 m × 0.3 m, are presented. Selected high-speed video frames visualizing the flame propagation are also presented. Three different bluff-obstruction scenarios are considered: 1) a reference case with a single smooth circular cylinder of diameter D = 0.0157 m, 2) a single cylinder identical to that in the reference case, mounted with a splitter plate of varying length from 5.13D to 0.26D, width 17.8D and thickness 0.06D, and 3) a single helically wired cylinder with wire diameter 0.1D and pitch 4D or 8D. All circular cylinders had a length of 17.8D and were mounted normal to the direction of the flow, spanning the channel cross-section 0.5 m downstream of the ignition point. The obstructions were inserted in the rig using a unique experimental setup. The peak overpressure generated by the explosion is of main interest. Both vortex shedding suppression techniques 2) and 3) yielded significant reduction in maximum overpressures when compared to the reference cylinder case 1). While all splitter plate configurations successfully reduced the maximum explosion overpressure, the splitter plates with length 1.02D and 0.51D were the most efficient, with an average reduction in overpressure of 32 ± 3%. The helical steel wire configurations also had a significant effect, with 25 ± 3% and 20 ± 3% reduction in the maximum overpressure for pitch 4D and 8D, respectively. The high-speed video visualization further buttressed the quantitative findings in the pressure measurements and clearly showed vortex shedding suppression. The current observations imply that the contribution from vortex shedding, i.e. apart from turbulence effects, to the overpressure generation in gas explosions is significant. The modelling community must consider this while preparing their simulators.