11th International Symposium on Applications of Laser Techniques to Fluid Mechanics,8–11 July 2002, Lisbon, Portugal, Paper No. 17-5 (COMPLEX WALL FLOWS)

1

STEADY FLUID FLOW INVESTIGATION USING PIV IN A MULTI-PASS

COOLANT CHANNEL

KEY WORDS: MULTI-PASS COOLANT CHANNEL, PARTICLE IMAGE VELOCIMETRY, LASER-2-FOCUS VELOCIMETRY.

ABSTRACTIn a stationary two-pass coolant channel system the fluid flow is investigated experimentally. One of the main ob-

jectives of this paper is to describe the capability of the measuring technique Particle-Image Velocimetry (PIV). Re-sults of heat exchange and pressure losses are not the objectives of this paper. The investigated system has an engine-near lay-out with an 180° turn and has smooth walls in the beginning of this project, later on ribbed walls will beinvestigated. As a first step, the system is analyzed in non-rotating mode. During future work it will rotate about anaxis orthogonal to the center-line of the straight passes. The results shown in this paper demonstrate the effect of the180° bend with isothermal flow condition excluding any buoyancy. Turbulent channel flow with a REYNOLDS numberof 25.000 and 50.000, derived with the hydraulic diameter of the first pass, was investigated.

At the test rig at DLR several test models were mounted to investigate pressure drop behavior and fluid motionseparately. Numerical investigations regarding flow structure and pressure drop have been carried out in order to vali-date a CFD code against the experimental data. The numerical solution is compared with streamline pictures, obtainedfrom flow visualization of all walls using oil flow technique, and near wall PIV results. Mid-span flow distribution isreceived from laser light sheet visualization technique with oil fog as indicator and also from PIV. Figure 1 shows acomparison of the measured mean flow velocity distribution at mid-span (cut 1 & camera position 1) with and withoutturning vane.

The results presented in this paper clarify the complex flow situation given by the two pass system with inherentturn. Especially in the bend region appear separation regions and vortices with high local turbulence. These very de-manding measuring task represents a benchmark test case for the used measuring techniques.

Figure 1. Comparison of the mean flow velocity distribution at mid-span (cut 1 & camera position 1) with and with-out turning vane

Dipl.-Ing. Martin ELFERT

German Aerospace CenterInstitute of Propulsion Technology

Linder Höhe, D-51147 Köln, GermanyFax Number: +49 2203 64395

E-mail address: martin.elfert@dlr.de

Dr.-Ing. Marc P. JARIUS

German Aerospace CenterInstitute of Propulsion Technology

Linder Höhe, D-51147 Köln, GermanyFax Number: +49 2203 64395

E-mail address: marc.jarius@dlr.de

zin

mm

210

215

220

225

230

235

240

245

zin

mm

210

215

220

225

230

235

240

245

2

1 INTRODUCTIONThe power generation industry and the aero-engine industry operate in a highly competitive market. Major objec-

tives are high requirements in terms of high technology development as well as cost and development time reduction.On the other hand, environmental and safety constraints are an increasingly stringent necessity, which enforces thedemand for new technologies. Currently the industry relies on expensive and time consuming rig test programs,whereas the existing numerical design tools have a number of deficiencies in accurately describing the complex multi-pass coolant channel flow. Under these conditions the Institute of Propulsion Technology is involved in national andEuropean research programs aimed to providing the industry with high quality experimental data from the flow fieldfor CFD validation.

In past projects the flow behavior in rotating passages was analyzed at DLR. Laser-2-Focus velocimetry (L2F)was used to obtain flow velocity components and fluctuations. Although time-consuming, this non-intrusive, single-point measurement technique worked very well within straight and smooth duct flows, which generally have a moder-ate degree of turbulence. However, state of the art, serpentine shaped, multi-pass systems are equipped with ribbedwalls, in order to improve heat exchange which is typical in realistic configurations. In this case L2F velocimetry wasnot able to measure accurate flow properties due to the increased turbulence intensities in the vicinity of the ribs as inthe bend region and further downstream. The dividing wall separating the two passages forces the flow into a sharpturn which generally results in a flow separation. The flow within the separation bubble itself is very unsteady.

An important work package within these research programs is to provide detailed information of the flow field in-side the multi-pass coolant channel using planar techniques such as Particle Image Velocimetry (PIV) [ADRIAN, 1991,WILLERT, GHARIB, 1991, RAFFEL et al, 1998] or Planar Doppler Velocimetry (PDV) [RÖHLE, 2000]. Modern planarmeasurement techniques such as PIV are capable of obtaining complete maps of flows even at high turbulence. As afirst step toward applying this technique, a multi-pass cooling system with and without turning vane is investigated instationary (e.g. non-rotating) mode using two-component PIV. The results are compared with results from L2F (multi-pass cooling system with turning vane), flow visualization and CFD. The high quality of the obtained new resultsencourage the application of two-component PIV to the rotating system. As a logical consequence, the application ofthree-component PIV will be necessary to obtain the complete flow field information within the complex flow pas-sages. This paper’s intention is to report on the status of applicability of PIV in a multi-pass coolant channel.

2 NOMENCLATUREdh m Hydraulic diameterf mm Lens focal lengthf# - Lens f-numberν m²/s Kinematic viscosityn rpm SpeedRe - REYNOLDS number, Re=ρv0d/ητ µs Pulse delayv m/s Volumetric absolute velocityv0 m/s Intake velocityx, y, z mm CARTESIAN co-ordinates

Abbreviationsp.s. pressure sides.s. suction sidet.e. trailing edget.v. turning vane

3

3 TEST FACILITY AND INSTRUMENTATION

3.1. Test FacilityA large-scale coolant multi-pass system rotating spanwise was investigated during the past experiments. Air is

supplied in the rig through a rotary sealing assembly, a second one is used for venting after the air has passed throughthe test section. Both are mounted to the end of the double hollow shaft. A schematic of the test rig is given in Fig-ure 2. More detailed information about the test facility are presented in [RATHJEN et al, 1999]. Reynolds number androtation number can be adjusted appropriately to the models inserted. For the PIV investigations a non-rotating test rigis installed at a laboratory and shown in Figure 3.

3.2. Test Model GeometryFigure 4 shows a schematic draw of the coolant multi-pass-model with turning vane. The test section consists of a

leading edge duct (first pass) with a trapezoidal cross-section extending radially outward, a 180° bend with a 90°turning vane and a second pass with a different trapezoidal cross-section extending radially inward. The ratio betweenthe hydraulic diameter of the second pass and the hydraulic diameter of the first pass is 1.5. The length of each pass is9.975 dh, where dh denotes the hydraulic diameter of the first passage. In the bend region the flow is not only directedthrough the 180° u-bend but also perpendicular to that following the turbine airfoil shape. The test section is 12 dhlong. The distance between the tip plate and the divider wall is about 1.875 dh. The divider wall has a thickness of 0.3dh. The model has a small plenum chamber and an inlet and outlet length, both are 3 dh long.

A schematic draw of a second cooling multi-pass-model without turning vane and with a modified second pass isshown in Figure 5. The length of the wall 1* is reduced to 1.3 dh, the angle is reduced to 8° and the wall 1* is nowparallel to 3*. The length of each pass is increased to 10.65 dh and consequently the distance between the tip plate andthe divider wall is reduced to 1.2 dh. A CAD-picture of the model geometry without turning vane is presented in Fig-ure 6.

3.3. Measuring Techniques

3.3.1. PIV – Particle Image Velocimetry (Flow)The flow field is measured first in steady mode by means of PIV. This modern technique is a scientific tool for

qualitative and quantitative investigations of flow fields. The principle of the PIV with laser, light sheet optic, laserlight sheet and test section is shown in Figure 7 and the existing test set-up in the laboratory is shown in Figure 2.PIV is based on the principle to capture the image of the flow field two times (pulse delay in the range of 5 µs). Tracerparticles (aerosol) with an average size of 0.1 µm are added to the flow. These particles reflect in the laser light sheet,which is formed with a light sheet optic consisting of a system of one spherical and two cylindrical lenses. It generatesa light sheet with a thickness of 1 mm and a divergent angle of 24°. The laser light sheet is adjustable and visualizesthe first pass as well as the second pass, due to a mirror. Illumination was provided by a standard, frequency-doubled,double-cavity Nd:YAG laser (NewWave, Gemini PIV) with a pulse energy of up to 120 mJ per pulse at 532 nm. Onthe recording side, a thermo-electrically cooled, interline transfer CCD camera (PCO, 1280 x 1024 pixel resolution)with a f = 55 mm, f # 2.8 lens (Nikon) was used. A bandpass filter with center frequency 532 nm and width 5 nm(FWHM) placed in front of the lens rejected most of the unwanted radiation. The pulse delay between the laser pulsesvaried between τ = 6 µs (investigation of mean flow) and τ = 12 µs (investigation of secondary flow). The high reso-lution CCD1 camera takes two pictures, depending on the pulsing laser. Each picture has a size of 2.5 MB. The CCDcamera, perpendicularly positioned to the light sheet, is a so-called cross correlation camera having double-framesingle-exposure evaluation. Now it is possible to calculate the value, direction and orientation of the absolute veloci-ties with the Cross Correlation Function (CCF). 50 pictures at each camera position are taken to calculate the meanvalue of the absolute velocity. For the secondary flow investigations (cut 1...6) into the first pass from the multi-passcoolant channel one camera positions is adjusted (Figure 8a) and for the investigations of the mean flow (cut 1 & 2)into the second pass of the multi-pass coolant channel five camera positions are necessary (Figure 8b).

1 Charge Coupled Device

4

3.3.2. OFV – Oil Flow Visualization (Wall Flow)To analyze the flow phenomena, several kinds of flow visualization techniques have been applied. Using the oil

flow visualization method, wall streamline patterns could be achieved within the duct. It is possible to show and lo-calize the boundary layer separation, the reverse flow and vortex regions next to the surfaces. To some extent it ispossible to draw conclusions from the two-dimensional method for application to the strongly three-dimensional char-acter of the flow field. The duct wall has been previously painted with a mixture consisting of oil of appropriate vis-cosity, oil acid and Titanium dioxide (TiO2) for contrast purpose. Then it has been exposed to the flow for a timeperiod (approx. 10 minutes) long enough to dry the oil paint. The duct is dividable in two halves and is opened afterthe test run for viewing. A detailed description of the range of application is given in [MALTBY, KEATING, 1962].

3.3.3. L2F – Laser-Two-Focus Velocimetry (Flow)For the non-intrusive measurement of flow velocities, angles and fluctuations inside the rotating duct the Laser-2-

Focus (L2F) technique [BEVERSDORFF, HEIN, SCHODL, 1992] has been applied using a newly developed optical unitto direct the laser beams into the rotating frame of reference. The optical device mounted in front of the rotor enablesthe system to stationary non-triggered signal processing. Following the beam path from the laser to the probe volume,the laser light is emitted by an Ar+ ion laser of 4 Watt power output which is connected to the launching unit mountedto the test rig by an optical fiber for vibration decoupling purpose. The L2F method is based on time of flight meas-urements of co-flowing particles. Small oil droplets with a size less than 0.5 µm were added to the flow before enter-ing the shaft. The flow-following behavior of these particles even at high accelerations has been successfully demon-strated and thereby a correlation between particle and flow velocity is assured [SCHODL, 1977].

3.3.4. LLSV – Laser-Light-Sheet Visualization (Flow)To obtain more information about the flow around the turning vane and the separation zone, a camera system has

been installed to the Perspex duct observing inside the second pass with the mid plane lighted with a laser light sheet.In order to visualize the flow structures, seeding of the flow with oil steam is provided by an oil fog generator. Withframe grabbing on a PC the video pictures were captured. The digitized pictures have been improved with regard toluminance and brightness.

4 RESULTSIn the following, some exemplary experimental results are presented. All results are presented in the form of

isolines and vectors. For a better understanding of the isolines, red areas represent high velocities, blue areas representareas with lower velocities. The shown vectors all have a uniform size to represent only the orientation of the flowfield. With regard to PIV processing, standard cross correlation algorithms were used with an interrogation windowsize of 32 x 32 pixel and a 50 % overlap. The velocity of the intake flow into the first pass amounts to v0 = 15.25 m/s,and consequently the REYNOLDS number is about 25.000 for the PIV measurements.

The comparison of the axial velocity component at cut 1 and 2 between L2F and PIV measurement in the secondpass at the position z = 193 mm is presented in Figure 9. The axial velocity component at the z-position is extractedoff the measured PIV results. For the case with t.v. the results shows relatively good agreement with the L2F results.The position of the jet and the wake is slightly different, which depends on deviations in model manufacturing. Meas-urements in the region of the separation bubble were not possible with the L2F system due to too high turbulence. Forthe case without t.v. the flow separation on the dividing wall is much larger then with t.v. and the value of the axialvelocity component inside the bubble is even higher. By means of the thickness of the separation bubble, the opencross section is contracted, due to that fact the value of the axial velocity component is increased. At mid-span theaxial velocity component is smaller and the separation region is nearly not remarkable.

The measured flow field of the secondary flow within the first pass in six cuts (z = 190...240 mm, step size10 mm) for a REYNOLDS number of 25.000 are shown in Figure 10. At a distance of 190 mm from the duct entry(Figure 10a, cut 1) a relatively big and strong vortex occurs (x = 16.0 mm, y = 13.0 mm). A smaller vortex exists atthe channel too (x = 8.0 mm, y = 5.0 mm). An interpretation of that phenomenon is, that the flow is disturbed due tothe inlet distortions, which are still visible at this position. The same result was found during CFD simulation wherethe complete inlet geometry with settling chamber was modeled (Figure 15) at cut 1. The mean value of the secon-dary flow velocity is in the range of about 1 m/s. At the position of cut 2 (Figure 10b, z = 200 mm) the big vortex

5

moved closer to the side wall (x = 17.5 mm, y = 11.0 mm) and the smaller vortex is already disappearing. The meanvalue of the secondary flow velocity is in the same range like before. If the position of the light sheet is varied to thepositions of cut 3, 4, 5 and 6 (Figure 10c, d, e and f) all vortices are disappeared. The mean value of the secondaryflow velocity is increasing more and more from cut to cut and is finally reaching a value of 13 m/s at cut 6 (Fig-ure 10f) and is becoming now the main stream velocity component due to the flow turning process. At cut 6 at theleading edge the beginning of the recirculation from the upper passage corner can be observed.

Within the second pass with turning vane the measured flow field of the main flow in two cuts 2 mm (right, wallsection) and 10 mm (left, mid-span) for a REYNOLDS number of 25.000 are shown in Figure 11. At the mid-spansection (left) a relative stationary separation bubble occurs at the upper surface (s.s.) of the t.v.. Behind the separationbubble exists a relative high flow deviation with high velocities about 24 m/s close to the wall. An other separationbubble with reverse flow is visible on the dividing wall. Due to that flow recirculation a jet flow with velocities about24 m/s exists on the pressure side at the t.e. of the t.v.. The thickness of the separation bubble is decreasing in down-stream direction. The form and extension of the separation bubble with a length of approximately 5 dh is similar tothose found by visualization (Figure 16a). The areas of high velocity are coinciding further downstream(z = 190 mm) where mixing of both jets leads to an uniform velocity profile. The mean value of the main flow veloc-ity further downstream is in the range of about 13...16 m/s. At the wall section (right) the mean value of the velocity isincreased to 25 m/s on the upper side of the t.v.. Turning of the flow leads to an impingement effect to the leadingwall of the second leg due to the inclination of the center-lines of both cross-sections. The separation bubble on thedividing wall is much larger in longitudinal extension than before. At the s.s. of the t.v. appears a boundary layer sepa-ration (x = 20 mm, z = 215...220 mm) due to side wall effects. The jet flow at the pressure side of the t.v. is still ex-isting. But the mean value of the main flow velocity distribution is decelerated further downstream to a range of about8...10 m/s.

The measured flow field of the main flow within the second pass without turning vane in two cuts 2 mm (right,wall section) and 10 mm (left, mid-span) for a REYNOLDS number of 25.000 are shown in Figure 12. At the mid-spansection (left) a separation bubble with reverse flow is visible on the dividing wall. The thickness of the separationbubble is decreasing more and more in downstream direction. The form and extension of the separation bubble with alength of approximately 2 dh is similar to those found by visualization (Figure 16b). At the top of the duct exists arelative high flow deviation with the highest velocities with a value of over 24 m/s. This area of high velocity(> 24 m/s) reaches up to z = 150 mm downstream. At the wall section (right) the area with high velocity over 24 m/sis increased and the extension of the separation zone at the dividing wall is larger as at the other cases.

The streamline patterns on the pressure side (Figure 13) and the suction side (Figure 14) obtained from oil flowvisualization experiments for both cases, with and w/o turning vane are compared with a numerical solution inclusivet.v. close to the wall. The streamlines on both walls show the combined effect of the 180° turn, the influence of theturning vane and clearly the extension and thickness of the separation bubble at the top of the dividing wall in thesecond channel. The main phenomena in this engine-near configuration are: (i) The separation bubble at the dividingwall is larger at the pressure side than at the opposite wall. (ii) The deflection of the flow within the bend due to aninclination of the orientation axis of the two passes forms a strong secondary motion after the bend running from theconcave wall (p.s.) around the rear wall to the s.s. of the 2nd pass. (iii) The bend induces a counter-rotating pair ofsecondary vortices, These are affected by the secondary motion described above. The vortex along the p.s. has anadverse orientation and is strongly weakened whereas the other along the s.s. is enlarged recognizable at the slope ofstreamlines against the pass axis. The saddle lines between both vortices were expected normally at the center of theside walls in a rectangular U-turn. In this case of an inclined-oriented two-pass system the saddle lines were shifted tothe p.s. and to the corner of the rear side, respectively. After the separation bubble at the top of the divider wall, al-most cross-oriented streamlines are visible over a curtain area at the pressure side confirming the dominance of the co-flowing secondary vortices of (ii) and (iii). Both vortex generation mechanisms depend on inertia effects in viscousflows.

A comparison of a calculated result (Figure 15a) and measured results with PIV with t.v. (Figure 15b) and withoutt.v. (Figure 15c) is presented. The relatively big vortex close to the dividing wall is visible in all results, but the posi-tion of this vortex is a little different for the case without t.v.. The core of the vortex is centred more into the middle ofthe channel and the orientation is changed.

6

Two captures of the flow around the bend with and without t.v. obtained by laser-light-sheet visualization isshown in Figure 16a & b. This technique confirms the already recognized and already described flow phenomena.

5 SUMMARYThis paper describes experimental investigations of a typical two-pass cooling channel system of a turbine blade

with respect to fluid flow phenomena affected by geometry effects and flow turning. A complex flow situation ispresent showing several kind of separation and vortices:(i) At the upper surface of the t.v. due to a to high angle of attack versus the incoming flow,(ii) further downstream at the s.s. of the t.v. due to too strong curvature,(iii) at the dividing wall due to the sharp edge where the flow is incapable to follow.

The jet region under the t.v. and the wake from the t.v. are leading to high velocity gradients and correspondinghigh shear stresses at the contact layer producing very high turbulence levels there. Higher turbulence levels than30 % leads to uncertainties in L2F results. With supposed levels of 50 % and more the PIV technique has no difficul-ties. But in order to give an accurate value of the turbulence level by PIV, you will need about 2000 pictures or more.This relatively big amount of data demand high storage capacities.

The performed flow visualization and simulation are very helpful in that case to understand the interaction of allapparent effects. The presented experiments were mainly intended to assess the feasibility of applying PIV in a multi-pass cooling channel used for applied cooling research. The results have shown that it is possible to apply PIV in amulti-pass cooling channel. The application of PIV is approved and will be adopted to the rotating system in the nearfuture. As a logical consequence, the application of three-component PIV will be necessary to obtain the completeflow field information within the complex flow passages.

6 REFERENCESADRIAN, R. J., 1991, Particle-Imaging Techniques for Experimental Fluid Mechanics, Annual Review of Fluid Me-

chanics 23.BEVERSDORFF, M., HEIN, O., SCHODL, R., 1992, An L2F-Measurement Device with Image Rotator Prism for Flow

Velocity Analysis in Rotating Coolant Channels. 80th Symp. On Heat Transfer and Cooling in Gas Turbines,AGARD PEP, Antalya, Turkey.

ELFERT, M., HOEVEL, H., TOWFIGHI, K., 1996, The Influence of Rotation and Buoyancy on Radially Inward andOutward Directed Flow in a Rotating Circular Coolant Channel, Proc. 20th ICAS Conference, Sorrento, Italy, 2490-2500

MALTBY, R. L., KEATING, R. F. A., 1962, The Surface Oil Flow Technique for Use in Low Speed Wind Tunnels,AGARDograph 70.

RATHJEN, L., HENNECKE, D.K., ELFERT, M., BOCK, S., HENRICH, E., 1999, Investigation of Fluid Flow, HeatTransfer and Pressure loss in a Rotating Multi-Pass Coolant Channel with an Engine Near Geometry, ISABE-Conference, Florence, ISABE- IS-216

RAFFEL, M., WILLERT, C., KOMPENHANS, J., 1998, Particle Image Velocimetry, Springer Verlag Berlin Heidelberg.RÖHLE I. 2000, “Doppler global velocimetry”, in RTO Lecture Series 217, Planar Optical Measurement Methods for

Gas Turbine Components, Neuilly-Sur-Seine Cedex, France.SCHODL, R., 1977, Entwicklung des Laser-Zwei-Focus-Verfahrens für die berührungslose Messung der Strömungs-

vektoren, Dissertation, Tech. Univ. Aachen, Germany.WILLERT, C., GHARIB, M., 1991, Digital Particle Image Velocimetry, Experiments in Fluids, No.10, Springer Ver-

lag.

7 Figures

Figure 2. Schematics of the test rig at DLR-Col

Inlet Outlet

9.97

5dh

1.87

5dh

0.15dh

12d h

0.3dh

0.12dh1.125dh

3dh

x/dh = 12.975

x/dh = 22.95x/dh = 0

25°

20°

35°

30° axis of rotatio

1.0dh

0.3dh

dh = hydraulic diameter of the firs

3∗

4

2

2∗

31∗

1

Figure 4. Schematic of test model geometry wvane

LLiigghhtt sshheeeett ooppttiicc

LLaasseerr

AAccrryylliicc mmooddeell

CCCCDD--ccaammeerraa

FFllooww mmeetteerr

ogne Figure 3. Test set-up for stationary PIV-measurements

1.2

d h 0.15 dh

7

10°

n

t pass

4∗

1.5 dh

Inlet Outlet

10.6

5 d h12

dh

0.3dh3 d h

x/dh = 22.95x/dh = 0

25°

20°

35°

8°

30° axis of rotation

1.0dh

0.3dh

dh = hydraulic diameter of the first pass

3∗

4

2

4∗

2∗

31∗

1

1.3dh

ith turning Figure 5. Schematic of test model geometry withoutturning vane

Leading wall

Trailing wall

F

MirrorLight sheet optic

8

igure 6. Test model geometry without turning vane Fi

Partfirst

gure 7. Principle of Particle-Image Velocimetry

Laser

Laser light sheet

Flow directionImage plane

Imaging optic

Tracer particle into light sheet

icle image from light pulse at t0

Particle image from second light pulse at t1

Figure 8. Camera and cgation into sec

a

vin

m/s

10 150

5

10

15

20

25

30

Wallsection, PIVWall section, PIVWall section, L2FWall section, L2F

PIV

L2F

Region ofBackflow

Figure 9. Comparison oposition z = 1

a b

ut positions. a. Secondary flow investigations into first pass (cut 1...6); b. Mean flow investi-ond pass (cut 1 & 2)

b

25

30

without tvwith tv

9

x in mm20 25 30

, corr.

vin

m/s

100

5

10

15

20

PIV

f the axial velocity component between L2F and PI93 mm. a cut 1; b cut 2; (according Figure 8b)

L2F

x in mm15 20 25 30

Mid-span, PIV without tvMid-span, PIV with tvMid-span, L2FMid-span, L2F, corr.

V measurement in the second pass at the

10

x in mm

zin

mm

0 5 10 15 20 250

5

10

15

20

25

30 `ìí=NI=ò=Z =NVM=ã ã

y

x in mm

zin

mm

0 5 10 15 20 250

5

10

15

20

25

30 `ìí=PI=ò=Z =ONM=ã ã

y

x in mm

zin

mm

0 5 10 15 20 250

5

10

15

20

25

30 `ìí=RI=ò=Z =OPM=ã ã

y

x in mm

zin

mm

0 5 10 15 20 250

5

10

15

20

25

30 `ìí=OI=ò=Z =OMM=ã ã

y

x in mm

zin

mm

0 5 10 15 20 250

5

10

15

20

25

30 `ìí=QI=ò=Z =OOM=ã ã

y

x in mm

zin

mm

0 5 10 15 20 250

5

10

15

20

25

30`ìí=SI=ò=Z =OQM=ã ã

y

a

fe

dc

b

Figure 10. Secondary flow velocity distribution in six cross cuts of the first pass measured with PIV (Re = 25.000)

Cut 1, z = 190 mm

Cut 5, z = 230 mm Cut 6, z = 240 mm

Cut 3, z = 210 mm Cut 4, z = 220 mm

Cut 2, z = 200 mm

11

Figure 11. Mean flow velocity distribution in two longitudinal cuts of the second pass with turning vane measuredwith PIV (Re = 25.000)

Mid-spany = 10 mm

Wall sectiony = 2 mmPos 1

Pos 2

Pos 3

Pos 4

Pos 5

x in mm5 10 15 20 25 30 35

50

55

60

65

70

75

80

85

x in mm5 10 15 20 25 30 35

50

55

60

65

70

75

80

85

90

95

100

105

110

115

120

125

130

135

140

145

150

155

160

165

170

175

180

185

190

195

200

205

90

95

100

105

110

115

120

125

130

135

140

145

150

155

160

165

170

175

180

185

190

195

200

205

zin

mm

210

215

220

225

230

235

240

245

v in m/s: 2 4 6 8 10 12 14 16 18 20 22 24

zin

mm

210

215

220

225

230

235

240

245

v in m/s: 2 4 6 8 10 12 14 16 18 20 22 24

12

Figure 12. Mean flow velocity distribution in two longitudinal cuts of the second pass without turning vane meas-ured with PIV (Re = 25.000)

Pos 1

Pos 2

Pos 3

Pos 4

Pos 5

Wall sectiony = 2 mm

Mid-spany = 10 mm

x in mm5 10 15 20 25 30 35

20

25

30

35

40

45

50

55

60

6565

70

75

80

85

90

95

100

105

110110

115

120

125

130

135

140

145

150

155155

160

165

170

175

180

185

190

195

200

zin

mm

200

205

210

215

220

225

230

235

240

245

v in m/s: 0 2 4 6 8 10 12 14 16 18 20 22 24

x in mm5 10 15 20 25 30 35

20

25

30

35

40

45

50

55

60

65

70

75

80

85

90

95

100

105

110

115

120

125

130

135

140

145

150

155155

160

165

170

175

180

185

190

195

200

zin

mm

205

210

215

220

225

230

235

240

245

v in m/s: 0 2 4 6 8 10 12 14 16 18 20 22 24

13

Suction side

Figure 13. Comparison between calculated and visualized wall velocities and streamlines on suction side, non-rotating (Re = 50.000)

Pressure side

14

Figure 14. Comparison between calculated and visualizerotating (Re = 50.000)

d wall velocities

and streamlines on pressure side, non-

15

Figure 15. Comparison of calculated and measured results. A Numerical simulation at cut 1 in first pass; b Measured

at cut 1 in first pass with turning vane; c Measured at cut 1 in first pass without turning vane

Figure 16. Pattern of laser light sheet visualization at the bend region. a With turning vane; b Without turning vane

a b

a b c