See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/260433770 DOUBLE PIPE FINNED HEAT EXCHANGER Article · January 2008 CITATIONS READS 0 4,055 2 authors: Arvind Gangoli Rao Y. Levy Delft University of Technology Technion - Israel Institute of Technology 105 PUBLICATIONS 1,559 CITATIONS 109 PUBLICATIONS 778 CITATIONS SEE PROFILE Some of the authors of this publication are also working on these related projects: Fly Green: the climate friendly propulsion technology in future (ClimProp) View project Jet Impingement Cooling View project All content following this page was uploaded by Arvind Gangoli Rao on 19 September 2018. The user has requested enhancement of the downloaded file. SEE PROFILE Proceedings of the 9th Biennial ASME Conference on Engineering Systems Design and Analysis ESDA08 July 7-9, 2008, Haifa, Israel ESDA 2008 59124 A SEMI EMPIRICAL METHODOLOGY FOR PERFORMANCE ESTIMATION OF A DOUBLE PIPE FINNED HEAT EXCHANGER G. Arvind Rao1, Yeshyahou Levy2 Turbo and Jet Engine Laboratory, Technion: - Israel Institute of Technology Haifa 32000, Israel. ABSTRACT Finned tubes are one of the most widely used means of passively enhancing the heat transfer in circular tube. Many investigators have proposed different correlations for predicting the performance of such heat exchangers based on their experimental results. However, the practical usage of such correlations is limited because of the variety of parameters that can influence Nusselt number and friction factor. Most of the correlations either have been developed with limited databases, or are geometry specific. Using CFD for analyzing performance of such heat exchangers is very computational intensive and hence cannot be used for design optimization purposes. On the other hand, empirical correlations have many limitations in terms of their applicability. The objective of the present article is to present a physically based model for evaluating heat transfer and frictional loss for an internally and / or externally finned double pipe heat exchanger that can be applied in a wide range of operating conditions of practical importance. This paper describes a simple semi-empirical-numerical methodology to evaluate heat transfer and pressure drop characteristics in a finned tube heat exchanger with internal and/or external fins. Conduction and turbulent forced convection are the prominent modes of heat transfer. In order to resolve the operational characteristics of double pipe finned heat exchangers, a numerical methodology is presented which uses well known existing correlations for flow in a smooth pipe and flow over a flat plate. The method of successive substitution is used to solve the problem numerically. The proposed methodology is applied to some simple cases and the results compare well with 1 2 data and correlations available in the literature. It is found that the addition of fins to such double pipe heat exchangers reduces the Nusselt Number; however the corresponding heat transfer rate increases owing to the increase in the overall heat transfer per unit area. Keywords: Double Pipe Heat Exchanger, Heat Transfer, Internally Finned Tubes, Externally Finned Tubes 1. INTRODUCTION Finned tubes are one of the most widely used means of passively enhancing the heat transfer in circular tube (one of the most widely used heat transfer surface in heat exchangers). Finned surfaces are applicable where an additional area can be provided for augmenting the heat transfer. An important application for such air to air heat exchangers is in the open regenerated Brayton-cycle [1] and alternative regenerator configurations [2] of the land based gas-turbine power-plant, wherein the compressor discharge air is preheated by the hightemperature exhaust gases. Similarly, such heat exchangers are used for regeneration in industrial furnaces. The thermodynamic efficiency of the whole cycle is improved because air enters the combustor at an elevated temperature, with out the need for fuel burn and therefore the amount of heat addition in the combustor is reduced. Finned tubes are used to reduce the size of a heat exchanger required for a specified heat duty, or to increase heat transfer rate of an existing heat exchanger design. An internally finned tube can provide a significant increase in the surface area, and can offer an appreciable enhancement of heat transfer Post Doctoral Fellow Corresponding author. Tel.: +972-4-8293807; Fax: +972-4-8121604. E-mail address: levyy@aerodyne.technion.ac.il (Y. Levy) 1 Copyright © 2008 by ASME rate [3]. Among several available techniques for augmenting heat transfer in heat exchanger tubes, the use of fins appear to be the most promising, as described by Webb [4]. It is evident that an efficient design of heat exchanger can substantially increase performance of the entire system. Finned tubes perform differently depending on whether the flow is laminar or turbulent. Several studies have been conducted to investigate the effect of fin characteristics on heat transfer. For both laminar and turbulent flow regimes, the finned tubes exhibit substantially higher heat transfer coefficients when compared with corresponding smooth tubes. The performance of finned pipe is mainly determined by the type of flow, fin efficiency (which determines the average heat transfer coefficient) and the friction factor, which is responsible for pressure / pumping loss. 1.1 Motivation Many investigators have proposed different correlations based on their experimental results [5-9]. But the practical usage of such correlations is limited because of the variety of parameters that can influence Nusselt number and friction factor. Most of the correlations have been developed with either limited databases, or are geometry specific. Moreover since most of the experimental test sections are small in length (typically less than 1m), the temp erature variation within these test sections is low and therefore the resulting correlations do not take into account the variation of fluid properties along the test section. However, when it comes to practical applications of such heat exchangers, where a large amount of heat is to be extracted, and length of the heat exchanger can be quite large. Hence, there may be a substantial change in the fluid temperature along the tube length and therefore variation of fluid properties can significantly affect the performance predictions. An additional motivation for carrying out the present research work is the fact that in the literature there is information on either internally finned tubes or externally finned tubes, but there is no published work on double pipe heat exchangers with both internal and external fins. Because of the above-mentioned limitations, the objective of the current research is to develop a physical model for evaluating heat transfer and frictional loss for an internally and / or externally finned double pipe heat exchanger, which can be applied in a wide range of operating conditions of practical importance. An additional motivation for present research is that most of the numerical investigations that have been carried out for finned pipe heat exchangers have used Computation Fluids Dynamics with various turbulence models for studying the flow pattern inside the heat exchanger [10, 11]. Even though such techniques are quite accurate and give an in-depth physical understanding of the problem, they are very time consuming, and hence cannot be used to evaluate performance of the complete heat exchanger, especially for industrial applications. Also such techniques can not be used for optimization exercises where a number of iterations are required before an optimum design can be achieved. On the other hand, empirical correlations have many limitations in terms of their applicable operational range. Hence there is room for a relative quick semiempirical methodology to evaluate performance of such heat exchangers that can also be used in design / optimization exercises. As mentioned by many authors [12,13], there is significant scope in optimizing the performance of finned tube heat exchangers. Schematic of the various fin arrangements investigated in the present paper is shown in Figure 1. [a] [b] [c] Figure 1. Fin arrangement in different types of heat exchangers, (a) Internally Finned Pipe (b) Externally Finned Pipe (c) Internally and Externally Finned Double Pipe Heat Exchanger 2. LITERATURE SURVEY As far as geometry of the internal fins is concerned, most internal fins are strips with surfaces positioned longitudinally along the tube axis. Many investigations, both experimental and numerical, have been conducted on different kinds of internally finned tubes using a variety of fluids (air, water, oil, ethylene, etc) [5-15]. These studies examined the overall performance in terms of circumferentially averaged friction factors and heat transfer coefficients and examined the effects of parameters like: number of fins, fin height, fin width, helix angle, etc. Because of the wide range of fin geometries and Reynolds numbers covered in their experiments, Jensen and Vlankancic [9] suggested different governing processes between ‘‘tall fin’’ and ‘‘micro-fin’’ tubes. While “tall fins” (as investigated in this paper), are used in industrial furnace, thermal power plants, etc, the ‘‘micro-fins’’ are generally used for cooling in electronic equipments. Most of the investigators obtained an increase in the Nusselt numbers from 15-180% as compared to the smooth tube. However this increase in heat transfer is counterbalanced by an increase in friction factor by 50-500 % [9]. Soliman et al. [18] analyzed fully developed laminar flow, while accounting for conduction in the tube wall and fins but keeping the outer surface of tube at a constant temperature. A comprehensive experimental and numerical investigation of laminar flow was performed by Shome and Jensen [19]. Rectangular fins were implicitly assumed in all these studies. 2 Copyright © 2008 by ASME Patankar et al. [10] analyzed fully developed turbulent flow and heat transfer characteristics for internally finned tubes and annuli using the mixing length model. The local heat transfer coefficients exhibited a substantial variation along the fin height, lowest being at the fin root and highest being at the fin tip. Literature survey for heat transfer through pipes by turbulent forced convection portrays that most of the investigators have used finite difference equations of conservation of energy, x-momentum and continuity equation and have emphasized on the importance of modelling the variation in transport properties for realistic results. Bankston and McEligot [20] predicted wall temperature and pressure drop characteristics for forced internal flows. They used power law index approximation for modelling variation in fluid transport properties. Malik & Pletcher [21] and Pletcher & Malik [22] worked on prediction of turbulent flow heat transfer in annular geometries with property variations in which they considered the flow as axisymmetric, steady and fully turbulent. 2.1 Internal Fins Masliyah and Nandakumar [12] have studied heat transfer in internally finned tube. The fins were of triangular shape and the number of fins was varied up to 24 and in the length up to 0.8 times of the tube radius. Finite element method was used to analyze laminar flow in an internally finned circular tube with uniform axial heat flux around the wall. They conclude that Nusselt number based on inside diameter was higher than that for a smooth tube without fins. They also found that for maximum heat transfer rate, there exists an optimum number of fins for a given configuration and application. Liu and Jensen [23] used an unstructured finite-volume method with a two-layer turbulence model to capture the near-wall turbulence in two spirally finned tubes. The circumferentially averaged friction factors and Nusselt numbers compared well with exp erimental data of Jensen and Vlakancic [9]. One of the commonly used correlations for internally finned tubes is that by Carnavos [7], who used 21 different types of internally straight finned tubes. For straight longitudinal fins, the correlation is given as, A 0 .4 fa Nu bh = 0.023⋅ Re 0bh.8 ⋅ Prbh ⋅ A fc fh = 0.046 0 .10 A fa A fn Re −bh0. 2⋅ A ⋅ n Aa Afn = nominal flow area based on tube ID as if the fins were not present, mm2 All flow quantities in the above equation are calculated at the fluid bulk temperature. The Carnavos correlation has been subsequently verified by many investigators like Patankar et al. [10] and Trupp and Haine [24]. Edwards et al. [25] experimentally investigated fully developed turbulent flow in longitudinally finned tubes with Laser Doppler Velocimeter (LDV) and found that the Carnavos correlation agreed well with their data (within ±10% accuracy). However they suggested that a Reynolds number exponent of 0.25 (instead of -0.2) would be more appropriate for evaluating the friction factor. 2.2 External Fins There is a dearth of literature when it comes to investigations in externally finned annuli. One of the most important works in this regard is by Braga and Saboya [5]. They performed experiments to determine heat transfer coefficient and friction factor for turbulent flow through annular ducts with 20 continuous longitudinal rectangular fins made of brass. They proposed the following Reynolds number dependent heat transfer coefficient (for air). h ⋅ Dh Nu = = 0.00529 Re 0.8680· (104 < Re < 5×104) (3) k They also proposed the following correlation for evaluating the friction factor f = 2.88467·Re- 0.2863 (104 < Re < 5.2×104) (4) where the pressure drop is evaluated as 1 L m& 2 ∆P = f ⋅ ⋅ 2ρ Dh π 2 2 4 ⋅ D o − ( Di + 2t1 ) ( ) (5) It should be noted that the above mentioned relation between pressure drop and friction factor as used by Braga and Saboya is different than the conventionally used relation given below ρ ⋅V 2 L ⋅ ∆P = 4 ⋅ f ⋅ (6) 2 D h 0 .5 (1) 0 .5 (2) where Aa = actual heat transfer area (mm2/mm); An = nominal heat transfer area (mm2/mm) Afa = actual free flow area, mm2; Afc = open core free flow area at fin ID, mm2 3. NUMERICAL MODELING The heat transfer effectiveness of finned pipes depends on factors like the fin conductance, fin dimensions and local heat transfer coefficient. The present problem cannot be solved analytically because of the system complexity involving simultaneous conduction and convection processes with variation in fluid properties, resulting in a non-linear system. The schematic of the axial discretization of a double pipe finned heat exchanger layout is shown in Figure 2. For better understanding and clarity of the posed problem, a 3-dimensional 3 Copyright © 2008 by ASME view of a small section of the heat exchanger is also shown in Figure 3. Annulus duct wall Figure 3 Annulus Flow (Fluid-2) External F in Core duct wall Axis Core Flow (Fluid-1) Annulus Flow (Fluid-2) Internal Fin External Fin th i elemental strip Figure 2 Schematic of the double pipe finned heat exchanger and its discretization. Annulus duct wall Hence, the thermal model comprises of conduction through walls and fins, and turbulent forced convection at the wall and fin surfaces. The governing equations for combined conduction and convection model mainly include heat balance equations for the discretized core and the annulus duct wall (refer figure 3). Equations 7-13 are the heat balance equations for the ith element of the discretized layout. Within an axially discretized element, the fluid and wall temperatures are assumed constant. The convection boundary conditions are air inlet temperatures and pressures, convective heat transfer coefficient, and the wall conductivity. Steady state forced convection is modelled on surfaces using the relevant convective heat transfer coefficients as described later. For the core duct wall, heat input is due to convection from the hot core flow and heat conducted from the inner fins. The heat is then convected out to the cold annulus flow and conducted out of the outer fins. Hence, the heat balance equation for the k th element of core duct wall is written as, q& cv , IW1 + q& cn, Ifin + q& cv ,OW1 + q& cn,Ofin = 0 (7) where: q& cv , IW q& cv , OW Outer Fin Cold Annulus flow Core duct wall (conducting) Inner Fin Hot Core flow Figure 3: Schematic of a segment in double pipe heat exchanger with fins 3.1 Modeling Heat Transfer As the Nusselt number for both annulus flow and core flow is high due to the turbulent flow, temperature variation of the core & annulus duct walls due to axial conduction of duct wall is negligible as compared to the axial temperature change due to forced convection. Hence, for the present case, axial conduction through duct walls is not considered and only radial conduction of heat through tubes is taken into account. In addition, due to relatively high turbulence level in the flow, the radial variation of temperature within the fluid is negligible. 1 1 = h IW 1 = h OW ⋅ A IW 1 1 ⋅ (T IW 1 ⋅ A OW ⋅ (T OW − T FI 1 ) − T FII (8.1) ) (8.2) In the above equations, h IW1 and h OW1 are the convective heat transfer coefficients at the core duct wall inner and outer surfaces respectively. There are several correlations available in literature to evaluate heat transfer coefficient for a flow through a smooth tube. One of the most widely used such correlation is the Dittus-Bolter Equation, and is given as [26] Nu = 0.023 ⋅Re0.8 ⋅Pra (9) Where, a = 0.4 (if the fluid is being cooled); a = 0.3 (if the fluid is being heated) and all the variables are evaluated at the bulk temperature. D Hence, hIW 1 = 0. 023 ⋅ Re 0F.81 ⋅ PrF0.14⋅ 1 , (9.1) k F1 For evaluating heat transfer coefficient in the annulus, (refer to Eqn.8 & Figure 3) the hydraulic diameter is used for evaluating the flow Reynolds Number. D hIWO = 0. 023 ⋅ Re 0h.,8F 2⋅ Pr F0.24 ⋅ h , (9.2) kF2 ρ ⋅ V ( D − D1 − 2 × tW1 ) where Reh,F2 = F 2 F 2 2 (9.2.1) µF 2 heat transfer by fins (used in Eqn. 7) can be calculated as, 1 1 q& cn,Ifin = − k Ifin ⋅ AIfin ⋅ (TFI − T Iw1 ) ⋅ ξ ⋅ − (10.1) −2ξ L +2ξ L 1 + e 1+e 1 1 q& cn, Ofin = − kOfin ⋅ AOfin ⋅ (TFII − TOw1 ) ⋅ ξ ⋅ − − 2ξ L 1 + e 1 + e +2ξL (10.2) 4 Copyright © 2008 by ASME ξ= where h fin ⋅ Pfin k fin ⋅ Ac / s , fin , Pfin 2 × L fin,cr , = Ac / s , fin = t fin × L fin, cr (10.2.1) The Harper and Brown [27] approximation as given by Holman [28], is used for evaluating the corrected fin height (Lfin,cr) for evaluating the heat transfer through longitudinal fins with uninsulated ends Lfin,cr = Lfin + tfin/2 (10.2.2) To evaluate heat transfer coefficient over surface of these fins (both inner and outer), the Nusselt number correlation for flow over a flat plate is used [29] as given below, Nu x,fin = 0.332 Re1/2⋅Pr1/3 for 100 < Re < 3×105 (10.2.3) Nu x,fin = 0.0296 Re0.8⋅Pr1/3 for 3×105 < Re < 107 (10.2.4) where all quantities are evaluated at film temperature. The transition Reynolds number (3×105) indicates the point when flow turns from laminar into turbulent. Finally, heat conducted by the core duct wall is given by L q&cn ,W 1 = q& cv ,IW 1 + q&cn , Ifin = q& cv ,OW 1 + q& cn ,Ofin = 2π ⋅ k W 1 ⋅ ⋅ (TIW 1 − TOW 1 ) N (11) If the annulus duct wall is not insulated, additional equations accounting for heat balance across annulus duct wall have to be solved, q& cv , IW 2 + q& cv ,OW 2 = 0 (12) where, q& cv , IW 2 q& cv , OW = h IW 2 2 = h OW ⋅A 2 IW 2 ⋅ A OW ⋅ (T IW 2 2 ⋅ (T OW − T FII 2 ) − T F∞ (12.1) ) (12.2) where h IW2 and h OW2 are the convective heat transfer coefficient at inner and outer surfaces of the annulus duct wall respectively, and can be evaluated by using Dittus-Bolter correlation given by eqn 9.1 & 9.2. In Eqn. (12.2), TF∞ is the ambient air temperature surrounding the heat exchanger. The heat transferred from annulus wall due to conduction can be calculated as, L q& cn,W 2 = q& cv ,IW 2 = − q& cv ,IW 2 = 2π ⋅ kW 2 ⋅ ⋅ (TIW 2 − TOW 2 ) N (13) To improve accuracy of the convective heat transfer mechanism, the fluid thermo -physical and transport properties are evaluated according to the local temperature and pressure, as described in the subsequent sections. 3.2 Modelling Pressure Loss The total pressure loss in the core flow is attributed to frictional loss at inner pipe wall surface and frictional losses at fin surfaces. ∆ PF1 = ∆ PW 1 + ∆ PIfin (14) where V 2 ⋅ ρ L ∆ PW 1 = 4 × fW 1 ⋅ F 1 F1 ⋅ N⋅D 2 1 (14.1) V 2 ⋅ ρ ∆ PIfin = f Ifin ⋅ F1 F1 ⋅ L Ifin H Ifin + t fin ⋅ N fin (14.2) 2 To evaluate friction losses due to the pipe surface, the Darcy- Weisbach friction factor correlation for evaluating friction factor for flow through a smooth pipe, given by Eqn. 15, is used. All variables are evaluated at the film temperature [29], ( (( ( f = 1.82 ⋅ Log ? ⋅ M ⋅ ) (? ⋅ R ⋅ T ) ) ⋅ Dh ) µ -1. 64 ) −2 4 (15) For computing frictional losses factor at the fin surface, the correlation given by Schlichting [29] for evaluating friction coefficient for flow over flat plate is used. f x , fin = 0 .646 (Re x, fin ) −0.5 100 < Re < 5×104 (16) f x , fin = 0. 0592 (Re x, fin) − 0. 2 5×104 < Re < 107 Here the transition Reynolds number has been reduced due to the accelerated transition from laminar to turbulent flow because of the pipe flow. Similar to the core flow, pressure losses in annulus flow can be attributed to frictional loss at the duct walls and the external fin surfaces. Hence ∆ PF 2 = ( ∆POW1 + ∆PIW 2 ) + ∆POfin (17) The Reynolds Number based correlations given by Eq (18) is used for evaluating the friction factor and its associated pressure loss in the annulus [30]. −2 ε Dh 2. 5 f = −2 log + (18) 3. 7 Re F 2 f where hydraulic diameter is given by , Dh = Dh = ( D2 − D1 − 2 × tW1 ) , and Reynolds number is given by Eqn (9.2.1). The above equation is solved iteratively for the friction factor (f). Friction due to external fins is evaluated in a similar manner to that by internal fins given by Eq. (16). 3.3 Fluid Property Variation It is well know that if fluid temperature variation is significant within a system, the fluid transport and thermodynamic properties vary to a large extent, especially for 5 Copyright © 2008 by ASME 3.4 Numerical Solution Method The present problem cannot be solved analytically because of the complex nature of system involving both conduction and forced convection processes, resulting in a non-linear system. Therefore zonal analysis method is used, wherein the layout is discretized into certain number of elements and governing equations are solved for each discretized axial element (refer to Figure 2). Various parameters like wall temperature, fluid temperature, fluid properties, heat transfer coefficient, friction factor coefficient, etc, are assumed constant within a discretized element. The Method of Successive Substitution [32] is used to solve the governing equations, which include heat balance equations for all elements and expressions for evaluating various terms involved in these equations. In this methodology, variables like the wall temperatures, fluid temperatures, etc, are assigned with an initial values and then proceeding through the system of equations (Eqn. 1-10), all variables are recalculated and successively substituted and iterated until satisfactory convergence is achieved. For the present exercise, a convergence value of 10-3 is used. there are no external fins in the heat exchanger and only the internal fins are used for enhancing the heat transfer. The comparison between Nusselt number and heat transfer coefficient obtained from present methodology with experimental results of Carnavos for the core flow is shown in Figure 4 and 5 respectively. As seen from the figures, the comparison between computed results and Carnavos correlation is very good, the error is less than 5%. Hence validity of the proposed methodology in evaluating the heat transfer and pressure drop characteristics for an internally finned pipe is established. It should be noted that Carnavos correlation is evaluated at fluid bulk temperature where as the film temperature is used for evaluating heat transfer in the semi-empirical methodology. 450 Simulation Series1 400 Carnavos Series2 350 300 Nuh high temperature flows. Therefore, fluid temperature variation in a practical heat exchanger is expected to be large and hence it is necessary that this variation be taken into account. Zografos et.al. [31] have provided the equations to compute the thermo physical and transport properties of seven commonly used fluids (air, liquid water, water vapour, carbon dioxide, Freon-12, engine oil, and mercury). They used the curve fitting process to fit the variation in fluid properties as, dynamic viscosity, constant pressure specific heat, thermal conductivity, and more with temperature, in the form of polynomials. In the present analysis, air is used as the working medium in both core and annulus flow. 250 200 150 100 50 0 1000 10000 100000 1000000 Re h Figure 4: Comparison of Nusselt number obtained from the present simulations with the Carnavos correlation 500 Simulation Series1 450 Carnavos Series2 400 4.0 RESULTS AND DISCUSSIONS H (Wm- 2K- 1) 350 The methodology described in the earlier section is used to study heat transfer and pressure loss characteristics of heat exchangers. As stated earlier, there are no results available in the literature for double pipe heat exchangers with both internal and external fins, therefore the results obtained from the proposed semi-empirical numerical methodology are compared with experimental results available in literature for heat exchangers with only internal fins or heat exchangers with only external fins. 300 250 200 150 100 50 0 1000 10000 100000 1000000 Reh 4.1 Heat Exchanger with Internal Fins only As stated earlier, many investigators have looked into the performance of internally finned tube with respect to its heat enhancement capability and increase in pressure drop due to the addition of fins. Both straight longitudinal fins and spiral fins have been investigated. The Carnavos correlation is used as a benchmark for comparison. In this case it is assumed that Figure 5. Comparison of the heat transfer coefficient obtained from the simulations with that obtained from the Carnavos correlation The comparison between friction factor and total pressure loss for internally finned tube calculated by the proposed methodology and that obtained by using Carnavos correlation 6 Copyright © 2008 by ASME is shown in Figure 6 and Figure 7 respectively. It can be seen that the discrepancy between the friction factor decreases with increase in the flow Reynolds number and the error never exceeds 6 % the predicted friction factor, in case for external fins, is more than that obtained in experiments. However it should be noted that the difference between the computed and experimentally obtained results diminishes with increase in Reynolds Number. The difference is mainly attributed in prediction of frictional drag at the fin surface. Since there is no other correlation / experimental data available for predicting frictional losses in an annular duct with fins, the above simulations can not be compared further. However it can be seen from Figure 9 that the total pressure loss predicted by the proposed methodology is in good agreement with that obtained experimentally by Braga & Saboya [30], the difference being less than 5% 0.008 Series1 Simulation 0.007 Carnavo Series2 s 0.006 f 0.005 0.004 0.003 0.25 0.002 Simulation FF1O_AVG, Braga&Sboya FF1O_BR, 0.001 0.2 0 1000 10000 100000 1000000 fh ReF1 0.15 Figure 6. Comparison of friction factor obtained from the simulations with the Carnavos correlation 0.1 4000 Simulation Series1 0.05 3500 Carnavos Series2 Ploss (Pa) 3000 0 0 10000 20000 30000 40000 50000 60000 70000 2500 Reh 2000 Figure 8. Comparison of friction factor obtained from simulations with experimental correlation provided by Braga & Saboya 1500 1000 500 2500 Simulation PRLOSS, 0 1000 10000 100000 Braga & Sboya PRLF_BR, 1000000 ReF1 -∆P2 Figure 7. Comparison of the pressure loss obtained from the simulations with that obtained from the Carnavos correlation for unit length of the pipe 2000 1500 1000 4.2 Heat Exchanger with External Fins only In this case, only external fins within the annular passage are considered for enhancement of the heat transfer (the internal fins within the core pipe are neglected in this analysis). Performance of heat exchanger with similar conditions as used by Braga and Saboya [30], the total pressure drop and heat transfer coefficient across the annulus, evaluated by the earlier described methodology, is shown in Figure 8 and Figure 9 respectively. It is seen that unlike in the case for internal fins, 500 0 0 10000 20000 30000 40000 50000 60000 70000 Reh Figure 9. Comparison between the numerically obtained loss in heat exchanger with experimental correlations of Braga & Saboya 7 Copyright © 2008 by ASME 500 F1(K) No Fins Inner Fins Inner(TF1) Outer Fins Outer(TF1) Both Fins Both(TF1) 495 490 TF1 (K) Figure 10 shows comparison between numerically obtained Nusselt number (by the proposed methodology) with that obtained experimentally by Braga & Saboya. It is seen that the two curves cross each other at low Reynolds Number, thus indicating that dependency of Nusselt Number on the Reynolds Number exponent is more than that given by Eqn (3). However, the overall comparison is good with the maximum error being less than 5 % 485 100 480 80 475 0 60 5 0.25 10 0.5 15 0.75 1.020 Nuh Length (m) 40 NU1O, Braga & Saboya NU_BR, Simularions 20 0 0 10000 20000 30000 40000 50000 60000 70000 Reh Figure 10: Comparison of the Nusselt Number from the simulations with Braga & Saboya Figure 11: Temperature variation of the core flow for different cases The variation of Nusselt Number for the core and the annular fluid with increase in the number of fins (both external and internal) is shown in Figure 12. It is observed that Nusselt number decreases for both fluids, however since there is a substantial increase in the over all heat transfer area, the effective heat transfer rate increases due to the simultaneous increase in the number of external and internal fins. 140 Core flow NU1O, 4.3 Internal and External Fins 120 Annular NU1I, 100 80 Nu From analysis of the results obtained in earlier sections, it is apparent that the proposed semi-empirical-numerical methodology can be used for evaluating performance of internally or externally finned tubes. In this section, the double pipe heat exchanger with both internal and external fins is analyzed by using the proposed semi-empirical-numerical methodology. Inlet conditions used for the test case are given in Table 1. The variation of the core fluid temperature for various cases (i.e. without fins, with only inner fins, with only outer fins, with both inner and outer fins) is shown in Figure 11. It is seen from Fig. 11 that the effectiveness of only inner fins or only outer fins in transferring heat from the core fluid to the annular fluid is almost the same. Therefore maximum heat is transferred in the case when there are both external and internal fins. 60 40 20 0 0 5 10 15 20 25 Number of Fins Figure 12: Variation of Nusselt number with external and internal fins The increase in heat transfer due to fins also results in enhanced pressure losses in the flow. The effect of fins (both external and internal) on the pressure loss characteristic of the double pipe heat exchanger is shown in Figure 13. Even though the pressure loss in core flow is less as compared to that in annular flow when there are no fins, the core fluid pressure losses increase more rapidly with the increases in number of fins 8 Copyright © 2008 by ASME 200 Core flow PRLOSS, Annular FF1I_C,flow Pressure loss (Pa) 160 120 80 40 0 0 5 10 15 20 25 Number of Fins Figure 13: Pressure Loss Characteristics with fins for core and annular flow The effect of fin conductance and fin height on the flow Nusselt number (for both core and annular flow) is shown in Figure 14. The Nusselt number increases with increase in fin conductance and reduces with increase in fin height, as expected. However increase in the fin height results in an increase in the overall heat transfer area and hence resulting in enhancement of the overall heat transfer rate. 120 100 Inner Fin (k =100) NU1I_100 Outer Fin (k = 100) NU1O_100 Inner Fin (k =10) NU1I, Outer Fin (k = 10) NU1O, Nu 80 60 40 20 0 0 5 10 15 20 25 Fin Height (mm) Figure 14: Variation of Nusselt number with fins conductance and fin height 5. CONCLUSIONS A new and fast calculation procedure / methodology is described to evaluate performance of a double pipe heat exchanger with internal and / or external fins. The methodology uses well-known and established correlations for the flow in smooth tubes and for flow over flat plates and hence is not design specific. In addition, the variation in the fluid properties due to variation in the temperature along the heat exchanger is taken into account. Hence the methodology can be applied to cases where the fluid undergoes large temperature changes. The predicted performance matches well with experimental results reported in literature for pipes with internal or external fins, and therefore we can extrapolate the validity of the developed model to the case of double pipe heat exchanger. Since the proposed model is based upon analytical formulations of semi-empirical correlations, the time required for numerical computations is very small and hence this methodology can be successfully used for optimization processes wherein a large number of design iterations are required before arriving at an optimal solution. Hence the present methodology conveniently fills in the gap between costly CFD simulations and easily available but inadequate empirical correlations for finned tube heat exchangers, and gives the designer a tool to optimize the design of such heat exchangers for practical applications. 6. ACKNOWLEDGMENTS The authors are thankful to the Ministry of Higher Education, Govt. of Israel and Israel science Foundation (ISF) for supporting the research. 7. NOMENCLATURE A area [m2] Bi Biot number [-] Cp specific heat at constant pressure [J/kg-K] dA area of elemental strip [m2] D diameter of duct [m] Dh hydraulic diameter [m] dx width of elemental strip [m] f friction factor [-] h convective heat transfer coefficient [W/m2-K] k thermal conductivity KT thermal conductivity of fluid [W/m-K] L length [m] M Mach Number [-] m& mass flow rate [kg/s] N number of elements [-] Nu Nusselt Number [-] P pressure [Pa] Pr Prandtl number [-] Q total enthalpy [J] q& heat transfer rate [W] R gas constant of air [= 287 J/kg K] Re Reynolds Number [-] t thickness [m] T temperature [K] V velocity [m/s] Greek scripts µ dynamic viscosity [Pa-s] ρ density [kg/m3] 9 Copyright © 2008 by ASME γ ratio of specific heats [-] ξ fin parameter [-] Subscripts b bulk quantity cv convective heat transfer cn conduction heat transfer cr corrected f evaluated at film temperature fin related to fin FI core fluid FII annulus fluid h hydraulic quantity I inner surface O outer surface w1 core duct wall w2 annulus duct wall x distance Superscripts a exponent in the Dittus-Bolter equation 8. [1] [2] [3] [4] [5] [6] [7] [8] [9] REFERENCES Cohen, H., Rogers, G. F. C., and Saravanamuttoo, H. I. H., 1996, Gas Turbine Theory, 4th Ed., Longman Group, Harlow, England. Dellenback, P. A., 2002, “Improved Gas Turbine Efficiency Through Alternative Regenerator Configuration, ” Journal of Engineering for Gas Turbines and Power, Vol. 124, pp. 441-446. Yu, B., Nie, J. H., Wang, Q. W. and Tao., W. Q., 1999, “Experimental Study on the Pressure Drop and Heat Transfer Characteristics of Tubes with Internal Wave-like Longitudinal Fins,” Heat and Mass Transfer, Vol. 35, pp 6573. Webb, R.L., 1994, Principles of Enhanced Heat Transfer, John Wiley & Sons, New York. Braga, C.V.M. and Saboya, F.E.M., 1999, “Turbulent Heat Transfer, Pressure Drop and Fin Efficiency in Annular Regions with Continuous Longitudinal Rectangular Fins,” Experimental Thermal and Fluid Science, Vol. 20, pp. 5565. Carnavos, T. C., 1979, ‘‘Cooling Air in Turbulent Flow with Internally Finned Tubes,” Heat Transfer Engineering, Vol. 1(2), pp. 41–46. Carnavos, T. C., 1980, ‘‘Heat Transfer Performance of Internally Finned Tubes in Turbulent Flow,’’ Heat Transfer Engineering, Vol. 1(4), pp. 32–37 Watkinson, A.P., Miletti, D.L., Kubanek, G.R., 1975, “Heat transfer and pressure drop in internally finned tubes in turbulent air flow,” ASHRAE Transactions, Vol.81, pp.330337. Jensen, M. K., and Vlakancic, A., 1999, ‘‘Experimental Investigation of Turbulent Heat Transfer and Fluid Flow in Internally Finned Tubes,’’ International Journal of Heat and Mass Transfer, Vol. 42, pp. 1343–1351. [10] Patankar, S.V., Ivanovic, M., Sparrow, E.M., 1979, “Analysis [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] 10 of Turbulent Flow and Heat Transfer in Internally Finned Tubes and Annuli,” Journal of Heat Transfer, Vol. 101, pp. 29-37. Said, N.M.A., and Trupp, A.C., 1984, “Predictions of Turbulent Heat Transfer in Internally Finned Tubes,” Chemical Engineering Communications, Vol.31, pp.65-69 Masliyah, J.H.; Nandakakumar, K., 1976, “Heat transfer in internally finned tubes”. ASME Journal of Heat Transfer, pp. 98. Al-Sarkhi, A., and Abu-Nada, E., 2005, “Characteristics of Forced Convection Heat Transfer in Vertical Internally Finned Tube,” International Communications in Heat and Mass Transfer, Vol. 32, pp. 557-564 Webb, R.L. and Bergles, A.E., 1983, “Heat transfer enhancement: Second generation technology,” Mechanical Engineering, Vol. 105, pp. 60-67. Bergles, A.E., and Webb, R.L., 1985, “ A guide to the literature on convective heat transfer augmentation,” in Advances in Heat Transfer (Eds. S.M. Shenkman, J.E. O’Brien, I.S. Habib and J.A.Kohler) ASME Symposium, Vol. HTDVol. 43, pp. 81-90. Soliman, H.M., and Feingoid, A., “Analysis of Fully Developed Laminar Flow in Longitudinally Internally Finned Tubes,” Chemical Engineering Journal, vol. 14, pp. 119-128. Vasil’chenko, Y.A. and Barbaritskaya, M.S. 1969, “Heat transfer in tubes with longitudinal fins,” Thermal Engineering, Vol.16, pp. 66-68. Soliman, H.M., Chau, T.S., and Trupp, A.C., 1980, “Analysis of Laminar Heat Transfer in Internally Finned Tubes with Uniform Outside Wall Temperature,” Journal of Heat Transfer, vol. 102, pp. 598-604. Shome, B., and Jensen, M.K., 1996, ‘‘Experimental Investigation of Variable Property/Mixed Convection Laminar Flow in Internally-Finned Tubes,’’ Journal of Enhanced Heat Transfer, Vol. 4, pp. 53–70. Bankston, C. A., and McEligot, D. M., 1970, “Turbulent and Laminar Heat Transfer to Gases with varying Properties in the Entry Region of the Circular Ducts,” International Journal of Heat and Mass Transfer, Vol. 13, No. 2, pp. 319–344. Malik, M. R., and Pletcher, R. H., 1978, “Computation of Annular Turbulent Flows with Heat Transfer and Property Variations,” Proceedings of 6th International Heat Transfer Conference, Vol. 2, Hemisphere, Washington, DC, pp. 537–542. Pletcher, R. H., and Malik, M. R., 1979, “Prediction of Turbulent Flow Heat Transfer in Annular Geometries,” Turbulent Forced Convection in Channels and Bundles (Eds. S. Kakac and D. B. Spalding), Vol. 1, Hemisphere, Washington, DC, pp. 185–205. Copyright © 2008 by ASME [23] Liu, X., and Jenson, M.K., 2001, “Geometry Effects on [24] [25] [26] [27] [28] [29] [30] [31] [32] Turbulent Flow and Heat Transfer in Internally Finned Tubes,” Journal of Heat Transfer, Vol. 123, pp. 1035-1044. Trupp, A.C., and Haine, H., “Experimental Investigations of Turbulent Mixed Convention in Horizontal Tubes with Longitudinal Internal Fins”, in Heat Transfer in Convective Flows (Ed. R.K. Shah), ASME Symposium, Vo1. HTDVol.107, ASME, New York, pp.17-25. Edwards, D.P., Hirsa, A., and Jensen, M.K., 1996, ‘‘Turbulent Air Flow in Longitudinally Finned Tubes,’’ ASME Journal of Fluids Engineering, Vol. 118, pp. 506– 513. Dirker, J., and Meyer, J.P., 2002, “Heat Transfer Coefficients in Concentric Annuli” Journal of Heat Transfer, Vol. 124, pp. 1200-1203. Harper, W.B., and Brown, D.R., 1922, “Mathematical Equation for Heat Conduction in the Fins of Air-cooled Engines”, NACA Rep. 158. Holman, J.P., 1968, Heat Transfer, 4th Edition, McGraw-Hill Kogakusha, Tokyo. F Sukhatme, S.P., 2000, A Textbook on Heat Transfer, 3rd Edition, Universities Press, Hyderabad. Fogiel, M., 1987, The Essentials of Transport PhenomenaII, Research and Education Association, New York. Zografos, A.I., Martin, W.A., and Sunderland, J.E., 1987, “Equations of Properties as a Function of Temperature for Seven Fluids”, Computer Methods in Applied Mechanics and Engineering, Vol. 61, pp.117-187. Stoecker, W. F., 1989, Design of Thermal Systems, McGraw–Hill, New York, pp. 111–126. 11 View publication stats Copyright © 2008 by ASME