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, Michigan State University, East Lansing, Michigan, USA b Institute of Industrial Science, University of Tokyo, Tokyo, Japan ABSTRACT: A method for the simultaneous measurement of velocity and passive scalar concentration fields by means of particle image velocimetry (PIV) and planar laser induced florescence (PLIF) techniques is described here. An application of the combined PIV-PLIF system is demonstrated by performing simultaneous velocity and concentration measurements 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; PLIF; PIV-PLIF system; jet flow; Reynolds stress; turbulence; flux correlation; concentration INTRODUCTION Simultaneous information concerning a passive scalar property and a velocity field is required in many fluid flow investigations, including mixing in combustion chambers and distributions of drugs in biomedical applications. The possibility of measuring velocity and a scalar at the same time, with a high spatial and/or temporal resolution is also of fundamental importance for the validation and development of models for 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; that is, 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 measurement of velocity and concentration fields. With the rapid development of modern optical techniques and digital image processing techniques, whole-field optical diagnostics, such as PIV and PLIF techniques, are assuming ever-expanding roles in the diagnostic probing of fluid mechanics. The Address for correspondence: Hui Hu, Ph.D., A22, RCE Building, Department of Mechanical Engineering, Michigan State University, East Lansing, MI 48824, USA. huhui@egr.msu.edu Ann. N.Y. Acad. Sci. 972: 254 259 (2002). 2002 New York Academy of Sciences.
HU et al.: TURBULENT JET MIXING FLOW 255 advances of PIV and PLIF techniques in recent years have lead them to be mature techniques for whole-field measurement of velocity and concentration and/or temperature in an objective plane or over the volume of an objective fluid flow. The development of a PIV-PLIF combined system is described here for the simultaneous measurement of velocity and concentration in the near field of a turbulent jet mixing flow and the determination of the correlation terms between the fluctuating velocities and concentration. EXPERIMENTAL SETUP AND TECHNIQUES The experimental set up contained a circular test nozzle (D = 30mm) installed in the middle of a water tank (600 600 1,000mm 3 ). Fluorescent dye (rhodamine B, concentration of about 0.3mg/liter) for PLIF or PIV tracers (hollow glass particles d = 8 12µm) 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 number (Re) values, and was measured by a flow meter. A cylindrical plenum chamber with comb structures was installed upstream of the test nozzle to insure the jet flow had fully developed. The turbulent levels of the core jet flows at the exit of test nozzles were about 3%. An overflow system was used to keep the water level in the test tank constant during the experiment. The investigation region is at the near field of the jet flow (Y/D < 5.0). The distance between the exit of the test nozzle and the free surface of the water in the test tank was about 30D. 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 was negligible, and the jet flow exhausted from the test nozzle was considered to be a free jet. During the experiment, the core jet velocities (U 0 ) at the exit of the test nozzle was set to be about 0.20m/sec. The Re value of the jet flow, based on the nozzle exit diameter and the core jet velocity, was about 6,000. Pulsed illumination laser sheets (thickness about 1.5mm) were generated by a double-pulsed Nd:YAG laser system (Quantel Inc.). The frequency of the doublepulsed illumination was 10Hz. The pulsed illumination duration was 4nsec, and power 200mJ/pulse. The time interval between the two pulses was adjustable, set at 3msec in the present study. A simultaneous image recording system was designed by using optics and two high-resolution CCD cameras (TSI PIVCAM10-30, 1K by 1K resolution). The emission peak of rhodamine B is about 590nm, and the wavelength of the illuminating laser light scattered by the PIV tracer particles is 532nm. Two kinds of optical filters were used to separate LIF lights from scattered laser light, and then recorded separately to obtain PLIF and PIV image simultaneously. A bandpass optical filter (532 ± 5nm) was installed at the head of camera #1: only the scattered laser light was transmitted to form the PIV image on the CCD sensor of camera #1, and LIF light was blocked out. A high pass filter (greater than 580nm pass) was installed in the head of the camera #2 to filter out the scattered laser light (wavelength 532nm). The LIF light (peak at 590nm) passed through the optical filter to generate the LIF image on the CCD censor of camera #2. Rather than tracking individual particles, an improved spatial correlation analysis method, named the hierarchical recursive PIV method, 1 was used for PIV image
256 ANNALS NEW YORK ACADEMY OF SCIENCES processing. The hierarchical recursive PIV method is actually a recursive operation process of conventional spatial correlation. The recursive operation starts with a large interrogation window size and search distance, the same as the conventional correlation analysis based PIV image processing methods. By using the results of the former iteration step as the approximate offset values in the next iteration step, the interrogation window size and search distance were reduced hierarchically. Whereas the conventional correlation method used 64 by 64 pixels or 32 by 32 interrogation windows, the hierarchical recursive PIV method could reduce the final interrogation window to 8 by 8 pixels with spurious vectors being less than 2%. In order to obtain a whole field quantitative concentration distribution in the objective flow from PLIF images, an improved whole-field calibration procedure 2 was conducted to account for the non-uniformity of the laser sheet. To improve the accuracy level of the PLIF measurements, the averaged background was subtracted from the LIF images. All the PLIF images were also normalized to account for the laser sheet intensity variations. A general mapping method 3 was used to obtain the spatial correlation between the PIV and PLIF images. The spatial resolution of PIV results was determined by the sizes of the interrogation window used for the correlation operation. The final interrogation window size for PIV image processing was 8 by 8 pixel; therefore, 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 was relatively straightforward to calculate the various ensemble-averaged velocity (U, V), turbulent velocity fluctuations ( ( uu), ( vv) ), mean concentration (c), concentration standard deviation ( ( cc) ), and the turbulent flux correlation terms ( uc, vc ) between the velocity and concentration. RESULTS AND DISCUSSIONS FIGURE 1 represents a typical pair of the instantaneous PIV and PLIF measurement results. Since the final interrogation window size was 8 by 8 pixels for PIV image processing, about 50,000 vectors could be obtained for every instantaneous FIGURE 1. Typical instantaneous measurements from the combined PIV PLIF system: (A) PIV measurements; (B) simultaneous PLIF measurements.
HU et al.: TURBULENT JET MIXING FLOW 257 PIV frame. The velocity vectors shown in FIGURE 1A display only 25% of the PIV velocity vectors. FIGURE 1B shows the instantaneous concentration field obtained by PLIF image processing, which is the simultaneous measurement result of the PIV results shown in FIGURE 1A. The contour levels given in the figure represent rhodamine B concentration levels normalized by the jet source concentration ξ 0 = 0.3mg/l. It is well known that the shear layers in jet flows are unstable via Kelvin Helmholtz instability. The instability grows downstream and rolls 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 of these processes can clearly be seen from the PIV-PLIF simultaneous measurement results shown in FIGURE 1. Two hundred and fifty PIV and PLIF image pairs, captured simultaneously at a frame rate of 10Hz, were used to calculate the ensemble-averaged values. The ensemble-averaged values were also normalized with the core jet velocity U 0 = 0.20m/sec and jet source concentration ξ 0 = 0.3mg/l. FIGURE 2 shows the profiles of ensemble-averaged velocity and mean concentration profiles in the three downstream locations of the turbulent jet flow. The measurements of Lemoine et al., 4 who used single point measurement techniques (LDV and LIF) downstream of a circular jet flow at the same Re value are also given in the figures. It can be seen that the present PIV and PLIF simultaneous measurements agree with the results of Lemoine et al. reasonably well. FIGURE 3 presents the distributions of various ensemble-averaged terms, which include ensemble-averaged velocity, ensemble-averaged concentration, turbulence intensity, concentration standard deviation and the radial and axial turbulent flux terms ( uc, vc ). The ensemble-averaged velocity and mean concentration distributions indicate that there is a high speed and high concentration region in the center of the jet turbulent flow, also called the potential core region. High turbulence intensity and high concentration fluctuation regions exist in the shear layers between the jet flow and ambient flows, whereas both the turbulence intensity and concentration fluctuation values are low in the potential core region. The potential core region extended to Y/D > 4.0 downstream, which is consistent with the finding that the length of the potential core region of a conventional circular jet flow ranges between FIGURE 2. Ensemble-averaged concentration and velocity profiles.
258 ANNALS NEW YORK ACADEMY OF SCIENCES 4D and 6D.5 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 vc and radial turbulent flux uc were found to be almost of the same order as the present measured results. FIGURE 3. Ensemble-averaged measurements from the combined PIV PLIF system: (A) mean concentration distribution; (B) mean velocity distribution; (C) concentration standard deviation distribution; (D) tubulence intensity; (E) radial turbulent flux term distribution; (F) axial turbulent flux term distribution.
HU et al.: TURBULENT JET MIXING FLOW 259 SUMMARY The development of a high resolution PIV-PLIF combined system that could provide simultaneous measurement of velocity and concentration in fluid flow, is described. The PIV tracer particles and fluorescent dye (rhodamine B) were premixed in an objective fluid 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 were presented as the measured results of the simultaneous PIV-PLIF system. REFERENCES 1. HU, H., et al. 2000. Improve the spatial resolution of PIV results by using hierarchical recursive operation. Proc. of 9th Int. Symp.on Flow Visualization, Edinburgh, Scotland, UK. 2. HU, H., et al. 2000. Simultaneous velocity and concentration measurements in a turbulent jet flow by using PIV PLIF combined system. Proc. of 4th JSME-KSME Joined Thermal Eng. Conf., Kobe, Japan. 3. SOLOFF, S.M., et al. 1997. Distortion compensation for generalized stereoscopic particle image velocimetry. Meas. Sci. Tech. 8: 1441 1454. 4. LEMOINE, F., et al. 1996. Simultaneous concentration and velocity measurements using combined laser-induced fluorescence and laser doppler velocimetry: application to turbulent transport. Exp. Fluids 20: 341 327. 5. HINZE, J.O. 1959. Turbulence. McGraw-Hill, New York.