Enhancement of heat transfer in a horizontal tube with wire coil inserts--a CFD analysis.

Sarada, S. Naga ; Radha, K. Kalyani ; Raju, A.V.S. 等

Introduction

Laminar flow heat transfer occurs in a variety of engineering applications and is of particular importance where viscous liquids are heated and cooled. Since the heat transfer coefficients in this type of flow are generally low, there is a need for augmentation. However, as most of the flow problems in industrial exchangers involve turbulent flow region, attention has been directed mainly towards turbulent flow heat transfer augmentation.

Wire coils are a type of inserted elements which present some advantages compared to other enhancement techniques, such as artificial roughness by mechanical deformation. They may be installed in an existing smooth tube heat exchanger. They keep the mechanical strength of the smooth tube. Their installation is easy and their cost is very low. The insertion of a device such as wire coil inside a smooth tube produces an increase in the heat transfer due to one or more of the following phenomena.

Turbulence promotion: Wires attached to the wall cause separation in the flow that increases its turbulence level.

Secondary flow promotion: Many inserted devices induce secondary flows which can favour thermal exchange. Helical wire coils produce a helicoidal flow at the periphery super imposed on the main axial flow. Due to the flow velocity increase and to the appearance of centrifugal forces, convection increases. This favours the convection in heating processes.

Hydraulic diameter reduction: Any inserted element in a smooth tube will reduce the cross-sectional area increasing the average flow velocity. The wetted perimeter also increases and the hydraulic diameter decreases.

Garcia and Solano [1] performed flow pattern assessment in tubes with wire coil inserts in laminar and transition regimes. They found that at low Reynold's numbers (Re < 500) the flow in tubes with wires is essentially similar to the flow in a smooth tube. Transition to turbulence is produced earlier in short pitch wire coils. The study carried out using three different wire coils describes the different flow patterns and their influence on the transition from laminar to turbulent flow. Hong and Bergles [2] studied the performance of twisted tape inserted tubes for laminar flow heat transfer and found that the nusselt number depends on the Re, Prandtl number, and tape twist ratio. They observed as much as threefold improvement in heat transfer rates using twisted tape inserted tubes.

Uttarwar and RajaRao [3] conducted experiments using seven wire coil inserted tubes of varying diameter and pitch of wire coil with Servotherm oil as working fluid for augmentation of laminar flow heat transfer in tubes. They found that as much as fourfold improvement in heat transfer coefficient can be obtained using these tubes. Lieke Wang and Sunden [4] discussed the selection of different tube inserts and made comprehensive comparison on the thermal and hydraulic performance for twisted tape inserts and wire coil inserts. The comparison was conducted in both laminar and turbulent regions. They found that in the turbulent region heat transfer enhancement ratio and overall enhancement ratio can be up to 3.5 and 2.0 respectively. Wire coil insert gives better overall performance if the pressure drop penalty is considered.

Sethu Madhavan and RajaRao [5] conducted experimental investigations for a copper tube fitted with helical wire coil inserts of varying pitch, helix angle and wire diameter. They found that preferred helix angle of the wire coil insert is in the vicinity of 50-[55.sup.0] for convective heat transfer to water.

S.V.Mokamati and R.C.Prasad [6] performed numerical simulation of heat transfer from tubes employing augmentation devices for heat transfer enhancement. A CFD simulation tool was developed with CFX software and the results obtained from the simulations are validated with the empirical correlations for a smooth tube heat exchanger. In this study, k-[omega] model better predicted the flow field and heat transfer field. The difficulties in simulating the augmented heat exchanger are discussed. The felt that for accurate results, the mesh need to be refined by six times which could not be handled by their available computer resources.

Smith Eiamsa-ard and Yuttana Ploychay [7] conducted experiments on a concentric tube heat exchanger. Hot air passed through inner tube while the cold water was flown through the annulus. A maximum percentage gain of 165% in heat transfer rate was obtained by using the helical insert in comparison with the plain tube.

Inserts in flow passage increase heat transfer rate at cost of increase in pressure drop thereby demanding for more pumping power. Hence it is necessary to design the device with an optimization between the enhanced heat transfer rate and large pressure drop.

Experiments with inserts are time consuming and some times difficult to conduct in a wide range of flow conditions. Numerical simulation provides an alternate method, which can be used to validate experimentally obtained data as well as generate new data for a variety of enhancement devices.

The objective of the present work is to determine enhancement of heat transfer while using tubes inserted with wire coils. The geometry and dimensions of the inserts used in this study is shown by figure 2 to 4. Copper is used as material of the pipe and Aluminium used as insert material.

Nomenclature

Re            Reynolds number

St            Stanton number with insert

q             Wall heat flux (W/[m.sup.2])

e1, e2        wire diameter (2mm, 3.4mm)

h             heat transfer coefficient (W/[m.sup.2]K)

p1, p2, p3    pitch of the wire insert coil (66mm, 38mm, 22mm)

[T.sub.o]     outlet temperature (K) of air

[T.sub.w]     Temperature on the outer wall of the pipe (K)

[T.sub.b]     Bulk temperature, K

[D.sub.e]     Equivalent diameter (m)

[St.sub.0]    Stanton number for bare tube

Experimental Setup and Procedure

Experiments are conducted with air to find the forced convection heat transfer coefficient for bare tube. The setup is shown in figure below. The apparatus consists of a blower unit fitted with a pipe in horizontal orientation. The length and the inside diameter of the pipe are 610mm and 27mm respectively. Nichrome bend heater encloses the test section to a length of 400 mm. Four thermocouples are embedded on the wall of the pipe and two thermocouples are placed in the air stream one at the entrance and the other at the exit of the test section to measure the temperature of flowing air. The test pipe is connected with an orifice to measure the flow of air through the pipe. Input to heater is given through dimmerstat. Experiments were carried out at constant heat flux condition. Line diagram of the experimental setup is shown in figure 1.

[FIGURE 1 OMITTED]

Numerical Simulation

A model pipe with 27mm inside diameter and a length of 610mm is considered for analysis. Three-dimensional numerical simulations of fluid flow and heat transfer for this pipe were performed over a range of Reynolds number. The model was developed in gambit 2.2.30 with fine mesh and exported into fluent 6.2.16. The commercial code fluent 6.2.16 has been used for the numerical solution of Navier-Stokes equations. The numerical method was based on a finite volume code, based on a set of governing equations and boundary conditions. The working fluid considered is air.

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

[FIGURE 4 OMITTED]

The types of models created with air as working fluid: 1. bare tube (without insert).

2. Coil wire insert with pitch (p1) = 66mm and wire diameter e1=2.0mm.

3. Coil wire insert with pitch (p2) = 38mm and wire diameter e1=2.0mm

4. Coil wire insert with pitch (p3) = 22mm and wire diameter e1=2.0mm

5. Coil wire insert with pitch (p1) = 66mm and wire diameter e1=3.4mm

6. Coil wire insert with pitch (p2) = 38mm and wire diameter e1=3.4mm

7. Coil wire insert with pitch (p3) = 22mm and wire diameter e1=3.4mm

Boundary Conditions

A free stream flow with uniform temperature and uniform velocity was applied to the inlet plane and zero relative static pressure was applied to the exit plane. Wall heat flux given is 1194 W/[m.sup.2] .

Turbulence Model Selection: K-[epsilon] model is used for the analysis. An unstructured mesh wad applied to the computational domain with a refined mesh density near the wall boundary. Models are meshed using GAMBIT.

The steps followed to create the model in Gambit

1. Create the 3-d model of cylinder in gambit for validation.

2. Create the inserts of coil wire from pro-e and import to gambit

3. Mesh the model with suitable mesh size.

4. Specify the boundary conditions.

5. Export the meshed model to fluent.

The steps followed to solve the problem in Fluent

1. Set up the problem in fluent and import grid.

2. Define solver properties (here defaults of segregated solver, implicit formulation, steady flow, absolute velocity formulation are selected.)

3. Define models (energy).

4. Define models viscous (k-[epsilon] turbulence model is selected).

5. Define boundary conditions (inlet velocity, temperature, heat flux, etc.)

6. Solve the controls for solution.

7. Provide an initial solution (initial values of the velocity field at inlet).

8. Set convergence criteria (it is a measure of how well the current solution satisfies the discrete form of each governing equation and it iterate until the residual for each equation falls below the specified value).

9. Iterate until convergence.

10. Save the data.

11. Analyze the results (plot xy).

Convergence Control

The computational domain was solved as a steady state conjugate heat transfer problem and the solution process was performed until the convergence and an accurate balance of mass and energy were achieved.

[FIGURE 5 OMITTED]

[FIGURE 6 OMITTED]

[FIGURE 7 OMITTED]

Results and Discussion

CFD Tool Validation with Bare Tube

The validity of the present model is established by comparing its results with available correlations for the heat transfer coefficient and the friction factor. Figure 8 shows the comparison of heat transfer coefficient obtained experimentally and by using CFD analysis for bare tube. The values are tabulated in table 2.

Temperature Distribution on Tube

Figure 9 to 11 show the Temperature Distribution on tube.

[FIGURE 8 OMITTED]

[FIGURE 9 OMITTED]

[FIGURE 10 OMITTED]

[FIGURE 11 OMITTED]

Figures 2 to 4 show the Gambit generated models of tube inserted with coils of pitch 66mm, 38mm and 22 mm respectively of wire diameter 2 mm.

Figures 5 to 7 show the meshed models of tube inserted with coils of pitch 66mm, 38mm and 22 mm respectively of wire diameter 2 mm.

Figure 8 shows the comparison between heat transfer coefficients obtained experimentally, analytically and by using Dittus Boelter equation for bare tube, it is observed that h value computational is 10.7% more than h experimental at Re =12395.

[FIGURE 12 OMITTED]

[FIGURE 13 OMITTED]

Figure 9 to 11 show the Temperature Distribution on tube

Figure 12 shows the computational heat transfer coefficient when coil wire insert of pitch 66mm and wire diameter 2 mm is used. It is observed that at Re =12395, h is increased by 93.77% when compared to bare tube.

[FIGURE 14 OMITTED]

[FIGURE 15 OMITTED]

Figure 13 shows the computational heat transfer coefficient when coil wire insert of pitch 38mm and wire diameter 2 mm is used. It is observed that at Re =12395, h is increased by 138.66% when compared to bare tube.

Figure 14 shows the computational heat transfer coefficient when coil wire insert of pitch 22mm and wire diameter 2 mm is used. It is observed that at Re =12395, h is increased by 190.6% when compared to bare tube.

Figure 15 shows the computational heat transfer coefficient when coil wire insert of pitch 66mm and wire diameter 3.4 mm is used. It is observed that at Re =12395, h is increased by 122.68% when compared to bare tube.

[FIGURE 16 OMITTED]

[FIGURE 17 OMITTED]

Figure 16 shows the computational heat transfer coefficient when coil wire insert of pitch 38mm and wire diameter 3.4 mm is used. It is observed that at Re =12395, h is increased by 215.2% when compared to bare tube.

Figure 17 shows the computational heat transfer coefficient when coil wire insert of pitch 22mm and wire diameter 3.4mm is used. It is observed that at Re =12395, h is increased by 311.08% when compared to bare tube.

[FIGURE 18 OMITTED]

Figure 18 shows the comparison of h value obtained experimentally, analytically for bare tube, by inserting wire coils of pitch 66mm, 38 mm and 22 mm for wire diameters of 2 and 3.4mm respectively.

Table 2 shows the percentage improvement in Stanton number for inserted tubes compared to bare tube which is in agreement with the experimental studies carried out by Sethu Madhavan and Raja Rao on turbulent flow heat transfer with helical wire coil inserted tubes [5].

The heat transfer enhancement ratios obtained from the present analysis for 2mm and 3.4mm wire diameters with 22mm pitch are 2.373 and 2.646 respectively which is also in agreement with the analysis made by Wang and Sunden [4].

Conclusions

Six kinds of inserts in a horizontal circular tube in the range of 10353<Re<12395 are considered. Comparison of enhancement of heat transfer coefficients obtained is made with respect to bare tube subjected to constant heat flux.

1. Heat transfer coefficient is increased by 190.6% when a coil wire insert of pitch 22mm, wire diameter 2.0mm is used at Re=12395 when compared to the experimental value of bare tube.

Heat transfer coefficient is increased by 162.4% when a coil wire insert of pitch 22mm, wire diameter 2.0mm is used at Re=12395 when compared to the computational value of bare tube.

2. Heat transfer coefficient is increased by 311.8% when a coil wire insert of pitch 22mm, wire diameter 3.4mm is used at Re=12395 when compared to the experimental value of bare tube.

Heat transfer coefficient is increased by 271% when a coil wire insert of pitch 22mm, wire diameter 3.4mm is used at Re=12395 when compared to the computational value of bare tube.

References

[1] A. Garcia, J.P. Solano, P.G. Vicente and A. Viedma, "Flow pattern assessment in tubes with wire coil inserts in laminar and transition regimes", International Journal of Heat and Fluid Flow (2006), doi:10.1016/j.ijheatfluidflow.2006.07.001.

[2] Hong, S.W., and Bergles, A.E., " Augmentation of laminar flow heat transfer by means of twisted tape inserts," ASME Journal of Heat Transfer, vol.98.1978.

[3] Uttarwar, M. Raja Rao, "Augmentation of laminar flow heat transfer in tubes by means of Wire Coil Inserts", ASME Journal of Heat transfer, vol. 107, November 1985.

[4] Lieke Wang and Bengt Sunden "Performance comparison of some tube inserts" International Conference of Heat Mass Transfer, Vol 29, No.1, pp 45-56, 2002

[5] SethuMadhavan and Raja Rao "Turbulent flow Vol.26, No12, pp1833-1845, 1983 heat transfer and fluid friction in helical wire coil inserted tubes" Int. J. Heat Mass Transfer.

[6] S.V. Mokamati and R.C. Prasad "Numerical simulation of fluid flow and heat transfer in a concentric tube heat exchanger".

[7] Smith Eiamsa-ard and Yuttana Ploychay "An Experimental study of heat transfer and friction factor characteristics in a circular tube fitted with a helical tape", .Proceedings of the 18th Conference of Mechanical Engineering network of Thailand, 18-20 October 2004, Khon Kaen.

[8] S. Bindu Madhavi, "CFD analysis for enhancement of heat transfer in horizontal tubes with helical wire inserts for tube flow in turbulent region", M. Tech thesis, JNTU College of Engineering, Hyderabad, 2007.

S. Naga Sarada *, K. Kalyani Radha * and A.V.S. Raju **

* Department of Mechanical Engineering

JNTU College of Engineering, Kukatpally, Hyderabad 500085, Andhra Pradesh, India.

E-mail: nagasaradaso@gmail.com, kalyaniradha@gmail.com

** Department of Mechanical Engineering

JNTU College of Engineering, Kakinada, Andhra Pradesh, India.

Table 1: shows the experimental parameters & operating
ranges used in of 22mm. heat transfer analysis.

Conditions                           Heat Transfer

Reynolds number(Re)               10353,11833, 12395
Net heat flow rate for Air (W)            64W
Surface Area ([m.sup.2])                 0.0536
Wall heat flux (W/[m.sup.2])            1194.02

Table 2: Improvement in Stanton Number

S.No.  GEOMETRY (mm)   St/[St.sub.0]   %improvement in St.

1      P=66,e=2            1.605              60.5%
2      P=38,e=2           1.8873               89%
3      P=22,e=2            2.137             113.8%
4      P=66 e=3.4          1.671              67.1%
5      P=38,e=3.4         2.1771             117.62%
6      P=22,e=3.4         2.4123             141.28%