EXPERIMENTAL STUDY ON HEAT TRANSFER AUGMENTATION FOR HIGH HEAT FLUX REMOVAL IN RIB-ROUGHENED NARROW CHANNELS

Transcription

1 JER-Tech--97- JER-Tech 97- JP979 m EXPERMENTL STUDY ON HET TRNSFER UGMENTTON FOR HGH HET FLUX REMOVL N RB-ROUGHENED NRROW CHNNELS July 997 Md.Shafiqul SLM*, RyutaroHNO, Katsuhiro HG Masanori MONDE* and Yukio SUDO VOL 8 m Japan tomic Energy Research nstitute

2 - Mi, (T9- (T9- This report is issued irregularly. nquiries about availability of the reports should be addressed to Research nformation Division, Department of ntellectual Resources, Japan tomic Energy Research nstitute, Tokai-mura, Naka-gun, baraki-ken 9-, Japan. Japan tomic Energy Research nstitute, 997 m

3 JER-Tech 97- Experimental Study on Heat Transfer ugmentation for High Heat Flux Removal in Rib-roughened Narrow Channels Md. Shafiqul SLM*, Ryutaro HNO, Katsuhiro HG + Masanori MONDE* and Yukio SUDO ++ Center for Neutron Science Tokai Research Establishment Japan tomic Energy Research nstitute Tokai-mura, Naka-gun, baraki-ken (Received June, 997) Frictional pressure drop and heat transfer performance in a very narrow rectangular channel having one-sided constant heat flux and repeated-ribs for turbulent flow have been investigated experimentally, and their experimental correlations were obtained using the least square method. The rib pitch-to-height ratios(p/k) were and while holding the rib hgight constant at.mm, the Reynolds number(re) from, to 98,8 under different channel heights of.mm,.97mm, and.mm, the rib height-to-channel equivalent diameter(k/de) of.,., and.9 respectively. The results show that the rib-roughened surface augments heat transfer - times higher than that of the smooth surface with the expense of.8- times higher frictional pressure drop under Re=-, p/k=, and H=.mm. Experimental results obtained by channel height, H=.mm shows a little bit higher heat transfer and friction factor performance than the higher channel height, H=.mm. The effect of fin and consequently higher turbulence intensity are responsible for producing higher heat transfer rates. The obtained correlations could be used to design the cooling passages between the target plates to remove high heat flux up to MW/m generated at target plates in a high-intensity proton accelerator system. Keywords : Narrow Channel, Heat Transfer ugmentation, Rib-Roughened Surface, Correlations, Turbulence ntensity, Cooling Passage, High Heat Flux, Target Plate + Department of dvanced Nuclear Heat Technology, Oarai Research Establishment + + Office of Planning * Saga University

4 JER-Tech 97- Md. Shafiqul slam* > 9-7^6^ lio.mm) O&f t t\ atbhls$(h)sr.mm,.97mm,.mmi:.,. : T9-H + + *

5 JER-Tech 97- Contents. ntroduction. Literature Survey. Heat Transfer ugmentation. Experimental Program. Test pparatus. Data nalysis 6. Results and Discussion 7. Frictional Pressure Drop Performance 7. Heat Transfer Performance 9 6. Concluding Remarks cknowledgements References Nomenclature 9 ppendix g " >CRFWJ=L» «. f-^ss^s 6. ifcfiji^ 9 6. >i;$*# at n fi ^ ^ 9 #»

6 JER-Tech 97-. ntroduction Surfaces roughened from one side with discrete small square ribs in very narrow channels whose heights are so small as approximately to mm that are similar to the gaps between the target plates could be introduced to remove high heat flux up to MW/m generated at target plates in a high-intensity proton accelerator system. The high-intensity proton accelerator of. GeV and m with a beam power of. MW is connected with a target and is now under designing at Japan tomic Energy Research nstitute (JER) for contributing to neutron science. The target consists of heavy metal plates such as, tungsten and these are cooled by water. n designing, heat transfer augmentation is very much important to keep the surface temperature of target plates below X from the view point of safety and extending its life. Surface roughness with repeated-rib promoters is a well established method. The ribs break the viscous sublayer and create local wall turbulence due to flow separation and reattachment between the ribs, which greatly augment heat transfer,usually with an increase in the flow friction and pressure drop. Developing and fully developed turbulence heat transfer and friction in narrow rectangular channels with rib turbulators on one-sided have been studied experimentally. The effects of rib height, rib spacing and rib shape on heat transfer coefficients and friction factors over a wide range of Reynolds numbers have been well established for the very narrow rectangular channels. Empirical friction and heat transfer correlations have been developed for heat transfer designers. However, heat transfer designers are continually seeking the high-performance enhanced surface. The concept of turbulence promoters was examined for this current study. t is of interest how the surface with rougheners can perform better than the surface without rougheners(smooth). The objective of this study is to investigate the following topics on heat removal phenomena of the solid targets. ) the effect of flow conditions such as, inlet water temperature and velocity ) the effect of gap size from mm to mm, ) heat transfer and pressure loss coefficients for single-phase flow under constant heat rate condition. Such a very narrow channel with one-sided repeated-rib promoters with onesided constant heat flux condition can play a predominent role in achieving this - l -

7 JER-Tech 97- goal. Twelve narrow rectangular channels were tested where four of them were smooth and the remaining eight were rib-roughened from one side. Half of the total test sections were tested under heating and remaining without heating conditions at varying channels heights. The rib pitch-to-height ratios (p/k) were and while holding the rib height constant at. mm, the Reynolds number (Re) from, to 98,8 under different channel heights of. mm,.97 mm, and. mm, the rib height-to-channel equivalent diameter (k/d e ) of.,., and.9 respectively. The various investigations carried out on internally augmented tubes including major information are listed in chap... Literature Survey mprovements of heat transfer performance through passive and active augmentation methods have been studied intensively for over three decades. Heat exchangers with single-phase tube-side flow are one of the important applications where passive surface-modification methods are popularly employed. Typical examples of passive augmentation are surface roughness, displaced promoters and vortex generators. Ribs, indentations, spiral flutes, and coil inserts are some common surface modification techniques that are effective for many applications, and can be used conveniently for retubing existing heat exchangers. With the passage of time, the above types of surface roughness are being utilized more and more for augmentation of heat transfer. These methods are used to improve heat transfer coefficients inside tubes or out side tubes or rods. ctive augmentation, which has also been studied extensively, requires the addition of external power to bring about the desired flow modification. Examples include heat transfer surface vibration, fluid vibration, and electrostatic field introduction, are costly and complex. Of these, the most well-known method for augmentation heat transfer is to roughen the surface using repeated-ribs, since these are relatively easy to manufacture, and its simplicity in application. Many investigations have been carried out both analytically and experimentally on tubes and annuli under air-flow conditions where very few studies are on narrow rectangular channels. There are almost no heat transfer data both analytically and experimentally on a very narrow rectangular channel with one-sided repeated-ribs and one-sided constant heat flux under water-flow conditions. o

8 JER-Tech 97- comprehensive list was prepared consisting of all the experimental investigations from which heat transfer and friction factor data could be extracted for different roughened surfaces for single-phase turbulent flow in internally augmented tubes up to now. No restrictions were placed on the flow parameters, namely, the Reynolds number and the Prandtl number, even though much of the data were obtained in the turbulent regime with either air or water as the test fluid. This file lists the authors of the investigation and the type of enhanced tubes along with number of tubes tested in the experiments, which are summarized in Tablei. The sources are listed in the reference section. n addition, the type of experiment, whether the fluid is heated or cooled, and the method of heat transfer are provided whenever possible for future reference. The objective of this list is to make clear that our experiment will be the first step to augment high heat transfer performance in a very narrow rectangular channel with one-sided repeated square ribs, and one-sided constant heat flux under water-flow conditions.. Heat Transfer ugmentation Different techniques are available to augment the heat transfer performance. The most well-known augmentative techniques are as follows : a. Surface promoters b. Displaced promoters c. Vortex flow Surface roughness is one of the promising techniques to be considered seriously as a means of augmenting forced-convection with single-phase flow heat transfer. nitially it was speculated that elevated heat transfer coefficients might accompany the relatively high friction factors characteristic of rough conduits. However, since the commercial roughness is not well defined, artificial surface roughness - introduced through knurling or threading or formed by repeated-rib promoters has been employed. lthough the augmented heat transfer with surface promoters is often due in part to the fin effect, it is difficult to separate the fin contribution. For the experimental data discussed in chap., the augmented heat transfer coefficient is referenced to the envelope diameter. n certain cases it may be desirable to leave the heat transfer surface intact and - -

9 JER-Tech 97- achieve the augmentation by disturbing the flow near the heated surface. Displaced promoters alter flow mechanics near the surface by disturbing the core flow and can be done to place thin rings or discs in a tube. t is seen that rings substantially improve heat transfer in the lower Reynolds number range where discs are not particularly effective. Examples are baffles and mixing elements. t has been established that swirling the flow will improve heat transfer in duct flow. Generation of swirl flow has been accomplished by coiled wires, spiral fins, stationary propellers, coiled tubes, inlet vortex generators, and twisted tapes. Virtually all of these arrangements have been shown to improve single-phase and boiling heat transfer at the expense of increased pumping power. Heat transfer coefficients are relatively high for vortex flow due to increased velocity, secondary flow produced by the radial body force when favorable density gradients are present, and a fin effect with certain continuous swirl generators. Of the several methods discussed above, surface promoters with repeated-ribs are the most popular as they are effective in enhancing the heat transfer, relatively easy to manufacture, and are cost effective for many applications, and was employed at the test section. The ribs break the viscous sublayer thickness and create local wall turbulence due to flow separation and reattachment between the ribs, which greatly augment the heat transfer. t is concluded that the roughness height should exceed the sublayer thickness before it becomes effective. This type of surface roughness is very much effective for augmenting the heat transfer coefficients with the expense of increased frictional pressure drop. The following chapter is concerned with the experiments for the prediction of turbulent friction factors and heat transfer performance in very narrow rectangular channels with repeated-ribs under onesided constant heat flux condition.. Experimental Program. Test apparatus The layout of the experimental apparatus and measuring equipment is shown in Fig.. t is seen that the closed-loop system incorporates the test section, flow meters, a storage tank, a circulating pump, and a D.C power supplier. Water from the pump was supplied to the test section through the flow meter, and then, - -

10 JER-Tech 97- it passed through a filter, a cooler, the storage tank and a preheater to the pump. The filter was used to filter out solid particles more than urn that may have found their way into the system. preheater was employed to control the temperature of the working fluid in the test section. The storage tank served as a container into which the working fluid from the test section was discharged, conditioned, and recirculated. The flow rate was measured using two flow meters in parallel and the flow rate through the test section was controlled by adjusting as required a number of valves installed in the circuit. The test section shown in Fig. was mm in length with a varying cross section. Heated length was mm, an entry and exit sections of mm long, were provided a smooth flow out of the test section. The test section consists of a heating surface made of copper plate with an electric heater, glass windows etc. nichrome plate (.9 //m) was used as a heater and received electricity from the D.C power supply. Heating of the copper test section was effected by the passage of a high current(-) at voltage(-6v) through the test specimen. The total heat input to the heating surface was calculated from a knowledge of the current which was measured by a precision ammeter together with the voltage across the heating surface, which was measured with a voltmeter. The ribs were square-type where cross sectional dimension was. mm x. mm and pitches were or mm. Figure shows the cross section of flow channel and schematic of flow channel with rib-roughened surface. The cross sectional dimensions can be varied according to channel heights. The width of the test section was mm and channel heights can be varied from. mm to. mm. The channel heights found variation from the inlet to the outlet sections and assumed linear variation. The heights were measured by a digital microscope and rechecked by a ultrasonic method. The measuring heights are listed in Figs. (-6). The major parameters along with *N.H /E.H. conditions of the test section are presented in Table. Surface temperature was measured using 6 copper-constantan thermocouples distributed across the span of the copper plates, mm far from the tip of the heating surface, and was recorded by the acquisition system consisting of a data logger and a personal computer. The difference between the two consecutive thermocouples was 8 mm. Three thermocouples of each inlet and outlet of the test section were provided to measure the inlet and outlet *N. H =Non heating E. H = electrical heating - -

11 JER-Tech 97- bulk temperatures of water. Eleven differential pressure taps were installed at opposite side of the heating surface to measure the static pressure drop across the test section. The difference between the two consecutive pressure taps was mm. The range and accuracy of the measuring instruments are given in Table and Table, respectively. The data, which were obtained under the following experimental conditions are listed in Table.. Data analysis The friction factor in a rib-roughened narrow rectangular channel can be determined by measuring the pressure drop across the test section and the flow rate of the water. The friction factor can then be calculated from : f=p/(x /Um /g).(d e /L) () The friction factor.f is based on isothermal condition (test section without heating). The present friction factor was compared with the friction factor for fully developed turbulent flow in smooth rectangular channels by means of the Blasius correlation : f/f o = f/[.79re" ] () n average outlet temperature of the water was used in the following equation () to obtain the net heat transfer rate. The average outlet temperature of the water was obtained by averaging three thermocouples located cross section of the test section at the outlet. The net heat transfer rate can be calculated from : Q=CpPv x (T U, -T in ) () The local heat transfer coefficient was calculated from the net heat transfer rate per unit surface area to the heating water, the corrected surface temperature (Twc) on each measured section, and local bulk mean water temperature as : h = Q/ s (Twc-T b ) () where the local temperature (T w ) was read from each cross-sectional thermocouple output. The corrected local surface temperature (Twc) for equation () was calculated by one dimensional steady state heat conduction equation as: T wc = T w -(Qx)/( s ) () where the axial and span wise heat conductions were not considered in this - 6 -

12 JER-Tech 97- study. The surface area, s was based on the smooth surface area, not including an increase of the area due to ribs. Equation () was used for the rib-roughened surface and the smooth surface heat transfer coefficient calculations. The maximum heat loss was taken place at the entry and exit of the test section. The local inner wall temperature (T w ) was at maximum ^ in the fully developed region at low Reynolds number and smaller channel heights of the test channel. The local inlet bulk water temperature was at best C depending on the test conditions. The local bulk water temperature,t b was calculated assuming a linear water temperature rise along the flow channel and is defined as: T b = T in + (T Ut -T in )x/l (6) The local Nusselt number of the present study was compared with the Nusselt number for fully developed turbulent flow in smooth surfaces correlated by the Dittus-Boelter as: NU/NUD.8 = [Q'.D e /{(T wc -T b )}]/(.Re 8 Pr ) (7) NUD-B =. Re 8 Pr (8) where Q* was surface heat flux supplied by the electrical input.. Results and Discussion. Frictional pressure drop performance Most of the experimental f data using water, air etc. as a working fluid show negligible effects on the Reynolds number for tubes or channels with surface roughness. Citings of such type of experiments are Webb et al. (97), Liou et al.(99), Han et al.(98) etc. For the rib-roughened surface, rib pitch / height plays predominent role to increase the friction factor for turbulent flows. Figure 7 shows the friction factor predictions from the experimental data with the data of Webb et al. for the k/d e =.9, k/d e =.,w/k=. channel geometry and then the k/d=., w/k=. tube geometry with varying p/k ratios. The results show that the friction factor is the highest at p/k= both the narrow channel and tube; then it decreases with increasing or decreasing pitch (p). Frictional pressure drop of the narrow channel causes lower pressure drop than that of the tube. Figures 8 and 9 compare the friction factor with the smooth surface along with - 7 -

13 JER-Tech 97- prediction and experimental results at different channel heights. The smooth surface friction data agrees well enough with the Blasius correlation. The experimental friction data are underpredicted by about 8 times lower than the prediction at p/k=o, Re= and, H=. mm where experimental results predict. times higher than the smooth surface. t is seen that the friction factor strongly depends on Re and decreases with increasing Re and p/k ratios. t is obvious that the roughness increases the friction factor with the expense of pressure drop along the flow direction. For the case of a very narrow rectangular channel with one-sided square ribs, the friction factor is expected to be high. The experimental results show very low friction data for the narrow channels which claim further investigation. Since there are almost no data to predict the frictional pressure drop in a very narrow rectangular channel having one-sided repeated-ribs, it is necessary to develop correlations to agree well with the experimental data. Following experimental correlations will be applicable for those narrow channels whose roughness will be one-sided repetitive nature. These correlations are related with the parameter of Re and expressed by the following equations : f =.7 Re" ; f =. Re" ; p/k=, p/k=, H=.mm H=.mm (9) () f=. Re" ; p/k=, H=. mm () f=. Re" ; p/k=, H=. mm () The solid lines passing through the data points are the least-square fits and the average deviations from the measured data are±., ±., ± and±% respectively. The correlation [] which is used to predict the evaluation results is expressed by the following equation : ( /f) H =. n (D/k) (p/k) () The friction factor affects very small on higher channel heights. t is expected that the friction factor approaches an approximately constant value as the Reynolds number increases which is consistent with the Moody diagram. t can be mentioned that the frictional pressure drop data were taken in the fully developed turbulent flow regime where the location falls under the following categories: x/de=~66 (TP) at p/k= to x/de=~7 (TP) at p/k= while H=.mm, x/de=7~8 (TP) at p/k= to x/de=8~9 (TP) at p/k= while H=.mm for the case of the rib-roughened surface, but for the

14 JER-Tech 97- smooth surface, x/de=~67 (TP), H=.mm and x/de=8~ (TP), H=.97mm. These are shown in Figs. (7-9).. Heat transfer performance Figures (-) show the surface temperature distribution along the length of the test section for the smooth and rib-roughened surfaces at different channel heights over a wide range of the Reynolds number. The results show that surface temperature first increases, then it becomes axially unchanging in the fully developed regime and then it decreases. The increasing and decreasing occurs due to not only entrance and exit effects but also heat losses. Surface temperature is higher at lower Reynolds numbers for the smooth cases but opposite for the roughened surfaces. More scattering data are found at higher Reynolds numbers. Figures (6-7) present the surface and bulk temperature distributions along the length of the test section for different channel heights at lower and higher Reynolds numbers. t is seen that at higher Re, the surface temperature is low and the difference between the surface and bulk temperatures is small, because more heat is taken away by the fluid which assists convection mechanism. These figures also reveal that there is a portion of the test section where the surface and bulk temperature distributions are parallel, yielding a uniform value of (T w -T b ) as expected for constant heat flux. t the inlet and exit of the test section, surface temperature changes remarkably due to entrance and exit effects. The variation of local Nusselt number along the length of the test section for different Re numbers are shown in Figs. (8-). The Nusselt number is large in the entrance and exit regions. t decreases with increasing axial distance approaching the fully developed values. t is also seen that at higher Re, the curve of the local Nu along the length of the test section is higher whereas at lower Re more uniform Nu data are found. Figure shows the Nusselt number predictions from the experimental data with the data of Webb et al. for the k/d e =.9, k/d e =., w/k=. channel geometry and then the k/d=., w/k=. tube geometry with varying p/k ratios. The results show that Nu is the highest at p/k= both the narrow channel and tube and then it decreases with increasing or decreasing p/k ratios. The solid line passing through the data points of Fig. is the least-square fits which is expressed by the following equation : NU/NU[>B =. (p/k)- () - 9 -

15 JER-Tech 97- The maximum scatter of the data points relative to the least-square line is ± 6.%. This correlation is much simpler than the Webb et al.'s correlation, fits much closer to the experimental data and basically shows the rib pitch/height effect on heat transfer only. The smooth surface Nu was calculated from the Dittus-Boelter equation. t is also seen from this figure that this correlation will not be applicable to the narrow channel because large variation of heat transfer performance occurs between the tube and narrow channel. The results would be explained as follows : Since repeated-rib surface is considered as a roughness geometry, it may be also viewed as a problem in boundary layer separation and reattachment. Separation occurs at the rib, forming a widening free shear layer where reattachment occurs down stream from the separation point. reverse flow boundary layer is originated at the reattachment point. The wall shear stress is zero at the reattachment point and increases from zero in the reverse flow and reattachment regions ; however, the direction of the shear stress is opposite in these regions. s reattachment length is a strong function of rib width(w), rib height(k), and channel height(h), it does not occur for p/k less than about eight. So, the maximum heat transfer coefficient occurs in the vicinity of the reattachment point even though heat transfer coefficients in the separated flow region are larger than those of an undisturbed boundary layer. From another point of view, the reason can be explained that the reattachment length behind the rib alters slightly with the rib shape as p/k> and for larger rib spacing, the redeveloping flow beginning at the reattachment point has a larger distance to develop into a thicker boundary layer before the succeeding rib is encountered, hence reduced heat transfer. Figure shows the heat transfer performance for the k/d e =.9, w/k=., H=. mm geometry and varying p/k ratios along with prediction. The prediction of heat transfer performance is calculated by the Webb et al.[] correlation and is expressed by the following equation: St = (f/ ) / [ +(f/) </ [. (k + ) 8 Pr (p/k) ] () From this figure, it is seen that the experimental results predict slightly lower heat transfer performance than the prediction and it augments heat transfer about - times higher than the smooth surface by means of.8- times frictional pressure drop with p/k= and Re= -. s shown in Fig., --

16 JER-Tech 97- the solid lines passing through the data points are the least-square fits and can be expressed by the following equations : Nu =.6 Re 7 Pr ; p/k=, H =. mm (6) Nu =. Re Pr ; p/k=, H =. mm (7) The maximum scatters of the data points relative to the least-square lines are ± and ± % respectively. Figure 6 shows the heat transfer performance for the k/d«=., w/k=. and H=. mm geometry with varying p/k ratios along with prediction and experimental results. The experimental correlations are expressed by the following equations: Nu =.8 Re 6 Pr ; p/k =, H=. mm (8) Nu =.6 Re 6 Pr ; p/k =, H=. mm (9) t is seen from this figure that higher channel heights do not play any significant role from the view point of heat transfer augmentation. The solid lines passing through the data points are the least-square fits and the maximum deviations are ± and ±% respectively. The heat transfer data were taken in the fully developed region where the location falls under the following categories: x/de=9~7 at p/k= to x/de=~76 at p/k= while H=.mm and x/de=7 ~6 at p/k= to x/de=~9 at p/k= while H=.mm for the case of the rib-roughened surface, but for the smooth surface, the position belongs to x'de=~6? while H=.mm to x/de=8~7, H=.mm respectively. They are shown in Figs.(-6). Such a high heat transfer augmentation performance in a very narrow channel with repeated-ribs would be responsible for creation of turbulence eddies, recirculation and overall increase of surface area. Due to the presence of these, the velocity profiles are relatively flat for the rib-roughened surface compared to the smooth surface. The flatter profiles enhance the transfer of momentum and energy, and hence increase surface friction as well as convection heat transfer rates. The causes for the disagreement of experimental results with the prediction seem to be that the prediction correlations are obtained for those experiments whose parameters are large, experimental conditions are different compared to this experiment. The present experiment is not only non uniform heating but also non symmetric rib-roughened surface from one side. n addition, the heat transfer performance reveals the similar tendency. This discrepancy suggests for -li-

17 JER-Tech 97- further investigation. 6. Concluding Remarks The thermal and hydraulic characteristics of the narrow rectangular channels with and without repeated-rib promoters in a single-phase turbulent flow for the target system have been studied experimentally. The friction factor distribution is not consistent with that of the existing works []. For the smooth channel flow, both the measured Nusselt number and friction factor agree well with available theoretical correlations. The main findings are :. Semi-empirical correlations for friction and heat transfer that are applicable to a very narrow rectangular channel with internally one-sided repeated-rib promoters are presented. These correlations could be useful in the design of the cooling passages between the target plates to remove high heat flux in a high-intensity proton accelerator system.. Relative to a smooth surface, the presence of periodic ribs at one-sided walls augments heat transfer - times higher by means of.8- times frictional pressure drop for the ranges of Reynolds number, Re= - s, dimensionless rib/pitch, p/k= and channel height,h=. mm investigated in this current study.. The highest heat transfer augmentation which causes the highest frictional pressure drop occurs at p/k=o compared to other larger or smaller p/k ratios.. Friction factor is a strong function of the Reynolds number from the transition to the fully developed regime and is expected to be a asymptotic value at higher Re.. Reattachment length is expected to be a strong function of rib width (w), rib height (k), and channel height (H) and be independent of the Reynolds number. 6. Higher channel heights do not play any significant role from the view point of heat transfer augmentation. Based on the results in this study, it is recommended for further study including the following:. Symmetric condition of ribs in the channel with uniform heating and ribs of --

18 JER-Tech 97- various shapes,. Non transverse ribs (helix angles other than 9 ), which induce a spiral component to the flow,. Closer rib spacings corresponding to the separation and recirculation mechanism,. Various rib heights to see the effects of its, and. Various channel heights to make clear the effectiveness of channel heights. cknowledgements n this study, the authors wish to express their sincere gratitude to Mr. Hideki ita, Mr. Kenji Sekita, and Mr. Fujisaki for their assistance to carry out the experiments. Without their contributions, the work would not have been possible. Finally with deep sincerity, the authors acknowledge profound indebtness to Dr. M. Z. Hasan, ssoc. professor, Mechanical Engineering Department, Saga University for his guidance, constant encouragement, valuable comments and suggestions. -

19 JER-Tech 97- References [] Bergles,.E., Some perspectives on Enhanced Heat Transfer-Second- Generation Heat Transfer Technology, Trans. SME, J. Heat Transfer, vol., pp. 8-96, 988 [] Rabas, T.J., Selection of the Energy-Efficient Enhancement Geometry for Single-Phase Turbulent Flow inside Tubes, in Heat Transfer Equipment Fundamentals, Design, pplications, and Operating Problems, SME HTD-vol. 8, pp. 9-,989 [] Chiou, J.P., ugmentation of Forced Convection Heat Transfer in a Circular Tube Using Spiral Spring nserts, SME Paper 8-HT [] Zhang, Y.F., Li, F.Y., and Liang, Z.M., Heat Transfer in Spiral-Coiled- nserted Tubes and its pplications, in dvances in Heat Transfer ugmentation and Mixed Convection, SME HTD-Vol.69, pp. - 6,99 [] Ravigururajan, T.S., and Bergles,.E., Study of Water-Side Enhancement for Ocean Thermal Energy Conversion Heat Exchangers, Final Report, College of Engineering, owa State University, mes, 986 [6] Li, H.M., Ye, K.S., Tan, Y.K., and Deng, S.J., nvestigation on Tube-Side Flow Visualization, Friction Factors and Heat Transfer Characteristics of Helical-Ridging Tubes, in Heat Transfer 98, Proc. 7th nt. Heat Transfer Conf., vol., pp. 7-8, 98 [7] Webb, R.L., Turbulent Heat Transfer in Tubes Having Two-Dimensional Roughness, ncluding the Effect of Prandtl number, Ph. D. thesis, University of Minnesota, Minneapolis, MN, 969 [8] Nakayama, M., Takahashi, K., and Daikoku, T., Spiral Ribbing to Enhance Single-Phase Heat Transfer inside Tubes, SME-JME Thermal Engineering Joint Conf., pp. 6-7, 98 [9] Yoshitomi, H., Oba, K., and rima, Y., Heat Transfer and Pressure Drop in Tubes with Embossed Spiral, Thermal and Nuclear Power(in Japanese), vol. 6, pp. 7-8, 976 [] Withers, J.G., Tube-Side Heat Transfer and Pressure drop for Tubes Having Helical nternal Ridging with Turbulent/ Transitional Flow of Single-Phase Fluid. Part. Single-Helix Ridging, Heat Transfer Eng., vol. no.,pp.8-8,98 --

20 JER-Tech 97- [] Withers, J.G., Tube-Side Heat Transfer and Pressure drop for Tubes Having Helical nternal Ridging with Turbulent/ Transitional Flow of Single-Phase Fluid. Part. Multi-Helix Ridging, Heat Transfer Eng., vol. no., pp.-,98 [] Gee, D.L., and Webb, R.L., Forced Convection Heat Transfer in Helically Rib-Roughened Tubes, nt. J. Heat Mass Transfer, vol., pp. 7-6, 98 [] Berger, F.P., and Whitehead,.W., Fluid Flow and Heat Transfer in Tubes with nternal Square Rib Roughening, J. Br. Nuclear Energy Soc, vol., no., pp. -6, pp.6-6, 977 [] Migai, V.K., and Bystrov, P.G., Heat Transfer in Profiled Tubes, Thermal Eng., vol. 8, pp. 78-8, 98 [] Kalinin, E.K., Drietser, G.., Yarkho, S.., and Kusminov, V.., The Experimental Study of the Heat Transfer ntensification under conditions of Forced One-and Two-Phase Flow n Channels, in ugmentation of Convective Heat and Mass Transfer, SME Symposium Volume, pp.8-9, 97 [6] Bolla, G., De Giorgio, G., and Pedrocchi, E., Heat Transfer and Pressure drop Comparison in Tubes with Transverse Ribs and with Twisted Tapes, Energia Nucleare, no., pp. 6-6, 97 [7] Ganeshan, S., and Rao, M.R., Studies on Thermohydraulics of Single and Multi-Start Spirally Corrugated Tubes for Water and Time- ndependent Power Law Fluids, nt. J. Heat Mass Transfer, vol., no.7, pp. -, 98 [8] Gupta, R.H., and Rao, MR., Heat Transfer and Frictional Characteristics of Spirally Enhanced Tubes for Horizontal Condensers, in dvances in Enhanced Heat Transfer, SME Symposium Volume, pp. -,979 [9] Nunner, W., Heat Transfer and Pressure Drop in Rough Tubes, UK tomic Energy uthority, Harwell, tomic Energy Research Establishment Lib/Trans. 786, 96 [] Sams, E. W., Heat Transfer and Pressure-Drop Characteristics of Wire- --

21 JER-Tech 97- Coiled Type Turbulence Promotors, TD-79, Pt., Book, Nov. 97, pp. 9-, 96 [] Kumar, P., and Judd, R.L., Heat Transfer with Coil Wire Turbulence Promotors, Can. J. Chem. Eng., vol. 8, pp. 78-9, 97 [] Novozhilov, J.F., and Migai, V.K., ntensifying Convective Heat Transfer within Tubes by Means of nduced Roughness, Teploenergetika, vol., no.6, pp.6-6, 98 [] Snyder, P.H., ntensifying Convective Heat Transfer for Moderate Prandtl Number and Reynolds number Flows within Tubes by Means of Wire Coil Type Turbulent Promotors, Westinghouse Research Laboratories Report 7-E9-RELQ-R, 97 [] Sethumadhavan, R., and Rao, M. R., Turbulent Flow Heat Transfer and Fluid Friction in Helical-Wire-Coil-lnserted Tubes, nt.j. Heat Mass Transfer, vol. 6, pp. 8-8, 98 [] Molloy.J., Rough Tube Friction Factors and Heat Transfer Coefficients in Laminar and Transition Flow, UKE ERE-R,967 [6] Ravigurujan, T.S., General Correlations for pressure Drop and Heat Transfer for Single-Phase Turbulent Flow in Ribbed Tubes, Ph. D. thesis, owa State University, mes, 986 [7] Mehta, M.H., and Rao, M.R., Heat Transfer and Frictional Characteristics of Spirally Enhanced Tubes for Horizontal Condensers, in dvances in Enhanced Heat Transfer, SME Symposium Volume, pp. -, 979 [8] Cunningham, J., and Milne, H.N., The Effect of Helix ngle on the Performance of Rope Tubes, Proc. Sixth nt. Heat Transfer Conf., vol. pp. 6-6, 978 [9] Mendes, P.R.S., and Mauricio, M.H.P., Heat Transfer, Pressure Drop, and Enhancement Characteristics of the Turbulent through nternally Ribbed Tubes, SME HTD Vol. 8, pp. -, 987 [] Yampolsky, J.S., Libby, P.., Launder, B.E., and LaRue, J.C., 98, Fluid Mechanics and Heat Transfer Spiral Fluted Tubing, Final Report, G Technologies Report G-78, 98 [] Scaggs, W.F., Taylor, P. R., and Coleman, H.W., Measurement and Prediction of Rough Wall Effects of Friction Factor in Turbulent Pipe Flow, Report TFD-88-, Mississippi State University,

22 JER-Tech 97- [] Newson,.H., and Hodgson, T.D., The Development of Enhanced Heat Transfer Condenser Tubing, Desalination, Vol., pp. 9-, 97 [] Sutherland, W.., and Miller, C.W., Heat Transfer to Superheater Steam-, mproved Performance with Turbulent Promoters, USEC Report, GEP-79, 96 [] Smith, J. W., and R.. Gowen, 96, Heat Transfer Efficiency in Rough Pipes at High Prandtl Number, lche J., vol., pp. 9-9,96 [] Takahashi, K., Kayama, W., and Kuwahara, H., Enhancement of Forced Convective Heat Transfer in Tubes Havina Three-Dimensional Spiral Ribs, Trans. JSME, vol., pp. -, 98 [6] Obot, N.T., and Esen, E.B., Heat Transfer and Pressure Drop for air Flow through Enhanced Passages, Report DOE/CE/99-8, Fluid Mechanics, Heat, and Mass Transfer Laboratory Report, FHMT Report 7, Clarkson University, Potsdam, NY, 99 [7] Kidd, G.., The Hear Transfer and pressure-drop Characteristics of Gas Flow inside Spirally Corrugated Tubes, J. Heat Transfer, vol. 9, pp. -9, 97 [8] rman, B., and Rabas, T.J., nfluence of Discrete and Wavy Disruption Shapes on the Performance of Enhanced Tubes, SME/ lch E Natl. Heat Transfer Conf. tlanta, G, 99 [9] Bejan,., 989, Convective Heat Transfer, John Wiley & Sons, New York. [] Berger F.P. and Hau K.-F, F.-L, 979, Local Mass / Heat Transfer Distribution on surfaces Roughened with Small Square Ribs, nt. J. Heat Mass Transfer, Vol., pp [] James, C.., Hodge, B.K., and Taylor, P.R., 99, Discrete Element Method for The Prediction of Friction and Heat Transfer Characteristics of Fully-Developed Turbulent flow in Tubes with Rib-roughness, HTD - Vol. 9, Turbulent Enhanced Heat Transfer SME. [] James, C.., Hodge, B.K., and Taylor, P.R., 99, Discrete Element Predictive Method for Two-Dimensional Rib Roughness in Fully- Developed Turbulent pipe flow, FED - Vol., ndividual papers in Fluids Engineering, SME. [] Han, J.C., Glicksman, L.R., and Rohsenow, W.M.,978, n nvestigation -7-

23 JER-Tech 97- of Heat Transfer and Friction for Rib- Roughened surfaces, Mass Transfer, Vol., pp. -6. nt. J. Heat [] Han, J.C., 98, Heat Transfer and Friction factor in Channels with Two Opposite Rib-Roughened Walls, Transactions of The SME, Vol. pp [] Han, J.C., Park, J.S., and Lei, C.K., 98, Heat Transfer Enhancement in Channels with Turbulence Promoters, Transactions of The SME, Vol. 7, pp [6] PNS Upgrade, 99 : Feasibility Study, NL-9/, rgonne National Laboratory, pril. [7] Liou, -M.T. and Hwang,-J.J., 99, Effect of Ridge Shapes on Turbulent Heat Transfer and Friction in a Rectangular Channel, nt. J. Heat Mass Transfer, Vol. 6, No. pp [8] Liou, -M.T. and Hwang,-J.J., 99, Turbulent Heat Transfer ugmentation and Friction in Periodic fully Developed Channel Flows, Transaction of The SME, Vol., pp [9] Rohsenow, W. M and Hartnett, J.P, 97, Hand book of Heat Transfer, McGraw-Hill, New York. [] Webb, R.L., Eckert, E.R.G., and Goldstein, R.J., 97, Heat Transfer and Friction in Tubes With Repeated-Rib Roughness, nt. J. Heat Mass Transfer, Vol., pp [] Zhang, Y.M. Gu, W.Z., and Han, J.C., 99, Heat Transfer and Friction in Rectangular Channels with Ribbed or Ribbed-Grooved Walls, Transactions of The SME, Vol. 6, pp. 8-6 [] slam. M.S., 99, Study of Heat Transfer Performance of a Tube having nternal Fins, Undergraduate thesis, Bangladesh University of Engineering and Technology. [] slam, M.S., Hino, R., Haga, K., Monde, M., and Sudo, Y., Study on Heat Transfer ugmentation in Rib-Roughened Narrow Channels, JSME Conference, Nagoya, March, 7-8,997, pp [] slam, M.S., Hino, R., Haga, K., Monde, M., and Sudo, Y., nvestigation on Heat Transfer ugmentation for High Heat Flux Removal in Rib- Roughened Narrow Channels, Report, Japan tomic Energy Research nstitute, JER-Tech,

24 JER-Tech 97- Nomenclature a,, a, a bi, b, b empirical constants empirical constants x flow cross sectional area, m s surface area, m C p specific heat, J/kg^ D, equivalent diameter, D e = WH / ( W+H), m f roughened friction factor g acceleration due to gravity, m/s h H k K forced convection heat transfer coefficient, W / m o C channel height, m rib height, m geometry constant k + roughness Reynolds number, k* = (k/d e ) Re (f/)* L Nu p channel length, m Nusselt number, Nu = hd e / rib spacing, m P pressure difference, N / m [ refer to eq.() ] Pr Prandtl number, Pr = C p fi q' heat flux, W / m Re Reynolds number, Re = D e u m / v St Stanton number, St = Nu / Pr Re T b bulk mean temperature, C T w, wall temperature, C u m w W x average fluid velocity, m/s rib width, m width of the channel, m local distance along the axial direction, m Greek symbol H dynamic viscosity, kg/m.s v kinematic viscosity, m /s p density, kg/m S copper material thickness, m X thermal conductivity, WmV increment -9-

25 JER-Tech 97- Subscripts b m we D-B bulk mean smooth surface corrected wall Dittus-Boelter --

26 (T) Thermocouple Pressure gauge Glass window Filter Diff. pressure gauge Rib-roughened surface CO DC power supply Water Cooler Heater element Thermocouple 9 \Thermal insulation > tn S T i?.mpa Pre-heater Water tank Rib-roughened surface Water pump (unit: mm) Fig. Schematic diagram of test apparatus Fig. Schematic drawing of test section

27 Outlet pt~. mm.68 mm. mm.8 mm i.9mm.9mm [. mm.9 mm i.mm.9 mm H=, mm. mm. mm to (S Cross section of flow channel Flow channel withrib-roughenedsurface [. mm. mm [ jw 8 i.mm [. mm [9 [. mm [. mm. mm.6 mm Fig. Schematic drawing of cross section of flow channel with rib-roughened surface nlet plt i. mm.8 mm Fig. Local channel height for two different channels (Smooth)

28 Outlet i.6mm. mm Outlet -it i.7mm[.mm li.6mm. mm.6 mm. mm i,mm [.8 mm >.8 mm. mm [~ [. mm).6mm > mm.6mm [~] [. mm. mm TP TP TP lmm [. mm 6. mm. mm 6mm mm 6mm [.8mm. mm ] CO to, mm] 8 i.7mm. mm. mm > i.mm 8 l9mm.mm,mm] "7" i 9 i.9mm.9mm 9 [. mm. mm~. mm.9 mm [To~l i.8mm.mm nlet TT[ i.mm.9mm nlet J t TT].8 mm.8 mm Fig. Local channel height for two different channels (Roughened surface, p= mm) Fig. 6 Local channel height for two different channels ( Roughened surface, p= mm)

29 Webb et al. (Tube): D=6.8 mm p/k= ; Re=69~ 8 =;Re=6- = ; Re=6~8 8 6 Webb et al.'s correlation O Expt. data p/k = x/de= 66 Expt. data p/k = x/de=~7 D Expt. data (no rib) x/de=~67 CO Present work (Narrow Channel) H=.mm: # p/k= ; Re= x/de= -6 6 = ;Re= x/de=~7 H=.mm: p/k= ; Re= x/de=7~8 = ; Re= x/de=8~9 o o o 8 o o o o S co " 8 e " 8 6 p/k= -^ H=.mm k=. mm w=. mm Present correlation p/k= i - J 6 8 Reynolds number 6 8 Q Rib pitch / height(p/k) ratio Fig. 7 Friction factor predictions from the experimental data with the data of Webb et al. for the k/d e =.9, k/d e =., w/k=. channel geomerty and for the k/d=., w/k=. tube geomerty with varying p/k ratios. Fig. 8 Comparison of friction correlations with Webb et al.

30 an " J H=. mm k=. mm w=. mm p/k= O Expt. data Expt. data D Expt. data Webb et al.'s correlation p/k= Blasius p/k= x/de=7~8 p/k= x/de=8~9 (no rib) x/de=8~ 6 8 Reynolds number 6 8 -in ~ Q. (D Re :7 Re :766 o Re :86 Re :7 "* S 9 * a!* Channel height Fluid Smooth heated :H=. : Water wall xial distance [mm] m o mm * o o O Fig. 9 Comparison of friction correlations with Webb et al. Fig. Local surface temperature distributions along the flow direction of the test section.

31 «Channel height Rib pitch Rib height Fluid : H=. mm : p= mm : k=. mm : water Q. # i D Re: Re: Re: Re: Re: Re: i i Q 8i a. fl fl 8 i Channel height :H=.97 mm Fluid : water Smooth heated wall xial distance [mm] a> a. B * * * 9z " o v " 6 * * * ore :76 Re :976 a Re :777 ore : 789 Re :68 Re : 899 Re. Re : Re :696 xial distance [mm] t s s CO Fig. Local surface temperature distributions along the flow direction of the test section. Fig. Local surface temperature distributions along the flow direction of the test section.

32 to i CO.. * d * «g '. Channel height Rib height Rib pitch Fluid *. m - :H=.mm : k=. mm : p=mm : water * * m B B < ore : 6 Re : 8 a Re : 686 Re : 8996 ore : Re: 6 Re : 9 a. CD CO. Re :67. Re :887. Re :77. o Re : 678. Re : 969. «Re : 88 ' '. ''Si Channel height Rib height Rib pitch Fluid : H=.mm : k=. mm : p=mm : water i xial distance [ mm] xial distance [mm] Fig. Local surface temperature distributions along the flow direction of the test section. Fig. Local surface temperature distributions along the flow direction of the test section.

33 - Q. «B CD Re : 6 Re :7977 Re : o Re :66 «Re :6 «6 * * * Channel height Rib height Rib pitch Fluid B 8 : H-.mm : k=. mm p=mm : water t E Re:7 Tiw o Tw o Local inner wall temperature Local surface temperature Fluid local bulk mean temperature Channel height : H=. mm Fluid : Water Smooth heated wall xial distance [mm] xial distance [mm] Fig. Local surface temperature distributions along the flow direction of the test section. Fig. 6 Local surface and bulk temperature distributions along the flow direction of the test section.

34 6 CD & 8 Tiw o Tw Tb o o Local inner wall Re temperature o o Local surface temperature f ~ Fluid local bulk mean temperature Channel height : H=. mm Fluid : Water Smooth heated wall O o O o Q. Tiw o Tw Tb Re: 7 Local surface temperature 8 8 Local inner wall temperature Fluid local bulk mean temperature TO xial distance [mm] Channel height : H=.97 mm Fluid : water Smooth heated wall xial distance [mm] Fig. 7 Local surface and bulk temperature distributions along the flow direction of the test section. Fig. 8 Local surface and bulk temperature distributions along the flow direction of the test section.

35 Tiw o Tw Tb Local surface temperature Re:767 Local inner wall temperature ; Re: 76 Channel height : H=. mm Rib pitch : p= mm Rib height : k= mm Fluid : water o,-, 6 " t 8 o Fluid local bulk mean temperature Channel height : H=.97 mm Fluid : water Smooth heated wall o o J ' Local surface, *,. temperature o * o ^ < # * o O \ m*o Local inner wall o temperature. Fluid local bulk temperatutre xial distance [mm] xial distance [mm] Fig. 9 Local surface and bulk temperature distributions Fig. along the flow direction of the test section. Local surface and bulk temperature distributions along the flow direction of the test section.

36 Re : 696 Local inner wall temperature.v Tiw o Tw ; Tb Local inner wall Re:6 temperature - _ kl ^ <a a. Channel height Rib pitch Rib height Fluid Local surface temperature Fluid local bulk temperatutre :H=.mm : p= mm : k=. mm : water P & # o ^ o o o o Local surface o ' _ temperature r o o o '.? ' " Fluid local bulk mean temperature [ Channel height: H=.mm Rib height : k=. mm Rib pitch : p=mm! Fluid : water - xial distance [mm] xial distance[mm] Fig. Local surface and bulk temperature distributions along the flow direction of the test section. Fig. Local surface and bulk temperature distributions along the flow direction of the test section.

37 Tiw. o Tw i Tb Local inner wall Re : 9 temperature 6 Tiw o Tw Re: 67 Local inner wall temperature co CO P CD CO CD # o - o. * ^oo o o # oo. Local surface temperature Fluid local bulk mean temperature T '' Channel height: H=.mm Rib height : k=. mm Rib pitch : p=mm Fluid : water 8 CD Q. CD o o Fluid local bulk mean temperature Local surface temperature Channel height: H=.mm Rib height : k=. mm Rib pitch : p=mm Fluid : water > s- n xial distance [mm] xial distance [mm] co S Fig. Local surface and bulk temperature distributions along the flow direction of the test section. Fig. Local surface and bulk temperature distributions along the flow direction of the test section.

38 7 E l- 6, 8 i Tiw o Tw Tbj Local inner wall p e. 88 temperature f.* -" ' " o o o * o # o Local surface o temperature Fluid local bulk mean temperature Channel height: H=.mm Rib height : k=. mm ; Rib pitch : p=mm Fluid : water Q. Q) Re :6 Local inner wall temperature o o o Local surface «* o temperature V Fluid local bulk mean temperature Channel height: H=.mm Rib height : k=. mm Rib pitch : p=mm Fluid : water o Tiw Tw Tb O n xial distance [mm] xial distance [mm] Fig. Local surface and bulk temperature distributions along the flow direction of the test section. Fig. 6 Local surface and bulk temperature distributions along the flow direction of the test section.

39 g Re : 6 Local inner wall temperature o Local surface temperature o o o Tiw o Tw Tb Re: 7 ; Q": kw/m Re : 766 ; Q":.kW/m Re : 86; Q":.kW/m Re : 7; Q": kw/m Channel height : H=. mm Fluid : Water Smooth heated wall CO CD a. Fluid local bulk mean temperature Channel height: H=.mm Rib height : k=. mm Rib pitch : p=mm Fluid : water. c " _ * *.. ' : '. - _ ' * * * so xial distance[mm] xial distance [mm] Fig. 7 Local surface and bulk temperature distributions along the flow direction of the test section. Fig. 8 Local Nusselt number distributions along the flow direction of the test section.

40 Channel Fluid Smooth height heated : H=.97 mm : water wall Channel height Rib pitch Rib height Fluid H=.mm p=mm k=. mm water CO en E ±s CD <O co ' O O ' H. * _ : : : Re : 7; Re: 89; o Re : 89; Re : 9; Re : 767; * o c Q": 7 kw7m Q": 8. kw/m Re: 7; Q": 87. kw/m Q": 9.7 kw/m Q"; 9kW/m Q": 8. kw/m oo!, S CD CO to o o a o Re :76 Q": kw/m o Re :976 Q": kw/m a Re :777 ; Q": kw/m Re:7B9; Q": 6 kw/m o Re.68 ; Q": 6 kw/m ore:899; Q": 6 kw/m Re:; Q": 6 kw/m Re:; Q": 6 kw/m Re :696 ; Q": 6 kw/m xial distance [mm] xial distance [mm] Fig. 9 Local Nusselt number distributions along the flow Fig. direction of the test section. Local Nusselt number distributions along the flow direction of the test section.

41 . E (/) ( z ill? 9 6; Q":kw/m 8 ; Q": kw/m 686; Q": kw/m 8996; Q":kW/m ; Q":89kW/m 6; Q":89kW/m 9 ; Q": 9. kw/m Channel height: H=.mm Rib height : k=. mm Rib pitch p=mm Fluid : water xial distance [mm] S Z C ± D DUUU! Channel height: H=.mm ; Rib height : k=. mm Rib pitch : p=mm Fluid : water * Re : 67 ; Q";. kw/m Re: 887; Q":. kw/m Re : 77 ; Q":. kw/m o Re: 678; Q": 9 kw/m Re : 969 ; Q":. kw/m Re: 88; Q": 77. kw/m ' O O O m ^ * "::?;::hf iiiiiiiiiih: xial distance [mm] S i Fig. Local Nusselt number distributions along the flow direction of the test section. Fig. Local Nusselt number distributions along the flow direction of the test section.

42 E C CD Channel height Rib height Rib pitch Fluid Re:6; Re:7977; Re : 6; : H=.mm : k=. mm p=mm : water Q": 67. kw/m Q": 99kW/m o Re:; Q" : 99kW/m Re :66; Q" :. kw/m Q" : 7. kw/m Webb et al. ( p/k= ; =; =; n Q (Tube): D= 6.8 mm Re = 8 to 697 Re = 78 to Re = 79 to 96 ) Present work ( Narrow Channel) H=.mm: p/k=,re = 77 to 7, x/de=9~7 =, Re=8 to 99, x/de=~76 H=. mm: p/k=, Re = 67 to 898, x/de=7~6 =, Re = 68 to 6, x/de= -9 Nu rib / Nu D. B =. (p/k)" O _ r, * _ xial distance [ mm] 6 Rib pitch / height(p/k) ratio Fig. Local Nusselt number distributions along the flow direction of the test section. Fig. Nusselt number predictions from the experimental data with the data of Webb et al. for the k/d e =.9, k/d e =., w/k=. channel geometry and for the k/d=., w/k=. tube geomerty with varying p/k ratios.

43 " 8 6 Webt) et al. correlator O Expt. data p/k =; X/D e = 9-7 Expt. data Expt. data p/k =; X/D e = -76 (no rib) ; X/D e = -6 " 8 6 Webb et al.'s correlation O Expt. data p/k= x/de=7~6 Expt. data p/k= x/de=~9 Expt. data (no rib) x/de=8 7 ) V) Z 8 6 Present correlation p/k= =^ c p/k= a / y Dittus- Boelter H=. mm k=. mm w=. J 6 8 Reynolds number 6 8 E c (D ( > " 86 J Present correlation 6 8 Reynolds number Dittus-Boelter H=. mm k=. mm w=. mm 6 8 Fig. Comparison of heat transfer correlations with Webb et al. Fig. 6 Comparison of heat transfer correlations with Webb et al.

44 Table nvestigations cited in this study a to Study. Lietal. [6]. Webb [7]. Nakayama et al. [8]. Yoshitomi et al. [9]. Withers [] 6 Withers [] 7. Gee and Webb [] 8. Berger and Whitehead [] 9. Migai and Bystrov [ ]. Kalinin etal. []. Bolla et al. [6]. Ganeshan and Rao [7]. Gupta and Rao [8]. Nunner[9]. Sams [] 6. Kumar and Judd [] 7. Novozhilov and Migai [] 8. Snyder [] 9. Sethumadhavan and Rao []. Chiou [] Molloy[]. Ravigururajan [6]. Mehta and Rao [ 7 ]. Ravigururajan and Bergles []. Zhang et al. [] 6. Cunningham and Milne [8] 7. Mendes and Mauricio [9] 8. Yampolsky et al. [] 9. Scaggs et al. []. Scaggs et al. []. Newson and Hodgson []. Sutherland and Miller []. Smith and Gowen []. Takahashi et al. []. Obot and Esen [6] Tubes tested Type ind rib ind ind ind ind rib rib ind / fl ind rib ind ind rib coil coil coil coil coil coil coil coil ind ind coil ind rib flute d d flute rib d d all Fluid W,W W W W W ir ir ir ir He, N Expt. condition E.H. E.H. E.H. E.H. T=Const T=Const E.H. E.H. T=Const W.cooled W.heated W W.heated W W.heated ir W.heated ir E.H. W E.H. ir E.H. W E.H. W W.heated Oil W.cooled ir E.H. ir W.cooled W W.heated W E.H. ir W.cooled W W.heated W Sublimation W W.heated W SO W SO W W.heated Stm E.H..W.Gly E.H. W E.H. ir E.H. f N-SO-f 8 HT Table Type Smooth Smooth Smooth Smooth Square ribs Square ribs Square ribs Square ribs Square ribs Square ribs Square ribs Square ribs Major parameters and conditions of the expt. H, p/k k/d e *N.H = Non heating, E.H. = electrical heating Expt. condition N. H N. H, E.H. N. H N. H, E.H. N. H N. H, E.H. N. H N. H, E.H. N. H N. H, E.H. N. H N. H, E.H. TO j i o CO to " E.H.= electrical heating, = air, Gly = glycerin, Stm = steam, W = water SO = isothermal, HT = heat transfer

45 Table Thermocouple range and accuracy o Table Range and accuracy of pressure gauges, flow meters, pump, voltmeter and ammeter Category Pressure Tap Flow meter Pump ccuracy ±.% FS ±.% FS ±.% FS ±.% FS ±.% FS ±.% FS ±.% FS ±.% FS ±.% FS ±.% FS Range -Kg f /cm -.Kg f /cm -Kg, /cm -.Kg f /cm -Kg f /cm -.Kg, /cm.kg,/cm -m /h -m /h -Kg f /cm Location Surface Heater mm mm mm mm mm mm 6mm 7mm 8mm 9mm mm mm mm mm mm mm 6mm 7mm 8mm 9mm mm mm mm mm mm mm Thermocouple no. TT TT TT TT TT TT6 TT7 TT8 TT9 TT TT TT TT TT TT TT6 TT7 TT8 TT9 TT TT TT TT TT TT TT6 TT7 TT8 TT9 ccuracy JS. JS. JS. JS. JS. JS. JS. JS. JS. JS. JS. JS. JS. JS. JS. JS. JS. JS. JS. JS. JS. JS. JS. JS. JS. JS. JS. JS. JS. Range - 6 C - 6 C - 6 C - 6 C - 6 C - 6 C - 6 C - 6 C - 6 C - 6 C - 6 C - 6 C - 6 C - 6 C - 6 C - 6 C - 6 C - 6 C - 6 C - 6 C - 6 C - 6 C - 6 C - 6 C - 6 C - 6 C - 6 C - 6 C - 6 C Voltage Current -6V - nlet Outlet TT TT TT TT TT TT JS. JS. JS. JS. JS. JS. - 6 C - 6 C - 6 C - 6 C - 6 C - 6 C

46 Table Experimental conditions Fluid Reynolds number Flow velocity nlet temperature Heat flux System pressure Flow direction Flow condition Heating condition Water, to 98,8 up to m/s upto C up to.7 MW/m up to.mpa Vertical Fully developed turbulent flow Uniform heating from one side ito

47 Tubes of Webb et at. (97): Tube designation : Webb et al. / ( k/d=.,p/k=, w/k=.) of Figure 7 and Figure ppendix Experimental data Major Parameters: Rib width Rib height Rib pitch : w=.8 mm : k=.766 mm : p= 7.66 mm The data points in the graphical comparisons presented in chapt. were carefully taken from the referenced articles/papers [ & ]. n this appendix, the referenced works are first identified by the tube designations and present works are identified by the channel number. The physical rib and tube/channel dimensions are given in a listed form and the experimental data both the tubes and channels are followed by the tabulated data points for the particular tube/channel as read from the figures of the given works. Tube diameter Experimental data : Tables -W and -W Narrow channel (997) : D=6.8 mm Shafiqul et al.: Channel - (k/d e =.9, p/k=io, w/k=.) of Figure 7, Figure 8, Figure and Figure Major parameters: Rib width Rib height Rib pitch Channel height : w=. mm : k=. mm : p= mm :H=.mm Experimental data : Tables -Eand-E

48 Table -W Friction data for Tube / OJ Point Re f T Table -W Heat transfer data for tube / Point Re Nu Table Point Table Point E Friction data for channel - (X/D e =-to-66) Re E Heat fr : transfer data for channel - (X/D e =9-to-7) Re Nu Point Re Nu Point Re Nu > tn

49 Tubes of Webb et al, (97): Table -W Friction data for tube / Tube designation : Point Re Webb et al. / (k/d=., p/k=, w/k=.) of Figure 7 and Figure Major Parameters: Rib width : w=.8 mm Rib height : k=.766 mm Rib pitch : p=.7 mm Tube diameter : D=6.8 mm Experimental data: Tables -W and -W > Narrow channel (997) Shafiqul et al.: Channel - (k/d e =.9, p/k=, w/k=.) of Figure 7, Figure 8, Figure and Figure Major parameters: Table -W Heat transfer data for tube / Rib width : w=. mm Rib height : k=. mm Rib pitch : p= mm Channel height : H=. mm Experimental data: Point Re Nu Tables -E and - E

50 Table - E Friction data for channel - (X/D.=-to-7) Tubes of Webb et al. (97); Point 6 7 Re h Tube designation: Webb et al. / (k/d=., p/k=, w/k=.) of Figure 7 and Figure Major Parameters: en Table Point E Heat transfer data for channel (X/D.= -to-76) Re Nu Point Re Nu Point Re Nu Rib width Rib height Rib pitch Tube diameter Experimental data: Tables -W and 6- W Narrow channel (997) : w=.8 mm : k=.766 mm : p= 9.6 mm : D=6.8 mm Shafiqul et al.: Channel - (k/d e =., p/k=io, w/k=.) of Figure 7, Figure 9, Figure and Figure 6 Major parameters: Rib width Rib height Rib pitch : w=. mm : k=. mm ; p= mm Channel height : H=.mm Experimental data: Tables - E and 6-E 8 o GO to

51 Table -W Friction data for tube / Point Re f T Table 6-W Heat transfer data for tube / Point Re Nu Table Point E Friction data channel (X/D,=7-to-8) Re f T Point Re Table 6 - E Heat transfer data for channel - (X/D,=7-to-6) f T Point Re Nu Point Re Nu o CO CO

52 Table 7 - E Friction data for channel - (X/De=8-to-9) Narrow channel (997) Shafiqul et al.: Channel - (k/d e =., p/k=, w/k=.) of Figure 7, Figure 9, Figure and Figure 6 Major parameters: Rib width Rib height Rib pitch Channel height Experimental data: Tables 7- E and 8- E : w=. mm : k=. mm : p= mm : H=.mm Point Re f T Table 8 - E Heat transfer data for channel - (X/D,=-to-9) Point Re Nu Point Re Nu Point Re Nu

53 Table 9 - E Friction data for channel - (X/D,=-to-67) Point Re f T Uh.-.6/Re *«. oo Narrow channel (Smooth, 997) Channel designation: Shafiqu! et al.: Channel - Figure 8 and Figure Major parameter: Channel height : H=.mm Table - E Heat transfei data for channel - (X/D e =-to-6) Experimental data: Tables 9-E and -E Point Re Nu

54 Table -E Friction data for channel - 6 (X/D,=8-to-) to Narrow channel (Smooth, 997) Channel designation: Shafiqul et al.: Channel -6 Figure 9 and 6 Major parameter: Channel height : H=.97 mm Experimental data: Tables -E and -E Point Table -E Heat transfer data for channel -6 (X/D e =8-to-7) Point Re Re f T Nu fblasius Point Uin,r=.*6/Re Re Nu J ER-Tec zr ^o s

55 m s. X.. «a ft a s * % ft qz ft ft * -b iv > >7 p. «$. ft % ft. a. «J. * si*. ft lit ft t ft ft * T -t 7 y 9" 9? y ^ y 9 ft ft it- ;>J V. MS w * * fi a%ti s ft ft {it fi t 9 y 7 m t i ft ^ ft - y y X 7 /< 7? h '" 7 t' T V T y yu 7 y y *; ;u h 7 # T - 7 K 'J * y 7. x /< T 'S b >L/ > X yu ^ - h i - y D y 'J - > 9 9 U v- *<V y y /P ;w t- y 7 - t y E ^ m kg s K mol cd rad sr Hz N Pa J W C V F n s Wb T H t m lx Bq Gy Sv f ll s"' mkg/s N/m Nm J/s s W/ C/V V/ /V V-s Wb/m Wb/ cd' sr lm/m* s- J/kg J/kg S ft t, U h S ^F, B 7>. t»> y h ;U y «ftw& d ^ min, h, d. L t ev u ev=.68 x "" J u=.66x"" kg % y?"x H D - /-; -,< - a \y y Y y 7 V y - y K S =. nm=-"'m b bar Gal Ci R b=fm! =-! 'm rad rem bar=.mpa= s Pa lci=.7xlo'"bq lr=.8xlo'c/kg rad = lcgy=" Gy rem= csv=! Sv ^ " " " * * ' ' "' HT -' "' "' "" "" "" - f *' H + T T -fe t f 7 T? y <) -f 9 a / i K 7 D t- h 'y -J- =i t- h d E P T G M k h (ii). * - li raiewa^j * us. SB da d c m u n tftir^ 98^fiJalcJ; o tztfl. ev. barli. P f a ^, / -y h, T-'K, barnitfj: N( = s dyn) kgf bf E MPa(=bar) kgf/cm atm mmhg(torr) bf/in (psi) x } tt t Pa-s(N'S/m )=P(*TX)(g/(cm-s)). x -.9 x "'.79 xlo".968 x -' lm /s='st(x K x " J 7.7 x ' 6.86 xlo".79 X J( = 'erg) kgf-m kw- h cald+ftjt) Btu ft lbf ev cal =.86 Jd+fttt) V *' tt * * x f x s x-'.77 x "'.679x-'.97 x ''.7666 x ' x s x x..9679x.86 x lo" x * x" 6.8x".69x".67 x " 6.8 x " 8.6 x " =.8J (»<tt) =.8 J ('O PS((ZJH;>J) = 7 kgf-m/s = 7.99 W.68x".677 xlo". x -".87 x lo".87x-".87 x "" Bq Ci Gy rad C/kg R tt Sv rem.7 x.77 x -»..8 x * 876 ft. (86 if fl

56 EXPERMENTL STUDY ON HET TRNSFER UGMENTTON FOR HGH HET FLUX REMOVL N RB-ROUGHENED NRROW CHNNELS