Experimental and computational study of heat transfer and flow structure of slotted impinging jet

Abstract

Keywords: Flow Structure; Heat Transfer Enhancement; Numerical Simulation; Slotted Jet; Thermochromic Liquid Crystal

Introduction

Impinging jet has evolved as a primary source for controlled cooling as well as heating of a surface due to its ability to improve the heat transfer rate for of drying of food products and electronic components cooling. The flow parameters of air jet are complex in nature and subsequently, heat transfer from the surface is significantly variable. The impinging jet has the advantage of high transfer rate than other heat transfer applications. Because of

this reason, heat transfer by jet impingement has attracted extensive research. Wang et al. [1] examined the flow and heat parameter of jets over a flat surface with orifice configurations of circular, elliptic, and square aspect ratios 3-5. They reported the existence of axis switching vortex rings for rectangular and elliptic jets before the impingement and it enhances turbulence and heat compared with uniform section air jet. They concluded that the rectangular orifice of aspect ratio

*Corresponding author. *E-mail address: illyasmech4@gmail.com This paper was recommended for publication in revised form by Associate Editor Erman Aslan Published by Yıldız Technical University Press, İstanbul, Turkey Copyright 2021, Yıldız Technical University. This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

5 provides better cooling performance. Zhu et al. [2] investigated the heat parameter of nozzles with velocity differences among adjacent jets. The results revealed that Reratio is the primary factor influencing the uniform heat parameter. They reported improved heat parameter of multi nozzles on comparison to single slot nozzle. Afroz and Sharif [3] numerically analyzed the heat parameter of air jets (Swirl number Sw = 0 -0.77) with varying H/D distances (0.5 – 4) and Blockage ratios (0.4 - 0.6). They stated that at the jet area, the velocity parameter is distributed radially and is substantial around the jet axis. Their results revealed that the swirling effect enhances heat transfer at a lower separation distance (H/D = 0.5) compared with uniform section air jet. The peak Nusselt value is initially reduced at higher H/D distances and increases with the increase in swirling effect. The analysis by Fenot et al. [4] comprehensively examined the heat and flow structure of quasi-annular jets with infrared thermography and PIV techniques. They considered the blockage ratio (ratio of inner diameter and outer diameter – Di/Do = 0.35 – 0.85) as the principal parameter and concluded that the existence of shear layer at jet main flow edge causing higher heat parameter. They determined that a higher blockage ratio enhances the entrainment of ambient air before the impingement. Yadav and Saini [5] numerically analyzed the performance of solar heater by jet cooling method (Re = 3500 - 17500). They reported increased heat transfer at the impinging jet area due to by the shearing of air and jet which breaks the boundary layer on the absorber surface. Mohammed and Razuqi [6] studied the heat parameter of a sink provided with air jet. Their study focused on the effect of fan location with constant heat flux varying between 10000–40000 kW/m2 and Reynolds number range of 3400 – 16000. They concluded that a decrease in thermal resistances with rise in Reynolds number for the cut-out template due to the higher turbulence of airflow. Badiger et al. [7] experimentally analyzed the influence of swirl on heat parameter of a jet on a flat vertical surface. They employed the twisted tape insert (Swirl number Sw = 0.52) with Reynolds number between 1000 and 2500. They reported that the position of impinging plate concerning the inner reaction zone of the inverse diffusion flame influences the intensity of heat parameter to the striking plate. Mohammed and Razuqi [8] analysed the heat parameter of a heat sink with air jet. Their study focused on a processor with a heat sink with closely packed rectangular fins with varying heat flux (10000-70000 kW/ m2) for a Reynolds number range between 4000 and 16000. The results reveal that the fins aligned in the core exhibited a larger temperature gradient compared to the fins that are inline around the core. They reported decreased thermal resistance with an increase in Reynolds value and increased Nusselt value when the Reynolds value and heat flux are increased. The study by Lee et al. [9] reported the impact of configuration of pipe on flow and heat parameter for Re = 23000

with L/D = 2 - 14. The study showed that the Nusselt value is high with larger pipe diameter at 0 ≤ r/D ≤ 0.5 and its effect is marginal at the vicinity of wall jet. The local change in heat parameter for circular air jet on a striking surface (Re = 12000 - 18000 and L/D = 0.5 - 8) is analyzed by Katti and Prabhu [10] revealing that stagnation Nusselt number increases with a low value of L/D below 1 because of accelerating flow and the existence of Nusselt number for L/D distance below 3. Yan et al. [11] carried out an analysis on a jet array with an elliptical cross-section with an aspect ratio of 0.25, 0.5,1,2 and 4 using the liquid crystal technique showing that axial change of impingement position is significant at aspect ratio > 1. A minimum stagnation zone in the Nusselt value was apparent by Lytle and Webb [12] for jet exit and surface distance, L/D < 0.25. They revealed that the inner peak disappears beyond L/D = 0.25 and radially leaving jet and stagnant ambient at decreasing L/D distance causes higher turbulence in the boundary layer resulting in outer peak Nusselt number. The velocity parameter at vicinity of the impinging plate and its impact on the Nusselt value of the laminar jet (Re = 100 - 500) issuing from rectangular slots (aspect ratio = 1 - 8.1) were numerically analyzed by Sezai and Mohamad [13]. They reported that the upstream flow from a rectangular slot creates a curvature of streamlines resulting in an off-center velocity peak near the impinging surface. Alekseenko and Bilsky [14] studied the spatial distribution of velocity considering impinging jets with a varying swirl at L/D = 3. Their analysis showed that the existence of vortex was apparent (S = 0.7 – 1). However, the development of the air jet and the heat parameter on wall may be affected by the upstream entry conditions of the slotted air jet, the underlying flow parameters has not yet been revealed in most of the heat parameter studies [2-12,22] of the impinging jet. Ianiro and Cardone [15] investigate and reported that the swirling jet produces a stronger axial recirculation region causing low heat transfer zones which demands a comprehensive study of flow characteristics. While the flow at the pipe outlet alters the jet spreading rate and turbulence characteristics before it impinges on the surface [16], it is essential to analyze those effects on flows and heat parameter of the impinging jet can be made. Therefore, in the present analysis, a comprehensive study on heat parameter and flow structure of slotted jet impingement is studied with different jet configurations. The study includes the jet configuration of the square, cross, and oval for L/D distances of 1 to 4. The performance of a slotted jet has also been compared with a circular jet. The analysis includes a flow parameter study on axial velocity, velocity vectors, and level of turbulence. However, the heat parameter studies of the air jet are employed thermocouples for measuring the surface temperature which can measure only on discrete points, the thermo-chromic liquid crystal sheet used in the present study enables to calculate the temperature variation over the surface. In addition, the reported studies of impinging jets show that the small change in nozzle exit flow structure

will have a significant influence on heat parameter on the target plate. While the studies on circular and swirling jets either the diameter of the pipe or the swirling effects are varied with the same insert, the present study focuses on jets with different configurations which provide the scope for analysing the flow parameters and its influence on heat parameter. While the current analysis provides information on the single jet, the inference from this analysis can be used to understand the heat and flow characteristics with multiple jet arrangements. This may provide the prospect of optimizing the jet arrangement with a combined geometry effect.

Experimental Setup

The physical setup consists of an air jet issuing facility and target surface as shown in Figures 1 and 2. The jet issuing facility has a 1.5 HP high-pressure centrifugal blower (Figure 1c) that delivers air to a 108 mm diameter converging pipe. The centrifugal blower is preferred for reasonable flow steadiness and a large volume of flow conditions. The converging pipe connecting the blower outlet minimizes both mean and fluctuating velocity variations in the flow without change in its total pressure [17]. The main piping network has a length of about a minimum of 20D

(Upstream length, LU =1130 mm) upstream of the venturi meter (Figure 1a) ensuring that flow has fully developed and about 15D (Downstream length, LD = 520 mm) at the downstream to let the effects of venturi contraction to settle. The venturi meter is tapped at its upstream and downstream to connect with a U-tube manometer. A Venturi meter (throat diameter = 20 mm) measures the air flow rate between the two main pipeline sections by a pressure differential. The flow regulating valve located at nozzle pipe controls the air flow rate. The vertical target surface was made of SS foil (SUS 304) having a size of 300 x 300 x 3 as shown in Figure 1b. The thin heater and SS foil are secured with screws over the acrylic plate measuring 450 mm by 450 mm by 12 mm. The power supply unit’s wide range of voltage and current variations. with fine control unit enables the imposition of the heat flux with a narrow working temperature limit (30°- 46°C) of the TLC sheet. A digital camera with a computer system is employed to capture the liquid crystal sheets. The profile of the slotted sections is shown in Figure 3. Thermochromic Liquid Crystal Sheet Thermo-chromic liquid crystal (TLC) sheets are supplied by M/s. Hallcrest exhibiting the RGB color spectrum is used in the analysis. The TLC utilized in this investigation has a bandwidth of R30C5W. Between 30 and 35 degrees Celsius is the temperature range in which the Hallcrest

Figure 3. The slot configurations (a) cross (b) oval (c) square.

RGB mode needs to be reduced to a single value to relate the temperature.

Figure 2. Schematic of an experimental setup. defines the TLC’s bandwidth. The TLC sheet’s operating temperature, however, is between the red start (30°C) and the clearing point (46°C); at lower temperatures, the sheet stays black. It is red between 30° C and 31°C and turns blue in the range between 35°C and 46°C. The clearing point temperature is referred to as turning black again. Thus the local spatial variation of temperature can be obtained by any digitizer and quantified by pixel-specific information. The pixel response corresponding to the temperature is generally available in the RGB color domain. The pixel response in

Calibration of TLC Sheet The aluminum plate, thin heater, and TLC that need to be calibrated are supported by an acrylic plate measuring 400 mm by 400 mm by 12 mm, which is known as the calibration test plate. As seen in Figure 4, an aluminum plate (0.3 x 0.3 x 0.03 m) has eight grooves in it where thermocouples can be inserted. Eight calibrated K-type thermocouples with thin leads will measure the temperature along the Aluminium plate. The NI 9213 data card is coupled with thermocouple leads and the input module of NI 9213 is connected to the computer system. The plate is heated incrementally between 30 C and 46 C, and thermocouples and a data collection system are used to record the average temperature. The power supply system’s precision control units make it possible to deliver heat in tiny, gradual increments. Once the temperature achieves a constant state, the thermocouples’ temperature and the

Figure 5. Normalized hue versus temperature for TLC calibration. TLC’s color are simultaneously recorded. The Lab View Vision Assistant module has the TLC picture loaded. Using the Vision Assistant module, the RGB (red, green, and blue) color is converted to the HSI (hue, saturation, and intensity) domain. Gonzalez et al. [18] stated that hue denotes pure color, saturation indicates how much white light has diluted color, and intensity is a measure of brightness. This leads to the plotting and presentation of the color value between TLC with temperature in Figure 5. Temperature is correlated with hue value using a polynomial fit generated from the calibration curve. The temperature is obtained by the values of crystal sheet using the correlation. Data Reduction Equations (1) and (2) are used to compute the convective heat transfer coefficient (h) and the Nusselt number (Nu) as given by Nuntadusit et al. [19] (1)

(2) The heat input is calculated using the equation given by Nanan et al. [20]

Where U0 is the velocity of air in the smooth pipe and γa is the kinematic viscosity of air. Uncertainty Analysis The Kline-Mc Clintock approach [21] is employed to perform an uncertainty analysis using the equation (5) (5) Convective heat transfer coefficient (h) and other directly measured values are included in the derived quantity (R), which denotes the quantity of interest that must be calculated using a suitable relation. ∆ϵ is the measurement error level. Table 1 provides the measured experimental value together with its error levels. Table 1. Uncertainty in experimental measurements Parameter

(3) The Reynolds number (Re) is computed with the equation given by Nanan et al. [20] (4)

The uncertainty in the Nusselt number for circular jet for X/D = 3.5 at L/D = 4 and for Re = 12700 is 3.85%.

Numerical Methodology

Computational Domain By solving the equations of fluid and energy with turbulence model as a closure, the numerical results are arrived

with the CFX tool. Figure 6a displays the analysis domain employed in this investigation. To carry out flow analysis with varying separation distance, L/D up to 4 is considered. The air is delivered to the impinging surface at a temperature of 30°C. As illustrated in Figure 6a, the boundary conditions are given with Re = 23000. The surfaces where the jet leaves the domain are represented by an unconfined wall with pressure, P = 1 atm. condition of uniform heat flux. Tetrahedral combined with prism are employed for meshing as shown in Figure 4b. The Table 2 shows the various turbulence models used in the previous literatures. The second-order high-resolution discretization scheme is used. However, this discretization scheme gives relatively

Figure 6. (a) Domain of impingement section (b) Mesh distribution with refined elements at the striking section.

Figure 7. Radial velocity component of circular jet for L/D = 2 at Re = 23000 for different turbulence models.

Figure 8. Radial velocity component of circular jet for L/D = 2 at Re = 23000 for varying grid intensity. precise results than the first-order scheme, it is relatively less

stable regarding convergence. To improve the convergence, The energy equation can be written as

the solution is primarily attained with first-order upwind discretization with a blend factor of 0 and residual of 10 . -6

The equations for incompressible flow are given by. The continuity equation is given by

Selection of Turbulence Model Figure 7 gives the turbulence model obtained in the current analysis comparing the deviation of the radial component of velocity with those experimentally obtained by O’Donovan and Murray [30]. SSG Reynolds stress and

Table 2. Turbulence models and other parameters in numerical analysis on impinging jet Previous studies

Experimental Results

Local Heat Transfer Visualization The TLC images showing the variation of Nusselt values for uniform cross section and slotted air jets are given in Figures 9 and 10. The Nusselt value variation at the mid surface for circular jet means that the existence

Figure 9. Nusselt number parameter for circular and slotted air jets for L/D = 1- 4 at Re = 12700.

of a core maintains its values of velocity uniform at the outlet and indicates lower jet spread before impingement as shown in Figure 9 at L/D = 1, conversely the Nusselt value distribution shows that slotted jet transits to turbulent jet with greater jet spread before impingement. However, the variation is marginal among the slotted je at L/D =1 at Re = 12700, the jet spread is notably higher for a square section at Re = 23000 as shown in Figure 10 (c) at L/D =1. In the case of slotted jet distinct zones appear at L/D = 1, indicating that the flow jet leaving the pipe takes the profile of the slot as shown in Figure 9 which is symmetry about both x and y axes. This fact is also reported by Ianiro and Cardone [34] in their study by multi-channel jet at Re = 28000. At L/D = 2 the change in flow pattern coupled with a decrease in Nusselt value is seen for the slotted jet

(Figure 9). The results of the circular jet show that the Nusselt number distribution has little dependence on the separation distance as the variation is marginal at L/D = 4 when compared with L/D = 1 and 2, the slotted jet however shows the marked variation of Nusselt value beyond the region X/D = 1.5 on either side. This is attributed to the merging of slotted jets downstream of flow exit with increased separation distance. The impingement area appears to be broadened with the air jet leaving the square section at all the L/D distances (Figure 10c) however the Nusselt value is not uniform. The Nusselt value further decreases beyond the region X/D = 2 for all the considered cases at L/D = 4, substantiating the weak flux in axial direction of the air jet in the impinged area. Conversely, the Nusselt value at the vicinity of the stagnation area is significantly high for the slotted air jet

Figure 10. Nusselt number distribution for circular and slotted jets for L/D = 1- 4 at Re = 23000.

even at increased separation distances, this is more intense for the square jet at Re = 23000 as shown in Figure 10c. This fact is reported by Gardon and Akfirat [36] in their study at Re = 22000 stating that increasing the turbulence enables the heat parameter to increase at the stagnation area though jet arrival velocity is falling.

Numerical Results

Axial Velocity Variation The axial velocity parameter for circular and slotted jets is shown in Figure 11. The axial velocity peak is observed for slotted jets and noticeably higher for square section corresponding to r/D ≈ 0.4 at L/D = 1 (Figure 11a) and the peak decreases with increased L/D distances. The velocity parameter of the circular jet reduces reaching its maxima with increasing r/D distances but not as rapidly as the slotted jet at all L/D distances. While the velocity peak in the slotted jet is due to the sharp edges creating intensive turbulence as observed by Parham et al. [37] for Re = 23460

at flow exit thereby increasing the velocity at a shorter L/D distance, whereas the lower turbulence circular impinging jet causing the existence of potential core results in a marginal change in velocity at the exit as shown by Viskanta [38] and Alimohammadi et al. [39]. The velocity peak decreases for the slotted jets at L/D = 2 (Figure 11b) nearby stagnation area and the decrease is because of low axial force with increase in separation distance as reported by Ianiro and Cardone [34]. The jet leaving the square section however produces higher axial velocity radially at 0 < r/D < 0.3 compared with cross and oval sections as well as a circular section at L/D = 2 as presented in Figure 11b. The reduction in velocity parameter at L/D = 3 is marginal comparing with L/D = 2 for the slotted jets and the decrease is significant at L/D = 4 (Figure 11d). On the contrary, while increasing the L/D distances the velocity parameter of the circular jet is not affected in impinging region. This is due to the velocity of the jet coming out of the circular section is uniform with an increase in L/D distance whereas higher turbulence and mixing slotted jet causes low arrival velocity of air jet at the impinging surface.

Figure 11. Axial velocity parameter for air jet for L/D = 1- 4 at Re = 23000.

Figure 12. Radial velocity parameter for air jet for L/D = 1- 4 at Re = 23000

Radial Velocity Variation The radial velocity distribution of slotted and circular jets is shown in Figure 12. The radial velocity near the impinging surface increases at 0 < r/D < 0.81 and 0 < r/D <

Comparison Of Results

Since the selection of a particular slot configuration considerably influencing the flow and heat parameter, the obtained values of axial velocity (u/U0) for the circular air jet at L/D = 2 is compared with previous studies as shown in Figure 16. The values of (u/U0) are varying between 0.1 and 0.98 in the area with r/D values between 0 and 1.03. The higher values of axial velocity are observed with r/D values between 0 and 0.37 for Fenot et al. [40]. The current values and the findings of Fairweather and Hargrave [41] and Sendaet al. [42] are fairly comparable in the stagnation as well in impinging regions. The relative error is 13.51% when compared to the mean values found in the literature, and it corresponds to the region where 0 < r/D < 0.9.

Conclusion

Experimental and numerical analysis were done on the slotted impinging jet heat transfer and flow properties. The analysis leads to the following conclusions, • The existence of potential core for circular air jet causing marginal variation of Nusselt value distribution at L/D = 4 compared to L/D = 1 and 2 whereas the slotted jet shows marked variation of Nusselt (136.8 at L/D = 1, 135.2 at L/D = 2 and 113.5 at L/D = 4) at X/D = 1.5.

The Nusselt value at 0 < X/D < 0.5 is higher in the range varying between 127.8 and 167.9 for the slotted jet even at increased separation distances compared with the circular jet. The impingement section is more extended for square jet corresponding to X/D = 3, 2.5, and 1.75 at L/D = 1, 2, and 3 respectively with relatively higher Nusselt number in the range of (Nu = 117 – 167.9 compared with circular, cross and oval jets The key finding on flow structure reveal that the square jet exhibits higher axial velocity peak (u/U0 = 2.124) in the area at r/D varies between 0 and 0.5 for Nozzle to plate distance = 1 and the peak decreases progressively with increasing separation distance and reaching its minimum at u/U0 = 1.729 for L/D = 4. The axial velocity parameter of the circular air jet is not affected by increased separation distances. The square jet exhibits relatively higher average radial velocity (ur/ U0 = 1.31) substantiating its higher momentum in the at r/D values between 0.5 and 1.5. The highest values of turbulence intensity (TI = 0.206) is observed at r/D = 0.5 for the square air jet at L/D = 3 The presence of reverse flow at the outer layer of the jet exit is observed for the square and cross jets at r/D = 0.4 and even at increased separation distances (L/D = 3 and 4) which enhances the entrainment of ambient air.

Nomenclature

Impingement surface area m2 Outlet diameter of pipe, m Dimensionless distance between outlet of air jet and target surface Air side heat transfer coefficient, W/m2 K Thermal conductivity, W/m K Nusselt number Reynolds number Dimensionless radius measured on the target plate Heat flux, W/m2 Temperature on target plate at r distance from the mid surface of the target plate, K Temperature of outlet air at smooth pipe, K Ambient temperature, K

Authorship Contributions

Conceptualization, Methodology, Resources, Formal analysis and investigation – S. Mohamed Illyas, B.R.Ramesh Bapu and A. Muthu Manokar. Writing, review and editing – S. Mohamed Illyas, B.R.Ramesh Bapu and A. Muthu Manokar.

Data Availability Statement

The data supporting the findings of study is provided in the article, according to the authors. The corresponding

author can provide the raw data supporting the findings of the study upon reasonable request.

Conflict Of Interest

No possible conflicts of interest exist that the author disclosed about writing or publication of this paper.

Ethics

Regarding the writing and publication of this manuscript, the author has not declared any potential conflicts of interest.

Share and Cite