1. Introduction
Heat exchangers, whether large or small, play an important role not only in industry but also in everyday life. A heat exchanger is any device that can transfer heat from one medium to another. To achieve a high heat-transfer rate over a smaller area, the impinging technique is quite interesting. The fluid will be impinged to attack the wall. This can be used to increase or decrease the temperature of a device, such as by transferring the heat out from electronic hardware. The heat produced from a CPU can be passed along the fins and out into the surrounding environment. The development is interesting in terms of the impinging jet characteristics and heat exchanger surfaces. In this research, our study was focused on the improvement of the heat exchanger surface. This has been developed from a flat plate to a dimpled surface for generating a vortex and high turbulence intensity on the surface. This can increase the heat transfer and reduce the pressure drop. In the conventional technique, ribs or baffles are set on the heat exchanger plate for increasing the heat transfer. However, this can also increase the pressure drop and the friction factor, leading to the production of a low thermal enhancement factor (TEF). The TEF indicates the heat-transfer performance as calculated from the Nusselt number (Nu) and friction factor (f). As such, the TEF represents the ratio of heat transfer and the power ability of the fluid to flow through the heat exchanger device. A higher TEF cannot be obtained at a higher heat-transfer rate with a high pressure drop.
Investigations have been conducted widely into the effect of those parameters of an impinging air jet on the heat transfer in many roughened-surface geometries. Recently, Ries et al. [
1] studied the heat transfer of a turbulent jet impinging on a 45° incline by using a direct numerical simulation (DNS) method. The results showed that the highest heat-transfer position was not the fluid impinging point (the stagnation point) but was in fact a little shifted to above the impinging point. Ries et al. [
2] investigated the flow and heat characteristics by experiment of an air jet impinging on a flat plate inclined at different angles. The highest Nu was found at the stagnation point for the flat plat inclined at 90°. Meola et al. [
3] investigated convective heat transfer coefficients on a flat plate. They were focused on the effects of shear layer dynamics. An infrared scanning radiometer and the heated thin foil technique were used to measure the temperature and determine the Nusselt number. Guerra et al. [
4] found that the local velocity and the temperature distributions were presented as well as longitudinal turbulence profiles on the impinged plate. Chaudhari et al. [
5] measured temperature and pressure on the impinged plate to find the heat transfer and pressure drop on this plate along in the radial line. The experiments were focused on the following parameters: Re in the range of 1500–4200. The ratio of the axial distance between the heated plate and the jet to the jet diameter was in the range of 0–25 in this study. The results of the maximum heat transfer coefficient as a Nusselt number (
) were 11 times higher with the synthetic jet than with natural convection. Draksler and Končar [
6] numerically analyzed heat transfer rates using the turbulence models for predictions of heat transfer and the flow characteristics of an axis-symmetric impinging jet. The results were validated based on the free jet impingement experiment. Nanan et al. [
7] studied the flow and heat-transfer characteristics of swirling impinging jets on an impinged plate. The results showed that the swirling impinging jets offered higher heat-transfer rates on impinged surfaces than the conventional impinging jets. Nuntadusit et al. [
8] investigated the effect of using multiple swirling impinging jets (MSIJs) and found that the Nu from using MSIJs was higher than from using multiple conventional impinging jets. Qiang et al. [
9] studied the low viscosity fluid (non-Newtonian fluid) impinging to the flat plate. The results indicate that non-Newtonian fluid is the optimum choice to obtain high heat transfer rate for laminar flow.
Tong [
10] studied the solving of Navier-Stokes equations by using a finite-volume for investigating the hydrodynamics and heat transfer of the impingement process. The results showed that a high maximum Nusselt number also provided the maximum pressure drop on the surface. M. Goodro et al. [
11] showed the effects of array jets impinging on the flat plate in Reynolds numbers ranging from 8200 to 30,500 and Mach numbers from 0.1 to 0.6. The heat transfer also significantly increased as the Reynolds number and Mach number increased. Pakhomov and Terekhov [
12] presented the results of numerical investigation into the flow structure and heat transfer of impact mist jets with a low concentration of droplets.
Ekkad and Kontrovitz [
13] focused on heat transfer distributions over a jet impingement target surface with dimples. They used the transient liquid crystal technique to measure the heat transfer. Results showed that the presence of dimples on the target surface, in-line or staggered with respect to jet location, produced lower heat transfer coefficients than the non-dimpled target surface. Lienhart et al. [
14] carried out an experimental and numerical investigation into how the turbulent flow over dimpled surfaces has an effect on the friction drag. The results showed that the friction factors of dimpled surfaces were higher than those of flat plates. Kanokjaruvijit and Martinez-Botas [
15] studied heat transfer in an inline array of round jets impinging on a staggered array of hemispherical dimples. They considered various parametric effects, such as Reynolds number, jet-to-plate distance, dimple depth and ratio of jet diameter to dimple-projected diameter. A transient wide-band liquid crystal method was used to measure the heat transfer rate. At a dimple depth of 0.15 and jet-to-plate distance twice that of the dimple diameter, the results showed that the heat transfer under these conditions was a significant 70% higher than with a flat plate. Xing and Weigand [
16] studied a jet array impinging on flat and dimpled plates by using the transient liquid crystal method. The best heat transfer performance was obtained with the minimum cross flow and narrow jet-to-plate spacing, whether on a flat or dimpled plate. The dimples on the target plate resulted in higher heat transfer coefficients than with the flat plate. Won et al. [
17] combined PIV experiments and modeling to obtain two-and three-dimensional (3-D) microjet flows. From the 3-D velocity fields, the researchers could quantify the flow physics around the impingement between the orifice and the bottom surface. Kwon et al. [
18] used a naphthalene sublimation technique for obtaining local and average Nusselt numbers on a dimple. The average Nusselt number increased as the turbulence intensity in the mixing layer over the dimple increased, which affected the dimple depth and the increase in Reynolds number. Turnow et al. [
19] studied the use of Large Eddy Simulation (LES) and Laser Doppler Velocimetry (LDV) to find vortex structures and heat transfer enhancement mechanisms of turbulence flow over a staggered array of dimples. LES was used to calculate the flow and temperature fields. The vortex structure of dimple packages was shown from LDV and LES. The heat transfer rate from a dimpled plate could be about 200% higher than that of a flat plate.
De Bonis and Ruocco [
20] applied the impingement technique for food drying or dehydration. The heat transfer rate, water activity and moisture depletion were measured in a food substrate using a turbulent air jet impingement for food heating. Parida et al. [
21] studied the jet impingement for cooling high heat flux (i.e., cutting-edge electronic technologies). The results showed that developments in cooling methodologies were required to avoid unacceptable temperature rise and to maintain a high performance. The overall improvement in quality of 150–200%, based on the maximum Nu recorded both experimentally and numerically, was affected by impingement and associated swirl. Na-pompet and Boonsupthip [
22] were interested in cooling rate impingements of food. They demonstrated the comparison between experimental data from related publications and numerical simulations. The results showed that a proper design of the plate position significantly enhanced the overall energy transfer on the target surface and improved the cooling rate of food model. Alenezi et al. [
23] studied the flow structure and heat transfer of jet impingement on a rib-roughened flat plat. The results indicates that at the smooth surface with the same rib position, the maximum
for each location was obtained when the rib height was close to the corresponding boundary layer thickness.
According to the previous work, the half-sphere dimpled surface of the heat exchanger plate has been studied and presented a moderate heat transfer rate. However, the half-sphere shape is difficult to machine, so the cylindrical dimpled surface was modified for use in this work. The cylindrical shape not only presents a high heat transfer rate but also presents the vortex of the fluid flow, which can occur around the dimple edge in order to promote heat transfer between the air and the test-section plate. In this research, the air is used as an impinging jet for Re = 1500 to14,600. A total of 15 case studies were investigated, based on different dimple sizes and positions on the impingement plate.
This research work aimed to continue the research carried out in previous work [
24]. The researchers aimed to develop a heat exchanger plate to improve the heat transfer from the impinging jet. The surface was designed and tested in many different cases. However, the experimental results failed to clearly explain the flow behavior on the test-section plate. This is because a sensor could not be set atop the plate as doing so would obstruct the flow. Instead, numerical studies were used to analyze and describe the experiment. In the simulation part, the friction factor (f) value on the test-section plate was the key variable utilized to present the flow obstacle of that device. In this work, the test-section plate with a higher f value required a higher pump power to drive the air flow through the device compared to the test-section plate with a lower f value. However, the TEF was the most important variable used to compare the heat exchange efficiency of the test-section plate. The TEF represents both the heat exchange efficiency and the pump power, which were analyzed in terms of Nu and f, respectively.
2. Experimental Setup
In the experiment, an air compressor with a tank was used to blow the air impinging on the test-section plate. The air continuously flowed through the orifice. The manometer was connected in parallel to the orifice for measuring the pressure drop. The pressure drop was converted to air flow velocity. The adjustable velocity valve was set in front of the test section. Ten Reynolds numbers of air flow velocity were analyzed in the range of Re = 1500 to 14,600. The air was then injected through the nozzle jet (diameter = 1 cm) impinging on the test plate (diameter plate = 30 cm). The test plate was covered by the flat plate on the top to protect the impinging jet air moving out. The three different distances between the nozzle jet and test plate were 2, 4 and 6 cm.
The heater was installed on the bottom wall of the test plate to maintain a uniform surface of heat flux. The electrical output power was controlled by a Variac transformer to obtain a constant heat flux. In
Figure 1, the diameter of the plate (D) is 30 cm. The diameters of the cylindrical dimples are 1 and 2 cm (d) with radial distances between dimples, E
r = 2 d and 3 d, and circumferential dimple distances, E
θ = 1.5 d and 3 d. The depth of the dimples is 0.5 cm. The confinement plates, which are 2, 4 and 6 cm (B), are above the test plate. The details of the geometry of the impingement plate with dimples and tested conditions are shown in
Table 1. There were 15 case studies, shown in
Table 2.
The 16 thermocouples were set in three different radii under the test plate for recording the average value of temperature at the surface. The experimental apparatus is shown in
Figure 2.
The thermocouple was put in the lower-test section plate. The two RTDs were used to measure the average temperature of the air outlet. The error of the temperature sensor was not over ±0.2 °C. All temperature sensors were connected to the signal converter. The data was reported to the monitor and recorded in the computer. The next section was heat at the test-section plate. This was generated from the heater, in which the constant heat flux could be controlled by variable voltage transformer. The last section was the insulator, which was installed at the bottom to control the heat direction so that it only flowed into the test-section plate. The uncertainty in velocity measurement was estimated to be less than ±7%, whereas that of wall temperature measurement was about ±0.5%. The maximum uncertainties of non-dimensional parameters were ±5% for Nusselt number, and ±5% for Reynolds number [
25].
3. Computational Models and Numerical Method
The computational models used in this research can be written as follows:
Equations (1) and (2) are based on the assumptions of a steady-state air flow and incompressible flow. The turbulent models used the Navier–Stokes equations with k-ε and RNG, and the energy equations for solving this present research problems.
Transfer of turbulent kinetic energy:
Transfer of turbulent energy dissipation rate:
Equations (1)–(6) [
26] were solved using a finite volume and were discretized by the Quadratic Upstream Interpolation for Convective Kinetics (QUICK) and the Semi-Implicit method for Pressure-Linked Equations (SIMPLE) algorithms [
27]. All the results converge with the residual values, being less than 10
−6.
The friction factor (f), Nusselt number (Nu), and thermal enhancement factor (TEF) are the key variables in this research. The friction factor can be calculated from the pressure drop on the test plate:
The Nusselt number can be written as:
and
stand for the values for the bulk fluid in the dimple plate, while
and
stand for the values for the bulk fluid in the flat plate [
28].
Reynolds Number in this research is calculated from diameter of jet (
):
6. Conclusions
Actual experiments were carried out and the results utilized as part of the analysis in the simulation part. In the simulation, the heat transfer behavior and friction factor value on the test plate were investigated for the thermal enhancement factor (TEF). TEF can be explained by the heat transfer efficiency of an impinging jet on the test plate. The optimum conditions in terms of the fluid pump power and heat transfer value could be obtained.
From the experiment and simulation results, one can conclude that the cylindrical dimpled surface enhances the TEF. The highest TEF (= 5.5) was found for the configuration with a dimple diameter of d = 1 cm, a radial dimple distance of Er = 2 d, a circumferential dimple distance of Eθ = 1.5 d, a distance between test plate and jet of B = 2, and Reynolds number of Re = 1500. The TEF which was obtained in this work was higher than that obtained in other research, which was about 4.0. This is because the friction factor ratio in this work (f/f0) was very low, even though the was not different from the in other work. The most suitable test plate conditions which could be used to obtain a high TEF are d = 1 cm, B = 2 d, Er = 2 d and Eθ = 1.5 d for Re=1500–14,600.