Swirling turbulent flow through a curved pipe

An experimental study of swirling turbulent flow through a curved bend and its downstream tangent has been carried out. This study reports on the recovery from swirl and bend curvature and relies on measurements obtained in the downstream tangent and data reported in Part 1 to assess the recovery. Unlike the nonswirling flow case, the present measurements show that the cross-stream secondary flow is dominated by the decay of the solid-body rotation and the total wall shear stress measured at the inner and outer bend (furthest away from the bend center of curvature) is approximately equal. The shear distribution is fairly uniform, even at 1 D downstream of the bend exit. At 49D downstream of the bend exit, the mean axial velocity has recovered to its measured profile at 18D upstream of the bend entrance. Furthermore, the mean tangential velocity is close to zero everywhere and the turbulent shear and normal stresses take another 15D to approximately approach their stationary straight pipe values. Therefore, complete flow recovery from swirl and bend curvature takes a total length of about 85D from the bend entrance. This compares with a recovery length of about 78D for bend curvature alone. The recovery length is substantially shorter than that measured previously in swirling flow through straight pipes and is a consequence of the angular momentum decreasing by approximately 74% across the curved bend. Consequently, the effect of bend curvature is to accelerate swirl decay in a pipe flow.List of symbols C _f total skin friction coefficient, = 2 _w / w ₀ ² - D pipe diameter, = 7.62 cm - De Dean number, = ^1/2 Re = 13,874 - M angular momentum - N _s swirl number, = D/2 W ₀ = 1 - r radial coordinate - R mean bend radius of curvature, = 49.5 cm - Re pipe Reynolds number, = DW ₀ /v= 50,000 - S axial coordinate along the upstream (measured negative) and downstream (measured positive) tangent - U, V, W mean velocities along the radial, tangential and axial directions, respectively - u, v, w mean fluctuating velocities along the radial, tangential and axial directions, respectively - u, v, w root mean square normal stress along the radial, tangential and axial directions, respectively - W ₀ mean bulk velocity, 10 m/s - w total wall friction velocity, = _w / - (w ) _s total wall friction velocity measured as S/D = -18 - turbulent shear stresses - pipe-to-bend radius ratio, = D/2R = 0.077 - axial coordinate measured from bend entrance - fluid kinetic viscosity - fluid density - _w total wall shear stress - azimuthal coordinate measured zero from pipe horizontal diameter near outer bend - angular speed of the rotating section     

Secondary flow patterns and mixing in laminar pulsating flow through a curved pipe

Mixing by secondary flow is studied by particle image velocimetry (PIV) in a developing laminar pulsating flow through a circular curved pipe. The pipe curvature ratio is η = r ₀/r _c  = 0.09, and the curvature angle is 90°. Different secondary flow patterns are formed during an oscillation period due to competition among the centrifugal, inertial, and viscous forces. These different secondary-flow structures lead to different transverse-mixing schemes in the flow. Here, transverse mixing enhancement is investigated by imposing different pulsating conditions (Dean number, velocity ratio, and frequency parameter); favorable pulsating conditions for mixing are introduced. To obviate light-refraction effects during PIV measurements, a T-shaped structure is installed downstream of the curved pipe. Experiments are carried out for the Reynolds numbers range 420 ≤ Re_st ≤ 1,000 (Dean numbers 126.6 ≤ Dn ≤ 301.5) corresponding to non-oscillating flow, velocity component ratios 1 ≤ (β = U _max,osc/U _m,st) ≤ 4 (the ratio of velocity amplitude of oscillations to the mean velocity without oscillations), and frequency parameters 8.37 < (α = r ₀(ω/ν)^0.5) < 24.5, where α² is the ratio of viscous diffusion time over the pipe radius to the characteristic oscillation time. The variations in cross-sectional average values of absolute axial vorticity (|ζ|) and transverse strain rate (|ε|) are analyzed in order to quantify mixing. The effects of each parameter (Re_st, β, and α) on transverse mixing are discussed by comparing the dimensionless vorticities (|ζ _P |/|ζ _S |) and dimensionless transverse strain rates (|ε _P |/|ε _S |) during a complete oscillation period.     

Velocity field in isothermal turbulent bubbly gas-liquid flow through a pipe

Velocity field was measured by laser Doppler velocimetry in isothermal, turbulent bubbly gas-liquid flow through a 26.6 mm inner diameter vertical pipe. The measurements were made about 33 diameters downstream from the pipe entrance, gas injection being just upstream of the entrance. The gas phase radial distribution at the measurement plane exhibited influence of the injection device in that higher gas fraction existed in the central region of the pipe. For comparison, velocity field was also measured in isothermal, turbulent single-phase liquid flow through the same pipe at the same axial plane. Measured were the radial distributions of liquid mean axial and radial velocities, axial and radial turbulent intensities, and axial Reynolds shear stress. The radial distributions of gas bubble mean axial velocity and axial velocity fluctuation intensity were also measured by LDV. A dualsensor fiberoptic probe was used at the same time to measure the radial distributions of gas fraction, bubble mean axial velocity and size slightly downstream of the LDV measurement plane.List of Symbols an average gas bubble diameter - f, f _TP friction factor, friction factor for gas-liquid flow - k _L liquid turbulent kinetic energy - , gas, liquid mass flow rate - R inner radius of pipe - r, {sitR}^* radial coordinate; nondimensional radial coordinate (=r/R) - Re _L liquid Reynolds number - U _G mean axial velocity of gas bubble - U _L mean axial velocity of liquid - U _LO mean axial velocity for flow at the total mass velocity with properties of the liquid phase - u _L ⁺ nondimensional mean axial velocity of liquid in wall coordinate - friction velocity - axial velocity fluctuation intensity of liquid - axial velocity fluctuation intensity of gas bubbles - V_L mean radial velocity of liquid - v _L radial velocity fluctuation intensity of liquid - (uv)_L single-point cross-correlation between axial and radial velocity fluctuations of liquid ( axial Reynolds shear stress) - T _in mean liquid temperature at test section inlet - x flow quality - y normal distance from wall - y ⁺ nondimensional normal distance from wall in wall coordinate (=yu/v_L) - _G gas phase residence time fraction - _L rate of dissipation in the liquid - _L Kolmogorov length scale in the liquid - _L liquid kinematic viscosity - _L characteristic turbulence length scale in the liquid - _G, _L density of gas, liquid - _m gas-liquid mixture density This work was partly supported by National Science Foundation, Thermal Transport and Thermal Processing Program, Chemical and Thermal Systems Division, under Grant No. CTS-9411898.     

Unsteady elastico-viscous flow in a curved pipe

Summary An analysis is presented of the flow of an elasticoviscous liquid in a pipe of circular cross-section, coiled in a circle oflarge radiusR, under the influence of a pressure gradient oscillating (about a zero mean) with frequencyn. Of particular interest is the secondary flow set up in the cross-section of the pipe due to centrifugal effects. For sufficiently small values ofn, it is found that the flow pattern approaches that which would be produced by a steady pressure gradient having the same instantaneous value (Thomas andWalters 1963), and of the type that one would be led to expect by centrifugal effects. For large values ofn, the flow in the interior of the pipe, away from the wall, is spectacularly reversed, flow now being directed towards the centre of curvature, elasticity having the effect of reducing the critical (high) frequencyn = n _c which onsets flow reversal.

---

Zusammenfassung Eine Analyse der Strömung einer elasto-viskosen Flüssigkeit in einer Röhre von rundem Querschnitt, die in einem Kreis mit einemgroßen RadiusR gewunden ist, wird unter dem Einfluß eines Druckgradienten, der mit der Frequenzn oszilliert (um einen Nulldurchschnitt), dargestellt. Von besonderem Interesse ist die Sekundärströmung, die in dem Querschnitt der Röhre unter dem Einfluß der Zentrifugalwirkungen entsteht. Für genügend kleine Werte vonn wird herausgefunden, daß das Strömungsbild sich demjenigen annähert, das von einem stationären Druckgradienten erzeugt werden würde, der denselben augenblicklichen Wert besitzt (Thomas andWalters 1963) und dem Typus angehört, den man aufgrund der Zentrifugalwirkungen erwarten würde. Für große Werte vonn wird die Strömung im Inneren der Röhre, d. h. von den Rändern entfernt, in spektakulärer Weise umgekehrt; jetzt wird die Strömung zum Krümmungsmittelpunkt hin gerichtet, und die Elastizität hat die Wirkung, die kritische (hohe) Frequenzn = n _c zu reduzieren, bei der die Strömungsumkehr beginnt.

---

---

With 2 figures     

Hierarchical structures in a turbulent pipe flow

A hierarchical structure (HS) analysis (β-test and γ-test) is applied to a fully developed turbulent pipe flow. Velocity signals are measured at two cross sections in the pipe and at a series of radial locations from the pipe wall. Particular attention is paid to the variation of turbulent statistics at wall units 10<y⁺<3000. It is shown that at all locations the velocity fluctuations satisfy the She–Leveque hierarchical symmetry (Phys. Rev. Lett. 72 (1994) 336). The measured HS parameters, β and γ, are interpreted in terms of the variation of fluid structures. Intense anisotropic fluid structures generated near the wall appear to be more singular than the most intermittent structures in isotropic turbulence and appear to be more outstanding compared to the background fluctuations; this yields a more intermittent velocity signal with smaller γ and β. As turbulence migrates into the logarithmic region, small-scale motions are generated by an energy cascade and large-scale organized structures emerge which are also less singular than the most intermittent structures of isotropic turbulence. At the center, turbulence is nearly isotropic, and β and γ are close to the 1994 She–Leveque predictions. A transition is observed from the logarithmic region to the center in which γ drops and the large-scale organized structures break down. We speculate that it is due to the growing eddy viscosity effects of widely spread turbulent fluctuations in a similar way as in the breakdown of the Taylor vortices in a turbulent Couette–Taylor flow at high Reynolds numbers.     

Structure of turbulent pipe flow

The problem of turbulent flow in a straight circular pipe is solved. We consider a system consisting of the equation of motion, the equation for the turbulence energy, the expression relating the turbulence coefficient with the turbulence scale, and the integral formula for determining the turbulence scale. A numerical solution is presented for this closed system of equations for turbulent flow. The results of calculations are compared with experimental data.     

Rotation effects on a fully-developed turbulent pipe flow

A fully-developed turbulent pipe flow is allowed to pass through a rotating pipe section, whose axis of rotation coincides with the pipe axis. At the exit end of the rotating section, the flow passes into a stationary pipe. As a result of the relaxation of surface rotation, the turbulent flow near the pipe wall is affected by extra turbulence production created by the large circumferential shear strain set up by the rapid decrease of the rotational velocity to zero at the wall. However, the flow in the most part of the pipe is absent of this extra turbulence production because the circumferential strain is zero as a result of the solid-body rotation imparted to the flow by the rotating pipe section. The combined effect of these two phenomena on the flow is investigated in detail using hot-wire anemometry techniques. Both mean and turbulence fields are measured, together with the wall shear and the turbulent burst behavior at the wall. A number of experiments at different rotational speeds are carried out. Therefore, the effects of rotation on the behavior of wall shear, turbulent burst at the wall, turbulence production and the near-wall flow can be documented and analysed in detail.     

Deposition of inertial particles from a turbulent pipe flow

An explicit formula is derived for the rate of deposition of large particles (droplets) on a tube wall in two-phase turbulent flow. Moscow. Translated from Izvestiya Rossiiskoi Akademii Nauk, Mekhanika Zhidkosti i Gaza, No. 2, pp. 68–75, March–April, 1998. The work was financially supported by the International scientific foundation INTAS (grant No. 94-4348) and by the Russian Foundation for Fundamental Research (project No. 97-01-00398).     

Direct numerical simulation of turbulent pipe flow

Fully developed turbulent pipe flow at low Re-number is studied by means of direct numerical simulation (DNS). In contrast to many previous DNS's of turbulent flows in rectangular geometries, the present DNS code, developed for a cylindrical geometry, is based on the finite volume technique rather than being based on a spectral method. The statistical results are compared with experimental data obtained with two different experimental techniques. The agreement between numerical and experimental results is found to be good which indicates that the present DNS code is suited for this kind of numerical simulations.     

Nuclear magnetic resonance study of turbulent flow in a pipe

Results are given from an investigation of longitudinal turbulent diffusion by the nuclear magnetic tracer method, and a technique is described for determining the velocity distribution function of the fluid particles in the pipe cross section.Translated from Izvestiya Akademii Nauk SSSR, Mekhanika Zhidkosti i Gaza, No. 6, pp. 105–110, November–December, 1971.     

Measurements of turbulent developing flow in a moderately curved square duct

Laser-Doppler measurements in the turbulent flow in a right-angled bend of square cross-section, radius/duct-width ratio 7.0, are presented and show the development of secondary circulation in cross-stream planes. Distribution of the streamwise and radial components of the mean velocity and turbulence intensity, and the corresponding Reynolds shear stress, are presented as contour plots and are intended for use in the further development of numerical flow prediction methods.     

Skin Friction following BLADE manipulation in a turbulent pipe flow

This study attempts to analyze the measured wall shear stress distribution downstream of single and tandem BLADEs in fully developed pipe flow. Previous works have indicated the adverse effect of overall drag increase with the single BLADE in both channel and pipe flows, and an even larger drag increase with the tandem BLADES. This is contrary to that observed for external boundary layer flow. Extensive comparisons are then made to the wall shear stress distribution following BLADEs in boundary layer flow, leading to the conclusion of little or no potential in the application of BLADEs alone to pipe flow.     

Fully developed flow in a curved pipe of arbitrary curvature ratio

It is generally assumed in curved pipe flow analyses that the curvature ratio, δ, of the pipe is very small, in which case the flow depends on a single parameter, the Dean number. This is not the case if δ is not very small. To determine the importance of this effect we have numerically solved the full Navier-Stokes equations, in primitive variable form, for arbitrary values of δ. A factored ADI finite-difference scheme has been used, employing Chorin's artificial compressibility technique. The results show that the central-difference calculation on a staggered grid is stable, without adding artificial damping terms, due to coupling between pressure and velocity. A spatially variable time step is used with a fixed Courant number.     

Numerical simulation of turbulent free-surface flow in curved channel

This paper presents first results of numerical simulation of turbulent free-surface flow. Simple implementation of surface capturing method is based on the variable density approach. The flow is treated as if there is only one fluid, but with variable material properties (density, viscosity). The switch in these values is done by a function resulting from the mass conservation principle. This approach simplifies the implementation of turbulence model. In this case the SST k−ω model was chosen in modification given by Hellsten.Numerical solution was carried out by finite-volume method with explicit Runge-Kutta time-integration. The artificial compressibility method was used for time-marching search for steady state solution. The whole model was tested on horizontally placed square-sectioned 90^∘ bend, which was partially filled by the water. The main goal of this study was to demonstrate the applicability of this model and solution method for capturing the water-air interface as well as for predicting the turbulent effects in both fluids.     

Heat transfer characteristics of pulsated turbulent pipe flow

Heat Transfer characteristics of pulsated turbulent pipe flow under different conditions of pulsation frequency, amplitude and Reynolds number were experimentally investigated. The pipe wall was kept at uniform heat flux. Reynolds number was varied from 5000 to 29 000 while frequency of pulsation ranged from 1 to 8 Hz. The results show an enhancement in the local Nusselt number at the entrance region. The rate of enhancement decreased as Re increased. Reduction of heat transfer coefficient was observed at higher frequencies and the effect of pulsation is found to be significant at high Reynolds number. It can be concluded that the effect of pulsation on the mean Nusselt numbers is insignificant at low values of Reynolds number. Received on 29 June 1998     

Large eddy simulation of fully developed turbulent flow in a curved channel

In this paper large eddy simulation of the fully developed turbulent flow in a curved channel is carried out. The computational results are presented and compared with the experimental results of Eskinazi and Yeh^[1]. It is shown that the numerical results of the present LES are reliable and the influence of the curvature on the turbulence feature is correctly revealed.