0


EDITORIAL

J. Fluids Eng. 114, 483 (1992) (1 page);   doi:10.1115/1.2910056
FREE TO VIEW
Abstract

TECHNICAL FORUM

J. Fluids Eng. 114, 484-485 (1992) (2 pages);   doi:10.1115/1.2910057
Abstract
Topics: Fluid mechanics
J. Fluids Eng. 114, 485-486 (1992) (2 pages);   doi:10.1115/1.2910058
Abstract
Topics: Fluids , Engineers

RESEARCH PAPERS

J. Fluids Eng. 114, 487-495 (1992) (9 pages);   doi:10.1115/1.2910059

The possible existence of a “law-of-the-wall” similarity velocity profile for 3-D boundary layers was investigated using nine different proposed relations with the data from nine experiments carried out in 3-D turbulent boundary layers. Both for pressure driven and shear-driven flows, the “law-of-the-wall” relation of Johnston for the local freestream velocity direction component best applies. Although not well described by any relation, the crosswise velocity component of pressure-driven flows and shear-driven flows is best represented by Mager’s relation and Chandrashekhar and Swamy equation, respectively.

J. Fluids Eng. 114, 496-503 (1992) (8 pages);   doi:10.1115/1.2910060

A finite-volume method is presented for calculating incompressible 3-D flows with curved irregular boundaries. The method employs structured nonorthogonal grids, cell-centered variable arrangement, and Cartesian velocity components. A special interpolation procedure for evaluating the mass fluxes at the cell-faces is used to avoid the nonphysical oscillation of flow variables usually encountered with the cell-centered arrangement. The SIMPLE algorithm is used to handle the pressure-velocity coupling. A recently proposed low diffusive and bounded scheme is introduced to approximate the convection terms in the transport equations. The computer code and the relevant data structure are so organized that most of the code except the implicit linear solver used is fully vectorizable so as to exploit the potential of modern vector computers. The capabilities of the numerical procedure are demonstrated by application to a few internal and external three-dimensional laminar flows. In all cases the CPU-time on a grid with typically 28,000 grid nodes was below half a minute.

J. Fluids Eng. 114, 504-511 (1992) (8 pages);   doi:10.1115/1.2910061

A new approach to the solution of the two-dimensional, incompressible, boundary-layer equations based on the Finite Element Method in both directions is investigated. Earlier Finite Element Method treatments of parabolic boundary-layer problems used finite differences in the streamwise direction, thus sacrificing some of the possible advantages of the Finite Element Method. The accuracy and computational efficiency of different interpolation functions for the velocity field are evaluated. A new element especially designed for boundary layer flows is introduced. The effect that the treatment of the continuity equation has on the stability and accuracy of the numerical results is also discussed. The parabolic nature of the equations is exploited in order to reduce the memory requirements. The solution is obtained for one line at a time, thus only two levels are required to be stored at any time. Efficient solvers for tridiagonal and pentadiagonal forms are used for solving the resulting matrix problem. Numerical predictions are compared to analytical and experimental results for laminar and turbulent flows, with and without pressure gradients. The comparisons show very good agreement. Although most of the cases were tested on a mainframe, the low requirements in CPU time and memory storage allows the implementation of the method on a conventional PC.

J. Fluids Eng. 114, 512-521 (1992) (10 pages);   doi:10.1115/1.2910062

This paper describes a study of changes in the vortex formation and the turbulent wake from a circular cylinder with a finite aspect ratio, placed on a ground plane. The experiment was carried out in an N.P.L. blow down type wind-tunnel, with a working section of 500 mm × 500 mm × 2,000 mm, and between the Reynolds number 2.5 × 104 and 4.7 × 104 . The surface-pressure distributions on the circular cylinder were measured and the drag coefficient was determined from these measurements. Vortices of two kinds generated in the flow-field around the cylinder were observed. The power spectrum, auto-correlation, space-correlation, velocity defects, and turbulent intensities in the turbulent wake behind a circular cylinder were also measured. It was found that the flow pattern changed rapidly above aspect ratio H/D = 4, with vortex shedding changing from symmetric “arch” type to antisymmetric “Karman” type.

J. Fluids Eng. 114, 522-529 (1992) (8 pages);   doi:10.1115/1.2910063

An experimental study has been performed to investigate the redeveloping turbulent boundary layer beyond separation-reattachment for a transitional separated flow. By considering the distribution of the intermittency, it has been confirmed that the turbulent structure changes gradually from a mixing layer to a turbulent boundary layer downstream of reattachment. The balances of the respective terms in the turbulent kinetic energy transport equation are evaluated from the energy dissipation rate obtained through numerical integration of the second moment of the energy spectrum. These terms together with those in the shear stress transport equation indicate the recovery process of the redeveloping boundary layer from nonequilibrium to equilibrium.

J. Fluids Eng. 114, 530-536 (1992) (7 pages);   doi:10.1115/1.2910064

Time-resolved measurements of the spanwise vorticity component, ωz , are used to investigate the motions in the outer region of turbulent boundary layers. The measurements were taken in very thick zero pressure gradient boundary layers (R θ = 1010, 2870, 4850) using a four wire probe. As a result of the large boundary layer thickness, at the outer region locations where the measurements were taken the wall-normal and spanwise dimensions of the probe ranged between 0.7 < Δy/η < 1.2 and 2.1 < Δz/η < 3.9, respectively, where η is the local Kolmogorov length. An analysis of vorticity based intermittency is presented near y/δ = 0.6 and 0.85 at each of the Reynolds numbers. The average intermittency is presented as a function of detector threshold level and position in the boundary layer. The spanwise vorticity signals were found to yield average intermittency values at least as large as previous intermittency studies using “surrogate” signals. The average intermittency results do not indicate a region of threshold independence. An analysis of ωz event durations conditioned on the signal amplitude was also performed. The results of this analysis indicate that for decreasing R θ , regions of single-signed ωz increase in size relative to the boundary layer thickness, but decrease in size when normalized by inner variables.

J. Fluids Eng. 114, 537-542 (1992) (6 pages);   doi:10.1115/1.2910065

By introducing the equivalent roughness which is defined as the distance from the wall to where the velocity gets a certain value (u/uτ ≈ 8.5) and which can be represented by a simple function of the roughness, a simple formula to represent the mean-velocity distribution across the inner layer of a turbulent boundary layer is suggested. The suggested equation is general enough to be applicable to turbulent boundary layers over surfaces of any roughnesses covering from very smooth to completely rough surfaces. The suggested velocity profile is then used to get expressions for pipe-friction factors and skin friction coefficients. These equations are consistent with existing experimental observations and embrace well-known equations (e.g., Prandtl’s friction law for smooth pipes and Colebrook’s formula etc.) as special cases.

J. Fluids Eng. 114, 543-553 (1992) (11 pages);   doi:10.1115/1.2910066

An integral method is presented for computing separated and reattached turbulent boundary layers for incompressible two-dimensional flows. This method is a substantial improvement over the inner-variable approach of Das and White (1986), which was based on a direct boundary layer scheme that had several shortcomings. In this new approach, the integral equations have been completely reformulated so that the theory now proceeds in an inverse mode using displacement thickness as input. This new formulation eliminates the need for the second derivative of velocity distribution, which in the past has always been a source of error in all previous inner-variable approaches. Other significant additions are: (a) a single pressure gradient-wake correlation from a large amount of experimental data; and (b) replacement of the wake parameter from the final equations with a more stable parameter, wake velocity. Derivations of integral equations and their final working expressions, in both dimensional and nondimensional forms, are presented in detail. Predictions by this theory for skin friction, freestream velocity, momentum thickness, velocity profile and separation, and reattachment points agree well with experimental data. Sensitivity studies display that the theory is stable against variations in initial conditions, input distributions, and the pressure gradient-wake correlation.

J. Fluids Eng. 114, 554-558 (1992) (5 pages);   doi:10.1115/1.2910067

Experimental results related to the interaction between a uniform flow and a two-dimensional counter flowing wall jet are presented for various ratios of the jet velocity to the freestream velocity. Both visual observations and wall pressure surveys were made in the jet penetration zone. Attempts were made to choose the proper scaling variables to suitably nondimensionalize the wall pressure distributions. The geometrical characteristics of the dividing streamline were determined for a range of test conditions. Limited tests were also carried out to check the influence of the size of the jet injection device on the flow characteristics.

J. Fluids Eng. 114, 559-565 (1992) (7 pages);   doi:10.1115/1.2910068

The growth and development of a horseshoe vortex system in an incompressible, three-dimensional turbulent junction flow were investigated experimentally. A streamlined cylinder mounted with its axis normal to a flat surface was used to generate the junction vortex flow. The flow environment was characterized by a body Reynolds number of 183,000, based on the leading edge diameter of the streamlined cylinder. The study included surface flow visualizations, surface pressure measurements, and mean flow measurements of total pressure, static pressure, and velocity distributions in three planes around the base of the streamlined cylinder, and in two planes in the wake flow. Some characterizations of vortex properties based on the measured mean cross-flow velocity components are presented. The results show the presence of a single large, dominant vortex, with strong evidence of a very small corner vortex in the junction between the cylinder and the flat surface. The center of the dominant vortex drifts away from both the body and the flat surface as the flow develops along and downstream of the body. The growth and development of the core of the large, dominant vortex are documented.

J. Fluids Eng. 114, 566-576 (1992) (11 pages);   doi:10.1115/1.2910069

The flow of a three-dimensional boundary layer approaching an upright wall mounted circular cylinder has been experimentally investigated by means of instantaneous flow visualization techniques using a laser sheet and time resolved measurements of the wall pressure, the gradients of which are related to the vorticity flux away from the wall. The mean separation point of the oncoming boundary layer is located on the plane of symmetry, 0.76 and 0.82 diameters upstream of the cylinder for the two investigated Reynolds numbers, based on the cylinder diameter, of 1.0 × 105 and 2.2 × 105 , respectively. The present flow visualization studies have shown that there is always a primary vortex present in the flow which induces an eruption of wall fluid. Very often, this eruption results in the formation of counter rotating or mushroom vortices. A secondary vortex further upstream has been observed occasionally. This vortex, as well as the vortices formed by the fast eruption of wall fluid evolve quickly in time and space and therefore cannot be obtained from time-average measurements. The primary vortex consists of several large scale structures which have originated in the oncoming boundary layer and which have acquired substantial additional vorticity. Point measurements indicate that the r.m.s. pressure fluctuations increase as separation is approached and reach a maximum near reattachment. A low degree of space-time correlation and longer integral time scales were also observed downstream of separation. A bimodal probability density function of the fluctuating pressure was observed in the vicinity of the mean separation point, close to the corner region and in the wake of the cylinder. Quasi periodic vortex shedding from the cylinder with a Strouhal number 0.13 was also observed.

J. Fluids Eng. 114, 577-584 (1992) (8 pages);   doi:10.1115/1.2910070

A computer model developed by Venkat and Spaulding (1991a) for unsteady flows over vibrating bodies is used to investigate the nonlinear characteristics of external flow over a flat plate, a section of which is subjected to time varying motion of various mode shapes (n). The Reynolds number, Re is fixed at 1000. For the first case, the Strouhal number, St and the vibration amplitude ratio, H0 are fixed at 0.25 and 0.025, respectively while for the second case, St and H0 are increased to 1.0 and 0.1, respectively. Simulations are performed for modes varying in the range 1<n<4. For n=1, upstream and downstream pressure wave propagation is very high compared to higher modes. The transfer of energy from the input frequency to the first harmonic is pronounced for higher modes. A source-sink pair exists over the vibrating section for even modes. For high St and H0 the pressure spectral amplitude of higher harmonics far downstream is quite large for n=4 compared to n=2 thus indicating more nonlinear interaction between the vibrating body and the external flow for large even modes. The pressure coefficient on either side of the vibrating section is controlled by the gradient of vorticity for odd modes and by the convective acceleration terms for even modes.

J. Fluids Eng. 114, 585-592 (1992) (8 pages);   doi:10.1115/1.2910071

An experimental investigation has been carried out in a curved duct of rectangular cross section in order to study the development of flow instability in such geometries. Hot wire anemometry was used to obtain detailed measurements of velocity on the symmetry plane of the duct for different curvature ratios. As the duct Dean number is increased, a centrifugal instability develops and the Dean vortices are seen to oscillate along the inner wall. To understand the contribution of these vortices to the laminar-turbulent transition, time histories and spectra of the flow were taken on the symmetry plane of the duct for different Reynolds numbers. These data reveal a time-periodic motion along the inner wall where the secondary flows originating from the side wall boundary layers collide. The bend angle where this instability develops depends on the Reynolds number while the frequency of the instability depends on the curvature ratio of the bend.

J. Fluids Eng. 114, 593-600 (1992) (8 pages);   doi:10.1115/1.2910072

This paper describes the relationship between hydraulic losses and secondary flow within sinuous conduits with complicated bends. It has been found that the nature of secondary flow present in the bends is quite sensitive to the geometric configuration of the bend and the actual aspect ratio of the conduit section. Indeed, many different secondary flow patterns have been found to exist as the bend geometry is altered. A wide range of experiments has been conducted for various aspect ratios of a rectangular conduit with different curvatures.

J. Fluids Eng. 114, 601-605 (1992) (5 pages);   doi:10.1115/1.2910073

In the quantification of air flow through penetrations in buildings, it is necessary to be able to characterize the flow without detailed knowledge of the geometry of the paths. At the conditions typical of buildings, the flow regime is partially developed laminar flow. This report develops a theoretical description of the hydrodynamic relationship based on a power-law representation between the air flow and applied pressure for laminar flow in short pipes. It is found that short pipes can be described with a simple power law dependence on pressure, but that the exponent of the power law is itself a function of pressure. The entry length of the flow is derived based on a formulation for short, sharp-edged pipes. The theoretical formulation is compared to measured data. A dimensionless quantity, S , is defined to account for the power law behavior and maps simply to the flow exponent. The exponent or S number can be used to infer many of the characteristics of the flow and may prove useful in the inverse problem of determining flow geometry from fluid properties and the measured pressure and flow.

J. Fluids Eng. 114, 606-615 (1992) (10 pages);   doi:10.1115/1.2910074

A two-component LDV system was used to investigate single and two-phase flow in a representative section at the impeller outlet of a centrifugal slurry pump. Measurements were performed at Qn and 0.5 Qn with water and dilute slurry flows (C = 0.04 and 0.16 percent by volume). The solids consist of 0.8 mm glass beads. Particle Reynolds number is 65. The point liquid and solid velocities were determined with a Doppler signal amplitude discrimination approach. A rotary shaft encoder was used to represent the velocity as a function of the impeller angular position, and to investigate the effect of a finite number of blades on the flow. The data on liquid and slip velocity distributions show large scale flow structures which dominate the two-phase turbulent flow. Overall, solid particles have a larger radial velocity than the carrier fluid at the impeller outlet, but they lag the fluid in the circumferential direction.

J. Fluids Eng. 114, 616-620 (1992) (5 pages);   doi:10.1115/1.2910075

The motion of small, monodisperse particles in fluid was studied in a horizontal, cylindrical container rotating about its axis. One instigation for the study was the common requirement for mixed-phase, chemical or biological reactors to maintain particles in suspension for extended periods. A cylindrical, rotating reactor can allow long-term particle suspension without particle collisions and resulting agglomeration. The purpose of this study was to verify parametric effects and optimize the time of particle suspension. The theoretical and experimental results were obtained for inert, constant-diameter particles of nearly neutral buoyancy. The centrifugal buoyancy and gravitation terms were both included in the equations of motion. Laser illumination, photography and computer imaging were used to measure experimental particle concentration.

J. Fluids Eng. 114, 621-625 (1992) (5 pages);   doi:10.1115/1.2910076

It has been shown by Thomas (1958) and Alford (1965), that axial flow turbo-machinery is subject to rotor dynamic destabilizing gas forces produced by the circumferential variation of blade-tip clearance when the rotor is whirling. However, the magnitude and direction of these forces have yet to be clarified. For example, it is still uncertain, under which circumstances the rotor whirl direction will be forward, and when it will be backward, with respect to the rotation. In the present paper, a simple analysis of the perturbed flow in an axial compressor stage with whirling rotor is presented, based on the actuator disc analysis of Horlock and Greitzer (1983), and the gas force on the rotor is calculated on this basis. It appears that in the normal operation range of an axial compressor, the whirl direction is predicted to be forward always. Backward whirl is predicted to take place only at very low flow rates, well below the normally expected stall limit. Experimentally, forces were indeed found in direction of backward whirl for low flow rates, and in direction of forward whirl for high flow rates, in the results reported by Vance and Laudadio (1984), as analyzed by Ehrich (1989). While this experimental evidence supports the present theory qualitatively, a direct comparison of the measured and predicted destabilizing force has yet to be carried out.

J. Fluids Eng. 114, 626-631 (1992) (6 pages);   doi:10.1115/1.2910077

The experimental analysis performed on several small size low area ratio aircraft fuel jet pumps in JP4 is outlined. The variables investigated were area ratio, nozzle and throat diameters, nozzle and suction pressures. The experimental values of head ratio were compared to a one-dimensional theoretical prediction method previously found to be applicable to moderate and high area ratio pumps. The results show the necessity of making some modifications in the model at low flow coefficient values. Measured wall static pressures were also compared with the results of an axisymmetric finite difference turbulent calculation; the comparisons are generally in good agreement. The development of cavitation and related parameters were also investigated. In order to enhance cavitation resistance, which is particularly important in the field of aeronautics, some studies were carried out on two stage jet pumps. The results obtained are outlined and discussed.

J. Fluids Eng. 114, 632-637 (1992) (6 pages);   doi:10.1115/1.2910078

Spectral analyses of all the forces and moments acting on a typical centrifugal pump impeller/volute combination are presented. These exhibit shaft frequencies, blade passing frequencies, and beat frequencies associated with a whirl motion imposed on the shaft in order to measure rotordynamic forces. Among other features the unsteady thrust was found to contain a surprisingly large blade passing harmonic. While previous studies have explored the magnitudes of the steady fluid-induced radial forces and the fluid-induced rotordynamic forces for this typical centrifugal pump impeller/volute combination, this paper presents information on the steady bending moments and rotordynamic moments due to the fluid flow. These imply certain axial locations for the lines of action of the radial and rotordynamic forces. Data on the lines of action are presented and allow inferences on the sources of the forces.

J. Fluids Eng. 114, 638-641 (1992) (4 pages);   doi:10.1115/1.2910079

The three-dimensional Navier-Stokes equation for the motion of ink both inside and outside the nozzle of a bubble jet printer is numerically solved, for the first time, to predict the bubble behavior and the drop ejection. The results of calculation for three types of ink agreed well with experimental data. The effect of initial bubble pressure, viscosity and surface tension on the volume and the velocity of the drop is numerically investigated. The three-dimensional calculation is very useful to the design of bubble jet printers because it saves a lot of time and cost to make and evaluate prototypes.

J. Fluids Eng. 114, 642-647 (1992) (6 pages);   doi:10.1115/1.2910080

Newtonian and non-Newtonian Couette flows through inelastic fluid saturated porous media due to a moving plate boundary have been investigated analytically. The momentum equation which includes both the viscous and inertia terms is solved to examine the effects of the pseudoplasticity, boundary friction, and porous inertia on the velocity profile and wall shear stress at the moving wall. Closed-form exact solutions are presented for three distinct cases of practical interest. A simple expression for the wall shear stress valid for all Newtonian and non-Newtonian cases of small Darcy number has been also derived through an integral treatment, and found to agree closely with the exact values obtained from direct numerical integrations of the momentum equation. The analytical results presented here can be used for possible applications to highly sensitive viscosity measuring devices.

J. Fluids Eng. 114, 648-656 (1992) (9 pages);   doi:10.1115/1.2910081

The present study concerns a particle-laden, swirling flow through a pipe expansion. A gas-particle flow enters the test section through a center tube, and a swirling air stream enters through a coaxial annulus. The swirl number based on the total inflow is 0.47. Numerical predictions of the gas flow were performed using a finite-volume approach for solving the time-averaged Navier-Stokes equations. The predicted mean velocity profiles showed good agreement with experimental results when using the standard k-ε turbulence model. The turbulent kinetic energy of the gas phase, however, is considerably underpredicted by this turbulence model, especially in the initial mixing region of the two jets. The particle dispersion characteristics in this complex flow were studied by using the Lagrangian method for particle tracking and considering the particle size distribution. The influence of the particle phase onto the fluid flow was neglected in the present stage, since only low particle loadings were considered. The particle mean velocities were again predicted reasonably well and differences between experiment and simulation were only found in the velocity fluctuations, which is partly the result of the underpredicted turbulent kinetic energy of the gas phase. The most sensitive parameter for validating the quality of numerical simulations for particle dispersion is the development of the particle mean number diameter which showed reasonable agreement with the experiments, except for the core region of the central recirculation bubble. This, however, is attributed again to the predicted low turbulent kinetic energy of the gas phase.

J. Fluids Eng. 114, 657-666 (1992) (10 pages);   doi:10.1115/1.2910082

The dispersion of particles in a plane mixing layer between two air streams is investigated using experimental and numerical techniques. The results show that large-scale spanwise vortices strongly influence the particle dispersion process. Particles with aerodynamic response times on the order of the large scale vortex time scales are found to concentrate near the outer edges of the vortex structures. Time average velocity measurements also demonstrate that these particles tend to move away from the center of the mixing layer. Substantial changes in the lateral particle dispersion are producible by controlled forcing of the vortex structures. Comparisons between the experimental particle dispersion patterns and numerical simulations show striking similarities. A two-part model involving stretching and folding is suggested as a particle dispersion mechanism.

J. Fluids Eng. 114, 667-671 (1992) (5 pages);   doi:10.1115/1.2910083

The turbulent dispersion of heavy suspended particles in turbulent shear flows is analyzed when crossing trajectory effects are important. A semiempirical expression for particle diffusion coefficient is developed via a comparison with experimental data of two-phase turbulent jet flows. This expression gives the particle momentum diffusion coefficient in terms of the gas diffusion coefficient, mean relatively velocity, and root mean square of the fluctuating fluid velocity. The proposed expression is used in a two-phase flow mathematical model to predict different particle-laden jet flows. The good agreement between the predictions and data suggests that the developed expression for particle diffusion coefficient is reasonably accurate in predicting particle dispersion in turbulent free shear flows.

J. Fluids Eng. 114, 672-679 (1992) (8 pages);   doi:10.1115/1.2910084

Traveling-bubble cavitation inception tests were conducted in a 30.48 cm water tunnel with a Schiebe headform. A computer code was developed to statistically model cavitation inception on a Schiebe headform, consisting of a numerical solution to the Rayleigh-Plesset equation coupled to a set of trajectory equations. Using this code, trajectories and growths were computed for bubbles of varying initial sizes. An initial off-body distance was specified and the bubble was free to follow an off-body trajectory. A Monte Carlo cavitation simulation was performed in which a variety of random processes were modeled. Three different nuclei distributions were specified including one similar to that measured in the water tunnel experiment. The results compared favorably to the experiment. Cavitation inception was shown to be sensitive to nuclei distribution. Off-body effect was also found to be a significant factor in determining whether or not a bubble would cavitate. The effect of off-body trajectories on the critical bubble diameter was examined. The traditional definition of critical diameter based on the minimum pressure coefficient of the body or the measurement of liquid tension was found to be inadequate in defining cavitation inception.

J. Fluids Eng. 114, 680-686 (1992) (7 pages);   doi:10.1115/1.2910085

Results of studies on the dynamics of “clouds” of bubbles via both an analytical technique using asymptotic expansions, and via numerical simulation using a three-dimensional boundary element technique (BEM) are reported. The asymptotic method relies on the assumption that the characteristic bubble size is much smaller than the characteristic inter-bubble distance. Results obtained from the two methods are compared, and are found to agree in the domain of validity of the asymptotic technique, which is for very low void fractions. Next, results of several numerical experiments conducted using the BEM algorithm are reported. The results indicate the influence of the mutual interaction on the dynamics of multiple bubble clouds.

TECHNICAL BRIEFS

J. Fluids Eng. 114, 687-689 (1992) (3 pages);   doi:10.1115/1.2910086

Experiments were carried out in applying the concept of passive device called BLADEs (boundary-layer alteration devices) to fully developed pipe flow to assess its feasibility as a drag reduction device. The results of both the volumetric flow rate measurement and the pipe wall pressure distribution taken far downstream show that there is a net increase in drag with the device. With BLADES in tandem arrangement, there is a further net increase in drag which is contrary to its counterpart in boundary layer flow. Although the wall shear stress measurement following the device indicates some reduction in local drag, its magnitude of reduction is much smaller than that seen in the equivalent boundary flow. All these results suggest little possibility of any useful application of BLADEs to pipe flow.

J. Fluids Eng. 114, 689-692 (1992) (4 pages);   doi:10.1115/1.2910087
Abstract
Topics: Force
J. Fluids Eng. 114, 692-694 (1992) (3 pages);   doi:10.1115/1.2910088

A technique to measure gas-phase turbulence modification by micron-sized particles with thermal anemometry is presented. Bridge output is first digitized and then spikes produced by particle impingement on the hot-wire probe detected using a slope threshold method and replaced by holding the last digital value before each spike. This procedure has negligible effect on flow statistics if spike duration is short compared to the time between spikes.

J. Fluids Eng. 114, 694-697 (1992) (4 pages);   doi:10.1115/1.2910089

A near-wall model for separated turbulent flows is presented and evaluated. The model is based on experimental observations, in particular, the wake-like behavior of turbulence away from walls, the influence of walls on this behavior, the diffusion/dissipation energy balance observed in the backflow portion of detached flow regions, and the near-wall behavior of eddy-viscosity. These lead to two analytically solvable ODE’s for the normal-to-wall behavior of turbulence kinetic energy and eddy-viscosity. The model is applied in conjunction with an outer eddy-viscosity model (such as a k–ε model) to predict eddy-viscosity distribution in separated flow regions. Predictions of two flow cases, using this approach, are shown.

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In