Jtam-A4.dvi JOURNAL OF THEORETICAL AND APPLIED MECHANICS 56, 3, pp. 751-763, Warsaw 2018 DOI: 10.15632/jtam-pl.56.3.751 EXPERIMENTAL AND NUMERICAL STUDY OF THE HEAT TRANSFER AND PRESSURE DROP IN TRIANGULAR CHEVRON CHANNELS Hossein Dolatabadi, Abolfazl Hajizadeh Aghdam Department of Mechanical Engineering, Arak University of Technology, Arak, Iran e-mail: abolfazl hajizade@yahoo.com Chevron channels are one of the popular techniques that are extensively used in manufac- turing of compacted heat exchangers. The present paper deals with the experimental and numerical analysis on the fluid flow and heat transfer in a triangular chevron channel. The studies are carried out for a uniform wall heat flux equal to 1350W/m2 using air as the working fluid. TheReynolds number varies from 1000 to 10000, phase shifts 0◦ ¬φ¬ 180◦, and channel heights are5¬D¬ 35mm.The study shows a significant effect of the optimum structure of the triangular chevronon theheat transfer rateand friction loss over the channel wall. The Nusselt number increases and the Thermal Enhancement Factor (TEF) decreases with the increasing Reynolds number. The best performance is noticed on the phase shift, φ=90◦. For triangular chevron surfaces, themaximumheat transfer is obtained 33.33%up to 37.5%more than for a smooth wall channel. Keywords: heat transfer, pressure drop, thermal enhancement factor, phase shift 1. Introduction Chevron channels are basic channel geometry in plate heat exchangers because of their efficient heat exchange capabilities. Today, to increase the heat flux rate in nuclear reactors, turbine blades and electronic equipment, some companies are using chevron surfaces. Chevron surfaces create turbulent sublayer of laminar flows, as a result the heat transfer becomes increased. When the friction factor increases, the fan power will increase for a stable velocity of the fluid flow. To generate more power and enhance system performance of designed turbines, the inlet temperature of the fluidmust be increased. In some cases high temperature of the fluid increases the temperature of turbine blades to themelting point. In such situationsmaking use of alloy or superalloysmightbea technical solution, but it is justifiedecconomically.To reduce temperature of hot surfaces, one has to use surfaces with spatial geometry including ribs or chevron surfaces. The experience proves that chevron surfaces lead to a suitable pressure drop. In solar systems, the fluid flow is laminar so the heat transfer is low, and the chevron surfaces create turbulent sublayers increasing the heat transfer. One can increase the heat transfer in fluids by: 1 – separation of fluids, 2 – making laminar sublayers turbulent, 3 – replacing flows in the hot surface. Thefirstmajor problem in designing heat exchanger is to savemore energy and increase the heat trasfer rate as well as reduce friction. Thus, chevron surfaces play an important here. The flow and heat transfer in chevron channels have been extensively investigated in the recent years, see for example Yang and Chen (2010), Vanaki et al. (2014), Sakr (2015). Khoshvaght-Aliabadi et al. (2016) found a significant improvement in the heat transfer co- efficient by using pure water instead of a water-ethylene glycol mixture. However, the Nusselt number increased considerablywithagrow in theweightpercentage of ethylene glycol in thewor- king fluid. The convective thermal resistance was noticeably reduced by using the SWMCHSs. As an example, a reduction of 113.8% was obtained for the water flow at the mass flow rate of 752 H. Dolatabadi, A.H. Aghdam 0.024kg/s in the SWMCHS with l=20mm and a=1.0mm, compared to the straight MCHS at the same conditions. Yongsiri et al. (2014) found that the ribs which induced recirculation gave a higher Nusselt number and friction factor than those which did not. Among the ribs examined, the ones with θ=60 yielded a comparable heat transfer rate 1.74 times of those in the smooth channel, and θ=120 yielded a thermal performance factor 1.21 whichwas higher than those given by others. Al-Shamani et al. (2015) showed interesting results of changing rib groove shapes in thermal and fluid fields. The results proved that trapezoidal grooves with increasing height in the flow direction (Trap+R-TrapG) gave the highest Nusselt number in comparison with other types of trapezoidal channels. 1) By changing types of nano particles (Al2O3, CuO, SiO2, and ZnO) the results revealed that SiO2 led to the highest Nusselt number, then followed byAl2O3, ZnO, and CuO, respectively, while pure water gave the lowest value of Nusselt number. 2) The Nusselt number increased with the increasing volume fraction of nanoparticles. 3) The Nusselt number increased gradually with a decrease in diameter of the nano particles. 4) The Nusselt number increased gradually with growth in the Reynolds number in the range of 10000-40000. Eiamsa-ard and Changcharoen (2011) found that by reducing the degree of sudden changes of the main flow, the flow separation suppresses and thus the corner separation of bubbles in front and rear surfaces of the rib decreases. The streamlines proved that among the ribs with concave surfaces unrecognized corner separation of bubbles appeared, whereas those with modified convex surfaces generated corner separation of bubbleswith size comparable to that of theunmodified square rib.All ribswith concave surfaces induceda larger recirculation zone than the others, resulting inhigh turbulence intensity.With good conformation of the streamlines, the rib with concave-concave surfaces gave the highest Nusselt number and the friction factor while those with convex-concave surfaces provided the lowest friction factor with moderate Nusselt numbers. Due to the prominent effect of a low friction factor, the rib with the convex-concave surface offers the highest TEFwith the maximum value of 1.19. Xie et al. (2014) studied the flow structure and heat transfer in a square passage with offset mid-truncated ribs. The heated surface of 135◦ V-shapedmid-truncated ribs provided the highest heat transfer enhancement, while the heated surface of 90◦ mid-truncated ribswith no staggered arrangement behaved best in reducing the pressure loss penalty. Moon et al. (2014) proved that the new boot-shaped rib gave the best heat transfer perfor- mance with an average friction loss performance, and the reverse pentagonal rib gave the best friction loss performance. Dellil et al. (2004) conducted a geometrical parametric study by changing the amplitude- to-wavelength ratio. Comparison of predicted results for a wavy wall with those for a straight channel indicated that the averaged Nusselt number increased until a critical value was re- ached where the amplitude wave was increased. However, that heat transfer enhancement was accompanied by an increase in the pressure drop. Characteristics of the fluid flow and heat transfer in a periodic fully developed region of a corrugated duct were numerically obtained using the finite element software by Islamoglu and Parmaksizoglu (2004). Both theheat transfer coefficient and thepressuredrop for the corrugated ducts were in good agreement. In addition, the finite element technique can be used to simulate heat exchanger channels. However, the present study is concerned with a special triangular chevron channel with the focus on variation in the phase shiftφ anddistance of platesD. The experimental andnumerical methods have been used to find the best φ andD with considering the maximumNu and TEF as well as the minimum friction factor. Experimental and numerical study of the heat transfer... 753 2. Experimental setup The experimental set up is shown in Fig. 1. The plate geometry is 125×80mm. A forced- -convection air flow was used in the experiments and an equal power input for heated length was established in the bottom walls. To measure the temperature distribution of each chevron wall, nine thermocouples (KTJmodel TA-288) were used with 4mmdiameter holes drilled into the side wall (number 8 in Fig. 1). These holes were located at axial distances of (T1), (T2), (T3), (T4), (T5), (T6), (T7), (T8), (T9). The accuracy of the thermocouples was ±1◦C. A U-manometer was located at the chevron surface tomeasure pressure dropwithin the triangular chevron channel (number 4 in Fig. 1). The accuracy of pressure transducer was±1mmC2H6O. The test section was also isolated to avoid thermal losses. The experimental procedure involved adjusting the flow rate to the desired value (number 1 in Fig. 1). After the fan was turned on and the desired Reynolds number was obtained, the power input of plate heaters (number 6 in Fig. 1) gradually increased andmaintained at 1350W/m2 to provide sufficient measurement while the fluid property varied.Theheat supplied into the chevronwalls was adjusted to achieve the desired level by using electric heaters, which were 1.314mm thick, 80mmwide and 150mm long. They were located at the back of the bottom triangular chevron plate. The voltage and current of the electric input to the plate type heaters were controlled by a DC power supply (P.A. Hilton model Ltd H111/07332) unit (number 5 in Fig. ). Temperatures were recorded at intervals of 15min until a steady state was reached. Steady state conditions were assumed to prevail when temperature measurements (number 8 in Fig. 1) on the plates were within±1◦C. These instruments were placed and fitted in a container made of the plexiglas insulator. Figure 1 shows the assembled configuration of the test module with the fluid entrance and exit from the channel. The chevron surfaces are fabricated fromblocks of aluminum1060 alloywhose physical properties are shown in Table 1. Fig. 1. Assembled configuration of the test module with triangular chevron surfaces A centrifugal fan which was made of stainless steel and had the inner diameter of 90mm (number 2 in Fig. 1) sucked the room air at the room temperature and exhausted into the atmosphere through the test section (thatmade of plexiglass with 8mm thickness (number 7 in Fig. 1)). The flow was controlled by a control valve placed inside the fan (number 1 in Fig. 1) and the air velocity was calculated by a velocity sensor (Dwyer AVU-3V model HP111 LH) as shown in Fig. 1 (number 3). The display on the DC power controller of the centrifugal fan (P.A.Hilton LtdH111/00629) is shown (number 9) inFig. 1. The accuracy of the velocity sensor was ±0.1m/s. 754 H. Dolatabadi, A.H. Aghdam Table 1.Physical properties of the aluminum 1060 alloy Properties Conditions T [◦C] Density [×1000 kg/m3] 2.7 25 Poisson’s ratio 0.33 25 Elastic modulus [GPa] 70-80 25 Hardness (HB500) 23 25 Thermal expansion 23.610−6/◦C 20-100 Thermal conductivity 234W/mK 25 2.1. Physical model Geometry of the experimental setup of the channel is shown inFig. 2.Thephysical properties of the air have been assumed to remain constant at the average bulk temperature. Impermeable boundary and no-slip wall conditions have been assumed over the channel walls as well as the chevron surfaces. The distance between the apexes of the triangle with the vertical axis coordinates (Dv =7.5mm)was constant. Length of the channel in front of the test section was Linlet =30cm and length of the channel behind the test section Loutlet =25cm. Length of the test section was (Ltest) 15cm and height and length of the ribs wereH =L=15mm. Location of thermocouples shown with a blue circle was HT =7.5mm of the base chevron surface. Fig. 2. Geometry of the experimental setup of the channel The next section creates the entering boundary condition with a heat flux rate of the heater 1350w/m2, The other wall is an insulating one and the inlet section with the input speed (velocity inlet) as well as the fluid outlet (pressure outlet). The inlet temperature of theworking fluid (air) was kept constant at 298.15K. 3. Performance parameters Parameters of interest in the present work are the Reynolds number, friction factor, Nusselt number and hydraulic diameter Dh = 2HavgW Havg +W (3.1) where Havg is average distance between the chevron surfaces, and W is width of the chevron surfaces. Re= ρuDh µ (3.2) Experimental and numerical study of the heat transfer... 755 where u is velocity of the fluid at the inlet of the test section,Dh is hydraulic diameter and µ is dynamic viscosity f =2 ∆p L (Dh ρu2 ) (3.3) The friction factor is computed by the pressure drop ∆P across length of the duct L, and ρ is density of flow. hx = q̈ Tx−Tb (3.4) where hx is the local convective heat transfer coefficient, Tx is the local temperature, Tb is the bulk temperature and q̈ is heat flux. The heat transfer was measured through the local Nusselt number which can be written as Nux = hxDh k (3.5) where Nu is the Nusselt number andK is conductivity. Then, the average Nusselt number can be obtained by Nu= 1 L ∫ Nuxdx (3.6) and Nus =0.024Re 0.8Pr0.4 fs =0.085Re −0.25 (3.7) where Nus and fs stand respectively for the Nusselt number and friction factor of the smooth duct. The Thermal Enhancement Factor (TEF) can be written according to Eiamsa-ard and Changcharoen (2011) TEF= Nu Nus ( f fs ) − 1 3 (3.8) 4. Numerical model Numerical analysis of the thermal behavior and flow dynamic characteristics of the chevron channel has been carried out to predict the heat transfer and pressure drop. The governing equations have been solved using a finite volume approach. The time-independent incompressi- ble Navier-Stokes equations and the turbulence model were discretized using the finite volume method. Many investigators predicted turbulent forced convection in a rectangular duct with periodic chevron shapes by utilizing different turbulencemodels, such as k−ε, k−ω, Reynolds stress model (k−ε) and large eddy simulations (LES)models. The k−εmodel made from the two-equation models when predicting flow patterns of revolving flows. In the present study, the k−εmodel is used for the turbulencemodeling, and the SIMPLE algorithm is used to handle the pressure-velocity coupling. The discretized nonlinear equations are implemented implicitly. To evaluate the pressure field, the pressure-velocity coupling algori- thm SIMPLE (Semi Implicit Method for Pressure-Linked Equations) is selected. The following assumptions are applied in the simulations: the flow is steady, fully developed turbulent and two dimensional, the thermal conductivity of the channel wall and chevron material do not change with temperature, and the channel wall and chevron material are homogeneous and isotropic with an enhanced wall treatment function. The solutions are considered to be converged when 756 H. Dolatabadi, A.H. Aghdam the normalized residual values are less than 10−6 for all variables, but less than 10−5 only for the continuity equation. A grid independence test has been performed for the channel to analyze the effects of grid sizes on the results, as shown inFig. 3. It is found that further increase in the grid beyond 51159 cells results in a variation in theNusselt number less than 1%, thus this grid number is taken as a criterion of grid independence. This fine mesh size is able to provide good spatial resolution for the distribution of most variables within the channel. Fig. 3. Investigation of the independency of mesh elements Figure 4 shows the grid size of mesh elements near the chevron walls. To investigate the independency of the mesh, the number of mesh elements increases near the chevron walls by y+ =2.014253. Fig. 4. Grid size of mesh element near the chevron walls 5. Results and discussion 5.1. Phase shift and Reynolds numbers variation in a triangular chevron channel Figure 5 shows experimental results of temperature variation of the fluid vs. dimensionless distance x/D for a phase shift of 0◦ and D = 5mm. The air temperature has increased by an increase in the dimensionless distance. The same results are for other phase differences. Figure 6 illustrates the average fluid temperature across the triangular chevron wall for different Reynolds numbers. The fluid temperature significantly decreases with the increasing Reynolds number.Thedifference between the numerical and experimental results is about 1%¬ error¬ 11%, which proves good validation between both results. Experimental and numerical study of the heat transfer... 757 Fig. 5. Experimental results of temperature variation of the fluid vs. dimensionless distance x/D for a phase shift of 0◦ andD=5mm Fig. 6. Variation and validation of the average temperature of the fluid vs. Re in experimental and numerical results at phase shifts 0◦, 90◦, 180◦ The numerical and experimental results of the variation of Nu for three phase shifts and within 3000¬Re¬ 10000 are shown in Fig. 7. The Nusselt number for the chevron channels is higher than for the plain channel because theNusselt number depends on the heat transfer rate. On the other hand, the values of the Nusselt number are found to increase with an increase in Re in all cases. Best results are obtained for the phase difference 90◦. The Nusselt number for φ=90◦ is about 4.8% higher than for φ=0. In fact, φ=90◦ makesmore turbulence interrupts in the development of a thermal boundary layer. Thevortices induced in andaround the chevron channels are thought to be responsible for the increase of turbulence intensity of the flowwhich leads to higher heat transfer rates. Figure 8 shows the experimental and numerical results of the effect of the Reynolds number and the relative phase shift on the friction factor. When the fluid passes through the chevron channel, a considerable pressure drop occurs. The value of the friction factor decreases with the increasingReynolds number in all cases, as expected, due to suppression of the viscous sub-layer with an increase in Re. Also, the pressure drop in the channel with a phase shift φ = 180◦ is 758 H. Dolatabadi, A.H. Aghdam Fig. 7. Variation of Nu vs. Re for phase shifts 0◦, 90◦, 180◦ in experimental and numerical results Fig. 8. Variation of f vs. Re for phase shifts 0◦, 90◦, 180◦ in experimental and numerical results greater than in other cases. As expected, the f from the φ=180◦ array is substantially higher than that with φ = 0◦ and φ = 90◦, whereas the φ = 0◦ yields the lowest φ. The f value of φ = 180◦ is found to be about 4% above φ = 90◦ and about 5.4% over the φ = 0◦. It can be interpreted that the φ= 180◦ causes a higher turbulence intensity in the flow due to more oscillating streamlines than for the other phase shifts (see Fig. 10). But it creates a “stronger” recirculation region and vortices inside the chevron surfaces, so it prevents the fluid from good mixing. Thus, it results in a less increase in the heat transfer in comparison with both φ=90◦ andφ=180◦. Using the phase shift, one can reduce the occurrence of the recirculation providing more surface sweep, therefore, the negative effect of the recirculation region on the heat transfer can be decreased. The shift between the numerical and experimental results for each phase shift is about 1%¬ error¬ 11%, which proves a good validation. Figure 9 shows the variation of TEFwith Re for all phase shifts. The combined effect of Nu and f values has been simplified by TEF, Eq. (3.8). In general, TEF tends to decrease with the Experimental and numerical study of the heat transfer... 759 rise of Re. It is worth noting that the TEF value of the phase shift 90 is the highest and found to be the best among other phase shifts. Fig. 9. Variation of TEF vs. Re for phase shifts 0◦, 90◦, 180◦ in experimental and numerical results For phase shifts 0◦, 90◦ and 180◦, themaximumTEFvalues are, respectively, about 1.05, 1.1 and 1.07. It is seen that the phase shift 90◦ gives the highest TEF at lower Reynolds numbers. At the given chevron surfaces, the phase shift 90◦ yields TEF around 3-7% higher than that of other two phase shifts. Because of considerably higher Nu and lower f, only for the phase shift 90◦ will be further investigated in the subsequent Section. The variation of the dimensionless parameters Nu, f andTEF versus Re for phase shifts 0◦, 90◦ and 180◦ are presented in Figs. 7, 8 and 9, respectively. For three cases and at all positions, it is observed that there is a relatively good concordance between the experimental data and the numerical results. For very low flow rates, as in the experiment, it is believed that the observed differences between the experimental and numerical datamay be partly caused by a small error in the DC source power, heater and the manometer velocity sensor in the system, which can slightly disrupt the turbulence flow. Another possible cause of these differences is the fact that the hot junction of the thermocouples, which are the top of the plate surface, covers a small circular area with a diameter of approximately 1mm and, therefore, the measured temperature is actually the average temperature of that small surface. Figure 10 shows the streamlines associated by amodified various phase shift at Re=10000. By focus on the upstream flow structures of various phase shifts, the phase shift is φ=90◦ and the size of bubbles decrease. The volume of the bubbles, induce larger recirculation zones which provide higher turbulent intensity. This may be related to a decrease of the laminar sub-layer and to themaximumheat transfer for φ=90◦.When the phase shift is close φ=90◦, it creates more vortex and decreases bubbles from triangular chevron surfaces. When the phase shift angle increases, the friction factor also increases because the fluid is oscillating between chevron surfaces. It is clear that the effect of the phase shift on temperature and flow development in the chevron channel is that there are smaller recirculation regions and more separation bubble regions formed in the adjacent inlet/outlet. This effect increases by closing to the phase shift φ=90◦, which means that the influence of the wall on the main stream becomes greater. This generates greater swirl flow in thewavywall due to transfer vortices of the bulk flow field in the wavy wall. This induces a higher temperature gradient near the wavy wall. Therefore, the net heat transfer rate from the wavy wall to the fluid is increased. 760 H. Dolatabadi, A.H. Aghdam Fig. 10. Streamline of fluids for phase shifts 0◦, 90◦, 180◦ 5.2. Effect of distance between triangular chevron surfaces Figure 11 demonstrates the variation of the average fluid temperature above 7.5mmfromthe triangular chevron plate according to numerical and experimental results. In Fig. 11, it is seen that the numerical calculations with the standard wall function are similar to the experimental data with the chevron channel for Re = 10000 and φ = 180◦. The wall temperature increases when the distance between the surfaces increases. It can interpreted that the fluid flow becomes more complex near the wall region whenD increases. As a result, the velocity of fluids decrease with groving D and the temperature near the wall becomes warmer. The average numerical and experimental Nusselt number for triangular chevron surfaces at Re = 10000 for φ = 180◦ is presented in Fig. 12. There is good agreement between numerical and experimental results with an error less than 10% for most of the results. It can be seen that with an increase in the distance between triangular chevron surfaces, the Nusselt number decreases. In fact, the volume of vortices in the laminar sublayer and the velocity of fluids decrease. Figure 12 shows Nu for triangular chevron surfaces. It is greater than for a smooth surface belowD=15mm.WhenD increases for a constantRe=10000, the velocity of the fluid decreases. As a result, Nu for triangular chevron surfaces is less than that for smooth surfaces above D=15mm. Figure 13 shows the variation of the experimental and numerical friction factor along the channel for Re= 10000 and φ=180◦. There is a good validation of the numerical and experi- mental results with an error less than 9% for most of results. The friction factor for triangular Experimental and numerical study of the heat transfer... 761 Fig. 11. Experimental and numerical variation of average temperature of the chevron plate vs. different distances Fig. 12. Experimental and numerical variation of Nu of the chevron plate vs. different distances chevron surfaces is the highest compared to the smooth surface. It is obvious that the friction factor decreases gradually for all configurations with an increase in the Reynolds number. Based on the same pumping power consumption, the TEF is shown and compared for dif- ferently distanced chevron surfaces tested, see Fig. 14. By considering the heat transfer and the pressure drop simultaneously at Re = 10000 and φ= 180◦, the working conditions should give a high TEF. It is seen from the figure that the TEF tends to decrease as D increases (for numerical results). Consistently, the investigation reveals a higher thermal enhancement factor atD=5mm. There appears transfer of additional heat by conduction which provides a better fluid mixing. As a result, to obtain the maximum TEF, one should to get the heat exchanger with the minimum distance between surfaces, see Fig. 14. 762 H. Dolatabadi, A.H. Aghdam Fig. 13. Experimental and numerical variation of the friction factor for the chevron plate vs. different distances Fig. 14. Variation of TEF for the chevron plate vs. different distances 6. Conclusions Atriangular chevron channel is a good alternative in highheat fluxapplications ormore efficient heat exchange devices used in a variety of engineering structures such as heating and air condi- tioning systems. In this paper, the effect of phase shift and distance between chevron channels is examined for Reynolds numbers ranging from 1000 to 10000. The aim of the stugy is to achieve to themaximumheat transfer and thermal enhancement factor aswell as theminimumpressure drop. The best phase shift angle is φ = 90◦. Also the distance between chevron surfaces D has been investigated. The results show that D = 5mm is the best distance between chevron surfaces in getting the maximum TEF and Nu, and the minimum f. Decreasing and holding the average base temperature of the chevron surfaces at a constant level, particularly at higher Reynolds numbers has been successfully achieved. Increasing the Reynolds number leads to a more complex fluid flow and the heat transfer rate.When the phase shift gets close to φ=90◦, Nu and TEF reach the maximum rate, whereas f the minimum nalue. The channels with the phase shift angle φ= 90◦ are the most attractive from the viewpoint of energy saving compa- red to others with the phase shifts φ = 0◦ and 180◦. For the triangular chevron surfaces, the Experimental and numerical study of the heat transfer... 763 maximum heat transfer is obtained from 33.33% up to 37.5%, which is higher than for smooth surfaces. References 1. Al-Shamani A.N., Sopian K., Mohammed H.A.,Mat S., Ruslan M.H., Abed A.M., 2015, Enhancement heat transfer characteristics in the channel with Trapezoidal rib-groove using nano fluids,Case Studies in Thermal Engineering, 5, 48-58 2. DellilA.,AzziA., JubranB.A., 2004,Turbulent flowand convectiveheat transfer in a chevron wall channel,Heat and Mass Transfer, 40, 793-799 3. Eiamsa-ard S., Changcharoen W., 2011, Analysis of turbulent heat transfer and fluid flow in channels with various ribbed internal surfaces, Journal of Thermal Science, 20, 3 4. Islamoglu Y., Parmaksizoglu C., 2004, Numerical investigation of convective heat transfer and pressure drop in a chevron heat exchanger channel,Applied Thermal Engineering, 24 5. Khoshvaght-Aliabadi M., Sahamiyan M., Hesampour M., Sartipzadeh O., 2016, Experi- mental study on cooling performance of sinusoidal-chevronminichannel heat sink,Applied Thermal Engineering, 92, 50-61 6. MoonM.-A.,ParkM.-J.,KomK.-Y., 2014,Evaluationof heat transfer performances of various rib shapes, International Journal of Heat and Mass Transfer, 71, 275-284 7. Sakr M., 2015, Convective heat transfer and pressure drop in V-chevron channel with different phase shifts,Heat Mass Transfer, 51, 129-141 8. Vanaki Sh.M., Mohammed H.A., Abdollahi A., Wahid M.A., 2014, Effect of nanoparticle shapes on the heat transfer enhancement in a chevron channel with different phase shift, Journal of Molecular Liquids, 196, 32-42 9. Xie G., Liu J., Ligrani P.M., Sunden B., 2014, Flow structure and heat transfer in a square passagewith offsetmid-truncated ribs, International Journal of Heat andMass Transfer,71, 44-56 10. YangY.-T.,ChenP.-J., 2010,Numerical simulationoffluidflowandheat transfer characteristics in channel with V chevron plates,Heat Mass Transfer, 46, 437-445 11. Yongsiri K.A., Eiamsa-ard P.B., Wongcharee K.C, Eiamsa-ard S., 2014, Augmented heat transfer in a turbulent channel flow with inclined detached-ribs, Case Studies in Thermal Engineering, 3, 110 Manuscript received September 25, 2017; accepted for print November 29, 2017