• Journal of Advanced Marine Engineering and Technology
• /
• v.30 no.2
• /
• pp.275-283
• /
• 2006

Experiment and three dimensional numerical investigations of incompressible turbulent flow through square channels with one- and two-sided ribbed walls are performed to determine pressure drop and heat transfer. The CFX(version 5.7) software package is used for the computation. The ribbed walls have a $45^{\circ}$ inclined square rib. Uniform heat flux is maintained on whole inner heat transfer channel area. The numerical results coincide with experimental data that obtained for $7,600{\le}Re{\le}24.900$, the pitch-to-rib height ratio (p/e) of 8.0. and the rib height-to-channel hydraulic diameter ratio ($e/D_h$) of 0.0667. The results show that values of local heat transfer coefficient and friction factor in the channel with two-sided ribbed wall are higher than those in the channel with one-sided ribbed walls.

• Proceedings of the KSME Conference
• /
• 2000.04b
• /
• pp.667-672
• /
• 2000

The Characteristics of the conical vortices on the roof surface of a low-rise building has been investigated by using a PIV(Particle Image Velocimerty) technique. The scaled model of TTU building with 1:92 scaling ratio was used. The Reynolds number based on the free stream velocity and the length of the model was $1.96{\times}10^5$. When the angle of attack for the building model is $45^{\circ}$, the conical vortices are occurred symmetrically and the center of vortices are changed with respect to the angle of the approaching flow. The rotating direction of the conical vortices found to be counter-rotating. The secondary vortex motions are investigated using the instantaneous flow field data.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.28 no.3
• /
• pp.338-348
• /
• 2004

The present study is conducted to investigate the effect of rib arrangements on an impingement/effusion cooling system with initial crossflow. To simulate the impingement/effusion cooling system, two perforated plates are placed in parallel and staggered arrangements with a gap distance of 2 times of tile hole diameter. Initial crossflow passes between the injection and effusion plates, and the square ribs (3mm) are installed on the effusion plate. Both the injection and effusion hole diameters are 10mmand Reynolds number based on the hole diameter and hole-to-hole pitch are fixed to 10,000 and 6 times of the hole diameter, respectively. To investigate the effects of rib arrangements, various rib arrangements, such as 90$^{\circ}$transverse and 45$^{\circ}$angled rib arrangements, are used. Also, the effects of flow rate ratio of crossflow to impinging jets are investigated. With the initial crossflow, locally low transfer regions are formed because the wall jets are swept away, and level of heat transfer rate get decreased with increasing flow rate of crossflow. When the ribs are installed on the effusion plate, the local distributions of heat/mass transfer coefficients around the effusion holes are changed. The local heat/mass transfer around the stagnation regions and the effusion holes are affected by the rib positions, angle of attack and rib spacing. For low blowing ratio, the ribs have adverse effects on heat/mass transfer, but for higher blowing ratios, higher and more uniform heat transfer coefficient distributions are obtained than the case without ribs because the ribs prevent the wall jets from being swept away by the crossflow and increase local turbulence of the flow near the surface. Average heat transfer coefficients with rib turbulators are approximately 10% higher than that without ribs, and the higher values are obtained with small pitch of ribs. However, the attack angle of the rib has little influence on the average heat/mass transfer.

• Korean Journal of Fisheries and Aquatic Sciences
• /
• v.28 no.2
• /
• pp.183-193
• /
• 1995

Assuming that fishing nets are porous structures to suck water into their mouth and then filtrate water out of them, the flow resistance N of nets with wall area S under the velicity v was taken by $R=kSv^2$, and the coefficient k was derived as $$k=c\;Re^{-m}(\frac{S_n}{S_m})n(\frac{S_n}{S})$$ where $R_e$ is the Reynolds' number, $S_m$ the area of net mouth, $S_n$ the total area of net projected to the plane perpendicular to the water flow. Then, the propriety of the above equation and the values of c, m and n were investigated by the experimental results on plane nettings carried out hitherto. The value of c and m were fixed respectively by $240(kg\cdot sec^2/m^4)$ and 0.1 when the representative size on $R_e$ was taken by the ratio k of the volume of bars to the area of meshes, i. e., $$\lambda={\frac{\pi\;d^2}{21\;sin\;2\varphi}$$ where d is the diameter of bars, 21 the mesh size, and 2n the angle between two adjacent bars. The value of n was larger than 1.0 as 1.2 because the wakes occurring at the knots and bars increased the resistance by obstructing the filtration of water through the meshes. In case in which the influence of $R_e$ was negligible, the value of $cR_e\;^{-m}$ became a constant distinguished by the regions of the attack angle $ \theta$ of nettings to the water flow, i. e., 100$(kg\cdot sec^2/m^4)\;in\;45^{\circ}<\theta \leq90^{\circ}\;and\;100(S_m/S)^{0.6}\;(kg\cdot sec^2/m^4)\;in\;0^{\circ}<\theta \leq45^{\circ}$. Thus, the coefficient $k(kg\cdot sec^2/m^4)$ of plane nettings could be obtained by utilizing the above values with $S_m\;and\;S_n$ given respectively by $$S_m=S\;sin\theta$$ and $$S_n=\frac{d}{I}\;\cdot\;\frac{\sqrt{1-cos^2\varphi cos^2\theta}} {sin\varphi\;cos\varphi} \cdot S$$ But, on the occasion of $\theta=0^{\circ}$ k was decided by the roughness of netting surface and so expressed as $$k=9(\frac{d}{I\;cos\varphi})^{0.8}$$ In these results, however, the values of c and m were regarded to be not sufficiently exact because they were obtained from insufficient data and the actual nets had no use for k at $\theta=0^{\circ}$. Therefore, the exact expression of $k(kg\cdotsec^2/m^4)$, for actual nets could De made in the case of no influence of $R_e$ as follows; $$k=100(\frac{S_n}{S_m})^{1.2}\;(\frac{S_m}{S})\;.\;for\;45^{\circ}<\theta \leq90^{\circ}$$, $$k=100(\frac{S_n}{S_m})^{1.2}\;(\frac{S_m}{S})^{1.6}\;.\;for\;0^{\circ}<\theta \leq45^{\circ}$$

• Journal of Power System Engineering
• /
• v.20 no.5
• /
• pp.66-71
• /
• 2016

The measurements of heat transfer and total friction factors for turbulent flows in the convergent rectangular channels with two opposite in-line ribbed walls are reported. The study has covered three different angled ribs ($30^{\circ}$, $45^{\circ}$ and $60^{\circ}$) and Reynolds number in the range of 22,000 to 75,000. The channel, composing of ten isolated copper sections in the length of test section of 1 m, has the channel convergence ratio of $D_{ho}/D_{hi}=0.67$. The results show that the ribs pointing downstream (${\wedge}-shaped$) is somewhat greater than the ribs pointing upstream (V-shaped) in the dimensionless Nusselt number and total friction factors.

• Korean Journal of Fisheries and Aquatic Sciences
• /
• v.28 no.2
• /
• pp.194-201
• /
• 1995

In order to make clear the resistance of bag nets, the resistance R of bag nets with wall area S designed in pyramid shape was measured in a circulating water tank with control of flow velocity v and the coefficient k in $R=kSv^2$ was investigated. The coefficient k showed no change In the nets designed in regular pyramid shape when their mouths were attached alternately to the circular and square frames, because their shape in water became a circular cone in the circular frame and equal to the cone with the exception of the vicinity of frame in the square one. On the other hand, a net designed in right pyramid shape and then attached to a rectangular frame showed an elliptic cone with the exception of the vicinity of frame in water, but produced no significant difference in value of k in comparison with that making a circular cone in water. In the nets making a circular cone in water, k was higher in nets with larger d/l, ratio of diameter d to length I of bars, and decreased as the ratio S/S_m$ of S to the area $S_m$ of net mouth was increased or as the attack angle 9 of net to the water flow was decreased. But the value of ks15m was almost constant in the region of S/S_m=1-4$ or $\theta=15-90^{\circ}$ and in creased linearly in S/S_m>4 or in $\theta<15^{\circ}$ However, these variation of k could be summarized by the equation obtained in the previous paper. That is, the coefficient $k(kg\;\cdot\;sec^2/m^4)$ of bag nets was expressed as $$k=160R_e\;^{-01}(\frac{S_n}{S_m})^{1.2}\;(\frac{S_m}{S})^{1.6}$$ for the condition of $R_e<100$ and $$k=100(\frac{S_n}{S_m})^{1.2}\;(\frac{S_m}{S})^{1.6}$$ for $R_e\geq100$, where $S_n$ is their total area projected to the plane perpendicular to the water flow and $R_e$ the Reynolds' number on which the representative size was taken by the value of $\lambda$ defined as $$\lambda={\frac{\pi d^2}{21\;sin\;2\varphi}$$ where If is the angle between two adjacent bars, d the diameter of bars, and 21 the mesh size. Conclusively, it is clarified that the coefficient k obtained in the previous paper agrees with the experimental results for bag nets.