Hajime Yokosawa

Measurement of turbulent flow in a square duct with roughened walls on two opposite sides

Abstract A fully developed turbulent flow along a square duct, two opposite walls of which had been roughened by square cross-sectioned ribs, was measured with a hot-wire anemometer. This paper presents the resulting velocities and stresses and compares them with measurements taken in a square duct with four smooth walls. Symmetrical results, with respect to the axes of symmetry of the duct cross section, were obtained in every measured quantity. Terms on both sides of the vorticity transport equation were calculated, and the balance of terms was discussed. As is well known, smooth-walled square ducts yield two secondary flow cells in any given quadrant of a cross section. But in ducts whose opposite walls have been roughened, we found a hitherto unobserved phenomenon: only one relatively large cell appeared in each quadrant of a ducts cross section.

Turbulent heat transfer in a square duct

Abstract The characteristics of the flow and temperature fields in turbulent flow through a straight duct with a square cross section are considerably complex owing to the momentum and heat transport by the secondary flow of the second kind. For a basic understanding of the turbulent heat transport process in such a complex turbulent flow, the detailed characteristics of the turbulent temperature field must be made clear. This paper presents experimental results concerning the fluctuating temperature intensity, cross-correlation coefficients between the fluctuating velocity and temperature, and the turbulent heat fluxes obtained for forced-convection heat transfer in a square duct. The contours of the turbulent heat flux in the streamwise direction are distorted greatly toward the duct corner, similar to those of the fluctuating temperature intensity. Similarity has been also found between the distributions of the turbulent heat flux in the transverse direction and the turbulent shear stress. Based on the results of the measurements, the eddy-viscosity, eddy thermal diffusivity, and turbulent Prandtl number have been also obtained. The distributions of the eddy-diffusivities and turbulent Prandtl numbers on the symmetric axis of a square duct are similar to those in a circular pipe. As the duct corner is approached, however, the eddy-diffusivities become smaller, and the turbulent Prandtl number becomes larger, and thus the assumption of the constant turbulent Prandtl number is invalid for turbulent heat transfer in a square duct.

Experimental Study on Convective Heat Transfer for Turbulent Flow in a Square Duct With a Ribbed Rough Wall (Characteristics of Mean Temperature Field)

We present experimental results concerning a time-mean temperature field obtained in forced convection heat transfer for a turbulent flow through a square duct with a ribbed rough bottom wall. The secondary flow pattern in the duct is reflected in the distribution of the local Nusselt number, the values of which on the smooth walls of the rough duct are 1.71-1.97 times those of the smooth duct. In the upper half cross section near the upper smooth wall opposite the bottom ribbed rough wall, the profile of the mean temperature distribution is similar to that of the primary flow velocity distribution, and the validity of the temperature inner law was confirmed

Secondary flow of the second kind in rectangular ducts with one rough wall

Abstract A study was made to clarify experimentally the influences of a rough wall in a rectangular duct on secondary flow of the second kind and to obtain fundamental data on flow characteristics. Measurements were conducted on turbulent air flows through a square duct and a rectangular (aspect ratio 2:1) duct having one rough wall using hot-wire anemometers. The secondary flow patterrn in the roughened ducts was remarkably different from that in the smooth ducts. In the roughened ducts, only one large longitudinal vortex appeared near the smooth wall on each side, and the secondary currents in the core region proceeded downward from the top smooth wall to the bottom rough wall along the midplate of the duct. The secondary flow was considerably intensified by the existence of the rough wall. The vorticity balance was examined by evaluating the production and convection terms in the vorticity transport equation using the measured turbulent stresses. The two terms almost balanced near the corner bisector of the smooth ducts. However, in the rough ducts, they did not necessarily balance near the corner region formed by the smooth and rough walls.

Influences of velocity gradient on hot-wire anemometry with an X-wire probe

The influence of velocity gradients on hot-wire anemometry with an X-wire probe were investigated. Analysis showed a strong possibility that the values obtained by the usual method of measurements with an X-wire probe involve serious experimental errors caused by the velocity gradients. In order to eliminate these errors, the authors devised an entirely new method of measurement on the basis of the analysis. In this method, two X-wire probes, one of which is the mirror image of the other, are used for measuring a quality at a location. Then an ultimate value of the quantity at the location is given as an arithmetic mean of the values obtained from the two probes. The validity of the analysis was confirmed by measurements of a turbulent flow in a square duct with the present method. The results showed that the present method is very effective for the measurement of quantities such as the secondary flow velocities and the turbulent shear stresses, quantities which are very sensitive to the influences of velocity gradients.

Performance characteristics of energy separation in a steam-operated vortex tube

Abstract Energy separation performance of a steam-operated vortex tube is experimentally investigated, and some reasonable criteria and expressions to estimate the energy separation performance are also introduced. The performance characteristics defined by the above expressions are the same as those of ideal gas in the high superheated region, and are well expressed independently of the degree of superheat, total mass flow rate and discharge resistance. When steam is in the wet region at the nozzle outlet, the performance considerably decreases because of the energy waste from moisture vaporization. And no energy is separated when the dryness fraction is less than approximately 0.98. Some technical data including the optimum operating conditions are also offered.

Turbulence kinetic energy in turbulent flows through square ducts with rib-roughened walls

Abstract Turbulence kinetic energy production and convection were examined in detail for the turbulent flows through various square ducts. The following ducts were used: a smooth duct, a duct with one wall roughened, and a duct with two facing walls roughened. Rough walls were produced by gluing square ribs to a smooth surface repeatedly. By the examinations the turbulent shear stresses were found to play the most important role in the production of the turbulence kinetic energy. Near the rough walls the turbulent normal stresses also have strong influences on the energy production, although their contributions are either positive or negative, depending on the relative streamwise locations with respect to a rib roughness element. The amount of turbulence energy convected by secondary flow is considerably smaller than that of the energy produced. The secondary flow has little influence on the balance of turbulence kinetic energy in the tested ducts.

FULLY DEVELOPED TURBULENT FLOW AND HEAT TRANSFER IN A SQUARE DUCT WITH TWO ROUGHENED FACING WALLS

An experimental and analytical study was conducted on a fully developed turbulent flow in a square duct, two opposite walls of which had been roughened. The flow was measured by means of a hot-wire anemometer. The calculation was performed with the k model and with the k and e model coupled to the algebraic stress model developed by Launder and Ying. The results of the calculation are compared with the measurements. In this duct, a result hitherto unobserved became apparent: namely, there were four relatively large longitudinal vortices. A prediction of heat transfer was also performed under conditions of a constant wall-temperature and a calculated flow field.

Turbulent flow through a rectangular duct with one roughened short-side wall.

An experimental study was conducted on a fully developed turbulent flow through a rectangular duct of aspect ratio 2:1, one of whose long-side walls was roughened. Detailed and reliable results of such quantities as the secondary-flow velocities and the turbulent shear stresses were obtained by using a well-devised technique of the hot wire anemometry developed by the authors. The secondary-flow pattern in the rough duct was remarkably different from that in the smooth duct. Only one large longitudinal vortex appeared in the vicinity of each short-side wall. on the other hand, in quite a large region near the core, the secondary-currents proceeded from the smooth long-side wall to the rough one along the midplane of the rough wall. Furthermore it was found that the production and convection terms in the vorticity equation, which had been believed to balance, did not balance in the rectangular rough duct.