High-Reynolds-number wake of a slender body

Abstract The high-Reynolds-number axisymmetric wake of a slender body with a turbulent boundary layer is investigated using a hybrid simulation. The wake generator is a $6:1$ prolate spheroid and the Reynolds number based on the diameter $D$ is $Re=10^5$. The transition of the wake to a state of complete self-similarity is investigated by looking for the first time into the far field of a slender-body wake. Unlike bluff-body wakes, here the flow is not dominated by vortex shedding in the near wake. Instead, the recirculation region is very small, the near wake is quasi-parallel and is characterised by the presence of broadband turbulence. Until $x/D\approx 20$, the wake decay of a slender body with turbulent boundary layer is very similar to the classic high-$Re$ behaviour, $U_d\sim x^{-2/3}$. Extrapolation of this observation to larger $x/D$ has led to the belief that these wakes decay following the asymptotic $-2/3$ decay law. Our results show, however, that this is not the case and the wake transitions to a faster decay rate once complete self-similarity is achieved. In this later region ($20 < x/D < 80$), mean and turbulence profiles are self-similar. Furthermore, despite the high global and local Reynolds numbers, the classic hypotheses that lead to the well-known decay exponents are not fulfilled. Instead, turbulent dissipation follows a non-equilibrium scaling and a new decay rate $U_d\sim x^{-6/5}$ is observed. The transition from $U_d\sim x^{-2/3}$ to $U_d\sim x^{-6/5}$ is preceded by the dominance of the azimuthal $|m|=1$ mode and the emergence of a large-scale helical structure.


Introduction
Turbulent wakes are ubiquitous in nature. From the wind fluctuations generated by buildings to the large-scale oceanic structures left by currents flowing past islands, turbulent wakes distribute energy and momentum across space in a wide variety of situations and environments. In this work, we focus on the wake generated by an axisymmetric slender body, which is particularly relevant for aerodynamic and hydrodynamic applications.
condition is chosen to approximate the flow at some distance from the body. The use of a temporal model reduces the computational cost very significantly but requires a good initial condition to achieve a quantitative match with experiments. Expanding on the results of Gourlay et al. (2001), Johansson et al. (2003) showed that the rate of decay of the wake might depend on the local Reynolds number. In their computational study, the wake transitioned from a decay rate of −2/3 to −1 when the value of the local Reynolds number, Re L = U d L/ν, went below approximately 500. In a later temporal study, Redford et al. (2012) performed two temporal simulations with different initial conditions. They showed that universality was achieved only after a long passage of time, and suggested that the universal regime might not be seen in practice. Interestingly, in the far wake, the flow statistics followed the high-Re decay despite the local Re falling below 500.
One year later, Nedić et al. (2013) introduced the concept of non-equilibrium dissipation to the study of axisymmetric wakes. Instead of assuming that turbulent dissipation scales as ε ∼ k 3/2 /L, where k is the turbulent kinetic energy (TKE) and L is a characteristic large-eddy length, a new scaling ε ∼ (Re/Re L ) n k 3/2 /L was proposed. This non-equilibrium dissipation scaling has been attributed to unsteadiness of the energy cascade and has led to a re-examination of the classical decay exponents. For example, the experiments and simulations of Dairay et al. (2015) found that the wake of a Re = 5 × 10 3 fractal plate exhibits a U d ∼ x −1 non-equilibrium decay. According to classic theory, this wake should have decayed as x −2/3 because Re L was high. However, this was not the case and only the non-equilibrium theory was able to explain it. Other works such as Obligado, Dairay & Vassilicos (2016) and more recently  have also observed this non-equilibrium region in the wake of square plates and disks.
Despite the use of slender bodies in many engineering applications, the majority of wake studies are devoted to the wake of bluff bodies. There are some particularities that make the study of high-Reynolds-number slender-body wakes especially challenging. Computationally, it is very costly to simulate turbulent boundary layers over long bodies and also resolve their far wake. The problem is tremendously stiff and the required spatial and temporal resolution is enormous. In a wind tunnel or in a tank, if the body is long, a significant amount of the measuring section can be taken by the wake generator. Furthermore, slender-body wakes are thinner than bluff-body wakes and potentially harder to probe and measure. These constraints have significantly limited the available studies on the topic.
The first study of this kind is Chevray (1968). In his work, a 6 : 1 prolate spheroid with a tripped boundary layer was used as a wake generator. The Reynolds number based on the diameter was Re = 4.5 × 10 5 and the domain spanned x/D = 18. In this region, the wake defect velocity showed a decay of U d ∼ x −2/3 and self-similarity was achieved only in the mean velocity. Years later, Jiménez et al. (2010) performed a study of the near wake of a standard submarine model (DARPA SUBOFF) with a tripped boundary layer at Re = 1.2 × 10 5 − 7.7 × 10 5 . This study, spanning 15 diameters, also showed the wake decay following U d ∼ x −2/3 . Only recently, high-resolution body-inclusive (BI) simulations have provided more data on the decay of the slender-body axisymmetric wake. Posa & Balaras (2016) observed a −2/3 decay for the wake of a SUBOFF at Re = 1.39 × 10 5 in a domain up to x/D = 12. Kumar & Mahesh (2018) simulated flow past an unappended SUBOFF at Re = 1.26 × 10 5 in a domain that spanned x/D = 15. They observed −2/3 decay followed by a transition to −1 at x/D ≈ 5, which they attributed to low-Re decay.
slender body wakes at high Reynolds numbers (in black). Slender-body wakes span up to x/D ≈ 20 from the end of the body, a much shorter domain than studies with other wake generators which are also shown. The figure also includes recent studies of a disk (in green), temporally evolving wakes (in blue) and a fractal plate wake (in orange). Overall, the picture of wake decay remains unclear. The disk wakes show an initial decay similar to U d ∼ x −1 and later on they transition to U d ∼ x −2/3 . The fractal plate decays following U d ∼ x −1 . The temporal simulations show significant variability owing to the use of different initialisation techniques and the transformation from t to x/D. Some also show long initial regions where the decay is faster than the classic high-Re U d ∼ x −2/3 . All four of the slender-body wakes show a decay close to the classic U d ∼ x −2/3 . This result of −2/3 decay of U d , which is derived by assuming full self-similarity of the flow, is somewhat puzzling since the length of the domain is limited in these studies and it is likely that only partial self-similarity is achieved.
The experimental evidence shows that the approach to self-similarity depends on the nature of the wake generator (Higuchi & Kubota 1990). It also shows that universality, if ever achieved, is not achieved at distances relevant for practical applications (Wygnanski, Champagne & Marasli 1986;Redford et al. 2012). All the available literature looking into the high-Re far wake is devoted to bluff-body wakes. The present work is the first look into the far wake of a slender body at high Reynolds number. The main goal is to study the transition to self-similarity and ascertain if the far wake follows the classic high-Reynolds-number decay laws, as suggested by the previous body of literature. Unlike the wake of bluff bodies, the near wake of a slender body with a turbulent boundary layer is not dominated by vortex shedding (Jiménez et al. 2010;Kumar & Mahesh 2018). In this work we also look into how this feature might affect the establishment and development of the wake. Some of the questions that we address are as follows. (i) Does the decay of a slender-body wake at high Reynolds number conform to classical theory? (ii) How is the transition to self-similarity different from the wake of a bluff body? (iii) Why have some previous experiments and simulations shown wake exponents close to low-Re values even at very high Reynolds numbers? (iv) Is there a possible relation between coherent structures and the wake decay?
To answer these questions, we simulate the wake of a 6 : 1 prolate spheroid with a tripped boundary layer. Unlike previous studies of slender-body wakes which reached up to a streamwise distance of x/D = 20, here, we span x/D = 80 at Re = 10 5 to investigate the far wake. To the best of the authors' knowledge, this is the highest Reynolds number in a far-wake study that resolves the flow at a slender-body wake generator.

Hybrid simulation
To study the evolution of near-body flow into the far wake in a computationally feasible manner we employ a hybrid simulation. The hybrid simulation combines two simulations: a BI one that resolves the flow around the wake generator and a body-exclusive (BE) one that resolves the far wake. The BE simulation can be a temporally evolving wake (Pasquetti 2011) or a spatially evolving wake similarly to the approach of VanDine, Chongsiripinyo & Sarkar (2018) who validated the method for a sphere wake. The use of the BI simulation avoids the drawback of regular temporal simulations, for which the choice of initial conditions introduces considerable variability in the subsequent wake evolution. In our implementation, the BI flow field provides the inflow for the subsequent BE simulation. The data from a selected cross-stream plane of the BI simulation are interpolated on to a new grid and fed into the BE simulation as an inlet boundary condition. This procedure allows us to relax the natural stiffness of the far-wake problem by using different spatial and temporal resolutions appropriate for the BI and BE stages. The grid cell in the BE simulation, which has to be sufficiently small to adequately resolve wake turbulence, is still much larger than that required to resolve the boundary layer. The consequently large time step in the BE simulation leads to considerable savings in computational time without compromising accuracy (VanDine et al. 2018).
The three-dimensional Navier-Stokes equations are solved in a cylindrical coordinate system using a staggered grid. An immersed boundary method (Balaras 2004;Yang & Balaras 2006) is used to represent the 6 : 1 prolate spheroid. Both the BI and BE simulations are performed with the same solver. The solver uses a third-order Runge-Kutta method combined with second-order Crank-Nicolson scheme to advance the equations in time, and second-order-accurate central differences for spatial derivatives. This solver has been extensively validated in both unstratified and stratified sphere wakes (Pal et al. 2017). A wall-adapting local eddy-viscosity (WALE) model is used (Nicoud & Ducros 1999). For both simulations, the boundary conditions are Dirichlet at the inflow, Orlanski for the convective outflow and Neumann at the lateral boundary. Further numerical details and validation are reported for a disk wake ) and a lower-Re slender-body wake with laminar boundary layer (Ortiz-Tarin, Chongsiripinyo & Sarkar 2019). The cylindrical coordinate system is (x, r, φ) with the origin at the body centre. The minor-axis diameter of the prolate spheroid is D and the major axis is aligned with the incoming velocity U ∞ . The body aspect ratio is 6 : 1 and the Reynolds number based on D is Re = U ∞ D/ν = 10 5 . To accelerate laminar-to-turbulent transition, the boundary layer is tripped by a small surface bump at 0.5D from the nose of the body. The bump has radial thickness of 0.002D (∼15 wall units) and streamwise extent of 0.1D. The location at which the flow time history from the BI simulation is extracted and subsequently fed as BI 10 5 5 2π −8 15 746 512 2560 BE 10 5 10 2π -80 479 256 4608 Table 1. Simulation parameters of the body-inclusive (BI) and body-exclusive (BE) simulations. inflow into the BE simulation is x/D = 6. This location is well away from the recirculation region which ends at x/D = 3.1. The simulation parameters are listed in table 1. The grid is designed to resolve the turbulent boundary layer and to resolve small-scale wake turbulence. The total number of grid points across BI and BE domains is approximately 1.5 billion. The turbulent boundary layer is resolved with x + = 40, r + = 1 and r φ + = 32. There are 10 points in the viscous sublayer and 130 across the buffer and log layers. In the wake, the peak ratio between the grid size and the Kolmogorov length (η = (ν 3 /ε) 1/4 ), in both BI and BE domains, is max( x/η) = 7.5, max( r/η) = 6 and max(r θ/η) = 5. The distributions of x/η and r/η in the domain are given in figure 2. The peak ratio of the turbulent viscosity over the molecular viscosity occurs at the tripping location and is ν sgs /ν ≈ 3. Everywhere else, ν sgs /ν < 1, confirming the excellent resolution of the large-eddy simulation grid of both BI and BE simulations.
Flow statistics are obtained by temporal and azimuthal averaging. The average, denoted by , is performed over 100 D/U ∞ after the flow has reached statistically steady state. Velocity is normalised with the free-stream velocity U ∞ and length with the body minor axis D. The mean velocities are (U, U r , U φ ), the root-mean-square (r.m.s.) turbulent fluctuations (u x , u r , u φ ) and the Reynolds shear stress u xr . The TKE is given by k = (u 2 x + u 2 r + u 2 φ )/2 and the turbulent dissipation rate by ε = 2ν s ij s ij where s ij is the fluctuating strain-rate tensor. The decay rate α of any quantity is computed as the best fit of a power law, f (x) = C(x − x 0 ) α , to the data. The procedure employed here is similar to that of Obligado et al. (2016). The exponent is fitted first and then the amplitude and the virtual origin are obtained. Similarly to the experimental findings of Jiménez et al. (2010), keeping the virtual origin x 0 fixed had negligible effects on α or the fit quality. The R 2 regression value exceeds 0.98 in all cases.

Flow visualisation
The visualisation of instantaneous vorticity and streamwise velocity in figure 3 reveals the main characteristics of the flow. The near wake is quasi-parallel and carries small-scale turbulence which has originated in the tripped boundary layer. The separation region, shown in red in figure 3(b), is small and extends for a distance of 0.1D from the end of the body, in accordance with the experiments of Chevray (1968). The wake topology is similar to that observed in previous slender-body studies such as Kumar & Mahesh (2018) and Posa & Balaras (2016). Disk, plates and other bluff body wakes have significantly larger recirculation bubbles, e.g. the disk of Chongsiripinyo & Sarkar (2020) has a recirculation bubble of length 2D. In addition, compared with the wake of bluff bodies, here the near wake is not dominated by the asymmetric shedding of the boundary layer, as was also observed by Kumar & Mahesh (2018). In a bluff-body wake, the asymmetric shedding of the boundary-layer vorticity leads to the dominance of a helical mode immediately after the recirculation bubble, e.g. Berger, Scholz & Schumm (1990), Johansson & George (2006b) and Nidhan et al. (2020). Figure 3(a,b) shows a quasi-cylindrical structure in the near wake. Only farther downstream beyond x ≈ 20, (figure 3c), does the wake start showing a large-scale sinuous structure.

The classic axisymmetric wake theory
The classic analysis of the axisymmetric turbulent wake starts with the Reynolds-averaged streamwise momentum equation, (4.1) By writing the left-hand side as a function of the momentum deficit U ∞ − U, assuming that the momentum deficit is U ∞ − U U ∞ and retaining only the first-order terms, one obtains where we have used the boundary-layer approximation and assumed high Re so that the viscous term is negligible. The analysis proceeds by assuming self-similarity of the velocity deficit and the Reynolds stresses, where η = r/L is a similarity variable that depends on a characteristic wake width L = L(x). Here U d and R s are the peak values of the velocity deficit and the Reynolds shear stress, respectively. Equation (4.2) then simplifies to where, for self-preservation, all the terms in brackets must have the same x dependence. The same procedure applied to the conservation of TKE leads to where the TKE, turbulent dissipation, production and transport terms are as follows, As in (4.4), for full self-similarity, all the bracketed terms in (4.5) must have the same x dependence. However, flows can be partially self-preserving, meaning that they show self-similarity only in the mean momentum equation or only up to certain orders of the turbulence statistics (George 1989). The integration of the momentum equation over a cross-section provides us with an additional constraint, namely, the momentum thickness θ(x), or its first-order approximationθ(x), is constant. Here, and θ ≈θ holds only when U d /U ∞ 1. After employing (4.3a,b) for the momentum deficit, the expression forθ can be readily integrated to obtain L 2 U d =θ 2 U ∞ , which directly relates L and U d .
At this point, to obtain the decay law of U d , more assumptions are required and classic theories start to differ. Tennekes & Lumley (1972) assume that R s ∼ U 2 d and, using (4.4), they obtain U d ∼ x −2/3 . Here R s can vary independently of U d and then the TKE equation comes into play. Townsend (1976) and George (1989) assume that dissipation follows the high-Re inertial scaling ε ∼ K 3/2 /L and that the turbulent production is P = −u xr (∂U/∂r) ∼ R s (U d /L). With these assumptions, they showed that K s ∼ R s ∼ U 2 d and also U d ∼ x −2/3 . In addition, George (1989) obtained the low-Re U d ∼ x −1 decay when the viscous term ∼ νU d /L 2 dominates the turbulent stress in the momentum balance and turbulent dissipation scales as ε ∼ νU 2 d /L 2 . 918 A30-8

Wake analysis
In the present study, the decay rates of near-wake statistics (x < 20) agree well with previous experiments and simulations of a slender-body wake. However, the behaviour after x = 20 has not been documented before. We find that after x = 20, the flow transitions from partial to full self-similarity, while showing notable departures from the classic wake decay. Figure 4 shows the streamwise evolution of several wake statistics. The peak defect velocity U d shows two distinct decay rates. The near wake, 5 < x < 15, decays in a similar manner to the classic high-Re result, U d ∼ x −0.66 . However, in the intermediate and far wake, 20 < x < 80, the decay rate increases to U d ∼ x −1.2 . Figure 4(b) shows the evolution of the wake width (L) measured in three different ways. L d is defined such that U(L d ) = U d /2, and L k is defined analogously using the TKE as k(L k ) = k(r = 0)/2. The displacement thickness (L θ ) is defined by L 2 θ = U d −1 ∞ 0 (U ∞ − U)r dr. For x > 6, the three measures of L show the same values of growth rate. The wake width grows as L ∼ x 0.33 in the region 15 < x < 20 and transitions towards L ∼ x 0.6 as the wake traverses 30 < x < 80.
Figure 4(c) shows the decay of the peak r.m.s. turbulent intensities together with the TKE and the Reynolds shear stress. The near wake has anisotropy with u x > u φ > u r and the individual decay rates agree well with Jiménez et al. (2010). In particular, u r , u φ ∼ x −0.21 and u x , u 1/2 xr ∼ x −0.26 . After x ≈ 40, the r.m.s. values become similar and they show a long region of constant decay rate. Here, the TKE and the Reynolds shear stress decay as k 1/2 , u 1/2 xr ∼ x −0.8 and the r.m.s. velocities as u r , u θ ∼ x −0.73 and u x ∼ x −0.86 . To proceed with the classic wake analysis, we move on to assess self-similarity. Figure 5 shows the profiles of velocity deficit, TKE, Reynolds shear stress (u xr ) and dissipation rate (ε), each scaled by its peak, as a function of normalised radial coordinate at different streamwise locations.
The U-deficit profile exhibits self-similarity from x ≈ 5 onwards. The u xr profiles collapse together in local scales after x ≈ 10. TKE shows self-similarity only after x ≈ 30 and ε after x ≈ 40. In the near wake, k and ε show an off-centre maximum, which is the imprint of the tripped boundary layer. Indeed, self-similarity of higher-order statistics require longer x, thus revealing stages of partial self-similarity (George 1989). Terms in  In addition to the assumption of self-similarity, the classic turbulent wake theory assumes that the defect velocity U d is small compared with the free-stream velocity U ∞ . This hypothesis simplifies the equations significantly, leads to the classic scaling laws, and to the integral constraint U d L 2 being constant at any cross-section along the wake. Figure 6(a) shows the evolution of the momentum thickness θ and the first-order approximation of the momentum thicknessθ. The full momentum thickness reaches a constant value of θ 0 = 0.075 after x = 10. The simplified momentum thickness does not become constant until x ≈ 60 and in the near wake it changes significantly. This explains why the growth rate of L and the decay of U d do not fulfil exactly U d L 2 = U ∞ θ 2 0 in the near wake. Note that, in the near-wake region 5 ≤ x ≤ 15, the simulation shows L ∼ x 0.26 and U d ∼ x −0.66 , i.e. U d L 2 is not constant.
To further understand the approach to similarity, the centreline evolution of the leading terms in the U equation are evaluated and plotted in figure 6(b). The second-order  term, U d (∂U d /∂x), remains non-negligible until x ≈ 15. Equation (4.2) becomes a better approximation of the exact U-equation with increasing x but complete balance between the decay of U and the radial derivative of u xr is not achieved until x ≈ 25. Note that the viscous term remains at least two orders of magnitude below the leading terms. At comparable Re, blunt-body wakes show a higher initial value of U d however, because their entrainment levels are higher, U d decays faster. Therefore, in the range x ≈ 10-20, U d of a blunt-body disk wake becomes half that of the spheroid wake as is evident after comparing the black and the green lines in figure 1. At x = 10 the disk of Chongsiripinyo & Sarkar (2020) exhibits a value of U d /U ∞ = 0.09 and at x = 30 the value is U d /U ∞ = 0.03. Here, at x = 10 the magnitude is U d /U ∞ = 0.22 and, at x = 30, U d /U ∞ = 0.06. This factor-of-two-larger value of U d for the prolate spheroid relative to the sphere wake at comparable x implies that a longer streamwise distance is required in the spheroid wake for the second-order terms to truly become negligible.

D s L/K s
The first-order approximation of the momentum conservation (4.2) and the first-order approximation of the momentum thickness definition (4.7a,b) have been used to explain the observed decay of U d in the near wake of slender bodies since the classic wake theories were established (Chevray 1968;Jiménez et al. 2010;Kumar & Mahesh 2018). They both rely on the assumption that U d U ∞ which, based on the present results, is not a good approximation in the near wake of a streamlined body.
After testing the validity of the assumptions involving the velocity deficit and the transverse Reynolds stress, we now move to the scaling of TKE and turbulent dissipation ε. Figure 7(a) shows the peak TKE (K s ) and the peak turbulent shear stress (R s ) scaled by U 2 d . Neither becomes constant, which implies that, at least until x = 80, U 2 d is not the proper scaling for k and u xr . Instead, they satisfy K s ∼ R s ∼ U ∞ U d (dL/dx). The r.m.s. velocities u r , u φ , u x also show the same trend individually. This scaling was observed in the wake of a fractal plate by Dairay et al. (2015) and it can be obtained from (4.4), introducing K s ∼ R s as observed in the simulation.
The inertial estimate, ε ∼ K 3/2 s /L, is tested for the normalised peak dissipation (D s ), black solid curve in figure 7(b). This curve does not asymptote to a constant, showing that inertial scaling does not hold in the region x < 80. An alternative is the non-equilibrium scaling proposed by Nedić et al. (2013), where Re k = √ kL/ν is the local Reynolds number. The non-equilibrium estimate, specifically with n = 2 in (5.1), is found to be a much better choice to scale D s beyond x = 40 as shown by the dashed curve in figure 7(b). The non-equilibrium scaling of dissipation has been observed in bluff-body wakes (Nedić et al. 2013;Dairay et al. 2015;Obligado et al. 2016;) and jets (Cafiero & Vassilicos 2019). It has also been found to hold in unsteady decaying turbulence (Goto & Vassilicos 2016), where it was shown to be related to the imbalance between the large-scale energy of the turbulence and the small-scale dissipation. Whereas in previous bluff-body studies n = 1 was the appropriate exponent to scale dissipation, here n = 2 is a better choice. The different exponent might be related to a different mechanism sustaining the large-scale wake motions, here the imprint of vortex shedding from the body is weak, large-scale coherent structures are only observed late in the wake, and the TKE and dissipation profiles show an off-centre peak whereas in bluff-body wakes such as those studied by Dairay et al. (2015) and Nidhan et al. (2020), the maxima occur at the centreline. Further investigation is still required for a more comprehensive understanding of the non-equilibrium scaling (Vassilicos 2015;Cafiero & Vassilicos 2019). In addition, assessment of whether the turbulent dissipation transitions to the classic inertial estimate at larger downstream distances will require longer spatial domains.
At this point, we have enough information to obtain power laws for {U d , L, K s , R s , D s } in the self-similar far wake. Let U d ∼ x −α . Then L ∼ x α/2 to satisfy the momentum integral constraint. Let K s ∼ x −β . Equation (5.1) with n = 2 leads to D s ∼ x −(β+3α)/2 . We have shown previously that K s ∼ R s ∼ U ∞ U d (dL/dx), so β = 1 + α/2. After balancing the TKE decay with the dissipation, equation (4.5) provides us with an additional constraint for α, whose solution is α = 6/5. Finally, which agree well with the growth rates observed in the simulation for the region 40 < x < 80, It is worth noting that the production, which scales as P s ∼ x −17/5 , drops from the leading-order TKE balance.
Here, it is necessary to remark on the near-wake behaviour. The observed power laws over 5 < x < 15 are U d ∼ x −2/3 and L ∼ x 0.26 . Our simulation and all the previous evidence (see figure 1) show a region of U d ∼ x −0.66 over 5 x 15. However, in this region, the profiles of u xr , k and ε are not self-similar. Thus, (4.4) and (4.5) cannot be invoked. Furthermore, R s U 2 d . The observed U d ∼ x −0.66 is probably a short-lived coincidence and not the asymptotic -2/3 decay. In this region, the first-order momentum integral constraint (U d L 2 ) is not constant and, hence, L x 0.33 . Instead, L grows at a slower rate, L ∼ x 0.26 .
There are three potential causes for the change of wake power laws from the near to far wake: (i) after a transient, the mean flow and turbulence achieve equilibrium, (ii) the local Reynolds number reduces significantly to bring viscosity into play, or (iii) there is a structural change in the wake. In figure 7, it was shown that turbulence and mean flow are not in equilibrium since k U 2 d , R s U 2 d throughout the domain. Regarding (ii), the flow does not show a transition to low local Reynolds number. As can be seen in figure 8, both Re L = U d L/ν and Re k are of the order of 10 3 . In their simulations, Johansson et al. (2003) observed that the flow transitioned to low-Re decay once the local Reynolds number went below 500, here the local Reynolds numbers are well above that threshold. In addition, the microscale Reynolds number, Re λ = √ kλ/ν where λ 2 = 15νu 2 x /ε, is above 100 and the energy spectra show a −5/3 decay throughout the domain (figure 10b).
Regarding (iii), the emergence or decay of coherent structures is a major structural change. The three-dimensional visualisation of figure 9 shows the emergence of a helical structure coinciding with the U d ∼ x −2/3 to U d ∼ x −6/5 transition. The near wake (figure 9(a), also evident in figure 3a,b) is quasi-parallel without any visible large-scale azimuthal asymmetry. This near-wake topology is similar to that found by Kumar & Mahesh (2018) and Posa & Balaras (2016). However, farther downstream (figure 9b) a sinuous single helix, originating at x ≈ 20, is observed. Previous simulations and experiments did not access locations downstream of x ≈ 20 and, hence, did not observe the emergence of helical structures in the intermediate wake region.
To further characterise the sinuous mode, figure 10(a) shows the cross-sectional area-integrated energy of the streamwise velocity fluctuations as a function of the azimuthal mode (m). Initially, at x = 3.5, the turbulence is nearly broadband. As the wake evolves, energy concentrates around the |m| = 1 mode and, progressively, the single helical mode dominates. The spectra (figure 10b) of u x confirms this finding. After x = 10, a peak at St = fD/U = 0.28 is visible and is sustained until the end of the domain. St = 0.28 corresponds to the wavelength of the undulation identified in figure 9(b).
We can further investigate the structure of the wake by looking into the spatio-temporal evolution of the azimuthal modes. Figure 11(a,c,e,g) show the time evolution of the Fourier coefficients of the axisymmetric m = 0 mode and the single helix |m| = 1 mode. The r − t variation of these modes is plotted at two streamwise locations: x = 3.5 and x = 30, representative of the near and the intermediate wake, respectively. The characteristic frequencies of these modes are also identified by performing a temporal Fourier transform.
The corresponding spectra are shown in figure 11 (b,d, f,h).
In the near wake at x = 3.5, which is 0.5D from the trailing end of the spheroid, the |m| = 0 axisymmetric mode is broadband and carries the small-scale turbulence of the boundary layer, see figure 11(a,b). The |m| = 1 mode (figure 11c), however, has distinctive low-frequency motions in the wake core. This low frequency is caused by the oscillation of the recirculation region, which is small and spans r ≈ 0.1 as was shown in figure 3(b). In the spectra of figure 11(d), two low-frequency peaks are observed: St = 0.16 and St = 0.28. At x = 3.5, both m = 0 and |m| = 1 have energy levels comparable with the other nearby modes, as was shown in figure 10(a).
At x = 30, the small scales of the boundary layer have been dissipated and the |m| = 1 low-frequency motion spans the whole wake width. A peak with a characteristic frequency of St = 0.28 is clearly identified in figure 11(h) and is visible in the periodic stripes of figure 11(g). This peak is also observed farther downstream, as was shown in figure 10(b) and corresponds to the undulation observed in the wake visualisations, figures 3(c) and 9(b). In this region, the asymmetric |m| = 1 mode contributes the most to the energy of the flow. Note that the origin of the St = 0.28 oscillation can be traced to the near wake. However, its effect as a coherent structure is not visible until x ≈ 20.
It is well established by experiments and linear stability analysis (Sato & Okada 1966;Wygnanski et al. 1986;Monkewitz 1988;Berger et al. 1990;Rigas et al. 2014), that the wake of axisymmetric bodies is dominated by a global instability with azimuthal wave number |m| = 1. In bluff-body wakes, the |m| = 1 mode dominates the flow starting right at the recirculation region where there is strong asymmetric shedding of the boundary layer (Rigas et al. 2014;Nidhan et al. 2020). In this wake, however, the recirculation region is much smaller, and the asymmetry induced by the separated flow is weaker. The asymmetry is particularly weak because the flow immediately behind the body is surrounded by an axisymmetric region of turbulence originating from the turbulent boundary layer. As a result, the large-scale symmetry breaking of the wake associated with the presence of a coherent helical structure, becomes dominant only farther downstream, at x ≈ 20, when the imprint of the boundary-layer turbulence becomes sufficiently weak. The dominance of the |m| = 1 mode in the intermediate wake coincides with the change of the decay rate and precedes the non-equilibrium scaling of dissipation. The relation between the non-equilibrium scaling of dissipation and the presence of coherent structures was hypothesised by Goto & Vassilicos (2016) in three-dimensional decaying turbulence and more recently by Cafiero & Vassilicos (2019) in turbulent jets. Here also, there is evidence of such a connection. In previous bluff-body wake studies Nidhan et al. 2020), the dominance of the helical |m| = 1 instability and the establishment of a region of non-equilibrium scaling occurs very early in the wake. In the slender-body wake, the helical mode develops farther downstream and so does the non-equilibrium region. Note the strong contribution to the energy of the |m| = 1 mode at x = 30 in figure 10(a) immediately before the non-equilibrium scaling becomes appropriate.
Farther downstream, near the end of the domain, as shown in figure 10(a), the dominance of |m| = 1 recedes. Therefore, the far wake might eventually transition to the classic U d ∼ x −2/3 . Simulations or experiments with longer domains are needed to clarify this.

Conclusions
The turbulent wake of a slender body has been studied using a high-resolution hybrid simulation. The wake generator is a 6 : 1 prolate spheroid with a tripped boundary layer and the diameter-based Reynolds number is Re = 10 5 . To the best of the authors' knowledge this is the highest Re considered in a far-wake study that solves the flow around the wake generator.
Previous investigations of slender-body wakes with turbulent boundary layers have reached a maximum downstream distance that does not exceed x/D = 18 (Chevray 1968;Jiménez et al. 2010;Posa & Balaras 2016;Kumar & Mahesh 2018). Until x/D = 18, the wake decay rate is similar to the classic high-Re behaviour: U d ∼ x −2/3 . However, in this study the domain extends until x/D = 80 and we show that, after x/D ≈ 20, the decay rate changes to U d ∼ x −6/5 .
In the near-wake region (x/D 20), the classic self-similar turbulence scalings (k ∼ u xr ∼ U 2 d and ε ∼ k 3/2 /L) that lead to the asymptotic U d ∼ x −2/3 decay do not hold. The variables have not reached full self-similarity, the magnitude of the second-order terms is still significant and the observed U d ∼ x −2/3 is probably a short-lived coincidence with the classic wake result. After x/D = 40, profiles of k and ε exhibit self-similarity. The transition to the observed far-wake decay law of U d ∼ x −6/5 is derived from the observed non-equilibrium scaling of dissipation ε ∼ (Re/Re k ) 2 k 3/2 /L, which was used previously to explain non-classical decay rates in bluff-body wakes (Nedić et al. 2013;Dairay et al. 2015;Obligado et al. 2016;. The change in wake decay cannot be attributed to the low-Re asymptotic result U d ∼ x −1 , because the local wake Reynolds numbers, Re = U d L/ν, Re k = √ kL/ν and Re λ = √ kλ/ν, where λ is the Taylor microscale, remain large throughout the domain. The decay rate transitions in the region where there is a structural change in the wake, dominance of a sinuous single helix (|m| = 1) azimuthal mode. In a bluff-body wake the strong asymmetric shedding of the boundary layer leads to a dominant helical mode immediately after the recirculation bubble (Berger et al. 1990;Johansson & George 2006b;Nidhan et al. 2020). Here, the separated region behind the slender-body is very small (approximately 0.1D), the near wake is quasi-parallel and dominated by the boundary-layer turbulence leaving the body, and the large-scale coherent helical structure is only visible farther downstream preceding the change of decay rate. This transition, which is not observed in the wake of disks or plates, allows us to elucidate the possible role of the |m| = 1 instability in the decay laws.
The appearance of a large-scale coherent structure in the wake and the dominance of the |m| = 1 mode precede the non-equilibrium scaling region. This finding supports the idea that there is a connection between the presence of coherent structures and the non-equilibrium scaling of dissipation as hypothesised by Goto & Vassilicos (2016) and more recently suggested by Alves Portela, Papadakis & Vassilicos (2018) and Cafiero & Vassilicos (2019). The |m| = 1 dominance is less evident towards the end of the domain (x/D = 80), indicating that the wake might eventually exhibit a further transition in behaviour, possibly to the classical power law.
The present work clarifies and expands the limited available literature on slender-body wakes at high Reynolds number. Streamlined bodies with turbulent boundary layers have many hydrodynamic applications and the characterisation of their wake has significant engineering implications. This is the first study which looks into the far wake of a slender body and describes its transition to complete self-similarity. The study is performed at Re in excess of any previous far-wake studies that include a wake generator, and the domain is significantly longer than any previous work with this type of body. The turbulent wake of a slender body is a more complex problem than that described by classical theory and further investigation is required to answer several remaining open questions. What happens farther downstream beyond 80D behind the body? Is there a change so that the classic decay laws are recovered? If so, do the coherent structures subside? More simulations and experiments are needed to definitively answer these questions.