Numerical Investigations On Augmentation Of Heat Transfer In Oil Coolers

Numerical Investigations On Augmentation Of Heat Transfer In Oil Coolers

D.Eswaraiap, S. Naga Sarada2,

1Department of Mechanical Engineering, GRIET, Bachupally, Hyderabad.-500090, Andhra Pradesh, India:

2Department of Mechanical Engineering, JNTUH College of Engineering, Kukatpally, Hyderabad 500085, Andhra Pradesh, India.

Abstract: Computational fluid dynamics (CFD) package (ANSYS CFX 11.0) was used to numerically study the heat transfer characteristics of oil cooler under laminar flow conditions with Reynolds number on tube side varying from 250 to 2400. The working fluid on tube side is ISOVG46 Turbinol and on shell side is water. Simulation is done for different flow rates of oil. Simulated results of Nusselt number and friction factor are in good agreement with the available experimental results and with the Sieder and Tate equation for plain tube. As the heat transfer rates for oil under laminar flow are very low, inserts are used to enhance the heat transfer rates on tube side. CFD investigations on the Nusselt number and friction factor of tubes equipped with louvered elliptical forward and backward strip inserts are carried out. As expected, the predicted results showed that the use of strip inserts led to higher heat transfer rates over the plain tube. The increase in Nusselt number and overall enhancement ratio are higher for backward insert and friction factor is higher for forward insert.

Key words: Heat transfer, enhancement, laminar, strip inserts, oil cooler, CFD

by Promvonge [2] to investigate the air flow

1. Introduction

   Heat transfer processes of viscous fluids usually take place in laminar or transitional regimes, where transfer rates are particularly low. Heat exchangers that work under these flow conditions are required to undergo enhancement techniques. Among the different techniques, which are effective to improve the thermo hydraulic behavior in the tube-side in single-phase laminar flow, the insert devices are important. The main advantage of inserts with respect to other enhancement techniques is that they allow an easy installation in an existing smooth- tube heat exchanger.

   A comprehensive experimental study was carried out on six wire coils inserted in a smooth tube, covering the laminar, transition and turbulent regimes by Garcia et al [1]. Experimental investigations were carried out

   friction and heat transfer characteristics in a round tube fitted with coiled square wire turbulators for the turbulent regime. Smith Eiamsa-ard and Pongjet Promvonge [3] experimentally investigated heat transfer and flow friction characteristics of insertion of a helical screw tape with or with out core rod in a concentric double tube heat exchanger.

   Wen et al. [4] carried out experimental study on the heat transfer enhancement and pressure drop in small tubes. Sreenivasulu and Prasad [5] presented a computational study on convective heat transfer in an annulus with its inner cylinder wrapped by a helical wire. Pavel et al [6] experimentally investigated the effect of metallic porous inserts in a pipe subjected to constant and uniform heat flux at a Reynolds number range of 1000-4500. Mehmet Sozen and T M Kuzay [7] numerically studied the

   enhanced heat transfer in round tubes filled with rolled copper mesh at Reynolds number range of 5000-19,000 with water as the energy transport fluid. Heat transfer, friction factor and enhancement efficiency characteristics of a circular tube fitted with conical ring turbulators and a twisted-tape swirl generator were investigated experimentally by Promvonge and Eiamsa- ard [8]. The effect of two tube inserts wire coil and wire mesh on the heat transfer enhancement, pressure drop and mineral salts fouling mitigation in tube of a heat exchanger were investigated experimentally by H. Pahlavanzadeh et al [9] with water as working fluid.

   Heat transfer and friction characteristics were investigated experimentally by using louvered strip inserts inserted in the inner tube of a concentric tube heat exchanger by Smith et al. [10]. The flow rate of the tube was in the Reynolds number range of 6000 to 42000.

   In the above literature review, most studies were focused on the turbulent flow heat transfer enhancement with either air or water as working fluids while numerical simulations are scarce. Therefore, the present study focuses on laminar flow heat transfer enhancement using louvered strip inserts with ISOVG46 turbinol (with high prandtl number of 220 at 700 C) as the working medium by numerical modeling using computational fluid dynamics.

2. Details of the experimental set up and procedure

   The experimental oil cooler test rig consists of oil circulation system, water circulation system, shell and tube heat exchanger, cooling tower and instrumentation system to measure oil and water flow rates, pressures and temperatures. Oil circulation system consists of an oil tank, strainer, canned type centrifugal oil pump, flow regulation valves and heating system. Heating system consists of an oil tank with four coiled ceramic

   insulated electric heaters. Similarly water circulation system consists of water sump, cooling tower, pump and flow regulation valves. Water pump with a maximum capacity of 10m3/hr was used for water pumping through the shell of the heat exchanger. The warm water from the outlet of the heat exchanger was cooled in an induced draught cooling tower and sent back to sump. Shell and tube heat exchanger consists of four tubes enclosed inside the shell. Hot oil is made to flow inside the tubes with water flowing in counter flow direction inside the shell. The outer surface of the shell is wrapped with insulation to minimize heat losses to surroundings.

   Fig 1: Schematic diagram of oil cooler test rig

   Schematic diagram of oil cooler test rig is shown in figure 1.

   Temperature Measurement was done by Resistance Temperature Detectors. RTDs were used for the measurement of inlet and outlet temperatures of water and oil. Along the length of tubes, exclusively made Copper-Constantan thermocouples (T-type) were used for the measurement of tube wall temperatures and bulk oil temperature inside

   the tank. Pressure transmitter was used to measure the pressure at a single point. Differential pressure transmitter was used to measure the pressure drop of oil across the tubes. In addition, two pressure transmitters were utilized for measurement of inlet pressures of oil and water. An oil flow rotameter and orifice plate was used to measure the oil flow rate. A water flow rotameter was used for measuring water flow rate.

   2.1 Procedure

   Oil is heated by switching on electric heaters. During heating of oil, the bye-pass valve is fully opened and the other valves are fully closed so that the oil temperature in the tank rises in short time. When the tank temperature reached the desired temperature of 700C, the inlet valve to the heat exchanger was gradually opened till the required flow rate was reached in the oil flow rotameter. Water pump was put on and the water flow rate was adjusted slowly to a value at which the water temperature rise is around 2 to 30C. Once the steady state was reached, data recording was initiated. This data consisted of wall temperatures, oil inlet and outlet temperatures, water inlet and outlet temperatures and pressure measurements. Experiments were conducted (plain tube experiment) with out inserts for different oil flow rates. All the experiments were carried out for laminar flow conditions with Reynolds number varying from 250 to 2450.

3. Data deduction

   Heat loads were evaluated by

   Qh = mh cph (Th.in ThOut) (1)

   Qc = mc cpc (Tcout Tc.in ) (2)

   V=mh/(h Ac) (3)

   Re= V Di/ (4)

   Nusselt numbers calculated from the experimental data for plain tube (tube side) were compared with the correlation recommended by Sieder and Tate

   Nu0=1.86(RePr(Di/L))1/3( / w) 0.14 (5)

   Properties of oil were considered at the local mean bulk temperature.

   Equation (5) gives theoretical Nusselt number.

   f the = 64/Re (6)

   f0 = P/ ((L/Di) ( V2/2)) (7)

4. Numerical Experiment

   The geometry used for numerical modeling is same as the experimental set up. The heat exchanger geometry is created and meshed using the software package ICEMCFD and analyzed using ANSYS CFX 11.0 for oil to water heat transfer through the metal wall. Tube wall thickness is considered while modeling the heat exchanger. Heat exchange from hot fluid inside the tubes to cold fluid in shell was modeled with convective heat transfer in the tube, conduction through the tube wall and convective heat transfer to the shell fluid. Structured grids were used to mesh the tube fluid volume and tube solid volume. For momentum equation, the inner and outer wall of the tubes was treated as no-slip ones. The inner and outer walls of shell were considered as no-slip with adiabatic condition. Hot fluid inlet is taken with mass flow inlet boundary condition. Hot fluid outlet was specified as atmospheric pressure. Cold fluid enters the heat exchanger in the opposite direction (counter flow arrangement) with mass flow inlet and pressure outlet condition. It has been found that the heat transfer coefficients predicted by the Sieder and Tate equation for plain tube are comparable with those calculated by the computer code CFX, with a maximum deviation of 9%.

   The meshed model of oil cooler (shell and tube heat exchanger) is shown in figure 2. Structured grids were used to mesh the tube fluid volume and tube solid volume. CFD analysis was carried out varying the inlet flow rate of oil. Before carrying out the actual analysis, a grid independency of the

   solution was established. The optimum mesh was chosen for further analysis and has 73600 nodes and 129523 elements for plain tube. The sensitivity of the results on the value of the imposed boundary condition, constant wall heat flux was studied. On changing the wall heat flux from 2000 to 6000 W/m2, changes in Nusselt number was marginal. The change is attributable to differences in the fluid temperature distribution along the tubes.

   1. CFD Modelling

      Fig 2: Meshed model of oil cooler

      Fig 3: plain tube velocity streamlines at inlet

      The mass flow rate of the cold water was kept constant as 0.3 kg /sec. Mass flow rate of hot fluid varied from 0.16 to 1.56Kg/sec, which is same as the one used in the experiment. A convergence criterion of

      1.0e-05 was used for continuity, momentum and energy equations. Inlet temperature of oil is kept constant at 347.5K. Streamlines plotted with velocity as variable are shown in figure 3. Velocity of oil at inlet varied from 0.12m/sec to 1.21m/sec with respect to variation in mass flow rates of oil for plain tube experiment.

      Nu0

      The results of the analysis of the CFD simulation are used to estimate overall heat transfer coefficient. It is found that the deviation between CFD and experimental results is within 5%.

      Nu(T he) Nu(E xp) Nu (C F D )

      30

      25

      20

      15

      10

      5

      0

      0

      1500

      Re

      2000 2500

      500 1000

      Fig 4: Verification of Nusselt number for plain tube

      Nusselt number obtained by CFD simulation is in close agreement with that of experimental and theoretical (Sieder and Tate equation) Nusselt number as shown in Figure 4.Figure 5 shows the validation of friction factor obtained from simulation. The maximum deviation in simulated Nusselt number and friction factor was observed to

      be + 4.9% and +1.25% respectively compared to the theoretical values.

      f0

      Enhancement of heat transfer on oil side can be obtained by using tube inserts. The inserts used in the present study are elliptical forward and backward louvered strip inserts. The strips are mounted on a core rod as shown in figure 6.These strip inserts are inserted in the four tubes of heat exchanger. The louvered strip insert material considered for CFD analysis is stainless steel with an inclined angle of 300 with axis of tube. Distance between two consecutive elliptical leaves (pitch) is taken as 90mm. Figure 6 shows elliptical forward and backward louvered strip inserts. The optimum mesh for oil cooler with inserts has 769939 nodes and 1433366 elements. Each of the run took about 2hrs with Intel Quad core processor, 2.83GHz with 6 GB RAM. Meshed model in the presence of strip inserts is shown in figure 7.

      f(T he) f (E xp) f (C F D )

      0.4

      0.35

      0.3

      0.25

      0.2

      0.15

      0.1

      0.05

      0

      0 1000 2000 3000

      Re

      Fig 5: Verification of friction factor for plain tube

      Fig 6: Louvered elliptical forward and backward strip inserts.

      Outlet temperature contours of louvered elliptical forward and backward inserts are shown in figure 9. Due to the presence of strip inserts, tube hydraulic diameter decreases. As the mass flow rate of oil is kept constant for analysis with/without strip inserts, velocity of oil increases inside the tubes due to the presence of tube inserts. This promotes oil side turbulence leading to enhancement of heat transfer.

      Fig 7: Cross-sectional view of elliptical louvered strip inserts -meshed model

      Fig 8: Outlet temperature contours of elliptical forward louvered strip inserts

      Fig 9: Outlet temperature contours of elliptical backward louvered strip inserts

      Inlet temperature of oil is kept constant at 347.5K. By comparing the oil outlet temperatures, as shown in figures 8 and 9, they are almost same for both forward and backward elliptical strip inserts. But pressure drop contours (figures 10 and 11) showed that pressure drop is less for

      backward louvered strips. Maximum pressure drop is found at lowest mass flow rate of oil. More pressure drop is observed for elliptical forward strip (558 Pa) compared to backward strip (407 Pa) as shown in figures 10 and 11 respectively.

      Fig 10: Contours of pressure drop for louvered elliptical forward strip inserts

      Fig 11: Contours of pressure drop for Louvered elliptical backward strip inserts

      Fig 12: Variation of Nusselt number for elliptical forward and backward leaves.

      Fig 13: Variation of friction factor for forward and backward elliptical inserts.

      Variation of Nusselt number with Reynolds number is shown in figure 12. It is seen that the Nusselt number increased with rise of Reynolds number for both inserts. This indicates an advantageous gain of using tube inserts over plain tube. The maximum increase in Nusselt number for elliptical forward and backward strips are 431% and 432% respectively compared to plain tube.

      Figure 13 shows the variation of friction factor with Reynolds number. It indicates the friction factor for backward elliptical inserts is slightly less than that of

      forward inserts. The friction factor is observed to decrease with increase of Reynolds numbers. The maximum increase in friction factor for elliptical forward and backward strip inserts are 49.22% and 43.72% respectively compared to plain tube.

      Overall enhancement ratio is used to evaluate the quality of the enhancement technique used.

      It is given by

      = (Nu/Nu0) (8) (f/f0)1/3

      Fig 14: Variation of overall enhancement ratio.

      Overall enhancement ratio is found to vary from 2.33 to 5.19 for elliptical backward louvered strip inserts which is higher than that of louvered forward strip inserts as shown in figure 14. This is due to lesser pressure drop associated with backward elliptical strip insert compared to forward elliptical strip insert.

5. Concluions

Computational Fluid Dynamic simulation of oil cooler using ANSYS CFX

1. was carried out with Reynolds number on tube side varying from 250 to 2400 for analyzing the flow behavior & enhancement of heat transfer using elliptical forward and backward louvered strip inserts. The plain tube CFD results like heat transfer coefficient, Nusselt number, friction factor were compared with theoretical and experimental results for validation. A good agreement was found between theoretical correlations and CFD simulated results.

   • Both the elliptical forward and backward louvered strip inserts found to provide a considerable enhancement in heat transfer with respect to plain tube.

   • The maximum %increase in Nusselt number of elliptical forward louvered strip with respect to plain tube is 431.37% and the overall enhancement ratio is found to vary from 2.02 to 4.65.

   • The maximum %increase in Nusselt number of elliptical backward louvered strip with respect to plain tube is 432.68% with the overall enhancement ratio varying from 2.33 to 5.19.

   • As the overall enhancement ratio of elliptical backward louvered strip insert is higher than that of forward insert, its use is recommended in oil coolers to enhance the heat transfer.

Nomenclature

Ac Cross sectional area for tube side fluid, m2

Cph Specific heat at constant pressure (oil)

Cpc Specific heat at constant pressure (water)

Di Inside diameter of tube, m D0 Outside diameter of tube, m

fthe Friction factor(theoretical),tube side without insert

f Friction factor, tube side with insert f0 Friction factor, tube side without

insert

1. Thermal conductivity of fluid , W/mK

2. Length of each tube, m

mh mass flow rate of hot fluid(Kg/sec) mc mass flow rate of cold fluid(Kg/sec)

Nu0 Nusselt number, tube side with out insert

Nu Nusselt number, tube side with insert

P Wetted perimeter, 4* D0, m

P Pressure drop, m Pr Prandtl number

Qh Heat duty for hot fluid (Oil), W

Qc Heat duty for cold fluid (Water), W Re Reynolds number

Thin Hot fluid inlet temperature, 0C Thout Hot fluid outlet temperature, 0C Tcin Cold fluid inlet temperature,0C Tcout Cold fluid outlet temperature,0C

V Mean velocity in tube, m/s

Density, Kg/m3

h Density of hot fluid, Kg/m3

Dynamic viscosity of oil at bulk temperature, kg/m s

w Dynamic viscosity of oil at wall temperature, kg/m s

Overall enhancement ratio

References

[1]Garcia, Vicente, Viedma Experimental study of heat transfer enhancement with wire coil inserts in laminar-transition- turbulent Regimes at different prandtl numbers,International Journal of Heat and Mass Transfer 48 (2005) 46404651

[2]P. Promvonge, Thermal performance in circular tube fitted with coiled square wires, Energy Conversion and Management 49 (2008) 980987.

1. Smith Eiamsa-ard, Pongjet Promvonge, 2007, Heat transfer Characteristics in a tube fitted with helical screw- tape with/without core-rod inserts, International Communications in Heat and Mass transfer 34, pp. (176-185).

2. Wen et al. Augmented heat transfer and pressure drop of strip- type inserts in the small tubes, Heat and Mass Transfer 40 (2003) 133141 DOI 10.1007/s00231-

002-0393-9 Springer.

[5]T. Sreenivasulu, B.V.S.S.S. Prasad Flow and heat transfer characteristics in an annulus wrapped with a helical wire, International Journal of Thermal Sciences, 48 (2009) 13771391.

1. Bogdan I. Pavel and Abdulmajeed A. Mohamad, Experimental Investigation of the Potential of Metallic Porous Inserts in Enhancing Forced Convective Heat Transfer, ASME J. Heat Transfer, August 2004, Vol. 126, pp. (540-545).

2. Mehmet Sozen and T M Kuzay, Enhanced heat transfer in round tubes with porous inserts, Int. J. Heat and Fluid Flow, April 1996, Vol. 17, pp. (124-129)

3. Promvonge and Eiamsa-ard, 2007. Heat transfer behaviors in a tube with combined

   conical-ring and twisted-tape insert, International Communications in Heat and Mass transfer 34, May, pp. 849-859.

4. Pahlavanzadeh, Jafari Nasr and Mozaffari, 2007. Experimental study of thermo-hydraulic and fouling performance of enhanced heat exchangers, International Communications in Heat and Mass transfer 34, May, pp. 907-916.

5. Smith Eiamsa-ard, somsak Pethkool, Chinaruk Thianpong ,Pongjet promvonge, Turbulent flow heat transfer and pressure loss in a double pipe heat exchanger with louvered strip inserts, International Communications in Heat and Mass Transfer 35 (2008) 120-129.

6. Krishna Reddy.K.,Experimental and Analytical studies on internally augmented oil cooler, M.Tech thesis, G.Pulla Reddy Engineering College,Kurnool,2005.