Experimental investigations on the effects of inclination angle and blowing ratio on the flat-plate film cooling enhancement using the vortex generator downstream

Applied Thermal Engineering 119 (2017) 573–584

Contents lists available at ScienceDirect

Applied Thermal Engineering journal homepage: www.elsevier.com/locate/apthermeng

Research Paper

Experimental investigations on the effects of inclination angle and blowing ratio on the flat-plate film cooling enhancement using the vortex generator downstream Liming Song, Chao Zhang, Yingjie Song, Jun Li ⇑, Zhenping Feng Institute of Turbomachinery, Xi’an Jiaotong University, Xi’an 710049, China

h i g h l i g h t s
The impact of VG on film cooling effectiveness is investigated experimentally.
The flow mechanism of VG enhancing film cooling performance is clear visualized by PIV system.
The effect of blowing ratio on the flow mechanism and film cooling effectiveness is studied.
The effect of inclination angle on the flow mechanism and film cooling effectiveness is studied.

a r t i c l e

i n f o

Article history: Received 7 December 2016 Revised 18 March 2017 Accepted 19 March 2017 Available online 21 March 2017 Keywords: Film cooling Vortex generator Blowing ratio Inclination angle Experimental investigations

a b s t r a c t Experimental studies were carried out to study the effects of inclination angle (a ¼ 20 ; 30 ; 40 ) and blowing ratio (M = 0.5, 1.0, 1.5) on the vortex generator (VG) to enhance the film cooling performance in a flat plate. Clear vortical structures visualized by Particle Image Velocimetry (PIV) system showed that an anti-counter-rotating vortex pair (ACRVP) was generated by VG. The secondary flow was entrained towards the wall and spread laterally because of the downwash effect of the ACRVP. The VG significantly improved the film cooling effectiveness. Especially when a ¼ 20 , M = 1.5, the area averaged film cooling effectiveness was improved as much as 248% by using VG. For the cases with VG, as the blowing ratio increases, the film cooling effectiveness increases in those cases with a ¼ 20 ; 30 due to the strong downwash effect of ACRVP. In the case with a ¼ 40 , the film cooling effectiveness was enhanced with the increase of the blowing ratio when M < 1, due to the strong downwash effect of ACRVP. However, when M > 1, the film cooling effectiveness decreased because the upward momentum of the secondary flow was so high that its flow rate exceeded the entrainment capacity of ACRVP, worsening the film attachment on the wall. Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction High thermal efficiency and power output is the ultimate goal of gas turbine design, and the greatest contribution is made by raising the turbine inlet temperature. Nowadays, the inlet temperature of advanced gas turbine has exceeded the melting point of turbine blade materials, therefore, sophisticated cooling technologies, such as internal cooling, impingement cooling, film cooling and combined cooling technologies, have been applied to lower down the blade temperature so as to ensure a long service time and operation safety [1,2]. Among those technologies, film cooling has been one of the most important cooling techniques for dec⇑ Corresponding author. E-mail address: [email protected] (J. Li). http://dx.doi.org/10.1016/j.applthermaleng.2017.03.089 1359-4311/Ó 2017 Elsevier Ltd. All rights reserved.

ades. Goldstein [3], Bogard et al. [4] and Han et al. [5] reviewed research achievements of film cooling in experiments and numerical computations over the last half century, and concluded that film cooling performance could be affected by various parameters, such as inclination angle of the coolant hole, blowing ratios, mainstream turbulence and blade surface curvature. The film cooling structure is intrinsically an inclined jet in a cross-flow (JICF). Abundant studies, both numerical researches (Fric and Roshko [6], Leylek and Zerkle [7], Tyagi and Acharya [8]) and experimental studies (Kelso and Perry [9], Blanchard et al. [10]), have explored the nature of JICF. These research results show that the typical feature of JICF affecting the film cooling performance is the counter-rotating vortex pair (CRVP) in the jet, namely, the so-called kidney-shaped pair of vortices. They are generated by the round jet when it penetrates into the crossflow. The

574

L. Song et al. / Applied Thermal Engineering 119 (2017) 573–584

Nomenclature Symbols D DR M T Tu U

diameter of the film cooling hole [mm] coolant-to-mainstream density ratio blowing ratio temperature [°C] mainstream turbulence intensity velocity component perpendicular to the wall [m/s]

Greek symbols a the inclination angle, angle between the film cooling hole and mainstream q density [kg/m3] k thermal conductivity [W/(m K)] g adiabatic film cooling effectiveness

CRVP imposes the upwash effect on the jet and promotes the liftoff tendency of the jet. Hence, how to cripple or suppress the strength of the CRVP has become the challenge in enhancing film cooling performance. The shaped hole, an exit hole shaped like a diffuser other than the ellipse, is the only major advancement which has been realized in industry. Bunker [11] reviewed the development of shaped film cooling technology and summarized that a better attachment and lateral spreading of the coolant would be achieved compared with traditional round hole. Besides, film cooling performance of shaped hole was little susceptible for blowing ratio and turbulence intensity of mainstream. Haven et al. [12] investigated the vortical structure of three shaped holes and found that anti-kidney pairs were generated, leading to a better jet attachment whatever the condition was. Considering the tremendous impact of exit hole shape, various novel hole designs were put forward, such as sharp edged diffuser hole [13], leaf-shaped hole [14] and trench hole [15]. In addition, various technologies for enhancing cooling effectiveness have been studied. Heidmann and Ekkad [16] put forward the ‘‘anti-vortex” film cooling hole concept – two side holes were drilled intersecting with the main hole, which effectively counteracted the detrimental vorticity associated with standard circular cross-section film-cooling holes. Furthermore, various novel hole designs were put forward, such as the sister hole structure. A numerical analysis was conducted by Ely and Jubran [17] on sister holes by using cylindrical holes with 55 inclination angle. Their results showed that the sister holes dramatically suppressed CRVP, resulting in the significant improvement in film cooling effectiveness. Wu et al. [18] applied Thermochromic Liquid Crystal (TLC) technique on analyzing the effects of side hole position and blowing ratio in film cooling performance, and found that the side holes could improve the film cooling performance by repressing the CRVP intensity of the main hole, and the downstream sister hole performs best at blowing ratio from 0.3 to 2.5. Besides, Kusterer et al. [19] introduced the Double-Jet Film Cooling (DJFC) technology through hole geometry, and established an anti-kidney vortex pair in the DIFC. Furthermore, Kusterer et al. [20] developed a Nekomimi (‘‘cat ears”) technology which is suitable for fabrication derived from the DJFC technology. Recently, many attempts on enhancing the film cooling performance through various flow control devices at the upstream or downstream of the film cooling hole are carried out, hoping to generate opposite vortex pairs to eliminate the negative effects of the

Subscript avg average value area the whole measured wall surface, 19; 3 6 Y=D 6 3 f secondary flow m mainstream spanwise direction parallel to Y direction streamwise direction parallel to X direction w the adiabatic wall

1 6 X=D 6

Acronyms ACRVP anti-counter-rotating vortex pair CRVP counter-rotating vortex pair CTA constant temperature anemometer PIV particle image velocimetry VG vortex generator

CRVP. Na and Shin [21] placed a step ramp facing backward at the upstream of the hole. The effects of the angle and sharpness of the ramp were investigated by numerical simulation. The results showed that the geometric modification changed the position of the interaction of boundary layer and coolant, eliminated the horseshoe vortex, resulting in a broad lateral extension, thus improved the film cooling effectiveness by two times or more. Sakai [22] experimentally studied the ribs installed in the secondary flow channel and three types of rear bumps in the downstream of the hole in detail. It was showed that a longitudinal vortex generating downward velocity component was formed at the trailing edge of the cylindrical bump, improving the film cooling performance. Funazaki et al. [23] put base-type double flowcontrol devices (DFCDs) at the upstream of the film cooling hole in a flat plate and achieved an enhancement in effectiveness. A research of DFCDs on the turbine blade was carried out recently by Kawabata et al. [24]. The vortex generator (VG) geometry, referred to as a microramp at times, has been recently addressed for boundary-layer flow control, such as the control of supersonic oblique shock applied by Babinsky et al. [25]. Rigby and Heidmann [26] first attempted to put the VG downstream and found that the design could cancel the upwash CRVPand generated the downwash anti-kidney vortex pair downstream through the numerical study. Zaman et al. [27] conducted an experiment on the effects of the height, location and radius of VG in preventing lift-off of a jet in crossflow, which demonstrated detailed flow field properties. Furthermore, Shinn and Vanka [28] performed the large eddy simulation on film cooling flows with VG and found that the flow mechanism of VG is to generate near-wall counter-rotating vortices, which is helpful in entraining coolant from the jet and transporting it to the wall. It is indicated in the previous works that VG shows an advantage of high film cooling effectiveness. However, there is a lack of detailed film cooling effectiveness distribution measurements in previous researches. Besides, as for the scope of authors, no one in their experimental studies has considered the effect of inclination angle on VG for enhancing the film cooling effectiveness. It’s of great importance to conduct more systematic and in-depth researches on VG concept. Therefore, in this work, a wind channel platform was built to study the impact of blowing ratio, as well as the inclination angle on VG to enhance the film cooling effectiveness. Three cases with VG at different inclination angles and other three cases without VG at different blowing ratios of M = 0.5, 1.0, 1.5, 18, were experimentally investigated in the current study. In

L. Song et al. / Applied Thermal Engineering 119 (2017) 573–584

the experiment, thermocouples were used to measure the film cooling effectiveness distribution of tested wall, and the technology of Particle Image Velocimetry (PIV) was employed to investigate the flow mechanism and capture the flow structure of X/ D = 1 and Y/D = 0 cross-section. In different experimental conditions, the Reynolds number of the mainstream was fixed at 11,000, and the density ratio was maintained at 0.85.

2. Experimental setup 2.1. Experimental apparatus and test section

2.2. Experimental method The blowing ratio M in the current study was defined as:

M¼

qf U f qm U m

This study focuses on the adiabatic film cooling effectiveness measurement, which was defined by the following equation:

g¼

Tm  Tw Tm  Tf

ð2Þ

where T m is the temperature of mainstream, T f is the temperature of secondary flow and T w is the local temperature on the tested wall. Besides, to have a better evaluation of the film cooling effectiveness, the spanwise averaged film cooling effectiveness was defined as:

gavg;spanwise ¼

The low-speed wind tunnel platform used to study the impact of vortex generator on the film cooling performance is shown in Fig. 1. The mainstream consisting of dry air was pumped into the wind tunnel by a blower (V max ¼ 2880 m3 =h). After flowing through the three-layer flow straightener, the mainstream entered into the test section. Meanwhile, the secondary flow supplied by the blower from the surrounding environment was also dry air, whose volume flow rate ranging from 0 to 50 m3/h was controlled by the valve to achieve the proper blowing ratio. Then the secondary air was heated by the electric heater (230 W). After entering into the rectangular plenum chamber, the secondary flow was injected into the test section through the embedded film cooling hole whose diameter was 20 mm. The test section consisted of a wide channel with a crosssection of 210 mm  210 mm and was 500 mm in length. A twolayer flat plate was installed in the bottom of the channel as shown in the top-right corner of Fig. 1. Three sides except tested plate wall were wrapped with high-transmittance spectralite for PIV measurement. To reduce the heat conduction existing on the tested wall, the tested plate was made up of a polyethylene (PE, k ¼ 0:04 W=ðm KÞ) plate and a polyurethane (PU, k ¼ 0:018 W=ðm KÞ) plate, both of them were 2 mm in thickness. The tested wall was assumed approximatively adiabatic for good heat insulation measures.

ð1Þ

where the inclination angle a is the angle between mainstream and secondary flow.

575

1 6D

Z

gdy

ð3Þ

The streamwise averaged film cooling effectiveness was defined as:

gavg;streamwise ¼

1 18D

Z

gdx

ð4Þ

And the area averaged film cooling effectiveness was defined as:

gavg;area ¼

1 108D2

ZZ

gdxdy

ð5Þ

The triangular pyramid vortex generator investigated is shown in Fig. 2. VG was located in the X/D = 1 downstream film cooling hole and shared the same geometric parameters in all tested cases. To investigate the impact of blowing ratio and inclination angle on the film cooling performance by using VG, three cases with VG and three cases without VG were tested. Six configurations in total were listed and numbered by the geometric difference for better description as shown in Table 1. Six different configurations with the inclination angle a from 20 to 40 were tested under the blowing ratio M ranging from 0.5 to 1.5. In the experiment, the high-precision K-type thermocouples point measurements were conducted to measure the temperature distribution of the tested wall. Measuring points with the diameter of 2 mm were arranged on the thick nether plate rather than on the thin upper plate, to avoid the leakage of the air flow. Besides, the holes were 20 mm in depth ensuring the adiabatic environment for the thermocouples. The detailed measuring points arrangement is shown in the top right corner of Fig. 2. The measuring points were distributed in the area with 1 6 X=D 6 19; 3 6 Y=D 6 3. To improve measurement efficiency, the measuring points were concentrated in the upstream zone where temperature changes severely and were sparse in the downstream zone where temperature changed gently. Considering the symmetry of geometry and temperature distribution, the distribution of bottom half plate was symmetrically imitated from the top half.

Fig. 1. The sketch of the wind tunnel platform.

576

L. Song et al. / Applied Thermal Engineering 119 (2017) 573–584

Fig. 2. Sketch of geometry and installation location of vortex generator.

Table 1 Operating condition parameters. Cases

a/deg

M

A20VG A30VG A40VG A20NVG A30NVG A40NVG

20 30 40 20 30 40

0.5, 0.5, 0.5, 0.5, 0.5, 0.5,

VG 1.0, 1.0, 1.0, 1.0, 1.0, 1.0,

1.5 1.5 1.5 1.5 1.5 1.5

Yes Yes Yes No No No

2.200

2.100

E/V

2.000

1.900

1.800

2.3. Experimental conditions and uncertainties analysis

1.700

1.600

In the experiment, the temperature data were all obtained by an 8-channel temperature collection module produced by DEWESOFT Company, whose sampling frequency was 200 kHz/ch at most, ensuring a high response. The streamline Constant Temperature Anemometer (CTA) was used to measure velocity profile and turbulence of mainstream. The CTA system could measure the velocity in a range from 0.5 m/s to 300 m/s with 1% uncertainty and has a high fluctuation frequency as much as 450 kHz to ensure the continuous signal output. Calibration for the velocity ranging from 1.0 m/s to 12 m/s was carried out, with an error of about 0.2840%. A related calibration curve is shown in Fig. 3. Temperature data were acquired when the experimental system turned into stable state, which was recognized by measuring the velocity profile of mainstream channel. 2D Particle Image Velocimetry (PIV) system for the flow characteristic measurements was constructed in the experimental system. Based on the theory that the luminosity of the every pixel in the picture represents the density of trace particle, PIV measurements were used to visualize the flow field, showing the mixing extent of mainstream and secondary flow. As shown in Fig. 4(a), PIV system consists of a double-pulsed YAG laser, a highresolution CCD camera, a Programmable Timing Unit (PTU) and a data collection system. The double-pulsed YAG laser could generate visible green laser sheet, whose wavelength was 532 nm. The pulse energy could reach to 200 mJ. The resolution of the highresolution CCD camera was 2048  2048, and its frame rate was 15fps. Used as the tracer particle with a diameter of 0.3–0.8 lm, smog was added in the secondary flow. In this experiment, for a better understanding of the flow mechanism of VG in enhancing the film cooling performance, the flow field in X/D = 3 and Y/D = 0 cross-sections of all tested conditions were visualized by the PIV system. The arrangement for Y/D = 0 cross-section measurement is shown in Fig. 4(b). For the X/D = 3 cross-section, the CCD camera could hardly remain perpendicular to the laser sheet. A method for separating the two laser proposed by Wang et al. [29] was used here, which markedly improved PIV system for the secondary flow field measurement.

2.0

4.0

6.0

8.0

10.0

12.0

U/m/s Fig. 3. The calibration curve of the streamline CTA.

(a)The main components of PIV system

The measurement holes at the upstream of the film cooling holes were drilled on the tested wall to measure the temperature of the mainstream with a K-type thermocouple. T f was in the same way measured in the hole drilled in the plenum chamber. Under our experimental conditions, T m was maintained at 22 °C and T f was held at 80 °C in steady condition. The density ratio DR was 0.85 acquired through the equation DR ¼ qf =qm . The velocity profile of mainstream and turbulence intensity measured by the hot

(b) PIV system arrangement for the Y/D=0 cross-section measurement

Fig. 4. PIV measurement system.

L. Song et al. / Applied Thermal Engineering 119 (2017) 573–584

1.1

tions. The Reynolds number calculated by the diameter of the film cooling hole was fixed at 11,000. To ensure the accuracy of the experiment, the classical measurement results carried out by Eckert et al. [30], Pederson et al. [31] and Schmidt et al. [32] were compared with the experimental results at M = 1.0 and the inclination angle a ¼ 30 , as shown in Fig. 6. It is demonstrated that the average streamwise film cooling effectiveness distribution agreed well with these classical results except for X/D = 15 by Schmidt, which is inferred to be caused by the incidental measuring error. The uncertainty of the film cooling effectiveness inferred from Eq. (1) was caused by the temperature measurements. After repeated measurements in the experiment condition at M = 1.0, a ¼ 20 for more than 20 times, it was concluded that the uncertainty of T m and T f was 0.5% and 2.7% respectively. The maximum uncertainty of T w was 5.7% at different locations on the wall. The error transfer formula was adopted:

0.12 0.10

1.0

U/Um

U/Um Tu

0.06

Tu

0.08 0.9

0.8 0.04 0.7

0.02

0.6 0.0

0.5

1.0

1.5

2.0

2.5

Z/D

3.0

3.5

0.00 4.0

Fig. 5. The velocity and turbulence intensity distribution of mainstream measured by the hot-wire anemometer (U m = 8.8 m/s).

anemometer is shown in Fig. 5. The velocity of mainstream was maintained at U m ¼ 8:8 m=s under different experimental condi-

0.50 Current Experiment Ecker[28] Pedersen[29] Schmidt[30]

ηavg,spanwise

0.40

577

s ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2  2  2 @ ln g @ ln g @ ln g dn ¼  dT 2m þ  dT 2f þ  dT 2w @T m @T f @T w

ð6Þ

where

@ ln g 1 1 ¼  Tm  Tw Tm  Tf @T m

ð7Þ

@ ln g 1 ¼ Tm  Tf @T f

ð8Þ

@ ln g 1 ¼ Tm  Tw @T w

ð9Þ

0.20

During the experiments, T m was maintained at 22 °C and T f was held at 80 °C in steady condition. The maximum T w was 52 °C. Analyzed from the Eq. (6), the maximum uncertainty of local film cooling effectiveness was 2.6%.

0.10

3. Results and discussion

0.30

3.1. Analysis of flow structure from PIV visualization 0.00 0.0

5.0

10.0

15.0

20.0

X/D Fig. 6. Comparison with the classical experimental results under M = 1.0 and inclination angle a = 30°.

Figs. 7 and 8 show the time-averaged flow field of the case A20NVG and A20VG in typical YZ cross-section and the middle XZ cross-section respectively at M = 1.5. As shown in Fig. 7(a), the representative CRVP vortical structure with the obvious upwash

1.5

Z/D

1.0

0.5

0.0 -1.0

-0.5

6HSDUDWLRQ $UHD

0.0

0.5

Y/D

(a) A20NVG

1.0

(b) A20VG

Fig. 7. Time-averaged flow field at X/D = 3 in the YZ cross-section at M = 1.5.

578

L. Song et al. / Applied Thermal Engineering 119 (2017) 573–584

(a) A20NVG

(b) A20VG Fig. 8. Time-averaged flow field in the middle XZ cross-section at M = 1.5.

effect on the secondary flow can be clearly observed. The existing separation area in 0 < X/D < 4 in Fig. 8(a) indicates that the secondary flow has lifted off the wall. As shown in Fig. 7(b), the CRVP is hardly visible while the anti-counter-rotating vortex pair (ACRVP) appears, generating the downwash effect and enhancing the lateral extension of film attachment. The flow field in Fig. 8 (b) demonstrates that the secondary flow is lower than that in the case A20NVG, and the disappearance of downstream separation area indicates a good attachment. However, the upstream separation area in 0 < X/D < 1 stays the same as that in the case A20NVG, showing that VG has little impact on the upstream flow field. Fig. 9 shows the flow field of nine configurations without VG in middle XZ cross-section at different blowing ratios. The separation area exists in nearly all the cases. The centerline film cooling performance could be briefly shown by the distribution of separation area. Three figures in the first row representing the case A20NVG show that at the same inclination angle a ¼ 20 , the secondary flow injects higher while the separation area gradually appears and extends as the blowing ratio increases. For A20NVG, the increase of M would bring a negative lift-off effect caused by augmenting momentum upwards. The variation tendency of the injec-

Fig. 9. Time-averaged flow field of configurations without VG at different blowing ratios in Y/D = 0 cross-section.

Fig. 10. Time-averaged flow field of configurations with VG at different blowing ratios in Y/D = 0 cross-section.

579

L. Song et al. / Applied Thermal Engineering 119 (2017) 573–584

tion height shows the same as M increases, which is reflected in the next two rows representing the cases of A30NVG and A40NVG. Attention is then paid to how the inclination angle influences the jet ejection at the same blowing ratio. At three different blowing ratios, the injection also becomes higher and the separation area adds as the inclination angle increases, which is evidently caused by the increase of momentum upwards. Besides, it is noted that the attachment of the secondary flow to the wall in the middle cross-section could only represent the centerline film cooling performance instead of the overall performance, so the spanwise distribution should be taken into consideration as well. Fig. 10 presents the effects of inclination angle and blowing ratio on the film cooling performance of the cases with VG. It demonstrates that the secondary flow ejection and the separation area distribution in X/D < 1 (the leading edge of VG) stays the same as those of the cases without VG at the same inclination angle and blowing ratio, indicating that VG has no impact on the upstream region. However, in the X/D > 3(the trailing edge of VG), the good attachment without separation area is observed on the wall and the secondary flow injects lower than that of in the cases without VG, due to the downwash effect of the ACRVP generated by VG. It’s noted that in the case A20VG, A30VG and A40VG at M = 0.5, the separation area appears on the top surface of VG in the midspan cross-section(Y/D = 0), showing the lack of film attachment. This is because the insufficient secondary flow could be totally entrained by the anti-counter-rotating vortex pair (ACRVP). Corresponding with Fig. 7(b), ACRVP entrains the secondary flow from the middle towards both sides, resulting in the lack of film attachment on the top surface of VG in the mid-span cross-section. However, when M = 1.5, the separation area appears in front of VG

instead of the top surface, because the upward momentum of the secondary flow is high enough to overcome the suppression of the mainstream. On the other hand, the flow rate of the secondary flow is sufficient beyond the entrainment capacity of the ACRVP, resulting in the adequate attachment on the top surface of VG in the mid-span cross-section. The obvious film attachment improvement in all cases with VG and the existing separation area at the low blowing ratio indicate that the ACRVP is strong enough to entrain the secondary flow corresponding with the Fig. 7(b). For the cases with VG, the increase of blowing ratio adds to the upward momentum of secondary flow deteriorating the lift-off tendency. On the other hand it contributes to the better film attachment on the wall. Therefore, the effect of blowing ratio on the film cooling performance should be taken into comprehensive consideration. As to the inclination angle, the cases with VG at M = 0.5, 1.0 have shown a similar flow field, indicating that the cases with VG are not sensitive to the inclination angle at the low blowing ratio because the ACRVP could totally entrain all the secondary flow towards the wall. But at M = 1.5, the secondary flow injects higher as the inclination angle increases, because the flow rate of the secondary flow exceeds the entrainment capacity of ACRVP and the effect of the upward momentum increment takes a leading role. 3.2. Film coverage contour Fig. 11 shows the film cooling effectiveness distributions on the adiabatic wall of A20NVG and A20VG at M = 0.5, 1.0, 1.5. It’s found from the experimental results of A20NVG that as the blowing ratio increases, the area of film coverage becomes narrower and the film

η 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35

10.0

X/D

15.0

-3.0 20.0

0.0

5.0

(a) A20NVG at M=0.5

10.0

15.0

X/D

(b) A20VG at M=0.5

0.0

5.0

10.0

X/D

15.0

3.0

Y/D

3.0

0.0

-3.0 20.0

0.0

0.0

5.0

(c) A20NVG at M=1.0

10.0

15.0

X/D

10.0

X/D

15.0

(e) A20NVG at M=1.5

3.0

Y/D

0.0

5.0

-3.0 20.0

(d) A20VG at M=1.0 3.0

0.0

-3.0 20.0

Y/D

5.0

-3.0 20.0

0.0

0.0

5.0

10.0

X/D

(f) A20VG at M=1.5

Fig. 11. Adiabatic wall film-cooling effectiveness distribution contour at inclination angle a ¼ 20 .

15.0

Y/D

0.0

0.0

Y/D

0.0

3.0

Y/D

3.0

-3.0 20.0

580

L. Song et al. / Applied Thermal Engineering 119 (2017) 573–584

cooling performance deteriorates. At M = 0.5, the film cooling performance is the best and gradually deteriorates along the mainstream direction for the aggravating mixing. When M becomes 1.0, as shown in Fig. 11(c), the centerline film cooling effectiveness increase at first and decrease then in 1 6 X=D 6 8 showing the reattachment of secondary flow. When M increases to 1.5, the lift-off tendency becomes stronger, resulting in the poor film cooling performance. However, for A20VG, the influence of blowing ratio on film cooling effectiveness distribution has changed a lot. At M = 0.5, compared with A20NVG, the overall cooling performance increases obviously, and significant lateral expansion appears while the centerline film cooling effectiveness decreases. Besides, the distribution characteristic has changed significantly. The best coolant coverage appears around the leading edge of VG while low film cooling effectiveness region appears around the trailing edge of VG, corresponding with the separation area in Fig. 10. It demonstrates that the mainstream is entrained into the ACRVP around the trailing edge of VG. When M increases to 1.0, the low effectiveness region disappears, and the overall film cooling performs better. But the film cooling effectiveness around the leading edge of VG decreases, because the injection with too high momentum in the Z direction could not cover the zone in front of VG. When M = 1.5, the film coverage obviously expands laterally, and the overall film cooling effectiveness is markedly enhanced. Fig. 12 shows the film cooling effectiveness distribution comparison of A30NVG and A30VG at different blowing ratios M from 0.5 to 1.5. Similar to A20NVG, the overall cooling performance of A30NVG decreases with the increase of the blowing ratio. Corre-

sponding with the case A20VG, the film cooling in the case A30VG performs better as the blowing ratio increases. The low effectiveness region still appears in A30VG at M = 0.5, proving that the mainstream entrainment happens around the trailing edge of VG. Fig. 13 shows the film cooling effectiveness distribution of A40NVG and A40VG at three different blowing ratios. The distributions of A40NVG shown in the first column demonstrate that the film cooling performance decays with the increase of blowing ratio. However, for A40VG, the film cooling performs the worst at M = 1.5, different from that in A20VG and A30VG. Because the combination of the large inclination angle and high blowing ratio causes that the flow rate of the secondary flow exceeds the entrainment capacity of the ACRVP. It is the strong upwards momentum that takes the leading role. Compared with A20VG and A30VG at M = 0.5, the low effectiveness zone disappears, and the film cooling effectiveness decreases along the mainstream direction, showing a disappearance of the mainstream entrainment. 3.3. Film cooling effectiveness To qualitatively describe the film cooling performance, the spanwise averaged film cooling effectiveness distributions are presented in Fig. 14. The film cooling effectiveness of A20VG, as shown in Fig. 14(a), is significantly higher than that of A20NVG at the all different blowing ratios. Especially at M = 1.5, A20VG performs the best and the spanwise averaged film cooling effectiveness

η 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35

10.0

X/D

15.0

-3.0 20.0

0.0

5.0

(a) A30NVG at M=0.5

10.0

X/D

15.0

(b) A30VG at M=0.5

0.0

5.0

10.0

X/D

15.0

3.0

Y/D

3.0

0.0

-3.0 20.0

0.0

0.0

5.0

(c) A30NVG at M=1.0

10.0

X/D

15.0

10.0

X/D

15.0

(e) A30NVG at M=1.5

3.0

Y/D

0.0

5.0

-3.0 20.0

(d) A30VG at M=1.0 3.0

0.0

-3.0 20.0

Y/D

5.0

-3.0 20.0

0.0

0.0

5.0

10.0

X/D

15.0

(f) A30VG at M=1.5

Fig. 12. Adiabatic wall film-cooling effectiveness distribution contour at inclination angle a ¼ 30 .

Y/D

0.0

0.0

Y/D

0.0

3.0

Y/D

3.0

-3.0 20.0

581

L. Song et al. / Applied Thermal Engineering 119 (2017) 573–584

η 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35

10.0

X/D

15.0

-3.0 20.0

0.0

5.0

(a) A40NVG at M=0.5

10.0

X/D

15.0

(b) A40VG at M=0.5

0.0

5.0

10.0

X/D

15.0

3.0

Y/D

3.0

0.0

-3.0 20.0

0.0

0.0

5.0

(c) A40NVG at M=1.0

10.0

X/D

15.0

10.0

X/D

15.0

3.0

Y/D

0.0

5.0

-3.0 20.0

(d) A40VG at M=1.0 3.0

0.0

-3.0 20.0

Y/D

5.0

-3.0 20.0

(e) A40NVG at M=1.5

0.0

0.0

5.0

10.0

X/D

15.0

Y/D

0.0

0.0

Y/D

0.0

3.0

Y/D

3.0

-3.0 20.0

(f) A40VG at M=1.5

Fig. 13. Adiabatic wall film-cooling effectiveness distribution contour at inclination angle a ¼ 40 .

improves at most 114% along the mainstream direction compared with A20NVG, showing the great advantage of VG in enhancing the film cooling performance. Besides, the spanwise averaged film cooling effectiveness in A20NVG drops along the mainstream direction. For A20VG, it first increases due to the downwash effect of the ACRVP and then decreases for the mixing of mainstream and secondary flow. It’s worth noting that in A20VG the obvious decline in the mainstream direction reflects that ACRVP is so strong that the mainstream and secondary injection mixes highly intensely. Fig. 14(b) shows the spanwise averaged film cooling effectiveness distribution of A30NVG and A30VG. The change rule of the spanwise averaged film cooling effectiveness is similar to Fig. 14 (a) and the overall cooling performance is inferior to Fig. 14(a). However, the effectiveness improvement of A30VG is more obvious than that of A30NVG, reaching up to 270% at some points. The impact of inclination angle on the effectiveness of the case without VG is more noticeable than the case with VG. The experimental results of cases with the inclination angle a ¼ 40 are shown in Fig. 14(c). It’s indicated that the effectiveness of A40NVG also drops along the mainstream direction and decreases with the increase of the blowing ratio. However, A40VG performs the best at M = 1.0 instead of at M = 1.5. When M = 0.5, the injection attaches well to the wall around the exit of film cooling hole, it performs better in the upstream region and decays along the mainstream direction. When M increases to 1.0, in the downstream of VG, the injection attachment improves and the spanwise film cooling effectiveness increases, owing to more

injection entrained into ACRVP. However, at M = 1.5, the film cooling performs the worst because the high momentum in the Z direction takes the leading role. To evaluate the lateral extension tendency, the streamwise averaged film cooling effectiveness distribution is shown in Fig. 15. The film coverage of the cases with VG is significantly wider than that of the cases without VG at the same blowing ratio and inclination angle, showing the lateral extension effect of ACRVP generated by VG. To briefly describe the lateral film distribution, the zone where gavg;streamwise > 0.1 is defined as film covered zone. Fig. 16 shows the averaged lateral extension width in six cases at M = 0.5, 1.0, 1.5. The increase of blowing ratio in the cases without VG brings about higher momentum in the Z direction and aggravates the lift-off tendency of secondary flow. Therefore, the film covered zone becomes narrower as the blowing ratio increases. The rule changes for the cases with VG. For A20VG and A30VG, the film covered zone becomes wider as the blowing ratio increases. Since the secondary flow could be fully entrained into ACRVP in the cases with VG, the more flow rate contributes to the wider lateral extension. But A40VG is the exception due to the large inclination angle which brings about the increase of upwards momentum. The area averaged film cooling effectiveness is applied to evaluate the overall performance presented in Fig. 17. As shown in Fig. 17(a), it’s revealed that VG significantly improves the film cooling performance, especially when M = 1.5, a significant improvement of the area averaged film cooling at most 249% is got in the case A40VG, compared with A40NVG.

582

L. Song et al. / Applied Thermal Engineering 119 (2017) 573–584

0.35

0.25 0.20 0.15 0.10

0.3

0.2

0.1

0.05 0.00 0.0

A20NVG M=0.5 A20NVG M=1.0 A20NVG M=1.5 A20VG M=0.5 A20VG M=1.0 A20VG M=1.5

0.4

ηavg,streamwise

0.30

ηavg,spanwise

0.5

A20NVG M=0.5 A20NVG M=1.0 A20NVG M=1.5 A20VG M=0.5 A20VG M=1.0 A20VG M=1.5

5.0

10.0

15.0

X/D

20.0

0.0 -4.0

-2.0

(a) α =20° 0.35

0.5 A30NVG M=0.5 A30NVG M=1.0 A30NVG M=1.5 A30VG M=0.5 A30VG M=1.0 A30VG M=1.5

0.4

0.20 0.15 0.10 0.05 0.00 0.0

4.0

°

ηavg,streamwise

ηavg,spanwise

0.25

2.0

(a) α =20 A30NVG M=0.5 A30NVG M=1.0 A30NVG M=1.5 A30VG M=0.5 A30VG M=1.0 A30VG M=1.5

0.30

0.0

Y/D

0.3

0.2

0.1 5.0

10.0

15.0

X/D

(b) α =30

20.0

0.0 -4.0

°

-2.0

0.0

2.0

Y/D

4.0

(b) α =30° 0.35

A40NVG M=0.5 A40NVG M=1.0 A40NVG M=1.5 A40VG M=0.5 A40VG M=1.0 A40VG M=1.5

0.30

0.20 0.15 0.10 0.05 0.00 0.0

A40NVG M=0.5 A40NVG M=1.0 A40NVG M=1.5 A40VG M=0.5 A40VG M=1.0 A40VG M=1.5

0.4

ηavg,streamwise

ηavg,spanwise

0.25

0.5

0.3

0.2

0.1 5.0

10.0

X/D

15.0

20.0

(c) α =40°

0.0 -4.0

-2.0

0.0

2.0

Y/D

4.0

(c) α =40

°

Fig. 14. Spanwise averaged film-cooling effectiveness distribution along the mainstream direction.

Fig. 15. Streamwise averaged film cooling effectiveness distribution along the lateral direction.

For the cases without VG, the impact of blowing ratio and inclination angle on film cooling performance is in agreement. As the blowing ratio or inclination angle increases, the injection tends to lift off and the area averaged film cooling effectiveness decreases in the linear tendency. The monotonous relations disappear in the cases with VG. With the increase of the blowing ratio, the area averaged film cooling

effectiveness in A20VG and A30VG increases, due to the strong downwash effect of ACRVP. While in A40VG, it first increases and then goes for the worse. That is because when M < 1, the secondary flow is totally entrained into the ACRVP and the film cooling effectiveness is enhanced, but when M > 1 the film cooling effectiveness

583

L. Song et al. / Applied Thermal Engineering 119 (2017) 573–584

l/Davg,streamwise

6.0

cooling effectiveness enhancement by using VG. The following conclusions can be drawn:

A20NVG A30NVG A40NVG A20VG A30VG A40VG

4.0

2.0

0.0 0.0

0.5

1.0

1.5

2.0

M Fig. 16. The averaged lateral extension width of six cases at M = 0.5, 1.0, 1.5.

decreases for the high the upward momentum of the secondary flow exceeding the entrainment capacity of ACRVP. As for the impact of inclination angle, when M = 0.5 or M = 1.0, the area averaged film cooling effectiveness stays the same as the inclination angle increases, in view of the experimental error. Though adversely at M = 1.5, the effectiveness gradually decreases as the inclination angle increases from a ¼ 20 to a ¼ 40 , the area averaged film cooling effectiveness only decreases 27.4%, while the decrease in the cases without VG reaches to 54.2%. The phenomenon reflects that the introduction of VG improves the robustness for the inclination angle in a wide range of blowing ratio. 4. Conclusions Detailed experimental studies on flow mechanism and thermal characteristics of film cooling with VG were carried out to investigate the impact of blowing ratio and inclination angle on the film

0.3

(1) Through the clear vortical structures visualized by PIV system, VG obviously generates the ACRVP with downwash and lateral extension effect, which significantly improves the film cooling performance in different conditions. Compared with the cases without VG, especially when a ¼ 20 , M = 1.5, the area averaged film cooling effectiveness improves as much as 248%. (2) For the cases A20VG and A30VG at M = 0.5, the mainstream can be entrained towards the wall because of the insufficient secondary flow, and low effectiveness area appears around the trailing edge of VG. However, for the case A40VG at M = 0.5, the low effectiveness area disappears, demonstrating that the ACRVP could not entrain mainstream in the case with the large inclination angle. (3) For the cases without VG, the film cooling effectiveness decreases as the blowing ratio increases. On the other hand, in the cases A20VG and A30VG, as the blowing ratio increases, the film cooling performs better, owing to the strong downwash effect of ACRVP. In the case A40VG, as the blowing ratio increases, when M < 1, the secondary flow is totally entrained into the ACRVP and the film cooling effectiveness is enhanced. However, when M > 1, the film cooling effectiveness decreases, because in the case A40VG with large inclination angle and high blowing ratio, the upward momentum of the secondary flow was so high that its flow rate exceeded the entrainment capacity of ACRVP, resulting in a lifting-off tendency of the secondary flow on the wall. (4) As the inclination angle increases, the film cooling effectiveness of the cases without VG decreases. For the cases with VG, when M = 0.5, 1.0, the film cooling effectiveness stays almost the same. Though the film cooling effectiveness decreases at M = 1.5, the area averaged film cooling effectiveness only decreases 27.4% when the inclination angle changes from a ¼ 20 to a ¼ 40 . A comparison is made between the cases without VG, a 54.2% decrease is observed.

0.3

A20NVG A30NVG A40NVG A20VG A30VG A40VG

0.2

M=0.5 M=1.0 M=1.5 M=0.5 M=1.0 M=1.5

ηavg,area

ηavg,area

0.2

NVG NVG NVG VG VG VG

0.1

0.0 0.0

0.1

0.5

1.0

M

1.5

(a)The impact of blowing ratio

2.0

0.0 10

20

30

α

40

(b)The impact of inclination angle

Fig. 17. Area averaged film cooling effectiveness distribution.

50

584

L. Song et al. / Applied Thermal Engineering 119 (2017) 573–584

It means that the application of VG improves the robustness of high film cooling effectiveness, enabling the high film cooling performance in a wide range of inclination angle.

[16] [17]

Acknowledgement [18]

This work was supported by the National Natural Science Foundation of China (Grant No. 51676149). [19]

References [1] X. Tan, J. Zhang, S. Yong, G. Xie, An experimental investigation on comparison of synthetic and continuous jets impingement heat transfer, Int. J. Heat Mass Transf. 90 (2015) 227–238, http://dx.doi.org/10.1016/j. ijheatmasstransfer.2015.06.065. [2] G. Xie, X. Liu, H. Yan, Film cooling performance and flow characteristics of internal cooling channels with continuous/truncated ribs, Int. J. Heat Mass Transf. 105 (2017) 67–75. [3] R.J. Goldstein, Film cooling, Adv. Heat Transf. 7 (1971) 321–379. [4] D.G. Bogard, K.A. Thole, Gas turbine film cooling, J. Propuls. Power. 22 (2006) 249–270. [5] J.-C. Han, S. Dutta, S. Ekkad, Gas Turbine Heat Transfer and Cooling Technology, CRC Press, 2012. [6] T.F. Fric, A. Roshko, Vortical structure in the wake of a transverse jet, J. Fluid Mech. 279 (1994) 1–47. [7] J.H. Leylek, R.D. Zerkle, Discrete-jet film cooling: a comparison of computational results with experiments, in: ASME 1993 Int. Gas Turbine Aeroengine Congr. Expo., American Society of Mechanical Engineers, 1993, pp. V03AT15A058–V03AT15A058. [8] M. Tyagi, S. Acharya, Large Eddy simulation of film cooling flow from an inclined cylindrical jet, J. Turbomach. 125 (2003) 734, http://dx.doi.org/ 10.1115/1.1625397. [9] R.M. Kelso, T.T. Lim, A.E. Perry, An experimental study of round jets in crossflow, J. Fluid Mech. 306 (1996) 111–144. [10] J.N. Blanchard, Y. Brunet, A. Merlen, Influence of a counter-rotating vortex pair on the stability of a jet in a cross flow: an experimental study by flow visualizations, Exp. Fluids 26 (1999) 63–74. [11] R.S. Bunker, A review of shaped hole turbine film-cooling technology, J. Heat Transf. 127 (2005) 441–453. [12] B.A. Haven, D.K. Yamagata, M. Kurosaka, S. Yamawaki, T. Maya, Anti-kidney pair of vortices in shaped holes and their influence on film cooling effectiveness, in: ASME 1997 Int. Gas Turbine Aeroengine Congr. Exhib., American Society of Mechanical Engineers, 1997, pp. V003T09A007– V003T09A007. [13] C. Heneka, A. Schulz, H.-J. Bauer, A. Heselhaus, M.E. Crawford, Film cooling performance of sharp edged diffuser holes with lateral inclination, J. Turbomach. 134 (2012) 41015. [14] K.-D. Lee, K.-Y. Kim, Performance evaluation of a novel film-cooling hole, J. Heat Transf. 134 (2012) 101702. [15] B. Kröss, M. Pfitzner, Numerical and experimental investigation of the film cooling effectiveness and temperature fields behind a novel trench

[20]

[21]

[22] [23]

[24]

[25]

[26]

[27]

[28] [29]

[30] [31]

[32]

configuration at high blowing ratio, in: ASME Turbo Expo 2012 Turbine Tech. Conf. Expo., American Society of Mechanical Engineers, 2012, pp. 1197– 1208. J.D. Heidmann, S. Ekkad, A novel antivortex turbine film-cooling hole concept, J. Turbomach. 130 (2008) 31020, http://dx.doi.org/10.1115/1.2777194. M.J. Ely, B.A. Jubran, A numerical study on improving large angle film cooling performance through the use of sister holes, Numer. Heat Transf. Part – Appl. 55 (2009) 634–653, http://dx.doi.org/10.1080/10407780902821532. H. Wu, H. Cheng, Y. Li, C. Rong, S. Ding, Effects of side hole position and blowing ratio on sister hole film cooling performance in a flat plate, Appl. Therm. Eng. 93 (2016) 718–730, http://dx.doi.org/10.1016/j. applthermaleng.2015.09.118. K. Kusterer, D. Bohn, T. Sugimoto, R. Tanaka, Double-jet ejection of cooling air for improved film cooling, J. Turbomach. 129 (2007) 809, http://dx.doi.org/ 10.1115/1.2720508. K. Kusterer, A. Elyas, D. Bohn, T. Sugimoto, R. Tanaka, M. Kazari, The Nekomimi cooling technology: cooling holes with ears for high-efficient film cooling, in: ASME 2011 Turbo Expo Turbine Tech. Conf. Expo., American Society of Mechanical Engineers, 2011, pp. 303–313. S. Na, T.I.-P. Shih, Increasing adiabatic film-cooling effectiveness by using an upstream ramp, J. Heat Transf. 129 (2006) 464–471, http://dx.doi.org/10.1115/ 1.2709965. E. Sakai, T. Takahashi, Y. Agata, Experimental study on effects of internal ribs and rear bumps on film cooling effectiveness, J. Turbomach. 135 (2013) 31025. K. Funazaki, R. Nakata, H. Kawabata, H. Tagawa, Y. Horiuchi, Improvement of flat-plate film cooling performance by double flow control devices: part I— investigations on capability of a base-type device, in: ASME Turbo Expo 2014 Turbine Tech. Conf. Expo., American Society of Mechanical Engineers, 2014, pp. V05BT13A023–V05BT13A023. H. Kawabata, K. Funazaki, Y. Suzuki, H. Tagawa, Y. Horiuchi, Improvement of turbine vane film cooling performance by double flow-control devices, J. Turbomach. 138 (2016) 111005. H. Babinsky, Y. Li, C.W. Pitt Ford, Microramp control of supersonic oblique shock-wave/boundary-layer interactions, AIAA J. 47 (2009) 668–675, http:// dx.doi.org/10.2514/1.3802. D.L. Rigby, J.D. Heidmann, Improved film cooling effectiveness by placing a vortex generator downstream of each hole, in: ASME Turbo Expo 2008 Power Land Sea Air, American Society of Mechanical Engineers, 2008, pp. 1161–1174. K.B.M.Q. Zaman, D.L. Rigby, J.D. Heidmann, Inclined jet in crossflow interacting with a vortex generator, J. Propuls. Power 26 (2010) 947–954, http://dx.doi. org/10.2514/1.49742. A.F. Shinn, S.P. Vanka, Large eddy simulations of film-cooling flows with a micro-ramp vortex generator, J. Turbomach. 135 (2013) 11004. S.S. Wang, H. Sun, J. Di, J.R. Mao, J. Li, L.X. Cai, High-resolution measurement and analysis of the transient secondary flow field in a turbine cascade, Proc. Inst. Mech. Eng. Part J. Power Energy 228 (2014) 799–812, http://dx.doi.org/ 10.1177/095765091453967. E.R.G. Eckert, V.L. Eriksen, R.J. Goldstein, J.W. Ramsey, Film Cooling Following Injection Through Inclined Circular Tubes, 1970. D.R. Pedersen, E.R.G. Eckert, R.J. Goldstein, Film cooling with large density differences between the mainstream and the secondary fluid measured by the heat-mass transfer analogy, J. Heat Transf. 99 (1977) 620–627. D.L. Schmidt, B. Sen, D.G. Bogard, Film cooling with compound angle holes: adiabatic effectiveness, J. Turbomach. 118 (1996) 807–813.