BOOKCOMP, Inc. — John Wiley & Sons / Page 1067 / 2nd Proofs / Heat Transfer Handbook / Bejan EXTENDED SURFACES 1067 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 [1067], (39) Lines: 787 to 804 ——— 0.927pt PgVar ——— Normal Page PgEnds: T E X [1067], (39) where each geometric variable is identified in Fig. 14.20. These equations quite accurately represent the experimental data continuously in the laminar, transition, and turbulent flow regimes (Manglik and Bergles, 1995). For tube fin heat exchangers, Wang (2000) has given a comprehensive review of 51 patents issued in the period 1981–1991 for a variety of louvered-, wavy-, and slit fin and punched tab (protruded-tab vortex generators) fin surfaces. Wang et al. (2001) have also provided an extended review and compilation of correlations for plain, wavy, louvered, and slit fins. The applicability of these predictive equations to both dry and wet (dehumidifying) conditions has been discussed, along with the issue of hydrophilic coatings. For the wavy fin tube fin coil, a numerical simulation for the air-side performance has been reported by Min and Webb (2001). Metwally and Manglik (2000) have analyzed periodically developed laminar forced convection in wavy plate fin cores. Their numerical model considers two-dimensional sinusoidal- wavy plate fins, which is applicable when the fin height fin spacing, and the results depicted in Fig. 14.21 show a strong influence of the wall waviness on the enhanced performance represented by the area goodness factor (j/f); the development and applicability of this figure of merit for compact heat exchangers are outlined in Shah and London (1978). Subsequent work on three-dimensional wavy plate fin channels (finite fin height) shows trends that are similar, albeit different in magnitude (Muley et al., 2002; Zhang et al., 2002). 14.4.2 Boiling Internally finned tubes are commonly in refrigerant evaporators as well as some other applications for flow boiling. A variety of different geometries are used, and some typical cases made up of large, medium, or low and microfin heights are shown in Fig. 14.22. As pointed out by Bergles (2000), increases in the average heat transfer coefficient of up to 200%, based on the smooth or “empty” tube of the same diameter, have been reported. Their boiling performance, particularly with refrigerants, has been investigated for almost five decades now, with some of the earliest studies reported by Boling et al. (1953), Katz et al. (1955), and Lavin and Young (1965). Figure 14.22 Some types of internally finned tubes used for refrigerant evaporators: (a) strip fin inserts; (b) star-shaped shrunk-fit fin inserts; (c) microfinned tubes. (From Bergles, 2000.) BOOKCOMP, Inc. — John Wiley & Sons / Page 1068 / 2nd Proofs / Heat Transfer Handbook / Bejan 1068 HEAT TRANSFER ENHANCEMENT 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 [1068], (40) Lines: 804 to 820 ——— 0.927pt PgVar ——— Long Page PgEnds: T E X [1068], (40) Extended reviews of this literature are given by Thome (1990), Webb (1994), Bergles (1998), and Kandlikar et al. (1999). In pool boiling of R-114, Hesse (1973) has shown that finned tubes have higher heat transfer coefficients compared with the performance of equivalent smooth tubes. There was, however, no change in the CHF (with q cr referenced to the smooth tube area), which may perhaps be due to the bubble interference in the interfin spaces. Based on an extended survey of finned tubes for pool boiling, Westwater (1973) has suggested that this situation may be remedied by keeping the fin spacing close to the departure diameter of the nucleate boiling bubble for the evaporating fluid; the issues of fin shape and boiling regimes that provide high enhancement levels have also been addressed. Several other studies have shown enhanced boiling of refrigerants and organics from finned surfaces (Bergles et al., 1981; Kakac¸ et al., 1988; Thome, 1990; Webb, 1994; Bergles, 1998). Of the three different types of internally finned tubes shown in Fig. 14.22, Schl ¨ un- der and Chawla (1969) and Pearson and Young (1970) have reported boiling data for R-11 and R-22, respectively, with different star-shaped shrunk-fit fin inserts (Fig. 14.22b). Additional data for R-122 and R-22 are given by Lavin and Young (1965), and for R-22 by Kubanek and Miletti (1979). In the former study, the effects of subcooling as well as boiling in different flow regimes are also discussed. However, in most heat exchangers for refrigeration and air-conditioning systems that are manufactured today, microfinned tubes (Fig. 14.22c) are invariably used Figure 14.23 Enhanced boiling heat transfer coefficients for R-11 flows in internally finned tubes reported by Kubanek and Miletti (1979). BOOKCOMP, Inc. — John Wiley & Sons / Page 1069 / 2nd Proofs / Heat Transfer Handbook / Bejan EXTENDED SURFACES 1069 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 [1069], (41) Lines: 820 to 832 ——— 0.927pt PgVar ——— Long Page PgEnds: T E X [1069], (41) (Webb, 1994). Some of the earliest experimental studies with such tubes were re- ported by Fujie et al. (1977), Itoh and Kimura (1979), Kubanek and Miletti (1979), and Shinohara et al. (1987). The R-11 boiling data of Kubanek and Miletti (1979) for finned tubes, which includes a tube with a star-shaped fin insert and a smooth tube, are shown in Fig. 14.23. Here the heat transfer coefficients are based on the same- diameter smooth tube surface area, and the enhanced performance is self-evident. Much of the current focus of work on microfin tubes is their performance with boil- ing of alternative (or chlorine-free substitutes) refrigerants. New data have been re- ported, for example, by Kedzierski (1993) for R-123 and by Eckels et al. (1994a) for R-134a. In a very recent new development, microfins machined in a herringbone fashion, instead of the conventional spiral or helical pattern, on the inner tube surface have been produced and tested with R-22 and R-407c (Ebisu and Torikoshi, 1998; Torikoshi and Ebisu, 1999). The data suggest a 80 to 100% increase in the average boiling heat transfer coefficients compared to that for conventional microfin tubes for the same conditions. As shown schematically in Fig. 14.24, Ebisu and Torikoshi (1998) attribute this improved performance to a better distribution of the liquid layer at the wall and its more circumferentially uniform thinning. Figure 14.24 Liquid film distribution and enhancement mechanisms in conventional spiral and herringbone microfin tubes. (From Ebisu and Torikoshi, 1998.) BOOKCOMP, Inc. — John Wiley & Sons / Page 1070 / 2nd Proofs / Heat Transfer Handbook / Bejan 1070 HEAT TRANSFER ENHANCEMENT 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 [1070], (42) Lines: 832 to 850 ——— 2.26314pt PgVar ——— Normal Page PgEnds: T E X [1070], (42) 14.4.3 Condensing Extended surfaces that include a variety of large-, medium-, and micro-sized fins are used extensively for condensation heat transfer enhancement in power, process, air-conditioning, and refrigeration applications. The heat exchangers in these duties involve both horizontal and vertical tube condensers, with fins on the inside or outside surfaces (or both) of tubes. In some applications, compact plate fin heat exchangers with offset strip fin cores have also been considered, and Fujii (1995) and Torikoshi and Ebisu (1999) have presented good review summaries of some of the newer developments. For integral fin tubes, besides the increased surface area, higher heat transfer coef- ficients are obtained because a relatively thin condensate film tends to be formed near the fin tips, and surface tension forces pull the condensate into the interfin grooved spaces, thereby promoting better drainage and reduction of liquid film resistance. The fin profile or shape plays an important role in promoting this surface tension– induced condensate film drainage, as was established almost a half-century ago by Gregorig (1954) in perhaps the first analytical study to address this issue. Since them, several surface tension–drained models have been proposed to predict the enhanced heat transfer coefficients and design fin profile shapes (Adamek, 1981; Honda and Nozu, 1987; Honda et al., 1988; Webb et al., 1985; Adamek and Webb, 1990; Honda and Kim, 1995; Honda and Rose, 1999). A fairly large body of data for condensation on horizontal tubes with external inte- gral fins is available in the literature, which has been reviewed insightfully by Marto (1988). This data set includes measurements for steam, refrigerants (R-11, R-22, and R-113), and several organic fluids. One of the first attempts to devise a correlation was made by Beatty and Katz (1948), who proposed the following equation for the average heat transfer coefficient based on the total area of the finned tube: h m = 0.689 k 3 l ρ 2 l i lg µ t ∆T 1/4 1.3η f A f A e π(D 2 o − D 2 r )/4D o 0.25 + A r A e D 0.25 r (14.28) Here the total outside surface area of a finned tube is taken to be the effective surface area A e . This semiempirical correlation is based on their data for methyl chloride, sulfur chloride, n-pentane, propane, and R-22 and the assumption that the conden- sate is readily drained by gravity. However, as pointed out by Webb (1994), this is perhaps not a generally valid assumption, as the surface tension force and not gravity primarily controls condensate film drainage from the fin surface. Webb (1994), and Shah et al. (1999) have compiled and comparatively reviewed the available correla- tions and models that account for surface tension–induced drainage. The role of fin geometry (profile and height) has been discussed by Honda et al. (1994) and Das et al. (1999). Comparing the performance of four different fin shapes, a profile “with a monotonically increasing radius of curvature near the fin tip and a constant thickness near the fin root” (Honda et al., 1994), tube C of the four fin tubes in Fig. 14.25 has been shown to provide the highest condensation enhancement with CFC-11 and HCFC-123. Furthermore, for a fixed fin profile and spacing, the effect of fin height BOOKCOMP, Inc. — John Wiley & Sons / Page 1071 / 2nd Proofs / Heat Transfer Handbook / Bejan EXTENDED SURFACES 1071 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 [1071], (43) Lines: 850 to 857 ——— 1.17499pt PgVar ——— Normal Page PgEnds: T E X [1071], (43) Figure 14.25 Cross-sectional fin profiles of various finnedtubes tested by Honda et al. (1994) for condensation of CFC-11 and HCFC-123. is depicted by the data of Das et al. (1999) in Fig. 14.26. Based on these assessments it appears that the Honda et al. (1988) and Rose (1994a,b)–Briggs and Rose (1994) correlations are preferred. Heat transfer enhancement in vertical condensers has been considered for several large-scale power and process industry applications. Several different types of finned tubes and other enhanced surfaces have been tested (Thomas, 1967, 1968; Alexander and Hoffman, 1971; Bergles, 1998). An excellent analysis for designing a fin profile Figure 14.26 Effect of fin height on enhanced steam condensation heat transfer. (From Das et al., 1999.) BOOKCOMP, Inc. — John Wiley & Sons / Page 1072 / 2nd Proofs / Heat Transfer Handbook / Bejan 1072 HEAT TRANSFER ENHANCEMENT 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 [1072], (44) Lines: 857 to 903 ——— 0.16516pt PgVar ——— Short Page PgEnds: T E X [1072], (44) to optimize the surface-tension-induced film condensation drainage is described by Mori et al. (1981). The preferred geometry according to this analysis is characterized by four factors: a sharp leading edge, gradually changing curvature of the fin surface from its tip to root, wide grooves or spaces between fins to collect condensate, and horizontal disks attached to the tube at periodic intervals to strip off the accumulated condensate. Their experimental data for R-113 showed further that such an arrange- ment indeed enhances the condensation performance. Refrigerant and steam condensation inside horizontal low-fin and microfin tubes have been studied extensively for air-conditioning and refrigeration and process ap- plications (Vrable et al., 1974; Royal and Bergles, 1978a,b; Luu and Bergles, 1979; Khanpara et al., 1986; Schlager et al., 1988, 1990; Shizuya et al., 1995; Kwon et al., 2000). Enhancements ranging from 100% up to 300% over the equivalent smooth tube performance are reported for steam, R-12, R-113, R-22, and R-410A, among others. For predicting the average steam condensation heat transfer coefficient, based on the inside or envelope diameter of the tube, Royal and Bergles (1978b) have pro- posed the following correlation: h m,i = 0.0265 k l d h G e d h µ l 0.8 · Pr 1/3 l 1 +160 h 2 f sd 1.91 (14.29a) where G e = G (1 −x m ) +x m ρ l ρ g (14.29b) Similarly, for refrigerants, Luu and Bergles (1979) give the following predictive equation: h m,i = 0.024 k l d h Gd h µ l 0.8 · Pr 0.43 l h 2 f sd −0.22 1 2 ρ ρ m 0.5 in + ρ ρ m 0.5 (14.30a) where ρ ρ m = 1 + x m ρ l − ρ g ρ g (14.30b) Much of the current work is now focused on the use of microfin tubes (Khanpara et al., 1986; Schlager et al., 1990; Shizuya et al., 1995; Chamra et al., 1996; Shikazono et al., 1998; Kwon et al., 2000). Also, the effort is now directed toward ascertaining their performance with a variety of alternative refrigerants or replacements for CFCs (Eckels et al., 1994b; Kedzierski and Goncalves, 1997; Tang et al., 2000). Three-dimensional finned surfaces have been shown to enhance condensation heat transfer from horizontal tubes over and above that normally obtained from BOOKCOMP, Inc. — John Wiley & Sons / Page 1073 / 2nd Proofs / Heat Transfer Handbook / Bejan EXTENDED SURFACES 1073 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 [1073], (45) Lines: 903 to 914 ——— 1.097pt PgVar ——— Short Page PgEnds: T E X [1073], (45) Figure 14.27 Three-dimensional finned surfaces for enhanced condensation: (a) notched fins (from Nakayama et al., 1975); (b) serrated-tip microfins (from Itoh et al., 1997). conventional low-fin tubes. The improvements in the envelope area–based heat trans- fer coefficients can be as much as seven times smooth tube values. A typical three- dimensional finned surface developed in Japan (Nakayama et al., 1975; Arai et al., 1977) is shown in Fig. 14.27a, and the notched fin profiles apparently provide better multidirectional condensate drainage from the fin tips. In an experimental study with circular pin fins, Chandran and Watson (1976) found the average heat transfer coef- ficients, based on the total surface area, to be as much as 20% higher than those for a smooth tube. Webb and Gee (1979) have proposed square-profiled spine fins on the basis of a theoretical value using the gravity drainage model. Similarly, for in-tube condensation, Itoh et al. (1997) have shown that microfins with serrated tips (uniformly spaced secondary discrete grooves on fin tips as shown in Fig. 14.27b) provide 30 to 60% improvements in the average heat transfer coef- ficients over same-sized conventional microfin tubes. The herringbone arrangement for microfins inside a tube proposed by Ebisu and Torikoshi (1998) and Torikoshi and Ebisu (1999), which essentially provide a cross-corrugated three-dimensional finned surface, has also been found to enhance R-22 and R-407C condensation heat transfer (see also the discussion in Section 14.4.2 on refrigerant boiling inside this herringbone microfin tube). BOOKCOMP, Inc. — John Wiley & Sons / Page 1074 / 2nd Proofs / Heat Transfer Handbook / Bejan 1074 HEAT TRANSFER ENHANCEMENT 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 [1074], (46) Lines: 914 to 924 ——— 0.627pt PgVar ——— Normal Page PgEnds: T E X [1074], (46) 14.5 DISPLACED ENHANCEMENT DEVICES 14.5.1 Single-Phase Flow The use of several different types of inserts, which are categorized as displaced en- hancement devices, is documented in the literature (Bergles et al., 1995; and Bergles 1998). They include static mixer elements (e.g., Kenics, Sulzer), metallic mesh, disks, rings, or balls, which tend to “displace” the fluid from the core of the channel to its heated or cooled wall, and vice versa; the heat transfer surface itself remains unaltered. The earliest set of data on displaced enhancement devices was perhaps reported by Koch (1958), who tested two such devices: suspended rings and disks as inserts, and tubes packed with Raschig rings and round balls. The disks were found to promote higher heat transfer with rather moderate increases in the friction factor penalty. In the case of rings and round balls, although the heat transfer improvements were comparable to that with disks, the friction factors were exorbitantly high (more than 1600%). Several studies (Van Der Meer and Hoogenedoorn, 1978; Marner and Bergles, 1978; Lin et al., 1978; Pahl and Muschelknautz, 1979) have reported the performance of different types of static mixers, and a comprehensive review of their characteristic features is given by Pahl and Muschelknautz (1979). Most of these de- vices are, however, effective only in laminar flows, as in turbulent flows, the pressure- drop penalties are extremely high (Bergles, 1998). The application of static mixers is generally restricted to chemical processing with heat transfer, where fluid mixing is the primary need. One of the newer displaced enhancement devices commercially available at the present time is the wire matrix insert shown in Fig. 14.28. A coiled-wire matrix, shaped in assorted size cloverleaf patterns, is metallurgically attached to a core rod, and this assembly, with different coil-matrix densities, is used as a tube insert. The wires tend to disrupt the core as well as boundary layer flows, thereby promoting better mixing and enhanced heat transfer. Oliver and Aldington (1986) have reported limited data for laminar flows of viscous liquids. For high-temperature gas applications, classical examples that continue to be em- ployed are bent tab, bent strip, punched-tab strip, and other types of inserts placed in Figure 14.28 Heatex wire matrix tube inserts. (Courtesy of Cal Gavin Ltd.) BOOKCOMP, Inc. — John Wiley & Sons / Page 1075 / 2nd Proofs / Heat Transfer Handbook / Bejan SWIRL FLOW DEVICES 1075 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 [1075], (47) Lines: 924 to 946 ——— 0.0pt PgVar ——— Normal Page PgEnds: T E X [1075], (47) the flue tubes of fire-tube boilers and hot-water heaters. Although often referred to as baffles or retarders or turbulators, these inserts are basically mixing devices that in- crease the convective heat transfer coefficient in turbulent flows (Koch, 1958; Evans and Churchill, 1963; Nirmalan et al., 1986; Junkhan et al., 1988). Similarly, spiral brush inserts in short channels with turbulent flows and high wall heat flux have been tested by Megerlin et al. (1974). Although the heat transfer coefficients improved by as much as 8.5 times that in a smooth tube, the pressure drop was exorbitantly high. The latter aspect of displaced enhancement devices in general has restricted their use in practical applications. 14.5.2 Boiling A limited number of studies have reported the use of displaced enhancement devices in boiling applications to increase CHF (Bergles, 1998). These have included rings, spacers, inserts (mesh or brush type), and so on, for both bulk and subcooled boiling conditions (Megerlin et al., 1974; Ryabov et al., 1977; Bergles, 1998). Much of this work was carried out in the late 1960s and early 1970s, driven primarily by the needs to address ways to increase CHF in nuclear power plants, and such devices have not received attention recently. 14.5.3 Condensing As in the case of boiling, displaced enhancement devices have not found much use in condensation applications. Only two rather dated studies (Azer et al., 1976; Fan et al., 1978) appear to have considered the use of Kenics static mixer inserts experimentally. The improvements in heat transfer coefficients, however, were once again accompanied by a very high pressure-drop penalty. 14.6 SWIRL FLOW DEVICES Swirl flow devices generally consist of a variety of tube inserts, geometrically varied flow arrangements, and duct geometry modifications that produce secondary flows. Typical examples of each of these techniques include twisted-tape inserts, periodic tangential fluid injection, and helically twisted tubes, shown schematically in Fig. 14.29. Of these, twisted-tape inserts have received considerable attention in the litera- ture, and their thermal–hydraulic performance in single-phase, boiling, and condensa- tion forced convection, as well as design and application issues, have been discussed in great detail (Manglik and Bergles, 2002a). 14.6.1 Single-Phase Flow Perhaps the most effective and widely used swirl flow device for single-phase flows is the twisted-tape insert, which has design and application literature dating back more than a century (Whitham, 1986). It has been shown to increase significantly BOOKCOMP, Inc. — John Wiley & Sons / Page 1076 / 2nd Proofs / Heat Transfer Handbook / Bejan 1076 HEAT TRANSFER ENHANCEMENT 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 [1076], (48) Lines: 946 to 946 ——— 3.097pt PgVar ——— Normal Page PgEnds: T E X [1076], (48) H A A A A Tape Direction of flow Tube ␦ ␦ Section AA Section AA Tangential flow injector d o d o d o axial flow pitch Tangential flow inlet Tangential flow inlet ()a ()b ()c 2H Figure 14.29 Typical examples of swirl flow devices: (a) twisted-tape insert; (b) altered tube flow arrangement; (c) twisted duct. (From Manglik and Bergles, 2002a.) the heat transfer coefficient with a relatively small pressure-drop penalty (Smithberg and Landis, 1964; Lopina and Bergles, 1969; Date and Singham, 1972; Hong and Bergles, 1976; Marner and Bergles, 1989; Manglik and Bergles, 1991;1992; Manglik and Yera, 2002). Frequent use of twisted tapes is in retrofit of existing shell-and- tube heat exchangers to upgrade their heat duties. Also, when employed in a new . enhanced steam condensation heat transfer. (From Das et al., 1999.) BOOKCOMP, Inc. — John Wiley & Sons / Page 1072 / 2nd Proofs / Heat Transfer Handbook / Bejan 1072 HEAT TRANSFER ENHANCEMENT 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 [1072],. Torikoshi, 1998.) BOOKCOMP, Inc. — John Wiley & Sons / Page 1070 / 2nd Proofs / Heat Transfer Handbook / Bejan 1070 HEAT TRANSFER ENHANCEMENT 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 [1070],. Bergles, 2000.) BOOKCOMP, Inc. — John Wiley & Sons / Page 1068 / 2nd Proofs / Heat Transfer Handbook / Bejan 1068 HEAT TRANSFER ENHANCEMENT 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 [1068],