Effect of oscillatory EHD on the heat transfer performance of a flat plate

(1)

Effect of oscillatory EHD on the heat transfer performance of a ﬂat plate

Wen-Junn Sheu

a

, Jen-Jei Hsiao

a

, Chi-Chuan Wang

b,⇑

a

Department of Power Mechanical Engineering, National Tsing Hua University, Hsinchu 300, Taiwan b

Department of Mechanical Engineering, National Chiao Tung University, Hsinchu 300, Taiwan

a r t i c l e

i n f o

Article history:

Received 29 August 2012

Received in revised form 5 February 2013 Accepted 8 February 2013

Available online 5 March 2013 Keywords: Electrohydrodynamics Ionic wind Oscillatory EHD Enhanced convection

a b s t r a c t

The present study investigates the effect of oscillatory EHD on the heat transfer performance of a ﬂat plate. A needle type electrode with positive polarity is used to generate ionic wind with applied voltage ranging from 4 to 9.5 kV. The wave forms of the input voltage are either steady or stepped and the fre-quency ranges from 0.5 to 2 Hz with the separation distance between the electrode and the test plate being 5 and 15 mm, respectively. It is found that at the same applied voltage, the heat transfer perfor-mance subject to the oscillatory EHD is always inferior to that of the steady EHD for all frequencies. A parabolic dependence of the ionic current with the supplied voltage is seen under steady EHD operation. However, the ionic current vs. supplied voltage shows a rather linear dependence pertaining to the oscil-latory EHD. For a smaller separation distance of 5 mm, the heat transfer performance for f = 0.5 Hz is slightly higher than that of f = 2 Hz. However, the trend is reversed when the separation distance is increased to 15 mm.

1. Introduction

The presence of a high electric ﬁeld between a discharged and a grounded electrode bring about an air motion which is normally regarded as the ionic wind. The pioneer studies employing the io-nic wind as a cooling method can be dated back to late 1960s[1,2]

and since then the use of electrohydrodynamics (EHD) for heat transfer augmentation applicable for single-phase or two-phase heat transfer had been reported considerably. An early review by Allen and Karayiannis [3] had outlined the basic enhancement mechanisms and a comprehensive review of past work. Recently, fan-less heat transfer augmentation such as cooling of LEDs under natural convection gains much attention. This is because the major concern of noise. In common implementation of heat transfer aug-mentation for LED cooling, passive methods incorporating natural convection heat sinks such as plate fin, pin fin and radial fin (e.g.

[4–6]) are mostly adopted. Active methods such as microjet array cooling, liquid-cooling, thermoelectric cooler, and oscillating heat pipes are also feasible techniques that efﬁciently dissipate heat out of the high power LEDs (e.g.[7–10]). Despite the foregoing ac-tive methods show effecac-tive heat removal in high power LEDs, con-cerns of noise and vibration resulting from the moving parts of these active methods remain. Therefore, rather than using mechanical devices to promote cooling, ionic wind featuring the

beneﬁt of forced convection but free of noise concerns is one of the potential candidates. This would certainly simplify the design and increase the reliability of the cooling module for LED devices due to the lack of moving parts.

There had been intensive studies in association with the EHD applicable for heat transfer augmentation subject to natural con-vection (e.g. [11–22]). These previous efforts investigated heat transfer augmentation in various configurations and electrode arrangements under steady EHD operation. It appeared that no ef-forts had been made concerning the influence of oscillatory electric field on the heat transfer performance. In this regard, it is the objective of this study to provide some preliminary results about the influence of oscillatory electric field on the heat transfer perfor-mance pertaining to natural convection of a plate heat sink.

2. Experimental setup

The schematic of the test facility is shown inFig. 1(a). It con-tains a functional generator (Tectronix AFG 3022), a high voltage power supply (positive polarity, up to 10 kV, from YOU-SHANG Technical Corp.), a power meter (DM-1250), an acrylic housing, and the test section. The test plate is an aluminum alloy with an effective thermal conductivity of 175 W m1_K1_{. The size of the} test section is 45 mm (W) 45 mm (L) 2 mm (H). A Kapton hea-ter is adhered to the bottom of the heat sink with a high thermal conductivity grease (k = 2.1 W m1_K1_{). The heater is of the same} size as the base plate for removing the spreading resistance. In addition, a VIP (vacuum insulation panel) insulation plate with a

⇑ Corresponding author. Address: E474, 1001 University Road, Hsinchu 300, Taiwan. Tel.: +886 3 5712121x55105; fax: +886 3 5720634.

E-mail addresses: ccwang@mail.nctu.edu.tw, ccwang@hotmail.com (C.-C. Wang).

Contents lists available atSciVerse ScienceDirect

International Journal of Heat and Mass Transfer

(2)

very low thermal conductivity of 0.004 W m1_K1_{is placed} be-neath the heater to reduce the heat loss. The heater is powered by a DC power supply. The needle electrode is made of tungsten with a radius of 3

l

m at the tip of the needle. The electrode is con-nected to a high voltage generator with the heat sink being grounded. The vertical separation of the electrode to the test plate is ﬁxed at 5 mm.

Five T-type thermocouples are equally attached to the base plate of the heat sink to obtain the mean temperature of the base plate (Tb). The thermocouples were pre-calibrated with an accu-racy of 0.1 °C. The average temperature is then used to estimate the surface temperature of the heat sink via Fourier’s law of con-duction. The ambient temperature is normally around 22–25 °C, and the system pressure is around 100 kPa. The measurements of thermocouples are collected via the data acquisition system MX100 from YOKOGAWA.

The high voltage generator is specially made with variable volt-ages of 0–10 kV. A voltmeter (YFE YF-3120) with an accuracy of 0.8% is bridged across the voltage generator and a microammeter (YOKOGAWA 73203) with a resolution of 0.1

l

A is connected be-tween the heat sink and the voltage generator to obtain the applied voltage and the corona current, respectively.

For visualization of the influence of ionic wind, a smoke gener-ator capable of generating water vapor is used. The generated water vapor is directed by a capillary tube with strainghter and is attached to the electrode as shown inFig. 1(b). The generated smoke is entrained by the supplied EHD flow along the flat plate. A He–Ne laser from Actor-Mate Corp. with a wave length of 532 nm and a maximum output power of 100 mW is used to visu-alize the generated ionic wind.

3. Data reduction

The total supplied heat input (Qt) is 3.5 W which can be ob-tained by the applied voltage and a given resistance of the heater. Notice that a VIP (vacuum insulation panel) is installed below the heater with effective thermal conductivity of 0.004 W m1_K1_, thereby eliminating the major heat loss below the heater. An insu-lation box made of bakelite with a low thermal conductivity of 0.233 W m1_K1_{is placed beneath the VIP. In addition, a total of} four T-type thermocouples are installed inside the bakelite. The measured average temperature in the backlite is then used to esti-mate the heat loss via Fourier’s law of conduction. The heat sup-plied to the heat sink (Qf) is thus obtained by subtracting the heat loss from the total power input:

Qt¼ V2 R ð1Þ Ql¼ kAdT dx¼ kA Tins;c1 Tins;c2 t ð2Þ Qf ¼ Qt Ql ð3Þ

The effective heat transfer coefﬁcient is calculated by Newton’s law of cooling:

Qf ¼ hAðTb T1Þ ð4Þ

The index of the EHD performance is characterized by the enhance-ment index for natural convection (Er) which represents the average heat transfer coefﬁcient with EHD divided by that without EHD:

Er¼ hehd

hnc ð5Þ

Notice that the subscript ehd and nc represents test conditions that are conducted under ionic wind and natural convection condition, respectively. The experimental uncertainty is estimated using the uncertainty propagation equation recommended by Kline and McClintock[23]. Normally, a total of 100 samples were taken during the experiment. The calculated uncertainty of heat transfer coefﬁ-cient varies from 1.74% to 6.15% while the enhancement index for natural convection (Er) varies from 2.46% to 6.59% with the conﬁ-dence level of 95%.

4. Results and discussion

The wave forms of the input voltage are either steady or oscil-latory (stepped function). The frequency ranges from 0.5 to 2 Hz and the input power is 3.5 W with the separation distance between the electrode and the test plate being 5 and 15 mm, respectively.

Fig 2(a) shows the enhancement heat transfer ratio, Er, vs. applied voltage amid the steady EHD and stepped EHD at a separation dis-tance of 5 mm andFig. 2(c) depicts the variation of thermal resis-tance against the supplied voltage. Note that the airflow induced by EHD is inclined to the flat plate. The airflow is not totally down-ward, it has an axial component that is parallel to the flat plate and is perpendicular to the buoyancy airflow direction. With this com-ponent, it distorted the rising airflow considerably and eventually improves the heat transfer performance as shows inFig. 2(a).The results clearly show that the enhancement ratio under steady EHD operation outperforms those of oscillatory operations. On the other hand, the enhancement ratios for the frequency ranging from 0.5 to 1.5 Hz are comparable whereas Erfor f = 2 Hz is appar-ently lower than those of lower frequencies. In fact, a maximum deviation of 7% in Erfor f = 2 Hz is found at a supplied voltage of 4 kV. The results are quite similar to those calculations made by Lai et al.[24]to some extent. Note that Lai et al.[24]conducted Nomenclature

A surface area, m2

Er enhancement index for natural convection, dimension-less

f frequency, Hz H height, mm i corona current,

l

A

h heat transfer coefﬁcient, W m2_K1

hehd average heat transfer coefﬁcient under EHD operation,

W m2_K1

hnc average heat transfer coefﬁcient under natural

convec-tion, W m2_K1

k thermal conductivity, W m1_K1 L length, mm

P power, W

Qt total heat transfer rate, W Qf heat supplied to the heat sink, W Ql heat loss from the bakelite, W R electrical resistance,X

T temperature, °C

Tb surface temperature of the base plate, °C T1 ambient temperature, °C

Tins,c1 measured temperature at the top of the bakelite, °C Tins,c2 measured temperature at the bottom of the bakelite, °C

DT temperature difference between LED and ambient, °C V voltage, V

(3)

a numerical simulation of the oscillatory EHD ﬂow in a positive wire-plate electrostatic precipitator. Two regimes of oscillation are observed in their simulation. The ﬂow is characterized by a

sin-gle eddy that is opposed to the wire and oscillates with frequencies between 1.2 and 4.3 Hz. The second regime is characterized by counter rotating eddies that oscillates amid 0.6 and 1.6 Hz. Despite

V

A

Function Generator

High Voltage Power Supply Voltage Meter Acrylicbox Sample

V

A

Power Supply Heater VIP Controlled Environmental Chamber 5 mm L = 5, 10 and 15 mm

(a) Schematic of the experimental setup for heat transfer measurement.

V

Function Generator

High Voltage Power Supply

He-Ne Laser

Smoke Generator

A

Voltage Meter

(b) Schematic of the flow visualization setup.

(c) Smoke generator

(4)

the conﬁguration is different between the current study and theirs, the characterized frequency seems to be in line with each other.

The higher heat transfer performance for the oscillatory EHD at a lower frequency of 0.5 Hz is also indirectly supported by the mea-sured I–V characteristics shown inFig. 2(b) where the measured

Applied Voltage (V) 3600 3800 4000 4200 4400 4600 Er (h EHD /hNC ) 1.4 1.5 1.6 1.7 1.8 1.9 Steady Step 2Hz Step 1.5Hz Step 1Hz Step 0.5Hz

(a) Enhancement index for natural convection vs. applied voltage.

Applied Voltage (V) 4800 5000 5200 5400 5600 5800 6000 6200 6400 6600 Current ( µ A) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 steady Step 2Hz Step 1.5Hz Step 1Hz Step 0.5Hz

(b) Ionic current vs. applied voltage.

Voltage-Thermal Resistance Applied Voltage(V) 3600 3800 4000 4200 4400 4600 Thermal Resistance(K/W) 26 28 30 32 34 36 Steady Step 2Hz Step 1.5Hz Step 1Hz Step 0.5Hz

(c) Thermal resistance vs. applied voltage

Voltage-Er

Voltage-Current

Fig. 2. Effect of steady and oscillatory EHD on the thermal resistance and the corresponding I–V characteristics for a horizontal separation distance of 5 mm.

Applied Voltage (V) 5000 6000 7000 8000 9000 10000 Er (hEHD /hNC ) 1.4 1.6 1.8 2.0 2.2 2.4 2.6 Steady Step 2Hz Step 1.5Hz Step 1Hz Step 0.5Hz

(a) Enhancement index for natural convection vs. applied voltage.

Applied Voltage (V) 5000 6000 7000 8000 9000 10000 Current ( µ A) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Steady Step 2Hz Step 1.5Hz Step 1Hz Step 0.5Hz

(b) Ionic current vs. applied voltage. Voltage-Thermal Resistance Applied Voltage(V) 5000 6000 7000 8000 9000 10000 Thermal Resistance(K/W) 20 22 24 26 28 30 32 34 36 Steady Step 2Hz Step 1.5Hz Step 1Hz Step 0.5Hz

(c) Thermal resistance vs. applied voltage. Voltage-Er

Voltage-Current

Fig. 3. Effect of steady and oscillatory EHD on the thermal resistance and the corresponding I–V characteristics for a horizontal separation distance of 15 mm.

(5)

ionic current for f = 0.5 Hz is slightly higher than that of f = 2 Hz. Since the induced ionic wind velocity is proportional to the square root of corona current [25], it therefore suggests a higher heat transfer performance at a lower frequency of 0.5 Hz. In the mean-time, it is interesting to note that the ionic current for the oscilla-tory EHD poses a rather linear relation with the applied voltage. Notice that the ionic current for the steady EHD operation is nor-mally proportional to the square of the supplied voltage provided the threshold voltage is exceeded. It is not fully understood why it reveals a rather linear dependence between the ionic current and supplied voltage. One of the possible explanations may be attributed to lesser contribution of the avalanche due to periodic shut-off of the electric ﬁeld.

Test results for a larger separation distance of 15 mm are de-picted inFig. 3. It is interested to note that the observed enhance-ment ratio for the oscillatory EHD shows an opposite trend when the horizontal separation distance is increased to 15 mm. Despite the heat transfer performance under oscillatory EHD is still inferior to that of the steady one at the same supplied voltage, the heat transfer performance of the higher frequency (f = 2 Hz) exceeds that of the lower frequency (f = 0.5 Hz) as shown inFig. 3(a). Expla-nations of this result are twofold. Firstly, as pointed out by Bafﬁgi and Bartoli[26] who conducted the heat transfer enhancement study for an inclined plate subject to air pulsating jet. Their results clearly showed that the downstream enhancement is more appar-ent than that of upstream. Therefore it explains partially about the rise of the heat transfer performance as the separation distance is increased. Note that the spark-over voltage is considerably in-creased with the separation distance, leading to a considerable rise of enhancement ratio (as high as 2.4 compared to only 1.8 in

Fig. 2(a)). Secondly, for a further interpretation of the observed re-sults, it is helpful to examine the associated flow pattern pertaining to the influence of oscillatory EHD.Fig 4represents the progress of the flow patterns amid various frequencies and that without EHD at a larger separation distance of 15 mm. Notice that the experi-ment is conducted without heat addition. With a steady EHD input,

as shown inFig. 4(a), the generated smoke is headed toward the test plate. On the other hand, with stepped oscillations of the sup-plied voltage, the induced airflow does not respond in phase with the oscillatory frequency. As clearly seen inFig. 4(b), a swinging motion is clearly seen where the smoke first moves toward the test plate and stick close to the plate for some time which is similar to that of a steady EHD operation, and eventually the airflow departs from the surface and again shows a parallel airflow as that without EHD. The oscillatory EHD thus characterizes a swinging flow pat-tern. The effective time, denoting the time for the directed airflow rested upon the test plate before it departs and swings back to be in parallel with the test plate, for f = 2 Hz is longer than that of f = 0.5 Hz. For instance, at a smaller frequency like 0.5 Hz, the effec-tive time for the induced airflow toward the test plate is only around 1.1 s in its two seconds operation as shown in Fig. 4(e). Conversely, the effective time for the induced airflow toward the test plate is about 1.4 s over its 2 s operation time when the fre-quency is raised to 2 Hz. In addition, the swinging motion for f = 2 Hz is faster than that f = 0.5 Hz. These observations can be eas-ily seen inFig. 4(b)–(e). As a consequence, the observed results suggested a longer encountered time for the induced airflow along-side the test plate, thereby a higher heat transfer performance is seen for f = 2 Hz when compared with that of f = 0.5 Hz.

5. Conclusions

This study examines the inﬂuence of oscillatory EHD on the heat transfer performance of a ﬂat plate with applied voltage being ranged from 4 to 9.5 kV. The frequency ranges from 0.5 to 2 Hz and the separation distance between the electrode and the test plate is 5 and 15 mm, respectively. Based on the foregoing discussions, the following conclusions are reached:

1. For the same applied voltage, it appears that the heat transfer performance subject to an oscillatory EHD is inferior to that of a steady EHD for all the tested frequencies.

(6)

2. A parabolic dependence of the ionic current with the supplied voltage is seen under a steady EHD operation. However, the ionic current vs. supplied voltage shows a rather linear depen-dence pertaining to an oscillatory EHD input.

3. For a smaller separation distance of 5 mm, the heat transfer per-formance for f = 0.5 Hz is slightly higher than that of f = 2 Hz. However, the trend is reversed when the separation distance is increased to 15 mm. This is due to a signiﬁcant difference in the induced airﬂow pattern and the effective time rested on the test plate caused by the oscillatory EHD.

Acknowledgment

This work is supported by the National Science Council of Tai-wan under contract of 100-2221-E-007-083 and 100-2622-E-009-005-CC2.

References

[1] R. O’brien, A. Shine, Some effects of an electric ﬁeld on heat transfer from a vertical plate in free convection, J. Heat Transfer 89 (1969) 114–116. [2] M.E. Franke, Effect of vortices induced by corona discharge on free-convection

heat transfer from a vertical plate, J. Heat Transfer 91 (1969) 427–433. [3] P.H.G. Allen, T.G. Karayiannis, Electrohydrodynamic enhancement of heat

transfer and ﬂuid ﬂow, Heat Recovery Syst. CHP 15 (1995) 389–423. [4] J.C. Shyu, K.W. Hsu, K.S. Yang, C.C. Wang, Thermal characterization of shrouded

plate ﬁn array on an LED backlight panel, Appl. Therm. Eng. 31 (2012) 2909– 2915.

[5] R.T. Huang, W.J. Sheu, C.C. Wang, Effect of orientation on the natural convective performance of square pin ﬁn heat sinks, Int. J. Heat Mass Transfer 51 (2008) 2368–2376.

[6] S.H. Yu, K.S. Lee, S.J. Yook, Optimum design of a radial heat sink under natural convection, Int. J. Heat Mass Transfer 54 (2011) 2499–2505.

[7] Y. Lai, N. Cordero, F. Barthel, F. Tebbe, J. Kuhn, R. Apfelbeck, D. Wurtenberger, Liquid cooling of bright LEDs for automotive applications, Appl. Therm. Eng. 29 (2009) 1239–1244.

[8] Y. Deng, J. Liu, A liquid metal cooling system for the thermal management of high power LEDs, Int. Commun. Heat Mass Transfer 37 (2010) 788–791.

[9] Z.M. Wan, J. Liu, K.L. Su, X.H. Hu, S.S. M, Flow and heat transfer in porous micro heat sink for thermal management of high power LEDs, Microelectronics 42 (2011) 632–637.

[10] J. Li, B. Ma, R. Wang, L. Han, Study on a cooling system based on thermoelectric cooler for thermal management of high-power LEDs, Microelectron. Reliab. 51 (2011) 2210–2215.

[11] S.M. Marco, H.R. Velkoff, Effect of electrostatic ﬁelds on free convection heat transfer from ﬂat plate, ASME Paper, no. 63, ht. 9, 1963.

[12] K. Kibler, H. Carter, Electrocooling in gases, J. Appl. Phys. 45 (1974) 4436–4440. [13] M. Franke, L. Hogue, Electrostatic cooling of a horizontal cylinder, J. Heat

Transfer 113 (1991) 544–548.

[14] M. Ohadi, D. Nelson, S. Zia, Heat transfer enhancement of laminar and turbulent pipe ﬂow via corona discharge, Int. J. Heat Mass Transfer 34 (1991) 1175–1187.

[15] B. Owsenek, J. Seyed-Yagoobi, R. Page, Experimental investigation of corona wind heat transfer enhancement with a heated horizontal ﬂat plate, J. Heat Transfer 117 (1995) 309–315.

[16] B. Owsenek, J. Seyed-Yagoobi, Theoretical and experimental study of electrohydrodynamic heat transfer enhancement through wire-plate corona discharge, J. Heat Transfer 119 (1997) 604–610.

[17] S. Bhattacharyya, A. Peterson, Corona wind-augmented natural convection. Part 1: Single electrode studies, J. Enhanced Heat Transfer 9 (2002) 209–219. [18] L. Zhao, K. Adamiak, EHD ﬂow in air produced by electric corona discharge in

pin–plate conﬁguration, J. Electrostat. 63 (2005) 337–350.

[19] R.T. Huang, W.J. Sheu, C.C. Wang, Heat transfer enhancement by needle-arrayed electrodes – an EHD integrated cooling system, Energy Convers. Manage. 50 (2009) 1789–1796.

[20] D.B. Go, S.V. Garimella, T.S. Fisher, R.K. Mongia, Ionic winds for locally enhanced cooling, J. Appl. Phys. 102 (2007). 053302-053302-8.

[21] D.B. Go, R.A. Maturana, T.S. Fisher, S.V. Garimella, Enhancement of external forced convection by ionic wind, Int. J. Heat Mass Transfer 51 (2008) 6047– 6053.

[22] I.Y. Chen, M.Z. Kuo, K.S. Yang, C.C. Wang, Enhanced cooling for LED lighting using ionic wind, Int. J. Heat Mass Transfer 56 (2013) 285–291.

[23] S.J. Kline, F.A. McClintock, Describing the uncertainties in single sample experiments, Mech. Eng. 75 (1953) 3–8.

[24] F.C. Lai, P.J. Mckinney, J.H. Davidson, Oscillatory electrohydrodynamic gas ﬂow, J. Fluids Eng. 117 (1995) 491–497.

[25] M. Robinson, Movement of air in the electric wind of the corona discharge, Trans. AIEE 80 (1961) 143–150.

[26] F. Bafﬁgi, C. Bartoli, Heat transfer enhancement in natural convection between vertical and downward inclined wall and air by pulsating jets, Exp. Therm. Fluid Sci. 34 (2010) 943–953.