Proceedings of International Symposium on Visualization and Image in Transport Phenomena, Turkey, -9 Oct. SIMULTANEOUS VELOCITY AND CONCENTRATION MEASUREMENTS OF A TURBULENT JET MIXING FLOW Hui HU a, Tetsuo SAGA b, Toshio KOBAYASHI b and Nobuyuki TANIGUCHI b a Department of Mechanical Engineering, A, Research Complex Engineering Building, Michigan State University, East Lansing, Michigan 88, USA. Email:huhui@egr.msu.edu b Institute of Industrial Science, University of Tokyo, Tokyo -8, Japan In the present paper, a method for the simultaneous measurements of velocity and passive scalar concentration fields by means of Particle Image Velocimetry (PIV) and Planar Laser Induced Florescence (PLIF) techniques is described. An application of the PIV-PLIF combined system is demonstrated by performing simultaneous measurements of velocity and concentration in the near field of a turbulent jet mixing flow. The distributions of the ensemble-averaged velocity and concentration, turbulent velocity fluctuation, concentration standard deviation and the correlation terms between the fluctuating velocities and concentration in the near field of the turbulent jet flow are presented as the measurement results of the simultaneous PIV-PLIF system. KEYWORDS: PIV technique, PLIF technique, PIV-PLIF combined System, Jet flow INTRODUCTION Simultaneous information of a passive scalar and velocity field is desirable in many fluid flow investigations like mixing in combustion chambers or distributions of drugs in biomedical applications. The possibility of measuring velocity and a scalar at the same time with high spatial and/or temporal resolution is also of fundamental importance for the validation and development of models of turbulence and turbulent mixing. For example, in a turbulent jet mixing flow, the species concentration field is determined by molecular diffusion and transported by the turbulent flow field. When considering the Reynolds-averaged scalar conservation equation, the effects of turbulent transport appear in terms of the correlation between the concentration and velocity fluctuations, i.e. expressions such as u 'c' and v 'c'. Experimental characterization of these correlation terms is needed for the development and validation of physical models. This requires the simultaneous measurements of the velocity and concentration fields. With the rapid development of modern optical techniques and digital image processing techniques, whole-field optical diagnostics, such as Particle Imaging Velocity (PIV) and Planar Laser Induced Fluorescence (PLIF) techniques, are assuming everexpanding roles in the diagnostic probing of fluid mechanics. The advances of PIV and
PLIF techniques in recent years have lead them to be mature techniques for the wholefield measurements of velocity and concentration or/and temperature in an objective plane or over a volume of an objective fluid flow. In the present paper, the development of a high-resolution PIV-PLIF combined system, which can achieve the simultaneous measurements of instantaneous spatial distribution of velocity and concentration in a fluid flow, will be described. The PIV-LIF combined system is applied to do simultaneous measurements of velocity and concentration in the near field of a turbulent jet mixing flow. As the measurement results of the simultaneous PIV-PLIF system, the distributions of the mean and RMS fluctuation of velocity and concentration, will be presented, together with the correlation terms between the fluctuating velocities and concentration in the near filed of the turbulent jet flow. EXPERIMENTAT SETUP AND TECHNIQUES Figure shows the schematically experimental set-up used in the present study. A test circular nozzle (D=mm) was fixed in the middle of a water tank (6mm*6mm*mm). Fluorescent dye (Rhodamine B, concentration is about.mg/liter) for PLIF or PIV tracers (hollow glass particles d=8 ~ μm) were premixed with water in a jet supply tank, and jet flow was supplied by a pump. The flow rate of the jet flow, which was used to calculate the representative velocity and Reynolds numbers, was measured by a flow meter. A cylindrical plenum chamber with comb structures was installed at the upstream of test nozzle to insure the jet flow to be fully developed and the turbulent levels of the core jet flows at the exit of test nozzles were about %. An overflow system was used to keep the water level in the test tank to be constant during the experiment. In the present study, the investigation region is at the near field of the jet flow (<.). The distance between the exit of the test nozzle and the free surface of the water in the test tank is about D. Therefore, the effect of the water free surface in the test tank on the vortical and turbulent structures in the near field of the jet flow is negligible, and the jet flow exhausted from the test nozzle is considered to be a free jet. During the experiment, the core jet velocities (U ) at the exit of the test nozzle was set to be about.m/s. The Reynolds numbers of the jet flow, based on the nozzle exit diameter and the core jet velocity is about 6,. Pulsed illumination laser sheets (thickness is about.mm) were generated by a double-pulsed Nd:YAG Laser system (Quantel Inc.). The frequency of the double-pulsed illumination is Hz. The pulsed illumination duration is ns, and power is mj/pulse. The time interval between the two pulses is adjustable, which is ms for the present study. A simultaneous image recording system was designed by using optics and two highresolution CCD cameras (TSI PIVCAM-, K by K resolution), which was shown schematically on the right upper corner of Fig.. Since the emission peak of Rhodamine B is about 9nm, and the wavelength of the illuminating laser light scattered by the PIV tracer particles is nm. Two kinds of optical filters were used to separate LIF lights from scattered laser lights, and then recorded separately to obtain PLIF and PIV image simultaneously. A bend pass optical filter (nm±) was installed at the head of the camera #, only the scattered laser light is transmissible to form PIV image on the CCD sensor of the camera #, and LIF light is blocked out. A high pass filter (>8nm pass) was installed in the head of the camera # to filter out the scattered laser light (wavelength nm). The LIF light (peak at 9nm) pass through the optical filter to
generate LIF image on the CCD censor of the camera #. Rather than tracking individual particles, an improved spatial correlation analysis method, named as Hierarchical Recursive PIV method (), was used in the present study to conducted PIV image processing. The Hierarchical Recursive PIV method is actual a hierarchical recursive process of conventional spatial correlation method. The recursive operation started with a large interrogation window size and search distance, which is as the same as conventional correlation analysis based PIV image processing methods. By using the results of former iteration step as the approximate offset values in the next iteration step, the interrogation window size and search distance were reduced hierarchically. The conventional correlation method always used 6 by 6 pixel or by interrogation windows, the hierarchical recursive PIV method can reduce the final interrogation window up to 8 by 8 pixel with spurious vectors being less than %. In order to obtain whole field quantitative concentration distribution in the objective flow from LIF images, an improved whole-field calibration procedure () was conducted to account for the laser sheet non-uniformity. In order to improve the accuracy level of the PLIF the measurement results, averaged background substracted from the LIF images. All the PLIF images were also normalized to account for the laser sheet intensity variations. A general mapping method () was used in the present to get the spatial correlation between the PIV and PLIF images. Since the spatial resolution of PIV results is determined by the sizes of the interrogation window used for correlation operation. The final interogation window size is 8 by 8 pixel for the present study, so the concentration data were also averaged over 8 by 8 subwindows during the PLIF image processing. Once the velocity and concentration fields were calculated, it is relatively straightforward to calculate the various ensemble-averaged velocity ( U, V ), turbulent velocity fluctuations ( ( u u ), ( v v ) ), mean concentration (C), concentration standard deviation ( c ' c ' ) and the turbulent flux terms ( u c, v c ) which is the correlation terms between the velocity and concentration. EXPERIMENTAL RESULTS AND DISCUSSIONS Figure shows a typical pair of the instantaneous PIV and PLIF measurement results. Since the final interrogation window size is 8 by 8 pixel for PIV image processing, about, vectors can be got for every instantaneous PIV frame. The velocity vectors shown in Fig. (a) displays only % of the PIV velocity vectors. Fig. (b) shows the instantaneous concentration field obtained by PLIF image processing, which is the simultaneous measurement result of the PIV results shown on Fig. (a). The contour levels given in the figure represent Rhodamine B concentration levels normalized by the jet source concentration ξ =.mg/l. It is well known that the shear layer origin form the exit of the test nozzle is unstable via Kelvin-Helmholtz instability for the circular jet flow. The instability grows downstream and rolled up into coherent vortex rings. The vortex ring structures merge as they move downstream and then break down into small vortex structures. The transition of the jet flow into turbulence occurs when the large vortex rings break down into small-scale vortices. All these processes can be seen clearly from the PIV-PLIF simultaneous measurement results shown in Fig.. In the present study, PIV and PLIF image pairs captured simultaneously at the frame rate of Hz were used to calculate the ensemble-averaged values. The ensemble-
averaged values were also normalized with the core jet velocity U =. m/s and jet source concentration ξ =. mg/l. Figure shows the profiles of ensemble-averaged velocity and mean concentration profiles in the three downstream locations of the turbulent jet flow. The measurement results of Lemoine et al. () by using single point measurement techniques (LDV and LIF) at D downstream of a circular jet flow at the same Reynolds level as the present case have also been given in the figures. It can be seen that the present PIV and PLIF simultaneous measurement results agreed with the measurement results of Lemoine et al. reasonable well. Figure shows the distributions of various ensemble-averaged terms, which include ensemble-averaged velocity ( U, V ), ensemble-averaged concentration (C), turbulent intensity ( ( u u + v v ) ), concentration standard deviation ( c ' c ' ) and the radial and axial turbulent flux terms ( u c, v c ). From the ensemble-averaged velocity and mean concentration distributions, it can be seen that there exits a hight speed and high concentration region in the center of the jet trublent flow, which is called potential core region. High turbulent intensity and high concentration fluctuation regions exist in the shear layers between the jet flow and ambinet flows, while in the potiential core region both turbulent intensity and concentration fluctuation values are low. The pontial core region extended to >. downstream, which is consistent with the result of that the length of potential core region of a conventional circular jet flow ranges about D to 6D (). Although the ensemble-averaged axial velocity component is much bigger than the radial velocity component in the near field of the turbulent jet flow, the axial turbulent flux v ' c ' and radial turbulent flux present measurement results. u ' c ' were found to be almost at the same order from the SUMMARY The development of a high resolution PIV-PLIF combined system, which can achieve the velocity and concentration simultaneous measurements of fluid flows, was described in the present paper. The PIV tracer particles and fluorescent dye (Rhodamine B) were premixed in an object flow and the objective fluid flow was illuminated by a double pulsed Nd:YAG laser. The LIF light and scattered illuminating laser light were separated successfully by using two kinds of optical filters, and recorded simultaneously by two high-resolution CCD cameras. The system was applied to measure the velocity and concentration distributions in the near field of a circular jet flow simultaneously. The distributions of the ensemble-averaged velocity and concentration, turbulent velocity fluctuation, concentration standard deviation and the correlation terms between the fluctuating velocities and concentration are presented as the measurement results of the simultaneous PIV-PLIF system. REFERENCES. Hu H. et al.. In Proc. of 9th International Symposium on Flow Visualization, Edinburgh, Scotland, UK, Aug. -... Hu H., et al.. In Proc. of th JSME-KSME Jointed Thermal Engineering Conference, Kobe, Japan, Oct. -6,.. Soloff S. M., Adrian R. J. and Liu Z. C. 997, Measurement Science and Technology, 8:-.. Lemoine F., Wolff M. and Lebouche M, 996. Experiments in Fluids. :-7.. Hinze, J. O. 99, Turbulence, McGraw-Hill Book Company
Laser sheet water tank low pass optical filter mixing region Beam splitter Double-pulsed Nd:YAG Laser CCD camera # mirror High pass optical filter for PLIF Overflow system A cylindrical plenum chamber synchronizer Test nozzle pump jet supply tank reserve tank for PIV CCD camera # Flowmeter Figure. Experimental system setup concentration.....9.9.8.8.7.7.6.6.............. m/s.... - - - - a. PIV measurement results b. simultaneous PLIF measurement results Figure. Typical instantaneous meansurement results of the PIV-PLIF combined system.. ensemble-averaged axial velocity V ensemble-averaged concentration, C D D D lemoine et al..8.6.8 D D D lemoine et al..6........ a. ensemble-averaged concentration ensemble-averaged axial velocity Figure. ensemble-averaged concentration and velocity profiles
.. mean concentration.9.9.8.8.7.7.6.6.............. m/s... - - - - a. mean concentration distribution concentration standard deviation... Turbulent intensity (u'u'+v'v') /..8.6....8.6....8.6.... - b. mean velocity distribution..8.6....8.6....8.6....8.6.. - - - c. concentration standard deviation c'c' distribution d. tubulent intensity (u' +v' ).. axial turbulent flux radial turbulent flux v'c' u'c'..6..8...6. -. -.6 -. -. -.8 -. -.6 -.... -....9.7....9.7....9.7.... - - - e. radial turbulent flux term c'u ' distribution f. axial turbulent flux term c'v' distribution Figure. Ensemble averaged meansurement results of the PIV-PLIF combined system