Computational Fluid Dynamics Analysis of an Enhanced Tube with Backward Louvered Strip Insert

energies

Article Computational Fluid Dynamics Analysis of an Enhanced Tube with Backward Louvered Strip Insert

Agung Tri Wijayanta 1,2, Budi Kristiawan 1,2, Pranowo 2,3, Agung Premono 4 and Muhammad Aziz 5,6,* 1 Department of Mechanical Engineering, Faculty of Engineering, Universitas Sebelas Maret, Kampus UNS Kentingan, Jl. Ir. Sutami 36A Jebres, Surakarta 57126, Indonesia 2 Research Group of Sustainable Thermoﬂuids, Universitas Sebelas Maret, Kampus UNS Kentingan, Jl. Ir. Sutami 36A Jebres, Surakarta 57126, Indonesia 3 Department of Informatics, Universitas Atma Jaya Yogyakarta, Jl. Babarsari 44 Yogyakarta 55281, Indonesia 4 Department Mechanical Engineering, Jakarta State University, Jl. UNJ Kampus A, Jl. Rawamangun Muka Jakarta Timur 13220, Indonesia 5 Institute of Industrial Science, The University of Tokyo, 4-6-1 Komaba, Meguro-ku, Tokyo 153-8505, Japan 6 Institute of Innovative Research, Tokyo Institute of Technology, 2-12-1 Ookayama, Meguro-ku, Tokyo 152-8550, Japan * Correspondence: [email protected]; Tel.: +81-3-5452-6196

Received: 3 August 2019; Accepted: 31 August 2019; Published: 1 September 2019

Abstract: A computational solution of an enhanced tube equipped with a backward louvered strip insert with various pitches was evaluated in this work. A k-ε renormalized group turbulence model has been applied for the turbulent model. Different pitches, S, of 40, 50, and 60 mm were investigated for the Reynolds number with a range of 10,000–17,500 using water as a working fluid. The extra louvered strip caused fluid flow disturbance, so that the flow pattern formed more turbulence. The turbulent flow was characterized by the flow pattern on the back of the inserts that form a vortex. The vortices formed caused a better heat transfer. The results of the computational analysis showed that the enhanced tube had a louvered strip with a pitch distance S = 40, 50, and 60 mm could increase the Nusselt numbers to 1.81, 1.75, and 1.72, and the friction factor to 7.59, 6.51, and 5.77 times greater than the plain tube, respectively.

Keywords: louvered strip insert; pitch; Nusselt number; friction factor; turbulent ﬂow; numerical investigation

1. Introduction An insert device involves a specific geometric form equipped into a tube, functioning as a turbulator. Therefore, it is expected that it can increase the flow turbulence, minimize the thickness of the thermal boundary layer, and also improve the heat transfer coefficient of the flow surface. Many researchers have evaluated the adoption of louvered strip inserts numerically and experimentally and measured its impact on the enhancement of heat transfer [1–12]. A louvered strip insert is categorized as a turbulator, having a flat, leaf-like shape, mounted on the wire, and made of thin metals. It has been clearly studied that the louvered strip inserts installed in a parallel-flow concentric tube heat exchanger resulted in higher heat transfer performance in comparison to the plain tube or without the insertion. A high recirculation flow can be created by the tube equipped with louvered turbulators, which is disturbing the boundary layer growth, resulting in the higher heat transfer due to the slightening of the boundary layer thickness. Compared to the plain tube, the addition of louvered strip inserts leads to a higher average Nusselt number, Nu, as well as friction factor, f. The characteristics of heat transfer and flow resistance of the tubular heat exchanger equipped with louvered strip elements have been studied

Energies 2019, 12, 3370; doi:10.3390/en12173370 www.mdpi.com/journal/energies Energies 2019, 12, 3370 2 of 12 by Eiamsa-ard et al. [1]. In their study, the effects of backward and forward arrangements and various inclined angles, θ, of 15◦, 25◦, and 30◦, have been clarified. Their study showed that the existence of louvered strip devices facilitates a higher heat transfer rate compared to the plain tube. The increase of the inclined angle results in the increase of Nu accordingly. The louvered strip can produce a powerful turbulence intensity, creating a rapid flow mixing, particularly at larger inclined angles. In addition, Fan et al. [2] have conducted a comprehensive numerical simulation in a circular tube equipped with louvered strip inserts, including the thermal-hydraulic performance of turbulent airflow. Slant angles and the louvered strip’s pitch were set as the parameters. Their calculation results showed clearly that these parameters strongly influenced the flow resistance and rate of heat transfer. A larger slant angle and smaller pitch of the louvered strips enhanced the heat transfer rate and increased the flow resistance. However, it seems that the influence of slant angles was more significant compared to the insert’s pitch. It is considered that it is because the slant angle is able to generate a relatively high radial flow velocity. The impacts of a louvered strip insert, which is equipped in a tubular heat exchanger, on the characteristics of both heat transfer and flow resistance with different nanofluids have been clarified by Mohammed et al. [3]. They set several slant angles and pitches and studied their impacts. Their results proved that the combination of the high slant angle α = 30◦ and small pitch S = 30 mm could enhance the heat transfer performance by about 367%–411% compared to the plain tube. A larger slant angle produced a higher intensity of turbulence resulting in the fast flow mixing. Furthermore, Huisseune et al. [4] observed the impacts of the delta winglet and louver geometry on the thermal-hydraulic performance of, e.g., a compound heat exchanger. It was clarified that a small fin pitch and large louvered angle led to a strong flow deflection, therefore leading to a significant contribution of the louvers. In addition, Wenbin et al. [5] experimentally modified and studied the performance of small pipe inserts with several different spacer lengths, arc radii, and slant angles; furthermore, the analyses also included their influences on the heat transfer, flow resistance, and thermal-hydraulic performances. In their study, a small pipe was formed into an S-shape and was installed on a core rode. It was found that tubes fitted with pipe inserts influenced both the heat transfer coefficient and the friction factor. Tabatabaeikia et al. [6] reviewed the heat transfer augmentation through the adoption of several different types of inserts. It is presented that among the available types of insert, under different arrangements and conditions, the louvered strip insert showed a better performance in backward flow than in forward flow. In addition, in comparison with any other kind of insertion, the adoption of the louvered strip insert clearly enhanced the heat transfer performance, although it also provides a higher friction factor. Jang and Chen [7] performed a numerical, three-dimensional analysis for heat transfer in laminar fluid flow inside a heat exchanger having louvered fins for various louver angles. Moreover, Park et al. [8,9] conducted experiments adopting a louvered fin heat exchanger in order to reveal the effect of unequal louver pitch and inclination angle on the thermal performance of the repeated frosting/defrosting cycles. Frost blockage was delayed when the unequal louver pitch was employed, compared to the equal louver pitch case. However, these works clarified that both heat transfer rate and pressure drop varied less in the repeated cycling for larger angles. Based on the mechanism of flow and heat transfer, Wei et al. [10] developed the mathematical model and continued to perform numerically a comprehensive airside, anti-freezing method during winter. Although using three-dimensional turbulent flow numerical simulations, Sadeghianjahromi et al. [11] developed both heat transfer and flow friction correlations for tubular heat exchangers having the louvered fins. They claimed that the numerical simulation data was exhibited within 15% of the wide ranges of their parameters. Zuoqin ± et al. [12] provided a similar tendency of thermal performance under different Reynolds numbers, Re, for louvered fin configurations. The enhancement mechanisms of fluid flow and heat transfer observed using a computational calculation method is important to be clarified as they are the key factors to further improve the heat exchanger performance. There are several reasons to study related heat transfer enhancement Energies 2019, 11, x FOR PEER REVIEW 3 of 12

98 studying and publishing the research works in [13,14], we were strongly motivated to continue the 99 work to a numerical analysis of the thermal hydraulic performance inside an enhanced tube equipped 100 with louvered strip insert with various pitches for turbulent single-phase flow. This is the reason that 101 this work is believed able to give a novel academic contribution, especially when compared with the 102 currently available and previously published studies.

103 Energies2. Computational2019, 12, 3370 Method 3 of 12 104 The geometry of the concentric tube, including the use of inner and outer tubes for hot and cold 105 numerically.water, respectively The computational, is represented calculation in Figure becomes 1, referring a proper to [9]. choice A computational to be used for domain predicting has been the 106 performancedeveloped, and based the on boundary the knowledge condition of mechanismshas been employed acquired in throughthe domain. experiments. Regarding Most the dimension, published 107 reportsthe inner on tube heat had transfer inside augmentation and outside bydiameter louvereds, di stripand d insertso, of 14.3 are and experimental 15.8 mm, respectively studies, which. On are the 108 lackingother hand, the evidence the outer to describe tube had the inside mechanism and outside of both diameter fluid flows, D andi and heat Do, transfer.of 23.4 and Hence, 25.4 after mm, 109 studyingrespectively. and publishingThe pipe length the research L is 2110 works mm. The in [ 13louvered,14], we strip were is strongly oval and motivated its dimension to continue is 10 mm the in 110 worklength, to a6 numericalmm in width analysis and of1 mm the thermalin thick hydraulicness that stick performances on the insidesolid pipe an enhanced with 2 mm tube ofequipped diameter. 111 withLouvered a louvered strip strips were insert arranged with variousin backward pitches with for turbulentvariations single-phase of pitch (S) flow.of 40, This50, and is the 60 reason mm with that a 112 thisslant work angle is believedα of 25°. toThe be hot able fluid to give has aan novel inlet academic temperature contribution, of 333.15 especiallyK with the when Re at comparedinlet velocity with of 113 the10, currently000–17,500 available. On the andother previously hand, the published cold fluid studies. has a temperature inlet of 300.15 K and a constant 114 mass flow rate of 0.1027 kg/s. In addition, the exit side was subjected to the pressure outlet. 115 2. ComputationalIn order to observe Method the thermal-hydraulic characteristics, covering the fluid flow and heat

116 transfer, water was employed as the working fluid, i.e. the hot and cold water flows circulate in inner The geometry of the concentric tube, including the use of inner and outer tubes for hot and cold 117 and outer tube loops, correspondingly. The used fluid properties depends of temperature. Table 1

water, respectively, is represented in Figure1, referring to [ 9]. A computational domain was developed,

118 lists the thermo-physical properties of the used hot and cold water [15], adopted as the inlet values − + and the boundary condition was employed in the domain. Regarding the dimension, the inner tube

119 during the computational process.

had inside and outside diameters, d and do, of 14.3 and 15.8 mm, respectively. On the other hand, the = i

.71. 8 1) ( /8) ( 12.7 1.07 fPr

(13) outer tube had inside and outside diameters, Di and Do, of 23.4 and 25.4 mm, respectively. The pipe ti insert. strip

/ 2/3 1/2 120 Table 1. Thermo-physical properties at the inlet.

N length L is 2110 mm. The louvered strip is oval, and its dimensions are 10 mm in length, 6 mm in

(/8) a emtyo ovrdsrp ihfradarneet b Louvered (b) arrangement, forward with strips louvered of Geometry (a) 2. Fig. fRePr width, and 1 mm in thicknessTemperature that sticksρ on the solidµ pipe with 2C mmp of diameter.k Louvered strips Fluid Pr were arranged backward with[K] variations[kg/m in3] pitch[kg/m (S) of⋅s] 40, 50,[kJ/kg and⋅K] 60 mm[W/m with⋅K] a slant angle α of

25◦. The hot ﬂuid had an inlet temperature of 333.15 K with the Re at inlet velocity of 10,000–17,500.

Cold water 300.15 997 8.52(10–4) 4.178 0.613 5.81 ehkvcreaini ntefr f(13): of form the in is correlation Pethukov On the other hand, the cold ﬂuid had a temperature inlet of 300.15 K and a constant mass ﬂow rate of

–4

0.1027 kgHot/s. water In addition, the333.15 exit side was983.3 subjected 4.67(10 to the) pressure4.185 outlet. 0.654 2.99

o 10 for . 10 × <5 Re

< 121 6 3 − +

1.(8(1) 112.7(/8)( fPr (12)

−

/ 2/3 1/2 u N

/) 1000) (/8)( R Pr fRe 122 123 (a)

124 sfollows: as (12) equation by given is correlation Gnielinski The ]. 28 [

db sn h neisiadPtuo orltosfrNusselt for correlations Pethukov and Gnielinski the using by ed veri ﬁ

pantb)wsas are u.Eprmna aao li uewas tube plain of data Experimental out. carried also was tube) (plain

h xeieto h etecagrwtotluee ti insert strip louvered without exchanger heat the of experiment the ovrf h aiiyo h xeietsse sdi hsstudy, this in used system experiment the of validity the verify 125To 126 (b)

127

Figure 1. (a) Backward louvered strip insert, and (b) geometrical deﬁnitions and computational domain cation Veri 3.1. ﬁ 128 Figure 1. (a) Backward louvered strip insert, and (b) Geometrical definitions and for backward arrangement. 129 computational domain for backward arrangement.

In order to observe the thermal-hydraulic characteristics, covering the fluid flow and heat transfer, .Rslsaddiscussion and Results 3. water was employed as the working fluid, i.e., the hot and cold water flows circulate in inner and outer tube loops, correspondingly. The used fluid properties depend on the temperature. Table1 lists the thermo-physical properties of the used hot and cold water [15], adopted as the inlet values during the

computational process.

tp (11)

/) (/ ff 10

∑

(3b)

1/3

, wo η T

, wo tp T /) (/ uNu Nu

lows:

(3a)

, bo

ttesm upn oe sfol- as power pumping same the at ] 21 , 20 [ index enhancement factor ,, ico ci TT

oprsnbtenNsetnme nacmn ne n friction and index enhancement number Nusselt between comparison

where,

a edtrie ythe by determined be can ) ( factor performance thermal The η − = − =

]. 27 [ Ref. on based calculated was certainty , , , ,, (2) cpccocioowobc ) ) ()( CTThATT T A h T T mC Q

rcinfco ee±46%ad±1.5,rsetvl.Teun- The respectively. 10.85%, ± and 4.63% ± were factor friction

– LMTD

erence. di temperature mean logarithmic the is T where Δ

ﬀ

netite fnndmninlprmtr o usl ubrand number Nusselt for parameters non-dimensional of uncertainties = − =

T ( temperature uid bulk mean the on based .Temaximum The ]. 12 )[

,, ,, D hphhihoiiLMT (1) () T ∆ UA T T mC Q ﬂ

i hropyia rpriso h ae r determined are water the of properties thermo-physical uid the All

ﬂ

xrse sfollows: as expressed

i (/2)(/) uLd ρ (10)

.Teha rnfrrt rmtehtwtri h ne uecnbe can tube inner the in water hot the from rate transfer heat The ]. 12 [

P ∆ n odwtrsd ufcs h qain aetefloigforms following the take equations the surfaces, water-side cold and

iin.Wt igepaefre ovcino ohtehtwater- hot the both on convection forced single-phase a With ditions.

µ (9)

h xeietldt eecletdudrtesed-tt con- steady-state the under collected were data experimental The

e R

i ud ρ

..Dt reduction Data 2.2. eedtrie by: determined were ) f ( factor friction and number Reynolds

i k

(8)

a aida 5,2° n 25°. and 20°, 15°, at varied was ) ( angle Slant mm. 40 at value

u N

ii hd

a taconstant a at was ) S = (pitch strip each between space The b. 2 Fig. in

tdit oerdmd rmselwr sshown as wire steel from made rod core a into tted was strip Louvered ﬁ lows:

epciey oevr h hcns fluee ti netws1mm. 1 was insert strip louvered of thickness the Moreover, respectively. i i sa fol- as is Nu ). Nu ( number Nusselt average the calculate to used was

rmml te ihmnradmjrae bu mad1 mm, 10 and mm 6 about axes major and minor with steel mild from i h of value The side. annulus the on resistance thermal convective the

ihfradarneet.Elpia ovrdsrpisr a made was insert strip louvered Elliptical arrangements. forward with ne ue odcac hra eitneo h ne uewl,and wall, tube inner the of resistance thermal conductance tube, inner

hw h emtia eal fnwdsg ovrdstrips louvered design new of details geometrical the shows a 2 Fig. he eitne nsre:tecnetv hra eitneo the of resistance thermal convective the series: in resistances three

ay±0.2%. ± racy

i h a eemndb h vrl hra eitne hc osssof consists which resistance, thermal overall the by determined was

⎣ ⎦

hrooperae a sdt edtetmeaue iha accu- an with temperatures the read to used was reader thermocouple −− ⎡ ⎤

(7)

oo p i 2 dh k U

.Mlicanldigital Multi-channel ]. 26 [ K 0.1 ± of accuracy an within diameter =

i ioi

1 d n(/) / ( ln ddd

eaueo h al h -yetemcul a ieso . min mm 0.3 dimension has thermocouple K-type The wall. the of perature i

qipdi h ue alo h ne uet eemn h tem- the determine to tube inner the of wall outer the in equipped

i.Smlry e -yetemculswere thermocouples K-type ten Similarly, uid. water hot and cold h ( tube inner the in cient coe transfer : ) ﬂ ﬃ

isboth uids working the of temperature outlet and inlet the measure

olwn xrsini sdt aclt h ovcieheat convective the calculate to used is expression Following ﬂ

ihteacrc f±05.Fu -yetemculswr sdto used were thermocouples K-type Four 0.5%. ± of accuracy the with

D LMT i i ave

T ∆ UA Q (6)

mtrwsue omauetepesr rpars h etsection test the across drop pressure the measure to used was ometer

emndby: termined o ae.Rtmtrhsa cuayo .%R.Adgtlman- digital A RD. 0.5% ± of accuracy an has Rotameter water. hot

U tube, inner the in cient coe transfer heat overall The sde- is , wrt fthe of rate ow the adjusting by obtained was tested number Reynolds ﬃ ﬂ

= ftehtwtrtn n eoeteilto h ne ie The pipe. inner the of inlet the before and tank water hot the of

(5)

ave

Q wrt ftehtwtr h oaee a lcdbtenteoutlet the between placed was rotameter The water. hot the of rate ow

ﬂ hc QQ

eprtr esrmns oaee a rvddt oto the control to provided was Rotameter measurements. temperature

xrse as: expressed wrt,pesr,and pressure, rate, ow of consists system measurement data The

ﬂ

ave Q ( rate transfer heat average The ewe o n odwtris water cold and hot between )

wdcntnl t013k/ rma vredwtrtank. water overhead an from kg/s 0.103 at constantly owed water ﬂ

a ananda nabettmeaueaon 8°.Tecold The °C. 28 around temperature ambient an at maintained was

c h loss Q (4) Q - Q =

uid cold of temperature The surrounding. the to disposed was water ﬂ

a pnlo,tu fe asn hog h etscin h xtcold exit the section, test the through passing after thus loop, open was ie sbellow: as given

loss

ow water Cold kg/s. 0.099 to 0.033 from varied was rate ow water Q ( loss heat The temperatures. water cold hog h nuainis insulation the through )

ﬂ ﬂ

, wo T b,o stebulk the is T temperatures. tube inner wall outer of average the is Hot °C. 60 at constant kept was water hot of temperature inlet The

.Ynnshe al. et Yaningsih I. Energies 2019, 12, 3370 4 of 12

Table 1. Thermo-physical properties at the inlet.

3 Fluid Temperature [K] ρ [kg/m ] µ [kg/m s] Cp [kJ/kg K] k [W/m K] Pr · · · Cold water 300.15 997 8.52 10–4 4.178 0.613 5.81 × Hot water 333.15 983.3 4.67 10–4 4.185 0.654 2.99 ×

Re can be calculated by incorporating the thermo-physical properties of the water and inner diameter of the inner tube, which can be deﬁned as:

ρuDi Re = (1) µ where µ, u, and ρ are the fluid dynamic viscosity, fluid velocity, and fluid density, respectively. Furthermore, considering the variations of the velocity, Re was set to 10,000–17,500. Furthermore, the heat transfer characteristic, represented as the Nu, can be defined as:

hD Nu = i (2) k where k and h are the thermal conductivity and convective heat transfer coefficient of the water, correspondingly. The characteristics of fluid flow were analyzed using the friction factor, f, that can be calculated as follows: ∆P f = (3) 2 (ρu /2)(L/di) f was approximated by using the pressure drop (∆P) between the tube inlet and outlet. In addition, three governing equations, including the continuity, momentum, and energy, were also employed for solving the numerical problem [16]. Each of them can be represented as follows. Continuity equation: ∆P f = (4) 2 (ρu /2)(L/di) Momentum equation: (ρVV) = p + µ( V) (5) ∇· −∇ ∇· ∇ Energy equation: ρCp( VT) = (k T) + (τ V) (6) ∇· ∇· ·∇ · where p, cp, k, and T are the pressure, fluid specific heat, fluid thermal conductivity, and temperature, correspondingly. Both subscripts of i and k show that the direction is toward i and k, respectively. The numerical model was considered to be steady, turbulent, and incompressible. Moreover, the tube walls were considered to be adiabatic, as well as the boundary condition between the fluid and insert. Both natural convection and thermal radiation were negligible and there is no slip at the tube walls. Computation was generated utilizing the commercial software of Ansys Fluent 19 [17], employing the finite volume method. It was also employed in order to solve Equations (4)–(6), in addition to the adoption of the k–ε renormalized group turbulence model. Furthermore, the standard pressure solver with the spatial discretization of the second-order upwind scheme was adopted in this numerical work for solving both energy and momentum equations. Moreover, the semi-implicit method for pressure-linked equations (SIMPLE) algorithm was adopted to approximate the pressure–velocity coupling. In order to validate the plain tube and all louvered strip inserts, a tetrahedron-cell mesh configuration was employed. The regeneration of the mesh is represented in Figure2, and in addition, the face sizing method in several areas was also adopted for the model. Energies 2019, 12, 3370 5 of 12 Energies 2019, 11, x FOR PEER REVIEW 5 of 12

Figure 2.2. Generated grid.

3. ResultsThe grid and systemDiscussion was modeled and generated using the tetrahedral, as shown in Figure2. In addition, the face sizing was also conducted to clarify that the grid was sufficiently dense and In this work, the comparison of the heat transfer and fluid flow characteristics between the fulfilled the grid independence. The selected generated grid systems included 5,601,970, 5,531,216, smooth tube calculated using the method used in this study and one obtained using the established 5,512,714, and 626,635 cells, which are utilized to calculate the louvered strip inserts with S equals correlations was initially conducted in order to ensure calculation accuracy. Calculated Nu was 40, 50, and 60 mm and the smooth-tube case, respectively. These good quality grid systems indicate evaluated and compared by using both the Petukhov and the Gnielinski equation [18,19], while f was that their solution is grid independent. The quality of grid systems was maintained with a value of analyzed using the Blasius equation [20]. skewness of about 0.2. The Petukhov Nu can be assumed as follow: 3. Results and Discussion 𝑓/8 𝑅𝑒 𝑃𝑟 𝑁𝑢 = / / (7) In this work, the comparison of the1.07 heat + transfer 12.7𝑓/8 and fluid𝑃𝑟 flow−1 characteristics between the smooth fortube 10 calculated4 < Re < 5 × using 106. the method used in this study and one obtained using the established correlations was initiallyOn the other conducted hand, the in order Gnielinski to ensure Nu equation calculation can accuracy.be represented Calculated as follow:Nu was evaluated and compared by using both the Petukhov and the Gnielinski equation [18,19], while f was analyzed using 𝑓/8𝑅𝑒 − 1000 𝑃𝑟 the Blasius equation [20]. 𝑁𝑢 = (8) 1 + 12.7𝑓/8/𝑃𝑟/ −1 The Petukhov Nu can be assumed as follow: for 1 × 103 < Re < 5 × 106. ( f /8) Re Pr Moreover, the Blasius correlatioNu =n is used to determine the f given by: (7) 1.07 + 12.7( f /8)1/2(Pr2/3 1) f = 0.3164Re−0.25 − (9) for 104 < Re < 5 106. for 4 × 103 < Re <× 105. Nu TheOn thecomparison other hand, of theNu Gnielinskiand f of the smoothequation tube can calculated be represented using the as follow: method in this study and one calculated with the established correlations( f / 8are)(Re shown1000 in) PrFigures 3 and 4, correspondingly. It can be observed that, the difference beNutween= the results −of this study with the established correlations(8) 1 + 12.7( f /8)1/2(Pr2/3 1) was very small under all evaluated cases. Hence, it could be concluded− that the present calculation shows very reasonable accuracy. In comparison to the established correlations, the results of the for 1 103 < Re < 5 106. calculation× are within× ±5.98% deviation of the Petukhov correlation and ±5.04% deviation of the Moreover, the Blasius correlation is used to determine the f given by: Gnelienski correlation for heat transfer (Nu), as well as ±2.85% deviation of the Blasius correlation for f. Furthermore, the differences in Nu and f were less than0.25 6% and 3%, respectively, as compared to f = 0.3164Re− (9) the experiments. Hence, it can be concluded that the numerical method in this present study provides afor confident 4 103 < levelRe < of10 certainty.5. × The comparison of Nu and f of the smooth tube calculated using the method in this study and one calculated with the established correlations are shown in Figures3 and4, correspondingly. It can be observed that, the difference between the results of this study with the established correlations was very small under all evaluated cases. Hence, it could be concluded that the present calculation shows very reasonable accuracy. In comparison to the established correlations, the results of the calculation are within 5.98% deviation of the Petukhov correlation and 5.04% deviation of the ± ±

Energies 2019, 12, 3370 6 of 12

Gnelienski correlation for heat transfer (Nu), as well as 2.85% deviation of the Blasius correlation for f. ± Furthermore,Energies 2019, 11, thex FOR di ﬀPEERerences REVIEW in Nu and f were less than 6% and 3%, respectively, as compared6 to of the 12 experiments. Hence, it can be concluded that the numerical method in this present study provides a conﬁdent level of certainty.

Figure 3. Validation of Nu for the plain tube. Figure 3. Validation of Nu for the plain tube.

Figure 4. Validation of f for the plain tube. Figure 4. Validation of f for the plain tube. In order to verify the heat transfer enhancement influenced by the presence of a backward louvered insert,In a order comparison to verify with the the heat plain transfer tube (without enhancement insertion) influenced was conducted, by the presence which is of also a backward shown in Figurelouvered5. It insert, can be a observed comparison that with the heat the transferplain tube characteristic, (without insertion) in terms ofwas the conducted,Nu, increased which following is also theshown increase in Figure of Re 5.for It mostcan be cases. observedNu was that significantly the heat transfer influenced characteristic, by the temperature in terms of distribution the Nu, increased of the fluidfollowing flowing the inside increase the enhancedof Re for tube.most Recases., and Nu therefore was significantly also the convective influenced heat by transfer, the temperature improved accordingly.distribution of These the fluid results flowi areng in coherenceinside the enhanced with previously tube. Re conducted, and therefore studies also [13 the–15 convective]. Figure5 also heat showstransfer, that, improved for backward accordingly. louvered Th stripese results inserts, are a pitch in coherenceS of 40 mm with provided previously the highest conductedNu and studies was followed[13–15]. Figure by those 5 also of 50 shows and 60 that, mm. for Furthermore, backward louvered using the strip backward inserts, louvereda pitch S insertof 40 mm also provided offered a betterthe highest result comparedNu and was to the followed plain with by thethose corresponding of 50 and 60 pitch. mm. The Furthermore, presence of ausing backward the backward louvered insertlouvered can insert increase also the offered fluid turbulent a better result intensity compar whiched is to near the toplain the tubewith wall.the corresponding In addition, the pitch. vorticity The aroundpresence the of inserta backward lead to louvered high convective insert can heat increase transfer the comparedfluid turbulent to the intensity backward which louvered is near insert.to the ThetubeNu wall.values In addition, for the backward the vorticity louvered around insert the insert withthe lead corresponding to high convective pitches heat of 40,transfer 50, and compared 60 were into athe similar backward range alonglouvered the increased insert. TheRe. ComparedNu values to for the the forward backward louvered louvered inserts [insert14], the with present the tubecorresponding designs provided pitches aof significant 40, 50, and improvement 60 were in a ofsimilar heat transfer.range along the increased Re. Compared to the forward louvered inserts [14], the present tube designs provided a significant improvement of heat transfer.

Energies 2019, 11, x FOR PEER REVIEW 7 of 12

Energies 2019, 12, 3370 7 of 12 Energies 2019, 11, x FOR PEER REVIEW 7 of 12

Figure 5. Nusselt numbers for various pitches. FigureFigure 5. 5.Nusselt Nusselt numbersnumbers for various various pitches. pitches. The comparison of the temperature distribution of the enhanced tube with backward louvered The comparisonThe comparison of theof the temperature temperature distribution distribution of the the enhanced enhanced tube tube with with backward backward louvered louvered insert and plain tube at a Re of 17,500 for various cross-sections at x is shown in Figure 6. It can be insertinsert and plainand plain tube tube at aatRe a Reof of 17,500 17,500 for for various various cross-sectionscross-sections at at x xis isshown shown in Figure in Figure 6. It6 .can It canbe be observed that for the tube with backward louvered strip inserts, under all evaluated conditions, the observedobserved that that for thefor the tube tube with with backward backward louvered louvered strip inserts, inserts, under under all all evaluated evaluated conditions, conditions, the the temperaturetemperature distribution distribution seemed seemed more more evenly evenly spre spread.ad. In addition, the the temperature temperature distribution distribution temperature distribution seemed more evenly spread. In addition, the temperature distribution became becamebecame more more homogeneously homogeneously distributed distributed following following the the increaseincrease ofof the the twist twist pitch. pitch. The The temperature temperature more homogeneously distributed following the increase of the twist pitch. The temperature contours contourscontours of the of backwardthe backward louvered louvered insert insert met met thethe samesame tendencytendency and and similar similar range range for forthe the of the backward louvered insert met the same tendency and similar range for the corresponding correspondingcorresponding pitches. pitches. In caseIn case of ofthe the tube tube havi havingng backward backward louveredlouvered insert, insert, a bettera better temperature temperature pitches. In case of the tube having backward louvered insert, a better temperature distribution can distributiondistribution can be can earned, be earned, resulting resulting in ina better a better te temperaturemperature gradient than than the the one one of ofthe the smooth smooth tube. tube. be earned,An improved resulting temperature in a better distributi temperatureon causes gradient a larger than temperature the one difference of the smooth between tube. the Antube improved inlet An improved temperature distribution causes a larger temperature difference between the tube inlet temperatureand outlet, distribution resulting in causes a better a convective larger temperature heat transfer di ﬀcoeffierencecient. between As shown the in tube Figure inlet 7, andfor all outlet, and outlet, resulting in a better convective heat transfer coefficient. As shown in Figure 7, for all resultingpitches, in athe better obtained convective highest temperature heat transfer gradient coeﬃ andcient. superior As shown temperature in Figure distribution7, for all are pitches, almost the pitches, the obtained highest temperature gradient and superior temperature distribution are almost obtainedsimilar highest to thetemperature other tube designs, gradient resulting and superior in the best temperature performance distribution for this geometry are almost (see also similar Figure to the similar to the other tube designs, resulting in the best performance for this geometry (see also Figure other6). tube designs, resulting in the best performance for this geometry (see also Figure6). 6).

Figure 6. Contour plots of temperature in cross sections x = 45, 390, 1740, and 2065 mm at Re = 17,500 for (a) plain tube, (b) S = 40 mm, (c) S = 50 mm, and (d) S = 60 mm.

FigureFigure 6. 6. ContourContour plots plots of of temperature temperature in in cross cross sections sections xx == 45,45, 390, 390, 1740, 1740, and and 2065 2065 mm mm at at ReRe == 17,50017,500 forfor ( (aa)) plain plain tube, tube, ( (bb)) SS == 4040 mm, mm, (c (c) )SS == 5050 mm, mm, and and (d (d) )S S= =6060 mm. mm.

Energies 2019, 11, x FOR PEER REVIEW 8 of 12

The temperature in each cross-section under Re of 17,500 is represented in Figure 7. It is obvious that the temperature distribution of the tubes having a backward louvered strip insert was improved. Therefore, it led to a better temperature gradient compared to the plain tube. A better temperature distribution results in a larger temperature difference between the tube inlet and outlet, as a result improving the convective heat transfer rate. For all the observed pitches, high-temperature gradient and a superior temperature distribution can be achieved, leading to the best performance for this Energiesgeometry.2019, The12, 3370 pitches S = 40, 50, and 60 mm resulted in very close values of temperature distribution.8 of 12 It becomes reasonable that the heat transfer rate of all pitch was also close (see again Figure 5).

Energies 2019, 11, x FOR PEER REVIEW 8 of 12

The temperature in each cross-section under Re of 17,500 is represented in Figure 7. It is obvious that the temperature distribution of the tubes having a backward louvered strip insert was improved. Therefore, it led to a better temperature gradient compared to the plain tube. A better temperature distribution results in a larger temperature difference between the tube inlet and outlet, as a result improving the convective heat transfer rate. For all the observed pitches, high-temperature gradient and a superior temperature distribution can be achieved, leading to the best performance for this geometry. The pitches S = 40, 50, and 60 mm resulted in very close values of temperature distribution. It becomes reasonable that the heat transfer rate of all pitch was also close (see again Figure 5).

FigureFigure 7. 7.Temperature Temperature in in the the cross cross sectionsection at Re = 17,50017,500..

TheFigure temperature 8 shows inthe each streamline cross-section patterns under forRe theof 17,500various is designs represented of backward in Figure 7louvered. It is obvious strip thatinserts. the temperatureIn the plain tube, distribution the concentration of the tubes of having the streamlines a backward around louvered the stripwall insertis shown was to improved. be weak Therefore,owing to the it led absence to a better of the temperature louvered tape. gradient For all compared pitches, a touniform the plain distribution tube. A better of the temperature streamlines distributioncan be achieved results owing in a to larger the generation temperature of the diﬀ swirlerence by between the louvered the tube tape inlet insert. and outlet, as a result improving the convective heat transfer rate. For all the observed pitches, high-temperature gradient and a superior temperature distribution can be achieved, leading to the best performance for this geometry. The pitches S = 40, 50, and 60 mm resulted in very close values of temperature distribution. It becomes reasonable that theFigure heat 7. transferTemperature rate inof the all cross pitch section was at Re also = 17,500 close. (see again Figure5). Figure8 shows the streamline patterns for the various designs of backward louvered strip inserts. In the plainFigure tube, 8 shows the concentration the streamline of patterns the streamlines for the various around designs the wall of isbackward shown tolouvered be weak strip owing to theinserts. absence In the of the plain louvered tube, the tape. concentration For all pitches, of the streamlines a uniform around distribution the wall of is the shown streamlines to be weak can be owing to the absence of the louvered tape. For all pitches, a uniform distribution of the streamlines achieved owing to the generation of the swirl by the louvered tape insert. can be achieved owing to the generation of the swirl by the louvered tape insert.

Figure 8. Streamline in cross sections x = 45, 390, 1740, and 2065 mm at Re = 17,500 for (a) plain tube, (b) S = 40 mm, (c) S = 50 mm, and (d) S = 60 mm.

FigureFigure 8. Streamline 8. Streamline in crossin cross sections sectionsx x= =45, 45, 390,390, 1740, and and 2065 2065 mm mm at atReRe = 17,500= 17,500 for ( fora) plain (a) plain tube, tube, (b) S =(b)40 S = mm, 40 mm, (c) S(c=) S50 = 50 mm, mm, and and (d (d) )S S= = 6060 mm.

Energies 2019, 12, 3370 9 of 12

Energies 2019, 11, x FOR PEER REVIEW 9 of 12 Figure9 shows the di fference in the velocity magnitude pathlines among various pitches. Having louveredFigure insert 9 shows strips the produced difference the fluid in the flow velocity to become magnitude clogged andpathlines tended among to form various certain patternspitches. Havingespecially louvered around insert louvers. strips Backward produced louvered the fluid stripflow arrangementto become clogged provides and atended vortex to e ffformect on certain fluid patternsflow which especially can create around a better louvers. flow Backward mixture in louver the fluided strip between arrangement the louvered provides strip anda vortex the tubeeffect wall. on fluidThe shortestflow which pitch canS =create40 mm a better produced flow mixture the most in significant the fluid between vortices. the This louvered vortex strip effect and grows the tube with wall.a decrease The shortest in pitch. pitch However, S = 40 mm the produced existence ofthe vortex most significant is not followed vortices. by theThis significant vortex effect growth grows of withswirl a around decrease the in strip. pitch. This However, condition the causedexistence that of thevortex results is not of thefollowed heat transfer by the significant rate for all growth pitches of(S swirl= 40, around 50, and the 50 mm strip.) were This close,condition although causedS = that40 mmthe results provided of the highestheat transfer value rate (see for also all Figures pitches5 (andS = 740,). The50, and presence 50 mm) of were louvered close, strip although inserts S can = 40 result mm provided in swirl and the fluid highest mixing, value so (see that also the Figures rate of 5heat and transfer 7). The increases.presence of louvered strip inserts can result in swirl and fluid mixing, so that the rate of heat transfer increases.

FigureFigure 9. Pathlines identifiedidentified byby velocity velocity magnitude magnitude for for (a )(aS)= S40 = mm,40 mm, (b) S(b=) 50S = mm, 50 mm, and (andc) S (=c)60 S mm= 60. mm. Figure 10 represents the f variation with Re for the plain tube and three different pitches. The valueFigure of 10f represents thethe decreasingf variation trendwith Re with for the the rise plain of Retube. Using and three a louvered different insert pitches. also o ffTheers valuea higher of ffrictional represents factor the decreasing compared totrend the with plain the one. rise The of fRevalues. Using for a louvered the louvered insert insert also withoffers the a highercorresponding frictional pitches factor ofcomparedS = 40, 50, to and the 60plain were one. about The 7.59, f values 6.51, andfor 5.77the louvered times ashigh insert as with the plain the correspondingtube, correspondingly. pitches of Compared S = 40, 50, to and the plain60 were tube, about the 7.59, flow disturbance6.51, and 5.77 of times backward as high louvered as the insertplain tube,increased correspondingly. the tangential Compared velocity to between the plain the tube, secondary the flow flow disturbance and the of surface backward along louvered the tube insert wall. increasedHence, it canthe betangential concluded velocity that the between backward the louvered secondary insert flow increases and thethe surface friction along factor the significantly tube wall. Hence,in comparison it can be to concluded the plain that tube, the which backward is attributed louvered to theinsert blockage increases and the subsequent friction factor diminishment significantly of inthe comparison flow momentum to the plain generated tube, which by the is backward attributed louvered to the blockage tape as and well subsequent as the pressure diminishment drop for the of thecorresponding flow momentum backward generated louvered by the tape backward tube design. louvered tape as well as the pressure drop for the corresponding backward louvered tape tube design. Moreover, as represented in Figures 5 and 10, the present calculation provides the correlation of the Nu and f to Re, respectively. The tendency is considered consistent with the experimental results from previous works [13–15]. It is reasonable that the backward louvered strip inserts can give a

Energies 2019, 11, x FOR PEER REVIEW 10 of 12

Energieshigher2019 friction, 12, 3370 factor compared to the forward design [14] because the fluid flow is directly blocked10 of 12 by the louvered strip inserts with the backward arrangement.

Figure 10. Friction factor for plain tube and tubes with various pitches. Figure 10. Friction factor for plain tube and tubes with various pitches. Moreover, as represented in Figures5 and 10, the present calculation provides the correlation of 4. Concluding Remarks the Nu and f to Re, respectively. The tendency is considered consistent with the experimental results fromThis previous study works deals [ 13with–15 the]. It modeling is reasonable and thatcalculatio the backwardn of an enhanced louvered tube strip having inserts cana louvered give a higher insert frictionwith various factor pitch compared ratios. to There the forward are several design import [14ant] because conclusions the fluid which flow can is directlybe drawn blocked as follows: by the louvered strip inserts with the backward arrangement. • Nusselt number (Nu) and friction factor (f) of the louvered insert having three different pitches 4. Concludingare higher compared Remarks to those of the plain tube. In addition, it is also clarified that as the pitch ratio decreases, Nu, f, and η increase accordingly. • TheThis presence study deals of the with louvered the modeling insert andwith calculation the smallest of anpitch enhanced S = 40 tubemm can having yield alouvered a significantly insert withhigher various heat pitch transfer ratios. rate There and arefriction several factor. important conclusions which can be drawn as follows: • The largest increase in the Nu and f was observed to be up to 1.81 and 7.59 times higher, Nusselt number (Nu) and friction factor (f ) of the louvered insert having three different pitches • respectively, compared to those of the smooth tube. are higher compared to those of the plain tube. In addition, it is also clarified that as the pitch ratio decreases, Nu, f, and η increase accordingly. Author Contributions: A.T.W. and M.A. conceptualized the main idea of the research work. B.K., P., and A.P. The presence of the louvered insert with the smallest pitch S = 40 mm can yield a significantly •contributed to the investigation and data analysis. M.A. contributed actively in supervising the research. A.T.W. wrotehigher the manuscript. heat transfer All authors rate and contributed friction factor. in writing and checking the final manuscript. The largest increase in the Nu and f was observed to be up to 1.81 and 7.59 times higher, respectively, •Funding: The authors greatly thank Universitas Sebelas Maret for funding this study (Research ID: 00310871042042019compared to) those. In addition, of the smooth the first tube. author expresses great gratitude to the Ministry of Research, Technology and Higher Education of the Republic of Indonesia that provides the Program of World Class AuthorProfessor Contributions: scheme B, 2019A.T.W. (Grant and No. M.A. T/48/D2.3/KK.04.05/2019) conceptualized the main for a idea valuable of the opportunity research work. to have B.K., a P.short-term and A.P. contributedresearch visit to to the The investigation University andof Tokyo, data analysis. Japan, for M.A. the completion contributed of actively the manuscript. in supervising the research. A.T.W. wrote the manuscript. All authors contributed in writing and checking the final manuscript. Acknowledgments: Deep thanks also to their former colleague, Wahyu Nur Utama, for his assistance during Funding:the investigationThe authorsand data greatly collection thank at the Universitas Laboratory Sebelasof Heat Transfer Maret for and funding Thermodynamics. this study (Research ID: 00310871042042019). In addition, the first author expresses great gratitude to the Ministry of Research, Technology andConflicts Higher of Education Interest: The of the authors Republic declare of Indonesia no potential that providesconflict of the interest. Program of World Class Professor scheme B, 2019 (Grant No. T/48/D2.3/KK.04.05/2019) for a valuable opportunity to have a short-term research visit to The UniversityNomenclature of Tokyo, Japan, for the completion of the manuscript. Acknowledgments: Deep thanks also to their former colleague, Wahyu Nur Utama, for his assistance during the investigationd inner and tube data diameter collection at[m] the Laboratory of Heat Transfer and Thermodynamics. ConflictsD ofouter Interest: tubeThe diameter authors [m] declare no potential conflict of interest. f friction factor h average convective heat transfer coefficient of the plain tube (W/m2⋅K)

Energies 2019, 12, 3370 11 of 12

Nomenclature d inner tube diameter [m] D outer tube diameter [m] f friction factor h average convective heat transfer coefficient of the plain tube (W/m2 K) · k thermal conductivity [W/m K] · k effective thermal conductivity [W/m K] eff · L inner tube length [m] Nu average Nusselt number Pr Prandtl number Re Reynolds number S pitch [m] t thickness [m] T temperature [K] u fluid velocity [m/s] w width of the delta-wing [m] W width of the aluminum strip [m] y twist pitch [m] ∆P pressure drop across the tube [Pa] η thermal performance factor µ dynamic viscosity [kg/m s] · ρ density [kg/m3] Subscripts i inner o outer p plain tube t tube with insert

References

1. Eiamsa-ard, S.; Pethkool, S.; Thianpong, C.; Promvonge, P. Turbulent flow heat transfer and pressure loss in double pipe heat exchanger with louvered strip inserts. Int. Commun. Heat Mass Transf. 2008, 35, 120–129. [CrossRef] 2. Fan, A.W.; Deng, J.J.; Nakayama, A.; Liu, W. Parametric study on turbulent heat transfer and flow characteristics in a circular tube fitted with louvered strip inserts. Int. Commun. Heat Mass Transf. 2012, 55, 5205–5213. [CrossRef] 3. Mohammed, H.A.; Hasan, H.A.; Wahid, M.A. Heat transfer enhancement of nanofluids in double pipe heat exchanger with louvered strip inserts. Int. Commun. Heat Mass Transf. 2013, 40, 36–46. [CrossRef] 4. Huisseune, H.; T'Joen, C.; De Jaeger, P.; Ameel, B.; De Schampheleire, S.; De Paepe, M. Influence of the louver and delta winglet geometry on the thermal hydraulic performance of a compound heat exchanger. Int. J. Heat Mass Transf. 2013, 57, 58–72. [CrossRef] 5. Wenbin, T.; Yong, T.; Bo, Z.; Longsheng, L. Experimental studies on heat transfer and friction factor characteristics of turbulent flow through a circular tube with small pipe inserts. Int. Commun. Heat Mass Transf. 2014, 56, 1–7. 6. Tabatabaeikia, S.; Mohammed, H.A.; Nik-Ghazali, N.; Shahizare, B. Heat transfer enhancement by using different types of inserts. Adv. Mech. Eng. 2014, 6, 1–13. [CrossRef] 7. Jang, J.-Y.; Chen, C.-C. Optimization of louvered-fin heat exchanger with variable louver angles. Appl. Therm. Eng. 2017, 91, 138–150. [CrossRef] 8. Park, J.S.; Kim, J.; Lee, K.S. Frosting behaviors and thermal performance of louvered fins with unequal louver pitch. Int. J. Heat Mass Transf. 2016, 95, 499–505. [CrossRef] 9. Park, J.S.; Kim, J.; Lee, K.S. Thermal and drainage performance of a louvered fin heat exchanger according to heat exchanger inclination angle under frosting and defrosting conditions. Int. J. Heat Mass Transf. 2017, 108, 1335–1339. [CrossRef] Energies 2019, 12, 3370 12 of 12

10. Wei, H.; Ge, Z.; Yang, L.; Du, X. Entransy based optimal adjustment of louvers for anti-freezing of natural draft dry cooling system. Int. J. Heat Mass Transf. 2016, 95, 499–505. [CrossRef] 11. Sadeghianjahromi, A.; Kheradmand, S.; Nemati, H. Developed correlations for heat transfer and flow friction characteristics of louvered finned tube heat exchangers. Int. J. Therm. Sci. 2018, 129, 135–144. [CrossRef] 12. Zuoqin, Q.; Qiang, W.; Junlin, C.; Jun, D. Simulation investigation on inlet velocity profile and configuration parameters of louver fin. Appl. Therm. Eng. 2019, 134, 468–481. 13. Yaningsih, I.; Wijayanta, A.T. Influences of pitch-length louvered strip insert on thermal characteristic in concentric pipe heat exchanger. MATEC Web Conf. 2016, 101, 03014. [CrossRef] 14. Yaningsih, I.; Wijayanta, A.T.; Miyazaki, T.; Koyama, S. Thermal hydraulic characteristics of turbulent single-phase flow in an enhanced tube using louvered strip insert with various slant angles. Int. J. Therm. Sci. 2018, 134, 355–362. [CrossRef] 15. Wijayanta, A.T.; Pranowo; Mirmanto; Kristiawan, B.; Aziz, M. Internal flow in an enhanced tube equipped with square-cut twisted tape insert. Energies 2019, 12, 306. [CrossRef] 16. Versteeg, H.K.; Malalasekera, W. An Introduction to Computational Fluid Dynamics—The Finite Volume Method; Prentice Hall: Harlow, UK, 2007. 17. ANSYS. ANSYS Fluent 19.0: Theory Guide; Fluent Inc.: Canonsburg, PA, USA, 2018. 18. Saha, S.K.; Tiwari, M.; Sundén, B.; Wu, Z. Advances in Heat Transfer Enhancement; Springer: Basel, Switzerland, 2016. 19. Cengel, Y.A.; Ghajar, A.J. Heat and Mass Transfer: Fundamentals and Applications, 5th ed.; McGraw-Hill: New York, NY, USA, 2014. 20. White, F.M. Fluid Mechanics, 7th ed.; McGraw-Hill: New York, NY, USA, 2011.

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).