betaFeedback

WorksPhysics and Astronomy

Flow past an inclined spheroid in homogeneous and stratified environments

Abstract

Large-eddy simulations (LES) are performed to study the flow past a 6:1 prolate spheroid placed at an angle of incidence of α=10∘α= 10^{\circ}. The diameter-based Reynolds number (Re=U∞D/νRe = U_{\infty}D/ν) is set to a value of 5×1035\times10^{3} and four values of diameter-based Froude numbers (Fr=U∞/NDFr = U_\infty/ND) are analyzed: Fr=∞,6,1.9,Fr = \infty, 6, 1.9, and 11. Visualizations of coefficient of pressure (CpC_p) and friction (CfC_f) contours reveal asymmetry in the Fr=∞Fr = \infty and 66 flows while, at Fr=1Fr = 1 and Fr=1.9Fr = 1.9, the flow over body does not have any visible asymmetry. This finding is further corroborated through the analysis of force coefficients on the body. The changes in the pressure coefficients (CpC_p), friction coefficients (CfC_f), and drag coefficients (CdC_d) with the Froude number are described in detail for α=10∘α= 10^{\circ}. We also present the analyses of forces on the body at α=0∘α= 0^{\circ} angle of incidence for comparison with the α=10∘α= 10^{\circ} cases.

INTRODUCTION

Despite their widespread presence in applications, slender body flows have received significantly less attention compared to the flows past bluff bodies. There are some particularities that make the study of slender body flows especially challenging. In a numerical simulation, resolving the boundary layer (BL) of a long body is computationally more expensive than resolving the BL of a blunt body. Additionally, in experimental studies, slender bodies can take up a significant portion of the measurement section and, since they normally generate thin wakes, they are potentially harder to probe and measure.

The first experimental study of flow past a slender body dates back to Chevray [5] who investigated the wake of a 6:1 prolate spheroid at Re=4.5×105Re = 4.5 \times10^5. The wake measurements spanned a streamwise distance of x/D=18x/D = 18. Han & Patel [11] conducted an experimental study of flow past a 4.3:1 spheroid at different angles of incidence (α\alpha) at Re≈2×104Re \approx2 \times10^4. They primarily focused on the flow separation pattern, identifying two regimes as α\alpha was changed: (i) closed or bubble separation and (ii) open or free-vortex type separation. Wang Wang et al. [26] extended the study of Han & Patel [11] to prolate spheroids with aspect ratios of 2:1, 3:1, and 4:1 for a wide range of α\alpha in the range [0∘,90∘][0^\circ, 90^\circ]. Experimental studies have also been performed at higher ReRe with tripped BL by Fu et al. (1994) and Chesnakas & Simpson [4]. Jiménez et al. [13] and Ashok et al. [1] studied the high ReRe wake of a DARPA SUBOFF at α=0∘\alpha= 0^\circ and in pitch configurations, respectively. While Jiménez et al. [13] focused on the self-similarity and scalings in the near to intermediate wake, Ashok et al. [1] characterized the wake asymmetries induced due to the pitch configuration.

Numerical simulations have also been used to study slender body flows. Constantinescu et al. [7] and Wikström et al. [27] conducted RANS and LES studies of flow past a 6:1 prolate spheroid at Re=4.2×106Re = 4.2 \times10^6 and α=10∘\alpha= 10^\circ, 20∘20^\circ. Tezuka & Suzuki [25] carried out a three-dimensional stability analysis study for the flow past a 4:1 spheroid for varying α\alpha. For α≠0∘\alpha\ne0^\circ, as ReRe was increased, they found transition from a symmetric to an asymmetric flow configuration in the longitudinal center-plane. The flow asymmetry at nonzero angle of incidence was later studied in more detail using DNS by Jiang et al. [12]. In the last few years, the wake of a DARPA SUBOFF has also been studied through LES by Posa & Balaras [23]; Kumar & Mahesh [14] at Re∼O(105)Re \sim O(10^5). Ortiz-Tarin et al. [19] was the first study of an unstratified slender body flow that probed far wake statistics, extending to x/D=80x/D = 80.

The LES study by Ortiz-Tarin et al. [20] is the first to take background stratification into account for the study of slender body flows. Their high-resolution LES study was conducted for a 4:1 spheroid at zero angle of attack, Re=104Re = 10^4 and Fr=∞Fr = \infty, 3, 1 and 0.5. They analyzed the laminar BL evolution, force distribution, and the near- and far-field characteristics of the steady lee waves. In the current work, we build upon their work studying the effect of varying FrFr on the flow past a 6:1 spheroid placed at a moderate angle of incidence (α=10∘\alpha= 10^\circ). We analyze the effect of stratification on: (i) the variation of CpC_p, CfC_f, (ii) the forces on the body, and (iii) the flow separation. We also present a brief analysis of forces on the body at α=0∘\alpha= 0^\circ for comparison with α=10∘\alpha= 10^\circ cases. To the best of our knowledge, this is the first study exploring the flow characteristics of an inclined slender body flow in stratified environments. In future work, we plan to extend this study to the analysis of vorticity and wake dynamics in these flows.

Numerical Methodology

The numerical solver used for these simulations have been extensively validated in the past for body-inclusive simulations of stratified wakes [20, 6, 19, 17, 16, 18]. For a detailed description of the solver, boundary conditions and immersed boundary method (IBM) approach, we refer the interested readers to Chongsiripinyo & Sarkar [6].

In the present work, Re=5×103Re=5\times10^{3} and four different Fr=∞,6,1.9Fr=\infty,6,1.9, and 1 are simulated for two angles of incidence α=0∘\alpha=0^\circ and 10∘10^\circ. For the stratified cases at α=10∘\alpha=10^\circ, radial and streamwise domains span 0≤r/D≤530\le r/D\le53 and −30≤x/D≤50-30\le x/D\le50, respectively. A large radial extent together with a sponge layer on the boundaries weaken the IGWs before they hit the end of the computational domain and hence control the amplitude of spurious reflected waves. Numbers of grid points in different directions are as follows: Nr=1000N_r=1000 in the radial direction, Nθ=128N_\theta=128 in the azimuthal direction, and Nx=3584N_x=3584 in the streamwise direction. For the unstratified wake at α=10∘\alpha=10^\circ, radial and streamwise domains span 0≤r/D≤210\le r/D\le21 and −11≤x/D≤47-11\le x/D\le47 while Nr=718N_r=718, Nθ=256N_\theta=256, and Nx=2560N_x=2560. At α=0∘\alpha=0^\circ, we simulate the wakes till a downstream distance of x/D=30x/D=30 and up to r/D=17r/D=17 and 53 for unstratified and stratified cases, respectively. The grid point distribution for α=0∘\alpha=0^\circ cases are as follows: Nr=1000N_r=1000 and 910 for stratified and unstratified cases, respectively, Nθ=128N_\theta=128, and Nx=3584N_x=3584.

Body Forces at α=0∘\alpha=0^\circ

Figure 1(a) and (b) show the variation of the pressure coefficient Cp=(P−P∞)/0.5ρU∞2C_p=(P-P_\infty)/0.5\rho U_\infty^2 and skin-friction coefficient Cf=Re0.5∣τx∣/0.5ρU∞2C_f=Re^{0.5}|\tau_x|/0.5\rho U_\infty^2, respectively, for different values of FrFr at α=0∘\alpha=0^\circ. In the reported CfC_f, the raw skin friction coefficient is multiplied by Re0.5Re^{0.5} to obtain an O(1)O(1) value. The θ=0∘\theta=0^\circ and 90∘90^\circ labels correspond to the horizontal x−yx-y and vertical x−zx-z plane, respectively. An averaging time window of 30D/U∞30D/U_\infty is used to obtain these results.

Variation of pressure and skin-friction coefficients

Figure 1. Variation of (a) pressure coefficient CpC_p and (b) skin-friction coefficient CfC_f for different FrFr at zero angle of attack (α=0∘\alpha= 0^\circ). The θ=0∘\theta= 0^\circ and 90∘90^\circ curves correspond to variations on the surface in the horizontal (θ=0∘\theta= 0^\circ) plane and the vertical (θ=90∘\theta= 90^\circ) planes on the spheroid. Plotted value of CfC_f is Re0.5Re^{0.5} times the friction coefficient.

For Fr≥1.9Fr\ge1.9 cases (figure 1(a)), the drop of pressure and its recovery primarily occur towards the beginning and the end of the body, respectively. For −2≲x/D≲1-2\lesssim x/D\lesssim1, CpC_p remains approximately constant for Fr≥1.9Fr\ge1.9. On the other hand, CpC_p in Fr=1Fr=1 case shows a monotonic decay till x/D≈2x/D\approx2 and a slight recovery for x/D≳2x/D\gtrsim2, indicating strong effect of buoyancy on the flow over the body. For Fr≥1.9Fr\ge1.9, CpC_p variations are qualitatively very similar. The differences between Fr=∞Fr=\infty and the critical FrcFr_c curves are less pronounced in the 6:1 spheroid (Frc=1.9Fr_c=1.9) than what was observed in the 4:1 spheroid (Frc≈1Fr_c\approx1) as reported by Ortiz-Tarin et al. [20]. Anisotropy between horizontal and vertical plane CpC_p curves appears at Fr=1.9Fr=1.9 and increases slightly at Fr=1Fr=1. Pressure variations (not shown here) confirm that the low pressure region in the tail of the spheroid (in the vertical plane) at Fr=1Fr=1 is imposed by the steady lee-wave field, indicating a strong influence of buoyancy on the flow at this value of FrFr.

Figure 1(b) shows the variation of CfC_f for different FrFr at α=0∘\alpha=0^\circ. Similar to the behavior of CpC_p, the variation of CfC_f for Fr=6Fr=6 and ∞\infty are very similar. The flow separates at x/D=2.34x/D=2.34 for the unstratified flow (marked by Cf→0C_f\to0). This value is in excellent agreement with the result of Patel & Kim [21]. Fr=1.9Fr=1.9 shows elevated and suppressed levels of CfC_f in vertical and horizontal planes, respectively, compared to Fr=6Fr=6 and ∞\infty. At Fr=1Fr=1, CfC_f in the vertical plane further increases compared to Fr=1.9Fr=1.9. Moreover, in the horizontal plane as well, Fr=1Fr=1 shows higher CfC_f than Fr=∞Fr=\infty and 6 for x/D≳0.5x/D\gtrsim0.5. Increased CfC_f in the vertical plane for Fr=1Fr=1 and 1.9 as a consequence of

α=0∘\alpha=0^\circ, FrFrCdC_dCdfC_d^fCdpC_d^pΔCd=Cd−Cd(∞)\Delta C_d=C_d-C_d(\infty)
Fr=∞Fr=\infty0.240.220.020
Fr=6Fr=60.260.230.030.02
Fr=1.9Fr=1.90.320.250.070.08
Fr=1Fr=10.520.290.230.28

Table 1. Drag coefficients (CdC_d) and corresponding pressure (CdpC_d^p) and friction contributions (CdfC_d^f) for α=0∘\alpha=0^\circ at different FrFr.

thinner BL over the body (not shown here for brevity) compared to Fr=6Fr=6 and ∞\infty. In the horizontal plane, for Fr=1Fr=1 (compared to Fr=∞Fr=\infty), BL thickens between x/D≈−2x/D\approx-2 to 0 and gets thinner beyond x/D≈1x/D\approx1, explaining the trend of CfC_f in the horizontal plane for Fr=1Fr=1.

Table 1 presents Cd=Fd/(0.5ρ0U∞2A)C_d=F_d/(0.5\rho_0U_\infty^2A) for different FrFr at α=0∘\alpha=0^\circ. Here, A=πD2/4A=\pi D^2/4. There is a more monotonic increase in CdC_d, CdpC_d^p, and CdfC_d^f with increasing stratification levels. Friction contributes more to the drag than pressure (except at Fr=1Fr=1), as expected for a slender body flow. However, compared to flow past a 4:1 spheroid [20], we find that the effect of stratification on the overall drag, quantified by ΔCd=Cd−Cd(∞)\Delta C_d=C_d-C_d(\infty), is weaker in the 6:1 spheroid. Between Fr=∞Fr=\infty and Fr=1.9Fr=1.9, CdC_d changes by ≈33%\approx33\% in the present case while there was a 100% increase in the 4:1 spheroid. This smaller increase in CdC_d for 6:1 spheroid is primarily due to a smaller increase in CdpC_d^p at Frc=1.9Fr_c=1.9 (compared to Fr=∞Fr=\infty), unlike in the 4:1 prolate spheroid. It is only when Fr=1Fr=1 that we see a sharp jump in CdpC_d^p leading to a ≈100%\approx100\% increase in CdC_d, consistent with the differences we observe in CpC_p curves (figure 1a) between Fr=1Fr=1 and Fr≥1.9Fr\ge1.9 cases.

It is a well-established observation that the amplitude of steady lee-waves decreases with increasing value of FrFr in the Fr≥1Fr\ge1 regime for a variety of wake generators, including 6:1 prolate spheroid [3, 15]. Specifically, Meunier et al. (2018) found that the lee-wave amplitude for a 6:1 spheroid (based on ∂w/∂z\partial w/\partial z) decayed as Fr−2Fr^{-2}. On the other hand, it should be noted that the definition of critical FrcFr_c given by Ortiz-Tarin et al. [20] is based purely on kinematic considerations, i.e., by equating half-wavelength of the lee-wave to the length of the body. As a result, Frc=L/πDFr_c=L/\pi D increases linearly with the aspect ratio. Thus, for the 6:1 spheroid, we see a weaker effect of stratification on the drag at its critical value of Frc=6/π=1.9Fr_c=6/\pi=1.9 compared to 4:1 spheroid for which Frc=4/π=1.27Fr_c=4/\pi=1.27 is lower. That the lee wave field in the Fr=1.9Fr=1.9 case is weaker than in the Fr=1Fr=1 case is also confirmed also by quantification of the pressure field, not shown here for brevity.

Body Forces at α=10∘\alpha=10^\circ

Introducing even a moderate angle of incidence, namely α=10∘\alpha=10^\circ, significantly changes the characteristics of flow on the body, reflected by significantly different trends of CpC_p, CfC_f, and force coefficients in α=10∘\alpha=10^\circ flows compared to α=0∘\alpha=0^\circ counterparts. In what follows, we discuss the trends of above-mentioned quantities when α=10∘\alpha=10^\circ and Fr=∞,6,1.9Fr=\infty,6,1.9, and 1. For α=10∘\alpha=10^\circ, a time averaging window of approximately 50D/U∞50D/U_\infty is used.

Coefficient of Pressure CpC_p

Figure 2(a-h) show the pressure contours on the leeside and windside of the body. We also present the variation of CpC_p along y=0y=0 line on the leeside and windside in figure 2(i,j), respectively. CpC_p obtained from potential flow solution is also presented for comparison in Figure 2(i,Figure 2j). On both sides, the agreement between the potential solutions and LES simulations is excellent till x/D≈1x/D \approx1. Beyond x/D≈1x/D \approx1 LES simulations deviate from potential solutions, presumably due to the three-dimensionality of BL evolution and effects of flow separation. Pressure drop on both sides primarily happens near to the nose of the body and in the ascending order of FrFr, i.e., the pressure drop in the nose for Fr=1>1.9>6>∞Fr = 1 > 1.9 > 6 > \infty (Figure 2(i)). It is also worth noting that the Fr=6Fr = 6 and ∞\infty cases show asymmetry in the leeside pressure contours about the y=0y = 0 plane as shown in Figure 2(a,Figure 2c).

Pressure contours on the leeside and windside of the spheroid, with surface pressure-coefficient variations

Figure 2. Pressure contours on the leeside (a,c,e) and windside (b,d,f) of the spheroid shown for all simulated FrFr at α=10∘\alpha= 10^\circ. Dashed lines in (a) and (c) correspond to y=0y = 0. Also shown is the variation of CpC_p (i, j) on the leeside and windside of the body surface in the y=0y = 0 plane. Potential solution for CpC_p is also shown in dashed line, obtained from Piquet & Queutey [22].

On the leeside, for Fr≥1.9Fr \ge1.9 cases, pressure recovers continuously, albeit slowly, after the initial drop in the nose (Figure 2i). On the contrary, pressure in the Fr=1Fr = 1 case, which shows the lowest initial drop (from the stagnation value at the nose) of all cases in the fore of the body (x/D<0x/D < 0) drops below that of the remaining cases for x/D>0x/D > 0. This is evident both from contours (Figure 2g) and CpC_p plots (Figure 2i). Mean pressure contours in the vicinity of the spheroid (not shown here for brevity), reveal a strong low-pressure region on the entirety of the leeside. This low-pressure region is imposed by the steady lee wave field at Fr=1Fr = 1. The differences between the unstratified case and strongly stratified cases (Fr=1.9Fr = 1.9 and 11) are even more pronounced on the windside compared to the leeside of the spheroid. Beyond x/D≈2x/D \approx2, CpC_p of Fr=1.9Fr = 1.9 and 11 fall significantly below those of Fr=∞Fr = \infty and 66 (Figure 2j), followed by a sharp recovery towards the end. The contours in Figure 2(b,Figure 2d,Figure 2f,Figure 2h) confirm this trend of higher and lower pressure at the head and tail, respectively, for the Fr=1Fr = 1 and 1.91.9 cases compared to Fr≥6Fr \ge6.

Coefficient of Friction CfC_f

Figure 3(a-h) show the contours of scaled shear stress Re0.5∣τx∣Re^{0.5}|\tau_x| on the leeside (left) and windside (right) for all cases at α=10∘\alpha= 10^\circ. In Figure 3(i,Figure 3j), we present the variation of CfC_f on the spheroid surface and in the y=0y = 0 plane on both sides. Similar to the pressure contours, a distinct asymmetry is present on the leeside of Fr=∞Fr = \infty and 66 flows as shown Figure 3(a,Figure 3c), to be discussed in more detail in the next subsection.

Contours of $\operatorname{Re}^{0.5}|\tau_x|$ and variation of $C_f$

Figure 3. Contours of Re⁡0.5∣τx∣\operatorname{Re}^{0.5}|\tau_x| (a-h) on the leeside and windside of the spheroid for all FrFr at α=10∘\alpha= 10^\circ. Dashed lines in (a) and (c) correspond to y=0y = 0. Variation of CfC_f (i, j) on the leeside and windside of the body at y=0y = 0 plane.

The shear stress contours show that flow separation primarily happens on the leeside at α=10∘\alpha= 10^\circ across all FrFr. The region of separated flow can be identified by the region of ∣τx∣→0|\tau_x| \to0, marked in green in Figure 3 contours. For Fr=∞Fr = \infty and 66, flow separates primarily from the two lateral sides on the lee of the body (Figure 3(a,Figure 3c)). There is a central region near y=0y = 0 which remains attached until nearly the tail. In the Fr=1.9Fr = 1.9 case, separation occurs from the sides as well as the central region around the y=0y = 0 plane (Figure 3e) while in the Fr=1Fr = 1 case, separation primarily happens in the central region (Figure 3g) and not at the sides. Thus, it can be inferred that the stratification level strongly influences the flow separation even at the moderate non-zero angles of attack of this study.

In the vertical-center plane (y=0y = 0), CfC_f varies similarly for all FrFr on both sides (Figure 3(i,Figure 3j)) till x/D≈0x/D \approx0. On the leeside (Figure 3i) and for x/D≥0x/D \ge0, the Fr=∞Fr = \infty and 66 cases show higher CfC_f than the strongly stratified cases of Fr=1.9Fr = 1.9 and 11. On the windside, beyond x/D≈1x/D \approx1, Fr≤1.9Fr \le1.9 cases show reduced CfC_f as compared to Fr≥6Fr \ge6 (Figure 3j). This region of elevated surface shear in the Fr=1.9Fr = 1.9 and 11 flows coincides with the region of steep pressure drop observed at x/D>1x/D > 1 (Figure 2j).

Force Coefficients

Figure 4(a,b,c) present the force coefficients (Ci=Fi/0.5ρU∞2AC_i = F_i/0.5\rho U_\infty^2 A) at α=10∘\alpha= 10^\circ, decomposed between pressure and friction contributions. Here Cx,CyC_x,C_y and CzC_z correspond to drag (CdC_d), lateral force, and lift (ClC_l) on the body, respectively. We first discuss CdC_d and ClC_l and follow by noting the unusual characteristics of CyC_y for α=10∘\alpha= 10^\circ. Figure 4(d) also shows CdC_d for α=0∘\alpha= 0^\circ for reference. In α=0∘\alpha= 0^\circ cases, Cl≈0C_l \approx0 and Cy≈0C_y \approx0 at all FrFr indicating no asymmetry in the flow around body.

Force coefficients decomposed between pressure and shear contribution

Figure 4. Force coefficients decomposed between pressure and shear contribution: (a) coefficient of drag CdC_d, (b) coefficient of lift ClC_l, (c) lateral force CyC_y at α=10∘\alpha= 10^\circ, and (d) CdC_d at α=0∘\alpha= 0^\circ.

For all FrFr, CdC_d, CdfC_d^f and CdpC_d^p (Figure 4a) at α=10∘\alpha= 10^\circ increase relative to the zero degree angle of incidence (Figure 4d). Similar to α=0∘\alpha= 0^\circ, there is a weak monotonic increase in CdC_d till Fr=1.9Fr = 1.9 and a large jump thereafter at Fr=1Fr = 1. This jump primarily comes from an approximately 100% increase in CdpC_d^p from Fr=1.9Fr = 1.9 to Fr=1Fr = 1. Figure 2(g,Figure 2h) show that the pressure contours at Fr=1Fr = 1 are quite different from those of Fr≥1.9Fr \ge1.9. The tail and nose on the leeside and windside, respectively, are at a lower and higher pressure (compared to Fr≥1.9Fr \ge1.9 cases) which leads to an enhanced CdpC_d^p. For ClC_l, primary contributor is the pressure rather than shear, as expected for a moderate angle of incidence. It is interesting to note that ClC_l decreases till Fr=1.9Fr = 1.9 and then increases significantly at Fr=1Fr = 1 case, resulting from an increase in ClpC_l^p. The reason for this increase is the large difference between the leeside and windside pressure in Fr=1Fr = 1 flow as shown in Figure 2(g,Figure 2h).

Figure 4(c) shows Cy≠0C_y \ne0 for Fr=∞Fr = \infty and 66 at α=10∘\alpha= 10^\circ. For Fr=∞Fr = \infty, Cy∼0.05C_y \sim0.05, and for Fr=6Fr = 6, Cy∼0.05C_y \sim0.05. This value is approximately 12% of the streamwise drag force in both cases. Intuitively, CyC_y should be equal to zero due to reflectional symmetry in the configuration about the y=0y = 0 plane. Non-zero CyC_y implies lateral asymmetry in the flow on the body. Pressure and friction contours in Figure 2(a,Figure 2c) and 3(a,c), respectively, show that this is indeed the case and the asymmetry originates on the leeside of the body at x/D≳1x/D \gtrsim1 (the y=0y = 0 intersection of the body surface is shown to better identify this lateral asymmetry). No asymmetry is present on the windside flow over the spheroid. When the value of FrFr is decreased to 1.9, Cy→0C_y \to0. Further decrease to Fr=1Fr = 1 also results in Cy=0C_y = 0. This indicates that the flow asymmetry is suppressed as the stratification is increased. Figure 2(e,Figure 2g) and Figure 3(e,Figure 3g) also confirm that for Fr=1.9Fr = 1.9 and 1, no asymmetry is visually evident on the leeside of the spheroid. Hence, two important findings from the analysis of CyC_y are: (i) the weakly stratified (Fr=6Fr = 6) and unstratified (Fr=∞Fr = \infty) flow exhibit lateral asymmetry and (ii) this asymmetry is suppressed as the strength of stratification is increased, i.e., at Fr=1.9Fr = 1.9 and 1.

Our finding regarding lateral asymmetry in the unstratified case is in accord with previous studies on flow past slender bodies (Tezuka & Suzuki, 2006; Ashok et al., 2015a,b; Jiang et al., 2015). It is also interesting to note that the CyC_y of Fr=∞Fr = \infty and Fr=6Fr = 6 are similar in magnitude but flipped in sign. We hypothesize that the Fr=∞Fr = \infty and Fr=6Fr = 6 flows might be locked in two different reflectional-symmetry-breaking states, with each state being equally probable. There can be a switching between these two states at a very large timescale. The existence of a long time scale for switching between different reflection-symmetry-breaking states have been extensively researched in flow past three-dimensional blunt bodies [10, 24, 8]. Interestingly, Jiang et al. [12], who also found lateral asymmetry in their flow (6:1 spheroid at α=45∘\alpha= 45^\circ), did not find a switch even after 600D/U∞600D/U_\infty. We aim to investigate the characteristics of the intermediate state for wake, besides looking at near-body flow. Hence, running the simulations for T∼1000D/U∞T \sim1000D/U_\infty would be prohibitively expensive and out of scope of the current work.

Conclusions

Large eddy simulations (LES) are performed to study the characteristics of flow over a 6:1 spheroid placed at an angle of incidence α=10∘\alpha= 10^\circ. We present results at Re=5×103Re = 5 \times10^3 and Fr=∞,6,1.9,Fr = \infty, 6, 1.9, and 11. Additionally, flows past a 6:1 spheroid at the same (Re,Fr)(Re, Fr) combinations but at α=0∘\alpha= 0^\circ are simulated to provide a basis for comparison. We find that the buoyancy effect introduced by stratification strongly modulates the pressure and friction forces on the body. The critical Froude number [20] given by Frc=L/πDFr_c = L/\pi D is defined kinematically by equating the body length to the half-wavelength of the generated lee wave. For the 6:1 spheroid, the effect of stratification on CpC_p, CfC_f, and CdC_d is more pronounced at Fr=1Fr = 1 than at the critical Froude number Frc≈1.9Fr_c \approx1.9. For both α=0∘\alpha= 0^\circ and 10∘10^\circ cases, CdC_d monotonically increases with decreasing FrFr. At α=10∘\alpha= 10^\circ, a distinct lateral asymmetry is visible in the CpC_p and CfC_f contours for Fr=6Fr = 6 and ∞\infty cases. This gives rise to a non-zero lateral force whose magnitude is approximately 12% of the streamwise drag at α=10∘\alpha= 10^\circ. Further increasing the stratification kills the asymmetry at Fr=1.9Fr = 1.9 and 11. At α=10∘\alpha= 10^\circ, we also find that the flow separation over the body is strongly dependent on the value of FrFr. At Fr=6Fr = 6 and ∞\infty, the flow separates from the lateral sides on the lee of the body while at Fr=1Fr = 1, the flow separates predominantly near the vertical center plane y=0y = 0. In the future studies, we plan to investigate flow separation and topology of shed vortices and their potential link to the flow asymmetry in α=10∘\alpha= 10^\circ cases as well as wake dynamics.

Acknowledgments

We gratefully acknowledge the support of the Office of Naval Research grant N0014-20-1-2253.

References

  1. [1]Ashok, A., Van Buren, T. & Smits, A. J. 2015a Asymmetries in the wake of a submarine model in pitch. J. Fluid Mech. 774, 416–442.DOI
  2. [3]Bonneton, P., Chomaz, J. M. & Hopfinger, E. J. 1993 Internal waves produced by the turbulent wake of a sphere moving horizontally in a stratified fluid. J. Fluid Mech. 254, 23–40.DOI
  3. [4]Chesnakas, C.J. & Simpson, R.L. 1994 Full three-dimensional measurements of the cross-flow separation region of a 6: 1 prolate spheroid. Exp. Fluids 17 (1), 68–74.
  4. [5]Chevray, R. 1968 The turbulent wake of a body of revolution. J. Basic Eng pp. 275–284.DOI
  5. [6]Chongsiripinyo, K. & Sarkar, S. 2019 Decay of turbulent wakes behind a disk in homogeneous and stratified fluids. J. Fluid Mech. 885, A31.DOI
  6. [7]Constantinescu, G. S., Pasinato, H., Wang, Y., Forsythe, J. R. & Squires, K. D. 2002 Numerical investigation of flow past a prolate spheroid. J. Fluids Eng. 124 (4), 904–910.
  7. [8]Dalla Longa, L., Evstafyeva, O. & Morgans, A. S. 2019 Simulations of the bi-modal wake past three-dimensional blunt bluff bodies. J. Fluid Mech. 866, 791–809.DOI
  8. [10]Grandemange, M., Gohlke, M. & Cadot, O. 2013 Turbulent wake past a three-dimensional blunt body. Part 1. Global modes and bi-stability. J. Fluid Mech. 722, 51–84.DOI
  9. [11]Han, T. & Patel, V. C. 1979 Flow separation on a spheroid at incidence. J. Fluid Mech. 92 (4), 643–657.DOI
  10. [12]Jiang, F., Gallardo, J. P., Andersson, H. I. & Zhang, Z. 2015 The transitional wake behind an inclined prolate spheroid. Phys. Fluids 27 (9), 093602.DOI
  11. [13]Jiménez, J. M., Hultmark, M. & Smits, A. J. 2010 The intermediate wake of a body of revolution at high Reynolds numbers. J. Fluid Mech. 659, 516–539.DOI
  12. [14]Kumar, P. & Mahesh, K. 2018 Large-eddy simulation of flow over an axisymmetric body of revolution. J. Fluid Mech. 853, 537–563.DOI
  13. [15]Meunier, P., Le Dizès, S., Redekopp, L. & Spedding, G. R. 2018 Internal waves generated by a stratified wake: experiment and theory. J. Fluid Mech. 846, 752–788.DOI
  14. [16]Nidhan, S., Chongsiripinyo, K., Schmidt, O. T. & Sarkar, S. 2020 Spectral proper orthogonal decomposition analysis of the turbulent wake of a disk at re= 50 000. Phys. Rev. Fluids 5 (12), 124606.DOI
  15. [17]Nidhan, S., Ortiz-Tarin, J. L., Chongsiripinyo, K., Sarkar, S. & Schmidt, P. J. 2019 Dynamic Mode Decomposition of Stratified Wakes. In AIAA Aviation 2019 Forum. Dallas, Texas: American Institute of Aeronautics and Astronautics.DOI
  16. [18]Nidhan, S., Schmidt, O.T. & Sarkar, S. 2022 Analysis of coherence in turbulent stratified wakes using spectral proper orthogonal decomposition. J. Fluid Mech. 934.DOI
  17. [19]Ortiz-Tarin, J.L., Nidhan, S. & Sarkar, S. 2021 High-reynolds-number wake of a slender body. J. Fluid Mech. 918.
  18. [20]Ortiz-Tarin, J. L., Chongsiripinyo, K. C. & Sarkar, S. 2019 Stratified flow past a prolate spheroid. Phys. Rev. Fluids 4 (9), 094803.
  19. [21]Patel, V.C. & Kim, S.E. 1994 Topology of laminar flow on a spheroid at incidence. Comp. Fluids 23 (7), 939–953.DOI
  20. [22]Piquet, J. & Queutey, P. 1992 Navier-stokes computations past a prolate spheroid at incidence - I. Low incidence case. Comp. Fluids 21, 599–625.DOI
  21. [23]Posa, A. & Balaras, E. 2016 A numerical investigation of the wake of an axisymmetric body with appendages. J. Fluid Mech. 792, 470–498.DOI
  22. [24]Rigas, G., Oxlade, A. R., Morgans, A.S. & Morrison, J. F. 2014 Low-dimensional dynamics of a turbulent axisymmetric wake. J. Fluid Mech. 755, R5.DOI
  23. [25]Tezuka, Asei & Suzuki, Kojiro 2006 Three-Dimensional Global Linear Stability Analysis of Flow Around a Spheroid. AIAA Journal 44 (8), 1697–1708.DOI
  24. [26]Wang, K.C., Zhou, H.C., Hu, C.H. & Harrington, S. 1990 Three-dimensional separated flow structure over prolate spheroids. Proc. R. Soc. A: Math. Phys. Eng. Sci. 429 (1876), 73–90.DOI
  25. [27]Wikström, N., Svennberg, U., Alin, N. & Fureby, C. 2004 Large eddy simulation of the flow around an inclined prolate spheroid. J. Turbul. 5 (1), 029.DOI

Paper details

vlnc:W_nf5n69px
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents