Advances in Mechanical Engineering-Experimental Visualization of the Flow Structure for Jet in...

8/9/2019 Advances in Mechanical Engineering-Experimental Visualization of the Flow Structure for Jet in Crossflow With a Cur…

1/12

Hindawi Publishing CorporationAdvances in Mechanical EngineeringVolume 2012, Article ID 950452, 12 pagesdoi:10.1155/2012/950452

Research ArticleExperimental Visualization of the Flow Structure forJet in Crossflow with a CurvedHole Passage

JunYu Liang and ShunKang

Key Laboratory of Condition Monitor and Control for Power Plant Equipments, Power and Mechanical Engineering,School of Energy, North China Electric Power University, Beijing 102206, China

Correspondence should be addressed to Shun Kang, [email protected]

Received 4 July 2012; Accepted 19 September 2012

Academic Editor: C. T. Nguyen

Copyright © 2012 J. Y. Liang and S. Kang. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

The objective of this paper is to investigate the influence of a hole curvature on the flow structure and characteristics downstreamof JICF (jet in cross-Flow) by means of smoke visualization and particle image velocimetry (PIV). The experiment was performedin a low speed wind tunnel with Reynolds numbers of about 480 and 1000, based on the hole diameter and main flow speed. Twogeometries were tested: a circular hole with 90◦ curvature and a circular straight hole for comparison, under blowing ratios 0.5and 1.0. The measurements were done in the symmetric plane and four cross-sections. The results show that the curved hole coulddecrease the mixing behavior of jet flow with the main flow as the hole leading edge also increases the chance of transportingthecoolant to the wall surface and the transverse coverage. The curved hole shows a high potential to increase the cooling eff ectivenessonce it is applied to the turbine blades.

1. Introduction

In order to raise the cycle efficiency, today’s gas turbineis heading towards high pressure ratio and high turbineinlet temperature. Consequently, thermal and mechanicalloads of the turbine components exposed to the hot gas willbe increased, leading to the necessity of applying efficientcooling techniques in order to guarantee the lifetimes. One

of the often used cooling techniques is film cooling, in whichcoolant air extracted from the compressor is transferred intothe cava of turbine blades and then ejected out throughdiscrete holes or slots, arranged in a certain way, aroundthe blades into the blade flow passage. The air forms a thinand low temperature film covering the blade and/or endwallsurfaces for protecting them from the hot main flow. Sincethe extraction of air from the compressor incurs a penalty to the thermal efficiency and the ejection of coolant air intothe blade passage and its mixing with the hot gas as well,introduce additional energy losses, it is therefore necessary to get more insight into the local flow field and then find away to optimize the cooling hole design.

The essential features of such film-cooling flow arepresent in a more generic flow situation of JICF. Extensiveinvestigations on this fundamental flow field by experimentaland numerical methods have been done for many years[1–7]. Margason [8] and many others had given a detailedsummary on the earlier studies. A detailed review on therecent progress in the study of JICF can be found fromKaragozian [9].

Interaction of the jet flow with the main cross-flow creates a localized, very complex large-scale vortex structurenear a jet hole, as shown in Figures 1 and 2 from Fric andRoshko [6] and New et al. [10]. It is known that the near-field entrainment in the vicinity of the jet exit is influencedby these vortices [11], while the far-field entrainment isdominated by the CVP system [12].

In film cooling applications, one of the targets in JICFinvestigation is reducing the coolant jet penetration andmaking it as close adherence to the wall over a long distanceas possible, which is of crucial importance for achievinghigher eff ectiveness of film cooling. As the rotation sense of a CVP is such that it lifts the jet flow off the blade surface,


2/12

2 Advances in Mechanical Engineering

C r o s s - fl o w

Wall

Wake vortices

Counter-rotatingvortex pair

Jet shear-layervortices

H or se shoe v or t ic e s

r-layerces

v or t ic e s

Figure 1: Sketch of four types of vortices associated with thetransverse jet near field (Fric and Roshko [6]).

Leading-edge vortices

Cross-flow

Jet

Wake vortices

Counter-rotatingvortex pair

Lee-sidevortices

Horseshoe vortex system

e vortices

Jet

Wake vortices

Counter-rotativortex pair

Lee-sidevortices

Horseshoe vortex sy

Figure 2: Sketch of large-scale vortical structures nearby a coolinghole with the shaded region of the cross-section of the jet cuttingalong the jet centerline in the main flow direction (New et al. [ 10]).

a certain research work has been done to weaken the CVP.Haven and Kurosaka [13] introduced a vortex pair insidethe jet hole passage, by installing two microvanes withinthe jet hole, which creates a vortex pair with a rotationsense opposite to the kidney vortices. Peterson and Plesniak [14] reported that interaction of in-hole vorticity createdby turning of plenum flow direction could weaken theCVP, resulting in a lower trajectory and increased spanwisespreading. Kang [15] proposed another way to weaken thestrength and scale of CVP by shaping the jet hole to createa secondary passage vortex pair (PVP) inside the jet hole.

The PVP, rotating in the opposite sense to CVP, would beformed due to the centrifugal force acting on the low speedflow inside the boundary layer of hole a surface, as sketchedin Figure 3. Stregth and scale of PVP will be increased withthe hole curvature and boundary layer thickness. The PVPin curved channels was first observed by Dean, so that it isalso named as Dean vortex pair characterized by the Deannumber [16]. Liang et al. [17–19] made a further numericalinvestigation in a curved hole with a square and circularcross-section by RANS and DES methods and confirmed thatthe curved hole weakens the counter-rotating vortex pair.However, there are still issues which remained unclear inthe detailed mechanisms, which request further experiment

and numerical work. Hence, this paper is to investigatethe influences of hole curvature on the characteristics andstructure of the vortex flow downstream in the JICF witha curved hole by means of PIV apparatus. The film coolingapplication will be taken as a background in the analysis anddiscussions. In addition, the experimental results may form

an additional reference and database for CFD validation.

2. Test Facility andExperimental Procedure

2.1. Test Rig. The experiments were performed in a low speed open-loop wind tunnel as sketched in Figure 4. Themain flow is supplied by a centrifugal fan with a capacity of maximum total pressure rise of 2200 kPa. The test sectionis 1600 mm long with a cross-section of 700 × 400mm2

(height × width). A PT100 thermal resistance and a pitottube are located at 500 mm downstream of the test sectioninlet to measure the main flow conditions. The free streamturbulence was measured with HWFA (Hot-Wire/FilmAnemometer). Sampling frequency is about 2 kHz. A flatplate with jet hole is placed at half height of the tunnel withits leading edge at 950 mm from the test section inlet andextended to the tunnel sides in the transverse direction and tothe tunnel exit in the streamwise direction. Air with particlesgoes throughthe flowmeter and enters a plenum, with a crosssection of 80 mm × 55 mm, placed inside the plate with amaximum thickness of 85 mm, as sketched in Figure 5. Theplate is comprised of a rectifier cone at the leading edge,an injection plate, made of wood. Center of the hole exitis located at 393 mm downstream of the leading edge. Acamera is placed on one side of the wind tunnel to measurethe symmetric plane. For measuring the cross sections, thecamera is placed at downstream of the wind tunnel outlet,

with a distance from the hole center of about 550 mm.The light source is supplied by Nd:YAG laser from BIG

SKY with the highest repetition rate of 20 Hz. The type of CCD camera is POWERVIEW Plus 4 MP from TSI withresolution of 2048 × 2048 pixel. The photo data captured by camera will be transported through a cable to PC. In orderto get the time averaged results, 50–100 pairs of two-framephoto for one case were captured to quantify the flow field.

2.2. Hole Configuration. Two types of hole configurations,straight and curved with 90◦ curvature, as sketched inFigure 6, were investigated. Hole diameter D is 6 mm. Thelength of straight hole L = 3.33D. Direction of the exit flow

of both geometries is upwards and normal to the main flow.

2.3. Test Settings. The smoke was only seeded for jet fluidsto visualize the jet; while in PIV measurements, the smokewas seeded both for jet and main flow fluids. Because of thelimitation of the sampling frequency of the PIV apparatusused and safety of the camera placed at downstream of thewind tunnel outlet, the main flow velocity is set to 1.2 m/sand 2.5 m/s, which corresponds to Reynolds number ReD,based on the hole diameter and mainstream velocity, of about 480 and 1000. Velocity profile over the flat plate wasmeasured at 393 mm downstream of the plate leading edge.Blowing ratios (BR), ranged from 0.5 and 1.5, were tested


3/12


4/12


Cross-flow Cross-flow

Straight hole Straight hole

Straight hole Straight hole

(a) ReD = 480 BR = 0.5

(b) ReD = 480 BR = 1

(c) ReD = 1000 BR = 0.5

(d) ReD = 1000 BR = 1

Figure 7: Instantaneous smoke visualizations in symmetric plane for straight hole configurations under BR = 0.5 and 1 and ReD =

480,1000.

Table 1: Operating conditions.

ReD De BR Tu δ (mm)

Straight hole480

——

0.51

0.24% 14

1000 —

—0.51

0.4% 12

Curved hole480

99198

0.51

0.24% 14

1000 207

4140.51

0.4% 12

downstream of the exit in the main flow direction markedwith an arrow and 3D in the normal direction. It can beclearly seen first from these pictures that in any of theconditions, the flow leaves the jet exit upwards to a certaindistance from the exit without visible vortex motion andthen turns towards the mainstream direction. Afterwards,the smoke trace is oscillating due to the Kelvin-Helmholtzinstability of the annular shear layer which separates fromthe edge of the jet flow, resulting in a series of jet shear-layervortices in the initial portion of the jet flow trajectory and awake vortex system under the trajectory, which is similar to

what reported by Fric and Roshko [6]. It is seen that is heightof the trajectory reduces with increasing Reynolds numberand decreasing blowing ratio.

The shear-layer vortices viewed in the symmetric planecorresponds to the cut-off view of the hairpin vortices asshown in Figure 9 which presents a snapshot of the coherentstructure of Q isosurface obtained with the DES method by Liang and Kang [19] under blowing ratio 0.5 and Reynoldsnumber 4000 for straight and 90◦ curved holes. The Q isdefined as

Q =1

2

Ωi jΩi j − Si jSi j

. (1)

It is seen from Figure 9(a) that when a hairpin vortex iswell formed and shaded downstream, a next hairpin vortex starts to be formed in series. These results may suggest

that the development of hairpin vortices is the underlying

mechanism of coolant liftup. For the curved hole as shownin Figure 9(b), the PVP (or Dean vortex pair) issued fromthe curved hole exit could be clearly observed and it has agreat influence on the formation and evolution of the hairpin

vortices. The hairpin downstream was skewed and weakeneddue to the interaction of PVP with shear layer vortices, so the

liftup of jet fluids would be suppressed and then increase the

cooling eff ectiveness.

A horseshoe vortex is seen upstream of the jet inFigure 9 which could not be resolved by the experimentalvisualization. Streamwise vorticity component of the hairpinvortices will form the CVP which is evolving in a periodic

unsteady manner with intermittent vortex structures asviewed from DES results. The wake vortices in Figures 7

and 8 extend upright from the wall towards the trajectory,

which could be the normal vorticity component of the

hairpin vortices. The wake vortices allow fluid to be drawn

from the boundary layer into the jet itself, with an efficient

entrainment of the boundary layer fluid into the downstream


5/12

Advances in Mechanical Engineering 5

Cross-flow Cross-flow

(a) ReD = 480 BR = 0.5

(b) ReD = 480 BR = 1

(c) ReD = 1000 BR = 0.5

(d) ReD = 1000 BR = 1

Curved holeCurved hole

Curved hole Curved hole

Small vortex scale

Greater adherence to wall

Figure 8: Instantaneous smoke visualizations in symmetric plane for curved hole configurations under BR = 0.5 and 1 and ReD = 480,1000.

X

Y

Z

F l o w

105 526 947 1368 1789

H e a d

L e g

ωx −2000 −1579 −1158 −737 −316

(a) Straight

X

Y

Z

F l o w

ωx 105 526 947 1368 1789−2000 −1579 −1158 −737 −316

PVP

(b) Curved

Figure 9: Instantaneous Q iso-surface for straight and 90◦ curved holes, BR = 0.5 and ReD = 4000 (Haller) [20].

and the wake region of the jet, as indicated by Fric andRoshko [6].

The shear-layer vortices with coherent counter-rotatingvortex pairs, named as wind-ward vortices and lee-wardvortices observed by New et al. [10] along the trajectory for blowing ratios from 2.3 to 5.8 for straight hole couldnot be observed from Figures 7 and 8 for blowing ratios0.5 and 1.0. However, this vortex structure could be clearly

observed under higher blowing ratio in this study asshown in Figure 10 which presents an instantaneous smokevisualization in the symmetric plane for a straight hole underblowing ratio 1.3 and Reynolds number 240. In Figure 10 achain of vortex pairs with coherent counter-rotating wind-ward vortices and lee-ward vortices is observed.

It was believed that the curved hole could weakenthe streamwise vorticity due to the secondary vortex pair,generated by the hole curvature, which is rotating in theopposite sense with CVP. As one can observe by comparingthe height of the trajectory from the wall is reduced by thecurved hole configuration under all the blowing ratios andReynolds numbers. The number of the hairpin vortex under

Re = 240 BR = 1.3

Wind-ward vortex

Lee-ward vortex

Wind-ward and lee-ward vortex

are coherent

Figure 10: Instantaneous smoke visualizations of straight hole insymmetric plane under blowing ratio of 1.3.

the same flow conditions is reduced by the curved hole, suchas from 7 to 4 by comparing Figures 7(a) and 8(a). Eventhe less visible vortex structure could be observed at thehigher Reynolds number condition in Figures 8(c) and 8(d).The jet flow breaks up in a short distance into small scale


6/12


ReD = 1000 BR = 0.5

Straight hole

Curved hole

Cross flow

4

3

2

1

Y / D

−2 −1 0 1 2 3 4 5 6

X/D

4

3

2

1

Y / D

−2 −1 0 1 2 3 4 5 6

X/D

(a)

ReD = 1000 BR = 1

Straight hole

Curved hole

Cross flow

4

3

2

1

Y / D

−2 −1 0 1 2 3 4 5 6

X/D

4

3

2

1

Y / D

−2 −1 0 1 2 3 4 5 6

X/D

(b)

Figure 11: Time averaged smoke visualizations in symmetric plane, ReD = 1000 and BR = 0.5,1.

vortices. These behaviors could enhance the transportationof jet fluids towards the wall surface and lead to greateradherence.

Besides, care should be taken in concluding the vortex structure presented in Figures 7 to 10. Since they areinstantaneous visualizations, the structure observed may be strongly associated with the time when a snapshot istaken. Hence, discussions on the time-averaged parameterare necessary which are given below.

3.1.2. Time-Averaged Presentation. All the time-averaged qu-antities to be presented hereafter are obtained over 50 to 100sequential shots, with the sampling frequency 7.24 Hz, of thePIV measurement results. Figures 11–13 show time-averagedsmoke photography, contours of the time averaged velocity magnitude V mag and vorticity ω y , in the symmetric plane forboth holes under blowing ratios of 0.5 and 1.0 for Reynolds

number 1000. It is seen from Figure 11 that, instead of the obviously observed vortices in the instantaneous smokevisualizations (Figures 7 and 8), jet smoke fluid in whiteis diff used along the main flow direction without visiblevortices. A dark area close to the wall downstream of the jet exit is clearly observed in the pictures for straight holecases but not for the curved hole cases. This indicates thatthe jet fluids issued from the curved hole could well adhereto the wall surface and could definitely improve the coolingeff ectiveness. It is seen from Figure 12, velocity contour withvectors, that the coming boundary layer flow is disturbed by the jet flow. Thickness of the boundary layer downstream thehole exit increases with the mainstream due to the motion

of CVP and the blowing ratio. Compared to the straight holecases, the thickness is reduced in the curved hole cases due tothe passage vortex pair, generated in the curved hole passage,with an opposite rotation sense to the CVP. Location of maximum velocity near the hole exit deviates from the holecenter towards the main flow direction due to the shearingaction of cross-flow. It can be observed that the maximumvelocity near the jet exit is increased with blowing ratio andreduced in the curved hole case, which can also be seen fromFigure 13. Figure 13 shows the normal velocity componentV z profile, extracted from Figure 12, at the location of Y/D =0.3 above the hole exit. It can be seen additionally thatthe location of the maximum velocity is hardly aff ectedby blowing ratio and is closer to the hole trailing edge at X/D = 0.4 in the curved cases, instead of X/D = 0.2 in thestraight cases. Haven and Kurosaka [13] studied the eff ectsof jet velocity profiles in top-hat and parabolic shapes under

blowing ratios from 2.3 to 5.8 by using laser-induced fluo-rescence and digital particle-image velocimetry techniques.They found that the thicker shear layer associated with theparabolic profile is able to delay the formation of leading-edge and lee-side vortices, compared to the top-hat profilewith thinner shear layer under the corresponding blowingratio. As a result, there is an increase in jet penetration and areduction in the near-field entrainment of cross-flow fluid by the parabolic profile. In the current study, the exit profile ismore three-dimensional, especially for the curved hole cases.The thinner boundary thickness at the lee side of jet flow willdefinitely decrease the jet penetration and consequently thelower jet trajectory, as viewed from Figures 7, 8, and 11.


7/12


0 0.11 0.22 0.33 0.44 0.55 0.66 0.77 0.88 0.97

ReD = 1000 BR = 0.5

Straight hole

Curved hole

4

3

2

1

0

0

−2 −1 0 1 2 3 4 5 6

X/D

−2 −1 0 1 2 3 4 5 6

X/D

Z / D

4

3

2

1

Z / D

V mag /V ∞

(a)

ReD = 1000 BR = 1

Straight hole

Curved hole

0 0.16 0.31 0.47 0.62 0.78 0.93 1.09 1.24 1.4

−2 −1 0 1 2 3 4 5 6

X/D

−2 −1 0 1 2 3 4 5 6

X/D

4

3

2

1

Z / D

4

3

2

1

0

0

Z / D

V mag /V ∞

(b)

Figure 12: Time averaged velocity in symmetric plane for both holes at ReD = 1000 and BR = 0.5 and 1.0.

2

1.5

1

0.5

0

−0.5

V z / V ∞

−0.5 −0.25 0 0.25 0.5

X/D

Curved BR = 0.5Curved BR = 1

Straight BR = 0.5Straight BR = 1

ReD = 1000

Figure 13: Profiles of normal velocity component V z near hole exitcenter in symmetric plane, Y/D = 0.3.

Figure 14 shows the contours of time averaged vorticity ω y in the symmetric plane, with positive vorticity turningin the counter clockwise. It is seen that vorticity near thewall for both holes is negative, associated with the boundary layer, except for a small range, a short distance downstreamof the hole exit under blowing ratio of 0.5. Along the jet

trajectory, vorticity is positive at the wind-ward side andnegative at the lee-ward side. The value of the vorticity

increases with blowing ratio for both straight and curvedholes and reduces from the straight hole to the curvedhole for the same blowing ratio, which is consistent withthe velocity profiles in Figure 14. It can be seen that thewind-ward side vorticity cannot be maintained as far as thelee-ward vorticity towards downstream, which is consistentwith the instantaneous smoke visualization observations inFigures 7 and 8 in which only vortices in clockwise couldbe clearly visible along the trajectory. Zones, with positivevorticity, located between the wall and the jet trajectory may indicate the wake vortisity.

3.2. Cross-Sections. In order to further understand the jet

flow structure in its cross sections, the smoke is again seededonly in the jet fluid and four cross sections X/D = 0,1,2,4were visualized and measured for both hole configurationsunder the studied Reynolds numbers and blowing ratios.

3.2.1. Instantaneous Visualization. Figures 15 and 16 presentthe instantaneous visualization pictures of the jet flow in thefour cross sections for straight and curved holes, respectively,at Reynolds number ReD = 480 and blowing ratios of 0.5 and 1.0. Scale of the space occupied by each of thepictures and their resolution are the same. It is seen thatthe jet flow is expanding in both vertical and transversedirections with distance towards downstream from X/D = 0


8/12


26 216 405

ReD = 1000 BR = 0.5ω y

Straight hole

Curved hole

−1300 −1111 −921 −542−732 −353 −163

4

3

2

1

0

Z / D

4

3

2

1

0

Z / D

−2 −1 0 1 2 3 4 5 6

X/D

−2 −1 0 1 2 3 4 5 6

X/D

(a)

ω y

Straight hole

Curved hole

4

3

2

1

0

Z / D

4

3

2

1

0

Z / D

−1700 −1426 −1153 −879 −605 −332 −58 216 428 763

ReD = 1000 BR = 1

−2 −1 0 1 2 3 4 5 6

X/D

−2 −1 0 1 2 3 4 5 6

X/D

(b)

Figure 14: Time averaged vorticity ω y in symmetric plane, ReD = 1000 and BR = 0.5,1.

X/D = 0 X/D = 1

X/D = 2 X/D = 4

CVP

(a) BR =

0.5

X/D = 0 X/D = 1

X/D = 2 X/D = 4

(b) BR =

1.0

Figure 15: Instantaneous smoke visualizations of jet flow in four streamwise sections for straight hole at Re D = 480, (a) BR = 0.5, and (b)BR = 1.0.

(hole exit center) to X/D = 4 for both straight and curvedholes. With increasing blowing ratio, that is, more coolantfluids are injected, the expanding area becomes larger andshifts farther away from the wall, especially for the straighthole case (Figure 15). The expanded area is larger for thecurved hole than that for straight hole in each of the sections,which lead to a favorable eff ect on the cooling eff ectiveness

by the curved hole. It is further seen from Figure 15 thatthe “kidney shaped” vortex structure is well observed in thecross sections of jet flow, which has been reported by many researchers, such as [4–6]. In the current visualizations, itis found that the structure of the jet flow from the exit till X/D = 1 is quite stable and symmetric at all the studied flow conditions for a straight hole. In section X/D = 2, the jet flow


9/12


X/D = 0 X/D = 1

X/D = 2 X/D = 4

PVP

PVP

Jet bifurcation

CVP

(a) BR = 0.5

X/D = 0 X/D = 1

X/D = 2 X/D = 4

(b) BR = 1.0

Figure 16: Instantaneous smoke visualizations of jet flow in four streamwise sections for curved hole at ReD = 480, (a) BR = 0.5, and (b)BR = 1.0.

X/D = 2

X/D = 1 X/D = 0

X/D = 4

ReD = 1000 M = 0.5

(a) Straight

X/D = 2

X/D = 1 X/D = 0

X/D = 4

ReD = 1000 M = 0.5

(b) Curved

Figure 17: Instantaneous smoke visualizations of jet flow in four streamwise sections at Re D = 1000 and BR = 0.5 for (a) straight hole and(b) curved hole.

oscillates in both of the normal and transverse directions andis very unstable in the section of X/D = 4.

It is observed from Figure 16 (curved hole) that thestructure of the jet flow is totally diff erent from that in

Figure 15 (straight hole) and exhibits completely diff erentflow characteristics, caused from the passage vortex pair(PVP) generated by the curved hole passage. Interactionbetween PVP and CVP leads to form a new vortex structure,as observed, which is distinct from the kidney shaped vortex structure in the straight hole cases. At section X/D = 0 (exitcenter), the jet fluids bifurcate in transverse direction sincethe PVP could provide a stronger transverse momentum. Atthe downstream sections X/D = 1, 2, 4, the PVP rotating inthe opposite sense to the CVP can be clearly observed andlocated above the CVP. The PVP and CVP are seen to becoherent in all the sections, except for X/D = 4 under BR =1.0 in which the vortex structure shows high turbulence.

With increasing Reynolds number, the instantaneoussmoke visualization pictures show highly unstable and lesswell-organized vortex structure for all the blowing ratiosand hole geometries studied, especially for the curved hole

cases, as one can see from Figure 17. Figure 17 showsthe instantaneous smoke visualization pictures in the foursections of both straight and curved holes under BR = 0.5for ReD = 1000.

3.2.2. Time-Averaged Presentation. Figure 18 is the time-averaged smoke visualization pictures for both hole in thesection of X/D = 1 under blowing ratios of 0.5 and 1.0for Reynolds numbers 480 and 1000. Although the vortex structure could not be obviously viewed from the averagedpictures as the instantaneous pictures in Figures 15 to 17, theCVP still can be clearly recognized, especially for the straighthole cases. It can be observed again that the jet issued from


10/12


3

2

1

Y / D

Z/D Z/D

Straight hole Curved hole

X/D = 1ReD = 480 BR = 1

−2 −1 0 1 2 0 1 2−2 −10.25

X/D = 1ReD = 480 BR = 0.5

3

2

1

Y / D

Z/D Z/D


−2 −1 0 1 2 0 1 2−2 −1

(a) ReD = 480

X/D = 1ReD = 1000 BR = 0.5

3

2

1

Y / D

Z/D Z/D


−2 −1 0 1 2 0 1 2−2 −10.25

3

2

1

Y / D

Z/D Z/D

Straight hole Curved hole X/D = 1ReD = 1000 BR = 1

−2 −1 0 1 2 0 1 2−2 −10.25

(b) ReD = 1000

Figure 18: Time averaged smoke visualizations in the section X/D = 1 for both straight and curved holes under BR = 0.5 (right) and 1.0(left), (a) ReD = 480, and (b) ReD = 1000.

3

2

1

3

2

1

Y / D

Y / D

−2 −1 0 1 2 0 1 2Z/DZ/D


ReD = 1000 BR = 0.5 X/D = 0

−2 −1

CVPPVP

−350−276−203−129 −55 18 92 166 239 313

ωx

(a) X/D = 0

3

2

1

3

2

1

Y / D

Y / D

−2 −1 0 1 2 0 1 2Z/DZ/D


ReD = 1000 BR = 0.5 X/D = 1

ωx

−2 −1

−600−474−347−221 −95 32 158 284 411 537

0.25

(b) X/D = 1

−400 −316 −232 −147 −63 21 105 189 274 358

3

2

1

3

2

1

Y / D

Y / D

−2 −1 0 1 2 0 1 2

Z/DZ/D

Straight hole

ωx

Curved holeReD = 1000 BR = 0.5 X/D = 2

−2 −10.25

(c) X/D = 2

3

2

1

0.25

3

2

1

Y / D

Y / D

−2 −1 0 1 2 0 1 2

Z/DZ/D

Straight hole

ωx

Curved hole

16 79 142 205 268−300−237−174−111 −47ReD = 1000 BR = 0.5 X/D = 4

−2 −1

(d) X/D = 4

Figure 19: Contours of time averaged vorticity ωx in four cross sections, ReD = 1000 and BR = 0.5.

the curved hole present a flat-top shape at both Reynoldsnumbers and both blowing ratios, leading to a much widertransverse spreading, compared to the straight hole cases.

In order to further understand the evolution of PVP andCVP and their interaction along the jet trajectory, contoursof time averaged vorticity ωx in the four sections X/D = 0,

1, 2, 4 for both straight and curved hole are presented inFigure 19 for Reynolds number 1000 and BR 0.5. For theother flow conditions, the contours are essentially the same.The positive values of ωx correspond to counter-clockwisevorticity. It is seen that two main isoline loops are observedin the sections of straight hole, corresponding to the CVP


11/12


2

1.5

1

0.5

Z / D

X/D

1 1.5 2 2.5 3 3.5 4

ReD = 480 ReD = 1000

Curved BR = 0.5

Straight BR = 0.5

Curved BR = 1

Straight BR = 1

Curved BR = 0.5

Straight BR = 0.5

Curved BR = 1

Straight BR = 1

Figure 20: Variation of CVP center in wall normal direction

along streamwise direction, read from the time-averaged sectionstreamline patterns.

with positive vorticity at the left hand side of hole centerand negative vorticity at the right hand side of hole center.Four main loops are observed in the sections of curved hole,corresponding to, except for CVP, the PVP with the signopposite to CVP as indicated in the plot of section X/D = 0.At the sections of X/D = 1 and 2, the PVP located above theCVP could be still observed. It is seen that the PVP attemptsto push the CVP downwards towards the wall surface or, in

other words, the PVP attempts to prevent CVP to elevatecoolant away from the wall surface. It can be noticed fromthat the interaction of PVP with CVP accompanies with amomentum exchange between the two vortex pairs. Both thestrength of PVP and CVP decrease gradually with distance.At the position of X/D = 4 the vorticity from PVP has beencompletely transported into the CVP, resulting in mixing of PVP with CVP and a significant reduction of the strength of CVP. CVP is well symmetric with respect to the symmetricline Z/D = 0 in all the sections, while for PVP a goodsymmetry can only be observed in sections X/D = 0and 1.

Figure 20 shows the variation of CVP center height

(Y/D), read from time-averaged section streamline patternsof one vortex due to its symmetry, with streamsiwe sections.As expected, the height is increasing with distance for all theconditions studied. Increasing blowing ratio and/or reducingof the Reynolds number will result in the increase of Y/D forboth holes. It may be noticed that high Reynolds numberscorrespond to low boundary layer thickness or displacementthickness. It then may be concluded that the evolution of CVP center height in the mainstream direction is aff ectednot only by the blowing ratios but also the boundary layerthickness. It can also be seen again that the height of thecurved hole is lower for all the flow conditions than that of straight holes.

4. Conclusions

The flow structure downstream of jet in cross-flow hasbeen investigated by means of PIV. Instantaneous andtime-averaged visualization pictures and quantities in thesymmetric plane and four cross-sections are presented anddiscussed for diff erent Reynolds numbers and blowing ratios,with the comparison of curved hole to straight hole.

The flow leaves the jet upwards to a certain distancefrom its exit without visible vortex motion and then turnstowards the mainstream direction with a series of hairpinvortices along the jet trajectory and a wake vortex systemunder the trajectory. Height of the trajectory reduces withincreasing Reynolds number and decreasing the blowingratio. With increasing Reynolds number, the smoke tracesbecome more unstable and disordered and break up intosmall scale vortices in a short distance downstream of the jetexit for all the blowing ratios and hole geometries studied,especially for the curved hole cases.

The PVP generated inside the curved hole rotating in

a sense opposite to CVP downstream. PVP could weakenCVP due the interaction between them and enhance thetransportation of jet fluids towards the wall surface andlead to greater adherence of the jet fluid to the wall in allconditions studied. The curved hole configuration shows itshigh potential in the film cooling applications.

Acknowledgment

The authors would like to acknowledge the financial supportreceived from the National Natural Science Fundamental of China.

References

[1] R. Fearn and R. P. Weston, “Vorticity Associated with a jet in across flow,” AIAA Journal , vol. 12, no. 12, pp. 1666–1671, 1974.

[2] G. Bergeles, A. D. Gosman, and B. E. Launder, “The near-fieldcharacter of a jet discharged normal to a main stream,” Journal of Heat Transfer , vol. 98, no. 3, pp. 373–378, 1976.

[3] Y. Kamotani and I. Greber, “Experiments on a turbulent jetin a cross flow,” AIAA Journal , vol. 10, no. 11, pp. 1425–1429,1972.

[4] J. Andreopoulos and W. Rodi, “Experimental investigation of jets in a crossflow,” Journal of Fluid Mechanics, vol. 138, pp.93–127, 1984.

[5] P. K. Snyder and K. L. Orloff , “Three-dimensional laser

Doppler anemometer measurements of a jet in a crossflow,”NASA TM 85,997, 1984.

[6] T. F. Fric and A. Roshko, “Vortical structure in the wake of atransverse jet,” Journal of Fluid Mechanics, vol. 279, pp. 1–47,1994.

[7] R. M. Kelso, C. Delo, and A. J. Smits, “Unsteady wakestructures in transverse jets,” in Proceedings of the AGARDSymposium on Computational and Experimental Assessment of

Jets in Cross Flow , Winchester, UK, April 1993, AGARD-CP-534.

[8] R. J. Margason, “Fifty years of jet in crossflow research,” inProceedings of the AGARD Symposium on Computational and Experimental Assessment of Jets in Cross Flow , Winchester, UK,April 1993, AGARD-CP-534.


12/12


[9] A. R. Karagozian, “Transverse jets and their control,” Progressin Energy and Combustion Science, vol. 36, no. 5, pp. 531–553,2010.

[10] T. H. New, T. T. Lim, and S. C. Luo, “Eff ects of jet velocity profiles on a round jet in cross-flow,” Experiments in Fluids,vol. 40, no. 6, pp. 859–875, 2006.

[11] S. L. V. Coelho and J. C. R. Hunt, “Dynamics of the near fieldof strong jets in crossflows,” Journal of Fluid Mechanics, vol.200, pp. 95–120, 1989.

[12] S. H. Smith and M. G. Mungal, “Mixing, structure and scalingof the jet in crossflow,” Journal of Fluid Mechanics, vol. 357, pp.83–122, 1998.

[13] B. A. Haven andM. Kurosaka, “Improved jet coverage throughvortex cancellation,” AIAA Journal , vol. 34, no. 11, pp. 2443–2444, 1996.

[14] S. D. Peterson and M. W. Plesniak, “Short-hole jet-in-crossflow velocity field and its relationship to film-coolingperformance,” Experiments in Fluids, vol. 33, no. 6, pp. 889–898, 2002.

[15] S. Kang, “Numerical investigation of the 3D flow structures

nearby cooling hole,” Asian Joint Conference on Propulsionand Power, AJCPP2006-22084.

[16] W. R. Dean, “Note on the motion of fluid in a curved pipe,”Philosophical Magazine, vol. 20, pp. 208–223, 1927.

[17] J. Y. Liang, S. Kang, and L. Ma, “Numerical investigationof the flow over curved rectangular cooling hole,” Journal of Engineering Thermophysics, vol. 31, no. 2, pp. 251–255, 2010(Chinese).

[18] J. Y. Liang and S. Kang, “Correlation and spectra analysis for jet in cross flow based on DES results,” in Proceedings of the4th Symposium on Hybrid RANS-LES Methods, Beijing, China,September 2011.

[19] J. Y. Liang and S. Kang, “Detached eddy simulation of the single cooling hole in flat plate,” Journal of Engineering Thermophysics, vol. 32, no. 3, pp. 395–398, 2011 (Chinese).

[20] G. Haller, “An objective definition of a vortex,” Journal of Fluid Mechanics, vol. 525, pp. 1–26, 2005.

[21] S. J. Kline and F. A. McClintock, “Describing uncertainties insingle-sample experiments,” Mechanical Engineering , vol. 75,pp. 3–8, 1953.

Date post:	01-Jun-2018
Category:	Documents
Upload:	kang2010
View:	217 times
Download:	0 times

Advances in Mechanical Engineering-Experimental Visualization of the Flow Structure for Jet in...

Documents