effects of spatial resolution on piv investigation of a turbulent orifice...

16th Int Symp on Applications of Laser Techniques to Fluid Mechanics Lisbon, Portugal, 09-12 July, 2012

- 1 -

Effects of spatial resolution on PIV investigation of a turbulent orifice jet

Giovanni Lacagnina1,*, Giovanni Paolo Romano1

1: Department of Mechanical and Aerospatial Engineering

SAPIENZA Università di Roma, Via Eudossiana 18, 00184 Roma, ITALY * correspondent author: [email protected]

Abstract Some of the most interesting turbulent phenomena are jets, the study of which is useful to reach a deep knowledge of mixing in fluids. Among all jets, there exists one, the orifice jet, which is still not fully understood because of its intrinsic complexity, e.g. the "vena contracta" phenomenon, though its wide range of applications ranging over underwater propulsion, pollution dispersion and studies on the cardio-circulatory system. In this work, the attention is focused to the large-scale statistics (mean field and higher statistical moments of the velocity) and to the small-scale statistics (velocity derivative statistical moments), with a special care regarding the effect of spatial resolution on Particle Image Velocimetry (PIV) measurements. To investigate the effect of spatial resolution on the measurements, framed areas of different size have been considered by changing the magnification factor (namely the object plane distance). Making reference to the orifice diameter D, squared regions 2D x 2D, 3D x 3D, 4D x 4D and 5D x 5D have been framed on the acquisition camera. Regarding to the large scales, first of all the occurrence of the vena contracta phenomenon for an orifice jet has been shown in several ways. In addition, it has been highlighted the small influence of spatial resolution on low order statistical moments, which slightly rises, due to local averaging, as the order of the statistics becomes higher. Then small scale features of the jet have been examined, particularly with interest to the evaluation of spatial derivatives of velocity components and then to the fulfillment of isotropy hypotheses of the jet. Based on those data, an incomplete local isotropy condition is attained, although restricted to the near zone, and an influence of spatial resolution high on the determination of local features of such a jet and moderate on the overall behavior of spatial derivatives and to the deductions about symmetry hypotheses. 1. Introduction In spite of the wide variety of applications of orifice jets, related to the ease of manufacturing, ranging e.g. over marine propulsion, pollutant discharge or studies on the cardio-circulatory system, the number of studies regarding them is limited (Quinn (1989), Boersma et al. (1998), Mi et al. (2001), Quinn (2005), Mi et al. (2007), Romano et Falchi (2010)). The reason presumably lies in the complexity of their velocity field, especially in the near field, so that the behavior of these jets is still not fully understood. The vena contracta occurrence can be elected as an example of it. In fact, due to the necessity for the flow to pass through the orifice, which has a limited diameter in comparison to the plenum size upstream, the streamlines of the jet tend to converge. Downstream of the plate they should recover the undisturbed direction, consequently they should change direction to get the original one, but they are not able to perform it abruptly. So the contraction persists further downstream until a minimum section, namely the vena contracta. This was probably first described by Evangelista Torricelli in his experiment on atmospheric pressure in 1643. With the aim to describe this phenomenon, a characteristic dimensionless quantity can be defined. Given the area of the vena contracta A and the area of the orifice A0, contraction coefficient results equal to

0AACc = .

For a sharp circular orifice its value is generally near to 0.6 (Michell 1890), but it is also dependent


- 2 -

on the specific shape of the contraction (Batchelor 2000) and of the orifice, increasing in all of 10 per cent moving in order from the circular shape, to the square, the triangular and the rectangular (King 1918). Regarding to the small scales, one of the most challenging problems in the study of jets is related to the comprehension of what happens in correspondence to them and then to the modeling of such phenomena in a realistic and reliable way. First of all A. N. Kolmogorov, in his work in 1941, introduced the hypothesis that, at sufficiently high Reynolds numbers, the flow displays statistically isotropic features in correspondence to small scales, regardless to the possible anisotropy of larger ones. An immediate application of this hypothesis can be found in the assessment of turbulent kinetic energy dissipation rate, present in the turbulent kinetic energy budget equation. Its value in general depends on 12 quantities among mean quadratic derivatives of velocity components, combinations of them and their statistical moments. Therefore an evaluation of all those terms is very challenging both experimentally and numerically due to the requirements of high temporal and spatial resolutions. Several ways of extending spatial resolution have been worked out. Keane et al. in 1995 and later Stitou et al. in 2001, e.g. , used the knowledge of the velocity field from cross correlation of large sub-regions and particle tracking PTV in order to achieve an high spatial resolution, Westerweel et al. in 1997 succeeded in it making use of a window offset in the analysis, whereas Scarano in 2002 and in 2003 developing an adaptive window shaping. Westerweel et al. and later Billy et al. in 2004 and Kahler in 2006 applied a two-point ensemble correlation at single-pixel resolution, while finally Lavoie et al. in 2007 compared PIV spatial resolution with the one of Hot-Wire-Anemometry, suggesting some spectral corrections in order to correct PIV results. In general, the turbulent kinetic energy dissipation rate is defined as (Hinze):

⎟⎟⎠

⎞⎜⎜⎝

⎛

∂∂

∂∂+

∂∂

∂∂=

i

j

j

i

j

i

j

iT x

uxu

xu

xuνε for i,j = 1,….,3

1

ui and uj are the generic velocity components, ν denotes the kinematic viscosity and the indexes i and j identify the orthogonal axes set. Writing explicitly its terms, equation 1 becomes:


- 3 -

⎪⎭

⎪⎬⎫⎟⎟⎠

⎞⎜⎜⎝

⎛∂∂

∂∂+

∂∂

∂∂+

∂∂

∂∂+

⎥⎥⎦

⎤⎟⎟⎠

⎞⎜⎜⎝

⎛∂∂+⎟

⎠⎞⎜

⎝⎛∂∂+⎟

⎠⎞⎜

⎝⎛∂∂+

⎢⎢⎣

⎡⎟⎠⎞⎜

⎝⎛∂∂+⎟

⎠⎞⎜

⎝⎛∂∂+⎟⎟⎠

⎞⎜⎜⎝

⎛∂∂+

⎪⎩

⎪⎨⎧

⎥⎥⎦

⎤

⎢⎢⎣

⎡⎟⎠⎞⎜

⎝⎛∂∂+⎟⎟⎠

⎞⎜⎜⎝

⎛∂∂+⎟

⎠⎞⎜

⎝⎛∂∂=

yw

zv

xw

zu

xv

yu

yw

zv

xw

zu

xv

yu

zw

yv

xu

T

2

2

222

222

222

νε

2

The intrinsic difficulty of evaluating this quantity is now clear. Making reference to the local isotropy hypothesis (Kolmogorov 1941), equation 2 gets a more approachable look associated to a single quantity, e.g. mean quadratic derivative of streamwise velocity with respect to the streamwise coordinate:

2

15 ⎟⎠⎞⎜

⎝⎛∂∂=xu

T νε

3

This advantageous hypothesis still awaits confirmations both numerical and experimental and certainly for orifice jets. There exists a different formulation, firstly expressed by Batchelor (1946) and Chandrasekhar (1950) and then proposed by George and Hussein more recently (1991), based on the less stringent hypothesis of axial symmetry to rotation. Thanks to this approximation, the functional dependency of equation 2 can be reduced to four quantities, e.g., in the plane (x,y), the derivatives of velocity components in the same plane.

⎥⎥⎦

⎤

⎢⎢⎣

⎡⎟⎟⎠

⎞⎜⎜⎝

⎛∂∂+⎟

⎠⎞⎜

⎝⎛∂∂+⎟⎟⎠

⎞⎜⎜⎝

⎛∂∂+⎟

⎠⎞⎜

⎝⎛∂∂−=

2222

822yv

xv

yu

xu

T νε

Nowadays is still not clear under which conditions a phenomenon shows local isotropy or axial symmetry conditions, albeit a good amount of works have confirmed the goodness of the second one in correspondence to several turbulent flows, such as quasi-homogeneous shear flows, boundary layers, pipe jets, circular jets, round plumes, plane jets, mixing layers and two dimensional cylinder wakes (George and Hussein 1991). This has never been studied on orifice jets. A possible way of testing the fulfillment of those conditions is making use of some non-dimensional parameters (George and Hussein 1991). They are:


- 4 -

2

2

32

2

22

2

1 ;;

⎟⎠⎞⎜

⎝⎛∂∂

⎟⎟⎠

⎞⎜⎜⎝

⎛∂∂

=

⎟⎠⎞⎜

⎝⎛∂∂

⎟⎠⎞⎜

⎝⎛∂∂

=

⎟⎠⎞⎜

⎝⎛∂∂

⎟⎟⎠

⎞⎜⎜⎝

⎛∂∂

=

xu

yu

K

xu

zw

K

xu

yv

K

4

2

2

62

2

52

2

4 ;;

⎟⎠⎞⎜

⎝⎛∂∂

⎟⎠⎞⎜

⎝⎛∂∂

=

⎟⎠⎞⎜

⎝⎛∂∂

⎟⎠⎞⎜

⎝⎛∂∂

=

⎟⎠⎞⎜

⎝⎛∂∂

⎟⎠⎞⎜

⎝⎛∂∂

=

xu

xw

K

xu

zu

K

xu

xv

K

5

292

2

82

2

7 ;;

⎟⎠⎞⎜

⎝⎛∂∂

∂∂

∂∂

=

⎟⎠⎞⎜

⎝⎛∂∂

⎟⎟⎠

⎞⎜⎜⎝

⎛∂∂

=

⎟⎠⎞⎜

⎝⎛∂∂

⎟⎠⎞⎜

⎝⎛∂∂

=

xu

xv

yu

K

xu

yw

K

xu

zv

K

6

211210 ;

⎟⎠⎞⎜

⎝⎛∂∂

∂∂

∂∂

=

⎟⎠⎞⎜

⎝⎛∂∂∂∂

∂∂

=

xu

yw

zv

K

xuxw

zu

K

7

As an example a PIV study allows to evaluate the parameters K1, K3, K4 and K9. If local isotropy conditions were fulfilled we should get, as demonstrated by Taylor in 1935, the following relations (George and Hussein 1991):

2;2;1;1 4321 ==== KKKK 8 2;2;2;2 8765 ==== KKKK 9 5.0;5.0;5.0 11109 −=−=−= KKK 10

On the other hand, if the axial symmetry conditions were fulfilled, it would be got (George Hussein 1991):

5.0;;;; 10987645321 −====== KKKKKKKKKK 11

71171 31

61;

31

31 KKKK −=+=

12

Starting from these guidelines, the first purpose of this paper is to investigate the behavior of small scales in the near field of a circular turbulent jet, aiming to define where the effects of inlet conditions are lost and useful simplifications can be applied. To do this, a high spatial and temporal resolution is required and this is a challenging problem for Particle Image Velocimetry (PIV). Here the problem of spatial resolution is considered directly by decreasing the size of the acquired flow region. 2. Experimental setup


- 5 -

The experimental setup can be split in the jet apparatus and the acquisition system. The first consists of a closed loop hydraulic circuit starting from an auxiliary tank, from which a centrifugal pump transfers the water to a secondary constant level head tank to avoid the artificial velocity fluctuations due to the pump, figure Figure 1. The jet will be generated making use only of the potential energy owned by the water for being high-located. Once flowed out of the head tank, the water enters through a honeycomb into the test tank which is subdivided by the orifice plate into a settling chamber (38 cm, upstream) and the test chamber (58 cm, downstream). Finally a second plate delimits a discharge chamber, downstream to the test one, through which the water comes back to the main tank, closing the loop. Experimental conditions are determined by the Reynolds

number defined by νDU 0Re = (where U0 is the exit bulk velocity, D is the diameter of the

orifice and ν the kinematic viscosity of fluid) set to 35000 and the Taylor microscale Reynolds

number defined by νλ

λ'Re u= (where u' is the longitudinal velocity rms value and λ is the

longitudinal Taylor microscale) which is ~ 10. Present study has been carried out by means of the PIV technique. The acquisition system is composed first of all of a light source, namely a double-pulse Nd-Yag laser having 200 mJ energy per pulse and a pulse duration of 8 ns, and of a digital high speed camera (2000 frames/s), with 1024 x 1024 pixel resolution and a 10 bit CMOS sensor. Between them is interposed a BNC 575 pulse generator, that allows the synchronization between laser illumination and camera recording.

Figure 1: Experimental setup

3. Results 3.1 Statistical moments One of the aims of present work was studying the effect of varying the spatial resolution on the features of the orifice jet. For this reason acquisitions with different framed areas have been performed. In detail 2D x 2D, being D orifice diameter (3 cm), 3D x 3D, 4D x 4D, 5D x 5D areas have been acquired at the same Reynolds number, i.e. 35000. Each case was studied acquiring and


- 6 -

analyzing 10000 couples of images. In table Table 1 the different experimental conditions are summarized.

Resolution Framed Area Spatial resolution ⎥⎦⎤

⎢⎣⎡cmpixel Spatial resolution ⎥

⎦

⎤⎢⎣

⎡sideframed

microscaleTaylor

1 2D x 2D ~ 170 λ/2D = 0.0217 2 3D x 3D ~ 110 λ/3D = 0.0145 3 4D x 4D ~ 85 λ/4D = 0.0108 4 5D x 5D ~ 65 λ/5D = 0.0087

Table 1: Parameters of present study In order to get a better knowledge of the near field of this jet, in figure Figure 2 the longitudinal mean velocity radial profiles have been plotted in correspondence to different downstream positions (0.1 D, 1 D, 2 D, 3 D, 4 D). First of all can be clearly seen the distinctive shape of the mean velocity exit profile of an orifice jet, with the two velocity peaks near the edges of the orifice (Mi et al. 2001). This is different from the top hat profile of a nozzle jet and the power law profile of a pipe

jet ⎟⎠⎞⎜

⎝⎛71 (Mi et al. 2001). Then the interaction between jet and surrounding fluid starts and

velocity tends to be distributed in the space and its profile to enlarge over and over.

-3 -2 -1 0 1 2 3-0.2

0

0.2

0.4

0.6

0.8

1

1.2

��

��

Axial velocity transversal profiles

0.1 D1 D2 D3 D4 D

Figure 2: Axial velocity transversal profiles in

correspondence to different positions downstream to the orifice

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 50.7

0.75

0.8

0.85

0.9

0.95

1

1.05

��

� ��

Axial velocity longitudinal profile in correspondence to the center of the jet

Res. 1Res. 2Res. 3Res. 4

Figure 3: Longitudinal velocity behaviour

To outline the effect of spatial resolution, figure Figure 3 shows the behavior of the longitudinal velocity along jet axis for the different resolutions. Here Uc and Umax denote the centerline mean velocity and its maximum value near to the exit. The position of the maximum velocity about one diameter downstream to the orifice is clearly seen, for all tested resolutions without significant differences. This is just an evidence of vena contracta phenomenon. Therefore, the centerline behavior of mean velocity does not seem to be influenced by the spatial resolution and this could be an indication of why this was not so much considered in the past. The effect of spatial resolution can be better pointed out by the transversal profiles of the statistical moments as the framed area varies, as presented in figures Figure 4 to Figure 10. Figures Figure 4 and Figure 5 show the transversal profiles of axial mean velocity respectively at 0.05 D and 2 D downstream to the orifice. In figure Figure 4, it can be noticed the slight difference between the results of the higher resolution setups (Resolution 1 and 2, Table 1) and the lower


- 7 -

resolution setups (Resolution 3 and 4). Data coming from the resolution 3 and resolution 4 setups are averaged onto a large framed area between the moving fluid inside the jet and the quite still ambient fluid.

-0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.50

0.2

0.4

0.6

0.8

1

1.2

��

��

Axial velocity transversal profile in correspondence to 0.05 D

Res. 1Res. 2Res. 3Res. 4Mi et al (2001)

Figure 4: Axial velocity transversal profile

0.05 D downstream to the orifice

-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1

0

0.2

0.4

0.6

0.8

1

1.2

��

��

Axial velocity transversal profile in correspondence to 2 D


Figure 5: Axial velocity transversal profile 2 D

downstream to the orifice

Present results are compared with the one by Mi et al. (2001), which is the outcome of a Hot Wire Anemometry (HWA) measurement on an orifice jet at Re=16000. Mi data, which are characterized by the higher spatial resolution of HWA 10-5 m, lie near to the ones among present results which have higher resolution (Resolution 1 and 2). Thus can be ensured the goodness of a proper PIV study, compared to HWA measurements, preserving the advantage of non-intrusivity. On the other hand, at the downstream profile presented in figure Figure 5, the effect of spatial resolution becomes less evident. The vena contracta phenomenon can also be observed in figure Figure 6 that represents the radial velocity transversal profile in the very near zone (0.05 D downstream to the orifice). In contrast with the results from all the other turbulent jets, the radial velocity preserves a univocal sign along each of the two sides of the jet which are determined by the axis in the center of jet. This means that the jet does not spread outwards, as the other jets, but rather reduces its section due to the contraction that persists downstream to the orifice plate because of the sudden transit through the small hole.


- 8 -

-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1-0.3

-0.2

-0.1

0

0.1

0.2

0.3

��

��

Radial velocity transversal profile in correspondence to 0.05 D


Figure 6: Radial velocity transversal profile 0.05 D downstream to the orifice

In the same manner seen before, the RMS shows a limited influence of the chosen spatial resolution. Figures Figure 7 - Figure 8 display the transversal profile of the root mean square of the axial component of velocity, in correspondence to a distance 0.1 D and 2 D downstream to the orifice. As for the mean velocity field, in figure Figure 7 the distinctive shape of the RMS exit profile, which is low at the center of the jet and high, with two sharp peaks, in correspondence to the edges of the orifice is noticed (Mi et al., 2001). Due to the different spatial resolution, these peaks appear highly sharp for the higher resolution data (Resolution 1) and then more and more smoothed as the spatial resolution decreases. A similar, but less evident, effect can be noticed in the results extracted in the radial profiles 2 D downstream, presented in figure Figure 8. Studying the higher order statistical moments the effect of spatial resolution remains similar. Figures Figure 9 and Figure 10, e.g., show the radial profiles of skewness and flatness nearly 2 D downstream to the orifice. The larger framed areas generate a result slightly different from the one resulting from the smaller areas, especially near the center of the jet. The larger is this zone the larger is the area on which each PIV resulting vector is averaged. The effect of this can be negligible for low order statistics but becomes more and more relevant as this order increases, because the differences in the results are raised to an increasing power, i.e. to three or four.


- 9 -

-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 10

0.05

0.1

0.15

0.2

��

� �

� �

Axial velocity RMS transversal profile in correspondence to 0.1 D


Figure 7: Axial velocity RMS transversal profile 0.1 D


-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 10

0.05

0.1

0.15

0.2

��

� �

� �

Axial velocity RMS transversal profile in correspondence to 2 D


Figure 8: Axial velocity RMS transversal profile 2 D


-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1-3

-2

-1

0

1

2

3

��

Skew

ness

Skewness transversal profile in correspondence to 2 D


Figure 9: Axial velocity Skewness transversal profile

2 D downstream to the orifice

-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 10

2

4

6

8

10

12

14

16

��

Flat

ness

Flatness transversal profile in correspondence to 2 D


Figure 10: Axial velocity Flatness transversal profile 2 D


3.2 Large scale and small scale route to isotropy In order to study the occurrence of large scale isotropy, a characteristic parameter can be evaluated.

It is the ratio of streamwise to vertical rms velocities, i.e. ⎟⎠⎞⎜

⎝⎛''vu ( Browne (1984), Djeridane(1996) ).

It is an indicator of the energy supplied to the large scales, due to the fact that the second order structure functions ( ) ( ) ( )xurxuu −+=2δ and ( ) ( ) ( )xvrxvv −+=2δ tend to 2u’ and 2v’ when

∞→r (Romano and Antonia (2001) ). If the phenomenon were isotropic at large scales this parameter should be equal to 1, being equivalent the transversal and the axial directions. Figure 11 shows the results of its evaluation.


- 10 -

0 2 4 6 8 10 12 14 160

0.5

1

1.5

2

2.5

��

� �

� �

� ��

��

Res. 1Res. 2Res. 3Res. 4Mi et al OP (2007)Romano Falchi OP (2010)Djeridane et al. LP (1993)Xu e Antonia SC (2002)Isotropy condition

Figure 11: Axial evolution of large scale isotropy parameter

For 5<Dx the flow field is observed to be not fully isotropic in correspondence to large scales.

Indeed the parameter tends towards a value near to 1.3 moving along jet axis. This fact is confirmed by the existing works on orifice jets (OP) of Mi et al. (2007) and Romano and Falchi (2010), both made with the PIV technique and having a pixel resolution of 100

cmpixel and 95

cmpixel respectively,

and also on long pipe jets (LP) (Djeridane (1996) ), made with Laser Doppler velocimetry technique, having a reference spatial resolution of ~ 10-4 m, and smooth contraction jets (SC) (Xu and Antonia (2002) ), made with Hot Wire Anemometry having a spatial resolution of ~ 10-5 m. No substantial influence seems to be related to the different spatial resolutions. Therefore, the large

scales, which mainly affect the results of ⎟⎠⎞⎜

⎝⎛''vu , have only a slight dependence on spatial

resolutions, as also observed from the analysis of low order statistical moments. With regard to the small scale isotropy, the isotropy parameters computable from the data available through a PIV study have been measured, namely K1, K3, K4, K9, and the results are shown in figures Figure 12 - Figure 15 compared with the ones of Romano and Falchi (2010) on the same jet in correspondence to a Re=20000 and having a pixel resolution of 95

cmpixel .

The parameter 2

2

1

⎟⎠⎞⎜

⎝⎛∂∂

⎟⎟⎠

⎞⎜⎜⎝

⎛∂∂

=

xu

yv

K compares the spatial evolution of longitudinal and transversal

velocities along their own direction. In an isotropic phenomenon all the directions are equivalent and therefore this ratio should be equal to 1, as given in equation 8. The results, shown in figure Figure 12, confirm this hypothesis, already in the near zone 5<

Dx , in correspondence to all

resolutions. In the very near zone spatial gradients are more intense and therefore the effects of spatial resolution are more pronounced there and tend to diminish moving downstream as the spatial gradients are smeared out by viscosity.


- 11 -

Figure 12: Axial evolution of small scale isotropy

parameter K1


parameter K3

The parameter 2

2

3

⎟⎠⎞⎜

⎝⎛∂∂

⎟⎟⎠

⎞⎜⎜⎝

⎛∂∂

=

xu

yu

K estimates the relative importance, with regard to the longitudinal

velocity component, of the transversal evolution with respect to the axial one. If the isotropy conditions were fulfilled this parameter should be equal to 2, as given in equation 8. From figure Figure 13 this condition doesn't seem to be fulfilled in the near zone as well as in the far, in agreement with Romano and Falchi results. Again the spatial resolution affects the very near zone and less all the rest of the jet.

2

2

4

⎟⎠⎞⎜

⎝⎛∂∂

⎟⎠⎞⎜

⎝⎛∂∂

=

xu

xv

K is corresponding and opposite to K3 in regard to axial evolution of transversal

velocity; the isotropic value is equal to 2 and again this condition doesn't seem to tend to be satisfied (figure Figure 14) in the studied zone in conformity with Romano and Falchi results. Once more the effect of a different spatial resolution is noticeable in the very near zone and to a lesser degree moving downstream along the axis of the jet.

The last studied parameter was29

⎟⎠⎞⎜

⎝⎛∂∂

∂∂

∂∂

=

xu

xv

yu

K . It depends on the cross derivatives of axial and

transversal velocities. The isotropy condition prescribes that its value should be equal to -0.5 and again doesn't seem to be satisfied, irrespective of the framed area, figure Figure 15.


- 12 -


parameter K4


parameter K9

4. Conclusions The near zone of a sharp edged round orifice jet has been investigated, focusing particularly on the influence of spatial resolution on the study of a jet based on the PIV technique. To this aim, both large and small scale features of the jet have been studied varying the size of the framed area through the object plane distance. The selected spatial resolutions are larger than what usually employed in PIV (except for the lower one). Moreover results on high order statistics and on spatial derivatives of orifice jets cannot be found in the literature. Regarding to the large scales, first of all the occurrence of the vena contracta phenomenon for an orifice jet has been shown in several ways. In addition, it has been highlighted the small influence of spatial resolution on low order statistical moments, which slightly rises as the order of the statistics becomes higher due to local averaging. This shows the importance of a proper choice of the setup, in particular as regard to the spatial resolution. Then small scale features of the jet have been examined, particularly with interest to the evaluation of spatial derivatives of velocity components and then to the fulfillment of isotropy hypotheses of the jet. Based on those data, an incomplete local isotropy condition is attained, although restricted to the near zone, and an influence of spatial resolution moderate on the overall behavior of spatial derivatives and on the deductions about symmetry hypotheses and high on the determination of local features of such a jet. 5. Bibliography • Batchelor G. K. (1946) The theory of axisymmetric turbulence. Proceedings of the Royal Society A, London 186: 480-502. • Batchelor G. K. (2000) An introduction to Fluid Dynamics. Cambridge Press, Cambridge, pp 387-392. • Billy F., David L. and Pineau G. (2004) Single pixel resolution correlation applied to unsteady flow measurements. Measurement Science and Technology 15:1039-1045. • Boersma B. J., Brethouwer G. and Nieuwstadt F. T. M. (1998) A numerical investigation on the effect of the inflow conditions on the self-similar region of a round jet. Physics of Fluids 10(4): 899-909. • Browne L. W., Antonia R. A., and Chambers A. J. (1984) The interaction region of a turbulent


- 13 -

plane jet. Journal of Fluid Mechanics 149: 355-373. • Chandrasekhar S. (1950) The theory of axisymmetric turbulence. Proceedings of the Royal Society A, London 242:557-577. • Djeridane T., Amielh H., Anselmet F. and Fulachier F. (1996) Velocity turbulence properties in the near-field region of axisymmetric variable density jets. Physics of Fluids 8(6): 1614-1630. • George W. K. and Hussein H. J. (1991) Locally axisymmetric turbulence. Journal of Fluid Mechanics 233: 1-23. • Hinze J. O. (1975) Turbulence Mc Graw Hill, New York, p 218. • Hussein H. J. (1994) Evidence of local axisymmetry in the small scales of a turbulent planar jet. Physics of Fluids 6(6): 2058-2070. • Kähler C. J. and Scholz U. (2006) Transonic jet analysis using long-distance micro PIV. 12th International symposium on flow visualization, Gőttingen, Germany. • Keane R. D., Adrian R. J. and Zhang Y. (1995) Super-resolution particle imaging Velocimetry. Measurement Science and Technology 6: 754-768. • King H. W. (1918) Handbook of Hydraulics. Mc Graw Hill, New York, pp 35-62. • Kolmogorov A. N. (1941) The local structure of turbulence in incompressible viscous fluid for very large Reynolds numbers. Doklady Akademii Nauk SSSR (reprinted in Proceedings: Mathematical and Physical Sciences (1991) 434: 9-13), 30: 301-305. • Lavoie P., Avallone G., De Gregorio F., Romano G. P. and Antonia R. A. (2007) Spatial resolution of PIV for the measurement of turbulence. Experiments in Fluids 43: 39-51. • Mi J., Nobes D. S. and Nathan G. J. (2001) Mixing characteristics of axisymmetric free jets from a contoured nozzle, an orifice plate and a pipe. Journal of Fluids Engineering - Transactions of the ASME, 2001 123(4): 878-883. • Mi J., Kalt P., Nathan G. J. and Wong C. Y. (2007) PIV measurements of a turbulent jet issuing from round sharp-edged plate. Experiments in Fluids 42: 625-637. • Michell J. H. (1890) On the Theory of Free Stream Lines Philosophical transactions of the Royal Society A 181: 389-431. • Quinn W. R. (1989) On mixing in an elliptic turbulent free jet. Physics of Fluids 1(10): 1716-1722 • Quinn W. R. (2005) Upstream nozzle shaping effects on near field flow in round turbulent free jets. European Journal of Mechanics B/Fluids 25: 279-301. • Romano G. P. and Antonia R. A. (2001) Longitudinal and transverse structure functions in a turbulent round jet: effect of initial conditions and Reynolds number. Journal of Fluid Mechanics 436: 231-248. • Romano G. P. and Falchi M. (2010) Recovering isotropy in turbulent jets. 8th International ERCOFTAC Symposium on engineering turbulence modelling and measurements, Marseille. • Scarano F. (2002) Iterative image deformation methods in PIV. Measurement Science and Technology 13: R1-R19. • Scarano F. (2003) Theory of non-isotropic spatial resolution in PIV. Experiments in Fluids 35: 268-277. • Stitou A. and Riethmuller M. L. (2001) Extension of PIV to super resolution using PTV. Measurement Science and Technology 12: 1398-1403. • Taylor G. I. (1935) Statistical theory of turbulence. Proceedings of the Royal Society of London. Series A, Mathematical and Physical Sciences 151 (873): 421-444. • Westerweel J., Dabiri D. and Gharib M. (1997) The effect of a discrete window offset on the accuracy of cross-correlation analysis of digital PIV recordings. Experiments in Fluids 23: 20-28. • Westerweel J., Geelhoed P. F. and Lindken R. (2004) Single-pixel resolution ensemble correlation for micro-PIV applications. Experiments in Fluids 37: 375-384. • Xu G. and Antonia R. A. (2002) Effect of different initial conditions on a turbulent round free jet. Experiments in Fluids 33: 677-683.

effects of spatial resolution on piv investigation of a turbulent orifice...

Documents