0

IN THIS ISSUE

TECHNICAL PAPERS

J. Fluids Eng. 2005;129(1):1-14. doi:10.1115/1.2375134.

Air in water flow is a frequent phenomenon in hydraulic structures. The main reason for air entrainment is vortices at water intakes, pumping stations, tunnel inlets, and so on. The accumulated air, in a conduit, can evolve to a different flow pattern, from stratified to pressurized. Among different patterns, slug is most complex with extreme pressure variations. Due to lack of firm relations between pressure and influential parameters, study of slug flow is very important. Based on an experimental model, pressure fluctuations inside a circular, horizontal, and inclined pipe ($90mm$ inside diameter and $10m$ long) carrying tow-phase air-water slug flow has been studied. Pressure fluctuations were sampled simultaneously at different sections, and longitudinal positions. The pressure fluctuations were measured using differential pressure transducers (DPT), while behavior of the air slug was studied using a digital camera. The objective of the paper is to predict the pressure variation in a pipeline or tunnel, involving resonance and shock waves experimentally. The results show that the more intensive phase interaction commences stronger fluctuations. It is shown, that the air-water mixture entering the pipe during rapid filling of surcharging can cause a tremendous pressure surge in the system and may eventually cause failure of the system (e.g., the maximum pressure inside the pipe would reach up to 10 times of upstream hydrostatic pressure as suggested by others too). Relations for forecasting pressure in these situations are presented as a function of flow characteristics, pipe geometry, longitudinal, and cross-sectional positions and head water.

Commentary by Dr. Valentin Fuster
J. Fluids Eng. 2006;129(1):15-22. doi:10.1115/1.2375126.

Cavitation is a serious problem in the development of high-speed turbopumps, and an inducer is often used to avoid cavitation in the main impeller. Thus, the inducer often operates under the worst conditions of cavitation. If it could be possible to control and suppress cavitation in the inducer by some new device, it would also be possible to suppress cavitation occurring in all types of pumps. The purpose of our present study is to develop a new, effective method of controlling and suppressing cavitation in an inducer using shallow grooves, called “J-Grooves.” J-Grooves are installed on the casing wall near the blade tip to use the high axial pressure gradient that exists between the region just downstream of the inducer leading edge and the region immediately upstream of the inducer. The results show that the proper combination of backward-swept inducer with J-Grooves improves the suction performance of the turbopump remarkably, at both partial flow rates and the design flow rate. The rotating backflow cavitation occurring at low flow rates and the cavitation surge which occurs near the best efficiency point can be almost fully suppressed by installing J-Grooves.

Commentary by Dr. Valentin Fuster
J. Fluids Eng. 2006;129(1):23-30. doi:10.1115/1.2375127.

We performed numerical simulation to investigate the effects of wall slip on flow behaviors of Newtonian and non-Newtonian fluids in macro and micro contraction channels. The results show that the wall slip introduces different vortex growth for the flow in micro channel as compared to that in macro channel, which are qualitatively in agreement with experimental results. The effects of slip on bulk flow behaviors depend on rheological property of the fluid. For Newtonian fluid, the wall slip always reduces the vortex length, while for non-Newtonian fluid, the strength of the slip determines whether the vortex length is reduced or increased. Analyses on the velocity and stress fields confirm the channel size dependent phenomena, such as the reduction of wall shear stress with the decrease in channel size. With the increase in average shear rate, the Newtonian fluid shows the reduction of wall shear stress that increases in the same trend with slip velocity-wall shear stress function, while for non-Newtonian fluid, the effect of the slip is suppressed by shear thinning effect and, therefore, the reduction of wall shear stress is less sensitive to the change in average shear rate and slip velocity-wall shear stress function.

Commentary by Dr. Valentin Fuster
J. Fluids Eng. 2006;129(1):31-39. doi:10.1115/1.2375128.

The tangential momentum accommodation coefficient (TMAC) is used to improve the accuracy of fluid flow calculations in the slip flow regime where the continuum assumption of zero fluid velocity at the surface is inaccurate because fluid “slip” occurs. Molecular dynamics techniques are used to study impacts of individual gas atoms upon solid surfaces to understand how approach velocity, crystal geometry, interatomic forces, and adsorbed layers affect the scattering of gas atoms, and their tangential momentum. It is a logical step in development of techniques estimating total TMAC values for investigating flows in micro- and nano-channels or orbital spacecraft where slip flow occurs. TMAC can also help analysis in transitional or free molecular flow regimes. The impacts were modeled using Lennard-Jones potentials. Solid surfaces were modeled approximately three atoms wide by three atoms deep by 40 or more atoms long face centered cubic (100) crystals. The gas was modeled as individual free atoms. Gas approach angles were varied from $10$ to $70deg$ from normal. Gas speed was either specified directly or using a ratio relationship with the Lennard-Jones energy potential (energy ratio). To adequately model the trajectories and maintain conservation of energy, very small time steps (approximately 0.0005 of the natural time unit) were used. For each impact the initial and final tangential momenta were determined and after many atoms, TMAC was calculated. The modeling was validated with available experimental data for He gas atoms at $1770m∕s$ impacting Cu at the given angles. The model agreed within 3% of experimental values and correctly predicted that TMAC changes with angle. Molecular Dynamics results estimate TMAC values from high of 1.2 to low of 0.25, generally estimating higher coefficients at the smaller angles. TMAC values above 1.0 indicate backscattering, which numerous experiments have observed. The ratio of final to initial momentum, when plotted for a gas atom sequence spaced across a lattice cycle typically follows a discontinuous curve, with continuous portions forward and backscattering and discontinuous portions indicating multiple bounces. Increasing the energy ratio above a value of 5 tends to decrease TMAC at all angles. Adsorbed layers atop a surface influence the TMAC in accordance with their energy ratio. Even a single adsorbed layer can have a substantial effect, changing TMAC $+∕−20%$. The results provide encouragement to continue model development and next evaluate gas flows with Maxwell temperature distributions involving numerous impact angles simultaneously.

Commentary by Dr. Valentin Fuster
J. Fluids Eng. 2006;129(1):40-47. doi:10.1115/1.2375133.

Turbulent flow around a rotating circular cylinder has numerous applications including wall shear stress and mass-transfer measurement related to the corrosion studies. It is also of interest in the context of flow over convex surfaces where standard turbulence models perform poorly. The main purpose of this paper is to elucidate the basic turbulence mechanism around a rotating cylinder at low Reynolds numbers to provide a better understanding of flow fundamentals. Direct numerical simulation (DNS) has been performed in a reference frame rotating at constant angular velocity with the cylinder. The governing equations are discretized by using a finite-volume method. As for fully developed channel, pipe, and boundary layer flows, a laminar sublayer, buffer layer, and logarithmic outer region were observed. The level of mean velocity is lower in the buffer and outer regions but the logarithmic region still has a slope equal to the inverse of the von Karman constant. Instantaneous flow visualization revealed that the turbulence length scale typically decreases as the Reynolds number increases. Wavelet analysis provided some insight into the dependence of structural characteristics on wave number. The budget of the turbulent kinetic energy was computed and found to be similar to that in plane channel flow as well as in pipe and zero pressure gradient boundary layer flows. Coriolis effects show as an equivalent production for the azimuthal and radial velocity fluctuations leading to their ratio being lowered relative to similar nonrotating boundary layer flows.

Commentary by Dr. Valentin Fuster
J. Fluids Eng. 2006;129(1):48-54. doi:10.1115/1.2375123.

Under certain opening conditions (partial opening) of a steam control valve, the piping system in a power plant occasionally experiences large vibrations. To understand the valve instability that is responsible for such vibrations, detailed experiments and CFD calculations were performed. As a result of these investigations, it was found that under the middle-opening (partial opening) condition, a complex three-dimensional (3D) flow structure (valve-attached flow) sets up in the valve region leading to a high pressure region on a part of the valve body. As this region rotates circumferentially, it causes a cyclic asymmetric side load on the valve body, which is considered to be the cause of the vibrations.

Commentary by Dr. Valentin Fuster
J. Fluids Eng. 2006;129(1):55-65. doi:10.1115/1.2375130.

In this study, a two-dimensional axisymmetric computational model of spark discharge in air is presented to provide a better understanding of the dynamics of the process. Better understanding of the modeling issues in spark discharge processes is an important issue for the automotive spark plug community. In this work we investigate the evolution of the shock front, temperature, pressure, density, geometry, and flow history of a plasma kernel using various assumptions that are typically used in spark discharge simulations. A continuum, inviscid, heat conducting, single fluid description of the flow is considered with radiative losses. Assuming local thermal equilibrium, the energy input due to resistive heating is determined using a specified current profile and temperature-dependent gas electrical conductivity in the gap. The spark discharge model focuses on the early time flow physics, the relative importance of conduction and radiation losses, the influence of thermodynamic model choice and ambient pressure effects.

Commentary by Dr. Valentin Fuster
J. Fluids Eng. 2006;129(1):66-79. doi:10.1115/1.2375124.

An Eulerian two-fluid computational fluid dynamics model has been developed for flows with microbubble drag reduction (MBDR). This paper focuses on recent validation studies for MBDR flows across a spectrum of Reynolds numbers. Direct numerical simulations and two sets of experimental flat plate boundary layer measurements are studied. In this paper, the interfacial dynamics and other models used are first presented, followed by detailed comparisons with the validation cases. Emphasis is placed on the modeling strategies required to capture measured volume fraction, bubble size, and bubble velocity distributions, as well as skin friction drag reduction.

Commentary by Dr. Valentin Fuster
J. Fluids Eng. 2006;129(1):80-90. doi:10.1115/1.2375129.

In transitional rough pipes, the present work deals with alternate four new scales, the inner wall transitional roughness variable $ζ=Z+∕ϕ$, associated with a particular roughness level, defined by roughness scale $ϕ$ connected with roughness function $▵U+$, the roughness friction Reynolds number $Rϕ$ (based on roughness friction velocity), and roughness Reynolds number $Reϕ$ (based on roughness average velocity) where the mean turbulent flow, little above the roughness sublayer, does not depend on pipes transitional roughness. In these alternate variables, a two layer mean momentum theory is analyzed by the method of matched asymptotic expansions for large Reynolds numbers. The matching of the velocity profile and friction factor by Izakson-Millikan-Kolmogorov hypothesis gives universal log laws that are explicitly independent of pipe roughness. The data of the velocity profile and friction factor on transitional rough pipes provide strong support to universal log laws, having the same constants as for smooth walls. There is no universality of scalings in traditional variables and different expressions are needed for various types of roughness, as suggested, for example, with inflectional-type roughness, monotonic Colebrook-Moody roughness, etc. In traditional variables, the roughness scale, velocity profile, and friction factor prediction for inflectional pipes roughness are supported very well by experimental data.

Commentary by Dr. Valentin Fuster
J. Fluids Eng. 2006;129(1):91-99. doi:10.1115/1.2375125.

Suction was applied asymmetrically to the exhaust of a rectangular subsonic jet creating a pressure field capable of vectoring the primary flow at angles up to $15deg$. The suction simultaneously creates low pressures near the jet exhaust and conditions capable of drawing a secondary flow along the jet shear layer in the direction opposite to the primary jet. This countercurrent shear layer is affected both by the magnitude of the suction source as well as the proximity of an adjacent surface onto which the pressure forces act to achieve vectoring. This confined countercurrent flow gives rise to elevated turbulence levels in the jet shear layer as well as considerable increases in the gradients of the turbulent stresses. The turbulent stresses are responsible for producing a pressure field conducive for vectoring the jet at considerably reduced levels of secondary mass flow than would be possible in their absence.

Commentary by Dr. Valentin Fuster
J. Fluids Eng. 2006;129(1):100-105. doi:10.1115/1.2375131.

This paper presents experimental and analytical results concerning the pressure drop and the core size in vortex chambers. The new formulation is based on the conservation of mass and energy integral equations and takes into account the presence of two outlet ports. The diminishing vortex strength is introduced through the vortex decay factor. The influence of vortex chamber geometry, such as diameter ratio, aspect ratio, and Reynolds number, on the flow field have been examined and compared with the present experimental data. It is shown that the presence of the swirl velocity component makes the pressure drop across a vortex chamber significantly different than the familiar unidirectional pipe flow. When the chamber length is increased, the vortex diminishes under the action of friction, producing a weaker centrifugal force which leads to a further pressure drop. It is revealed that by increasing the Reynolds number, the cores expand resulting into a larger pressure coefficient. For a double-outlet chamber where the flow is divided into two streams, the last parameter is found to be less than that of a single-outlet.

Commentary by Dr. Valentin Fuster
J. Fluids Eng. 2006;129(1):106-115. doi:10.1115/1.2375132.

The unsteady viscous flow and heat transfer in the vicinity of an axisymmetric stagnation point of an infinite rotating circular cylinder with transpiration $U0$ are investigated when the angular velocity and wall temperature or wall heat flux all vary arbitrarily with time. The free stream is steady and with a strain rate of $Γ$. An exact solution of the Navier-Stokes equations and energy equation is derived in this problem. A reduction of these equations is obtained by the use of appropriate transformations for the most general case when the transpiration rate is also time-dependent but results are presented only for uniform values of this quantity. The general self-similar solution is obtained when the angular velocity of the cylinder and its wall temperature or its wall heat flux vary as specified time-dependent functions. In particular, the cylinder may rotate with constant speed, with exponentially increasing/decreasing angular velocity, with harmonically varying rotation speed, or with accelerating/decelerating oscillatory angular speed. For self-similar flow, the surface temperature or its surface heat flux must have the same types of behavior as the cylinder motion. For completeness, sample semi-similar solutions of the unsteady Navier-Stokes equations have been obtained numerically using a finite-difference scheme. Some of these solutions are presented for special cases when the time-dependent rotation velocity of the cylinder is, for example, a step-function. All the solutions above are presented for Reynolds numbers, $Re=Γa2∕2υ$, ranging from 0.1 to 1000 for different values of Prandtl number and for selected values of dimensionless transpiration rate, $S=U0∕Γa$, where $a$ is cylinder radius and $υ$ is kinematic viscosity of the fluid. Dimensionless shear stresses corresponding to all the cases increase with the increase of Reynolds number and suction rate. The maximum value of the shear stress increases with increasing oscillation frequency and amplitude. An interesting result is obtained in which a cylinder rotating with certain exponential angular velocity function and at particular value of Reynolds number is azimuthally stress-free. Heat transfer is independent of cylinder rotation and its coefficient increases with the increasing suction rate, Reynolds number, and Prandtl number. Interesting means of cooling and heating processes of cylinder surface are obtained using different rates of transpiration.

Commentary by Dr. Valentin Fuster

TECHNICAL BRIEF

J. Fluids Eng. 2006;129(1):116-119. doi:10.1115/1.2375135.

The Rayleigh-Taylor instability of a Newtonian viscous fluid overlying Walters $B′$ viscoelastic fluid is considered. For the stable configuration, the system is found to be stable or unstable under certain conditions. However, the system is found to be unstable for the potentially unstable configuration. Further it is found numerically that kinematic viscosity has a destabilizing effect, whereas kinematic viscoelasticity has a stabilizing effect on the system.

Commentary by Dr. Valentin Fuster