• Home
• Program
• Contact

10:45   Compressible Flows 2 10:45 15 mins #201 MULTI-POINT VELOCITY MEASUREMENTS IN GRID TURBULENCE INTERACTED WITH A SPHERICAL SHOCK WAVE Kento Inokuma, Tomoaki Watanabe, Koji Nagata, Yasuhiko Sakai Abstract: Interaction between a spherical shock wave and grid turbulence is studied with wind-tunnel experiments, where a spherical shock wave is ejected from an open end of a shock tube into the grid turbulence. We measure streamwise velocities of the grid turbulence by three hot-wire probes placed at different spanwise locations. The interaction with the shock wave results in amplification of the velocity fluctuations and in increase in the small-scale fluctuations of the grid turbulence. The correlation coefficient between streamwise velocities at different spanwise locations is increased after the interaction, confirming that integral length scale of turbulence is increased in the tangential direction of the shock wave surface. 11:00 15 mins #269 Direct numerical simulations on effects of turbulent Mach number in interaction between planar shock wave and turbulence Kento Tanaka, Tomoaki Watanabe, Koji Nagata Abstract: We performed direct numerical simulations on interaction between planar shock wave and local turbulence. The shock wave appears as discontinuous pressure increase and the strong one is deformed by the turbulence with high turbulent Mach number. In addition, it is expected that the shock wave is locally broken in this case. On the other hand, weak shock wave is hardly deformed when the shock wave interacts with low turbulent Mach number turbulence. These results show that the shock wave deformation and collapse induced by the interaction with turbulence are related with density fluctuations in compressible turbulence. 11:15 15 mins #77 Reynolds and Mach number effects on the skin-friction decomposition in turbulent boundary layers Yitong Fan, Weipeng Li, Sergio Pirozzoli Abstract: Mean skin-friction drag, which is a wall property though, is seriously influenced by the turbulent boundary-layer dynamics, and thus can be decomposed into several physics-informed constituents across the boundary layer. So far, plenty of investigations have been concentrated on the decomposition of skin-friction drag coefficient under incompressible conditions. Whereas much fewer discussions with regard to the supersonic turbulent flows can be found in the open literature. This paper provides a skin-friction decomposition method for the compressible turbulent boundary layers (TBLs), following the work by Renard and Deck and our previous studies. With the direct numerical simulation solutions of both incompressible and compressible TBLs, contributions to the skin-friction drag generation (C_{f}) of molecular viscous dissipations (C_{f1}), turbulent kinetic energy (TKE) production (C_{f2}), spatial growth of the flow (C_{f3}), and the streamwise heterogeneity (C_{f4}) have been separated and quantified. We investigate the Reynolds-number and Mach-number dependence of the mean skin-friction drag generation by assessment of turbulent statistics in this newly developed method. Profiles of pre-multiplied integrands in C_{f1}/C_f and C_{f2}/C_f are shown in Fig. 1 in the PDF file, giving their variances with Reynolds and Mach number across the TBL. For incompressible TBLs, constant distributions of C_{f1}/C_f^{1.5} and logarithmic variations C_{f2}/C_f^{1.5} with Reynolds number, in coincidence with Renard and Deck's findings, indict approximate identities for each term. Such phenomena (trends) are also observed in the supersonic cases, though values of C_{f1}/C_f^{1.5} and C_{f2}/C_f^{1.5} gets larger at higher Mach numbers. Compressibility transformations of velocity and Reynolds stress are applied to scale the decomposed constituents, finally leading to the friction decomposition into incompressible and compressible parts. Results show that the incompressible constituents collapse well onto the results of incompressible TBLs, and the compressible have non-negligible contributions confined within the near-wall region to both C_{f1} and C_{f2}, accounting for approximately 18% at M=2 in total. 11:30 15 mins #187 Direct Numerical Simulation of a BZT dense gas compressible shear layer Aurélien Vadrot Abstract: Dense gases are single phase vapors which are characterized by long chains of carbon atoms and by a medium to large molar mass. Because of their characteristics, dense gases exhibit an unusual behavior in the vicinity of the critical point. Their fundamental derivative of gas dynamics becomes lower than unity: $\Gamma = 1+\frac{\rho}{c}\frac{\partial c}{\partial \rho}\Big|_s <1$, where $\rho$ the density, $c=\sqrt{\partial p/\partial \rho |_s}$ the speed of sound, $p$ the pressure and $s$ the entropy. Consequently, the sound speed decreases during isentropic compression unlike in perfect gases. Also, in the vicinity of the critical point, the sound speed of dense gases drops significantly. Depending on the chosen gas and on the position in the $p-v$ diagram, with $v$ is the specific volume, the sound speed can be around ten times lower in dense gases than in air, which leads to a strong increase of compressibility effects. \\ Among dense gases, Bethe-Zel'dovich-Thompson (BZT) gases display a so-called inversion zone, that is a region where the fundamental derivative of gas dynamics becomes negative. In that case, expansion shock-waves are allowed in agreement with the second law of thermodynamics and the intensity of shock-waves is reduced. This feature could help reduce energy losses in supersonic flows as shock-waves are responsible for a significant fraction of them. \\ Dense gases are mostly used as heat transfer fluids in energetic exchangers. They are especially suited to Organic Rankine Cycle (ORC) systems, which collect heat from hot sources at moderate temperature (solar, geothermal or industrial sources) in order to produce electricity. Dense gases are used in such ORC systems because of their low boiling temperature, well adapted to the hot source temperature level. \\ However, the use of such gases raises modelling issues when numerically designing ORC turbines. A better knowledge of their turbulent behavior in the vicinity of the inversion region is needed to assess potential differences with perfect gases induced by their uncommon thermodynamical properties. Direct Numerical Simulation (DNS) has been used in few previous works to better understand the development of turbulence in BZT dense gas flows for flow configurations such as freely decaying and forced Homogeneous Isotropic Turbulence and turbulent channel flow. The present study is devoted to the DNS of a 3D BZT dense gas compressible shear layer. If homogeneous isotropic turbulence symbolizes the inter-blade space in the turbine, the shear layer there stands for the wake. Yet so far no DNS of a 3D dense gas compressible shear layer has been achieved. That is the purpose of the present study. \\ The shear layer is composed of two streams of gas at two different speeds facing to each other (Figure \ref{fig1}). At the interface, the mixing is occurring. The characteristic non-dimensional number of this type of flow is the convective Mach number: $M_c=(U_1-U_2)/(c_1+c_2)$, which links the speed of the streams ($U_1$ and $U_2$) with their sound speed ($c_1$ and $c_2$). The 3D DNS of the dense gas shear layer has been achieved at $M_c=1.1$. After a validation of the results for the perfect gas with available literature, a comparison is made between the dense gas and the perfect gas during the self-similar time period, where the momentum thickness evolves linearly with time. We especially compare the turbulent kinetic energy balance that details the physical quantities at stake in the shear layer. The production, the dissipation, the transport, the pressure-dilatation and the compressible dissipation terms are compared between the perfect gas and the dense gas. In a second step, the development of turbulence is investigated through spectral analysis of the turbulent kinetic energy balance. This study is expected to shed some light on the specific features present in turbulent dense gas flows. 11:45 15 mins #262 AN LES INVESTIGATION OF HIGH-SPEED TURBULENT GAS JETS Francesco Bonelli, Annarita Viggiano, Vinicio Magi Abstract: The aim of this work is to investigate compressibility effects on the behavior of turbulent gas jets by employing an in-house Large Eddy Simulation solver (Flow- Large Eddy and Direct Simulation, FLEDS) [1]. It is well known that compressibility, induced by high speed, affects turbulence by reducing turbulent mixing in both wall bounded and free shear flows but with different characteristics [2-4]. Indeed, in the case of wall bounded flows compressibility effects become relevant only for Mach numbers higher than 5 [2], whereas in the case of free shear flows, according to some researchers [2,3], such effects are not negligible starting from Mach number equal to about 1-1.5. However, in the past many studies [2-4] have been focused on boundary layers or, in the case of free shear flows, on mixing layers, but very few works have dealt with free jets [5]. In order to fill this lack, in the present work the authors have studied four jets with Mach number equal to 0.8, 1.4, 2.0 and 2.6 at Reynolds number equal to 10000. The results have shown that the potential core length increases with Mach number (Figure 1, left), thus confirming that compressibility reduces turbulent mixing, however, downstream of the potential core the velocity decay rate and the spreading rate are not affected by compressibility except for the case at Mach number equal to 2.6. Thus, compressibility has relevant effects on turbulence transition, but once the flow is turbulent such effects are less evident. This behavior is confirmed by the streamwise profiles of the normalized turbulent kinetics energy (TKE) (Figure 1 right) evaluated at the radial distance where the TKE peak is reached. The peak value decreases slowly up to Mach number equal to 2.0, whereas there is a larger reduction going from Mach number 2.0 to 2.6. The authors acknowledge the CINECA Awards N.HP10BUD7FQ, under the ISCRA initiative, for the availability of high performance computing resources and support. References [1] F. Bonelli, A. Viggiano, V. Magi, How does a high density ratio affect the near- and intermediate-field of high-Re hydrogen jets?, International Journal of Hydrogen Energy 41 (33): 15007-15025, 2016. [2] D. C. Wilcox, Turbulence Modeling for CFD, DCW Industries Inc., La Canada, California, 1994. [3] P. Bradshaw, Compressible turbulent shear layers, Annual Review of Fluid Mechanics 9 (1): 33-54, 1977. [4] S. K. Lele, Compressibility effects on turbulence, Annual Review of Fluid Mechanics 26 (1): 211-254, 1994. [5] J. Bellan, Large-Eddy Simulation of Supersonic Round Jets: Effects of Reynolds and Mach Numbers, AIAA Journal 54 (5): 1482-1498, 2016. 12:00 15 mins #429 Turbulent inlet effects on the cooling efficiency of an impinging jet - A compressible DNS study Gabriele Camerlengo, Jörn Sesterhenn Abstract: Impinging jets on a hot surface provide a highly efficient cooling mechanism [1]. Among other applications, they are essential for the cooling of turbine blades, for stationary and aeronautical uses, and electronic components. Their broad relevance led to an increased research interest during the last decades. Nevertheless, many aspects of the underlying physics for heat or mass transfer from the lower wall are still unclear as they occur at very small time and length scales, which are often not detectable in experiments or are not at all resolved in RANS or LES computations. Fortunately, the Reynolds numbers of these jets are typically in a range accessible for direct numerical simulations (DNS) by modern supercomputers. We present a DNS of a fully turbulent compressible impinging jet flow with Reynolds and Mach numbers of 8000 and 0.8, respectively. The jet is vertically confined between two isothermal walls and issues from a pipe of diameter D through an orifice in the uppermost wall. The lowermost wall serves as impingement plate. The temperature of the walls is 80 K higher than the average total temperature of the jet at the pipe outlet. Following the configuration of experiments whose data we have access to, the distance between the orifice and the impinging plate is set to 5D. Characteristic non-reflecting boundary conditions are applied at the lateral outflow. In previous computations carried out at our research group [2], an identical configuration with laminar inflow conditions was studied. Certainly, the inflow in engineering configurations will not be laminar. Therefore, we prescribe turbulent inflow conditions by coupling the impinging jet simulation with an upstream pipe of length 3D. The inflow of the pipe is enforced by copying time-dependent density and velocity profiles from an auxiliary fully developed turbulent pipe flow simulation. When compared with synthetic turbulence generation methods, this procedure, which has never been used before for the simulation of an impinging jet flow, offers the advantage of not requiring any external calibration parameter and of giving a very accurate representation of all turbulence scales. Results are compared with the laminar inlet ones reported previously by Wilke and Sesterhenn [2]. The presented analysis focuses on the heat transfer at the wall and its spatial distribution, whose peculiar shape is determined by the vortex dynamics in the proximity of the wall. As expected, it is observed that the presence of a turbulent inlet affects significantly the free jet region of the flow (Fig. 1a). The different evolution of the free jet affects subsequently the wall region, where a general increase of the turbulence intensity emerges. This leads in turn to an increase of the heat transfer at the wall, when compared to the laminar inlet case. Fig. 1b shows the average dimensionless heat flux (Nusselt number Nu) distribution at the wall as a function of the distance from the jet axis r for both cases. It can be seen that the there is no proportional relation between the two curves, since the heat flux remains relatively higher over a greater area when turbulent inlet conditions are prescribed. In order to characterize the mechanisms responsible for the enhanced heat transfer and its modified spatial distribution, instantaneous and statistical quantities, such as turbulent heat transfer and Reynolds stresses, will be presented and discussed side by side with the ones obtained in the laminar inlet case. References: [1] H. Martin. Heat and mass transfer between impinging gas jets and solid surfaces. 13 of Advances in Heat Transfer, pages 1–60. 1977. [2] R. Wilke and J. Sesterhenn. Statistics of fully turbulent impinging jets. Journal of Fluid Mechanics, 825:795–824, 2017. 12:15 15 mins #209 A COMPARATIVE STUDY OF RICHTMYER-MESHKOV INSTABILITY AND TURBULENT MIXING Ping Li, Tao Wang, Bing Wang, Jianyu Lin, Jingsong Bai Abstract: When a shock wave accelerates a perturbed interface between two different fluids, the perturbed interface is unstable and grows, and the Richtmyer-Meshkov instability (RMI) occurs, which can induce turbulent mixing at late times. The RMI and turbulent mixing are of significance in many applications ranging from man-made to natural phenomena, such as inertial confinement fusion, scramjet engine and supernova explosion. In this paper, a comparative study of three-dimensional multi-mode RMI and turbulent mixing in air/SF6 configuration is performed by using direct numerical simulation (DNS), large-eddy simulation (LES) and implicit large-eddy simulation (ILES), which are all based on our in-house finite volume code MVFT. In MVFT code, the piecewise parabolic method (PPM) and volume of fluid (VOF) are used to solves the compressible multicomponent Navier-Stokes (NS) equations. Figure 1 shows the structures of turbulent mixing zone (TMZ) at 1.5ms obtained by using DNS, LES and ILES. Figure 2 and 3 display the TMZ width and total turbulent kinetic energy vs. time, respectively. The RMI and induced turbulent mixing evolves similarly for different numerical method. However, the dissipation effect of numerical algorithm is extremely insufficient, so the ILES results show the rollingup of spike and bubble is absolutely weak, the structure scales of TMZ are smallest, the turbulent field is strongest, and the TMZ width is widest. For LES, the viscosity is relatively too high, the TMZ width grows slowest, the structure scales of TMZ are largest, the merger of spikes and bubbles are intensive, which lead to high mixing rate. The DNS should provide more useful and physical results. 12:30 15 mins #520 HOT-WIRE MEASUREMENTS OF THE EVOLUTION OF TOTAL TEMPERATURE AND MASS FLOW PULSATIONS IN 2D AND 3D SUPERSONIC BOUNDARY LAYERS Aleksey Yatskikh, Alexander Kosinov, Nikolai Semionov, Yury Yermolaev, Gleb Kolosov, Vasiliy Kocharin Abstract: This paper is devoted to the experimental study of the growth of mass flux pulsations and disturbances of total temperature during the laminar-turbulent transition of flat-plate and swept-wing boundary layers with the Mach number of the oncoming flow M = 2 and 2.5.