A REVIEW ON PIV WITH IMAGE MEASURING TECHNIQUES*

Chen Gang, Le Jing, Xu Lianfeng, Shao Jianbin, Jin Shanghai and Li Jianzhong

Xi’an University of Technology, China

Address for correspondence: Professor GANG CHEN, The Institute of Hydraulics,

Xi’an University of Technology, Xian, Shaanxi Province, 710048, China

Tel/Fax: +86-29-3283506, E-mail: chen_g@mail.xaut.edu.cn

Abstract: Particle Image Velocimetry (PIV) and its image measurement techniques have been rapidly developed with the drastic development of image processing method and computers. As a tool for measuring the whole instantaneous flow field without contacting the flow, PIV and related image measurement techniques have been extensively applied to automated measurements of multiphase flows, turbulence and thermal flows with reliable accuracy. Related techniques and their applications are briefly reviewed in this paper.

Keywords: image measurement, Particle Image Velocimetry (PIV), multiphase flow

1  INTRODUCTION

It is well known that particles and dye materials seeded in a flow can visualize the fluid flow. Pathline lengths or moving distances of all particles in a whole flow field can be measured with particle images at a time interval to derive velocity vector field. With this simple concept, an advanced tool for automated measurement of fluid flow, called Particle Imaging Velocimetry (PIV in acronym) has been developed with the noticeable development of modern computer techniques. PIV has been used to measure instantaneous velocity vector fields from slow flows to supersonic flows during the past ten years (Adrian, 1991, Raffel & Kompenhans, 1995, Willert, et al, 1996, Dracos, 1996, Raffel, et al, 1998). In contrast to the conventional methods for one point measurement such as the Pitot tube, the hot wire anemometer and the laser Doppler velocimeter, PIV can carry out two-dimensional and three-dimensional instantaneous velocity measurement with contact free. The velocity vector map obtained by PIV enables extraction of physical information such as pressure field, vorticity field, etc.

Combining PIV algorithm and other image processing techniques, PIV measurement can be extended to research on the moving-boundary flows, multiphase flows and atomization in high-speed flows. A survey article of particle velocimetry confirms that PIV has been rapidly advanced in its fundamentals and applications to multiphase flows, thermal flows, turbulence structures, etc (Adrian, 1996). In the present paper, PIV with image measurement techniques is concisely reviewed.

2  CURRENT IMAGE MEASUREMENT TECHNIQUES

2.1  Classification of piv and their features

In the previous researches, the method acquiring velocities at grids using high-density distribution patterns of particle images is referred to PIV, and the method using each particle tracking for low particle number density is referred to PTV. In this paper, PIV is used as a general term of the velocimetry using particle images. When particles have a good behavior of traceability to a fluid flow, the particle velocities usually represent the local fluid velocities. If the particles do not follow a flow, the particle velocities do not represent the local fluid velocities.

It is known that there is not a general-use type of standard PIV system. Each type of PIV needs the fittest hardware and software to measure a flow field. Even though many types of PIV work in the present days, they include the common processes as the following operations: seeding the flow of fluid for visualization, illuminating the measurement space, photographing the visualized flow or recording the flow images digitally using a video system, and finally processing the recorded images to calculate flow velocity vectors. These operations consist of the following elements: tracer particles whose size is small enough to follow the concerning fluid flow with or without turbulence (several microns in diameters), light source for which a high power laser is usually used to illuminate the two-dimensional flow space as a thin light sheet, a charge-coupled device camera, i.e., CCD camera for recording the images and a digital image processor consisting of computer and its software. According to principles of flow velocity calculation based on the image processing, a classification of PIV is shown in Table 1.

Besides the PIV systems in Table 1, Holographic PIV (HPIV) (Meinhart, et al, 1993, Barnhart, et al, 1995, Sheng & Meng, 1997, Meng, et al, 1997) and Image-shifting PIV (IPIV) (Raffel & Kompenhans, 1995), both of which belong to the optical method, have made an important contribution to the recent flow measurement field. HPIV can track millions of particles in three-dimensional space, which is much more than several hundred particles seeded in other PIV techniques. The particle size for HPIV and IPIV must be restricted to a few microns to avoid overloading the fluid. An advanced algorithm is proposed to process larger set of data in HPIV at high speed than the cross-correlation methods (Sheng & Meng, 1997). IPIV techniques are especially effective for high-speed flows, which demand short time intervals of the order of a few microseconds between the exposures. This may overcome the demerit of high-framing-speed video techniques, which drastically impair the spatial resolution in the PIV recordings. Most widely used image-shifting method adopts a rotating mirror system (Raffel & Kompenhans, 1995). It permits shift velocities exceeding 500m/s without any noticeable reduction in the optical quality of the images (Raffel & Kompenhans, 1995).

2.2  Other topics of PIV and image measurement

               Table 1  Some current PIV techniques

It is necessary to extract physical information from velocity data. Velocity vectors are obtained at particle positions when the tracking techniques are applied. They are rearranged at needed grid points in order to calculate physical properties such as stream function, vorticity and pressure. Three methods have been proposed by Yamamoto et al (1996a), who introduced some numerical methods used in modern computational fluid dynamics (CFD) to the rearrangement processing. Recently a hybrid system combining PIV and CFD has also been reported by his group (Ido, et al, 1997). Their hybrid system was developed to restore the velocity vectors in the flow regions in which velocity information lacks. The original velocity vectors obtained from PIV measurement are not changed in this disposal process. For this purpose, any basic relations and models in fluid dynamics, such as continuity equation and Laplace equation, and some numerical methods in CFD can be adopted justifiably in the hybrid system.

Visualization techniques are often concerned in image measurement. Visualization techniques, such as the methods using oil film, dye materials and tracer particles, are often utilized to show physical structures of flows including vortex sizes, separation positions on solid walls, and also distributions of bubbles and particles. Visualized images can be analyzed using image processing techniques to obtain some quantitative information.

3  APPLICATIONS OF PIV AND IMAGE MEASUREMENT

3.1  Multiphase flows and waves

Multiphase flows can be diagnosed by video cameras and image processing. Flow images may be recorded by ordinary charge coupled digital cameras and high-speed video cameras with frame rates up to several thousands fps, and the recorded images can be analyzed by PIV and image processing techniques (Hewitt, et al, 1990, Reese, et al, 1995, Crowe et al, 1998). Multiphase flow parameters, such as void fraction in gas-liquid flow, particle concentration in gas-solid and liquid-solid flow can be extracted by PIV techniques. For instance, a paper on simultaneous velocity measurements of both components of a gas-liquid two-phase flow by PIV was reported (Hassan, et al, 1992). This process contains a separation treatment of overlapped deformable bubbles and drops with edge detection (Yamamoto, et al, 1997, Canny, 1986), and also includes separation between tracer particles and bubbles (drops and particles) in multiphase systems (Song, et al, 1996). PIV measurement of three-dimensional distribution of void fraction in bubble plume flow was reported by Murai et al (1997). Yamamoto et al (1997) also investigated particle number flow rate measurement with PIV. Recently, PIV and image measurement techniques were also used to investigate high-efficiency sand transportation with spiral air flows in pipeline. Particle moving behaviors and sand plug features can be analyzed by image measurement techniques (Miyazaki, et al, 1999a and 1999b).

Shockwave in air-water flow can be shown by visualization. Photographs allow a comparison between aerated and non-aerated flows across a shock, and show a detail of the shock front in the direction of flow (Reinauer and Hager, 1996). For studying the effects of surface disturbances on the entrainment of bubbles by a liquid jet, a digital CCD camera with a resolution of 780X480 pixels for recording images and a strobe for illumination were used, and bubble size distributions were obtained by analyzing the recorded images using some image processing algorithms ( , 1998). Turbulence generation due to breaking water waves was investigated using the techniques of visualization and PIV. The periodicity of the breaking waves, mean velocities and turbulence intensities based on repeated measurements were obtained (Chang and Liu, 1999). The phenomenon of a liquid jet released under gravity and falling through or impacting onto another liquid before colliding with an obstructing solid surface can be studied with visualization techniques, and the volume of air entrained can be estimated from photographs by measuring the sizes of the air bubbles using image processing algorithms (Storr and Behnia, 1999).

3.2  Thermal flow

Thermo-sensitive tracer particles are usually used for simultaneous measurement of velocity and temperature in a thermal flow. Temperature field is obtained by calculating the intensity of colors on the same particle images used to calculate velocity. Optical properties and time response of the tracer particles should be checked beforehand. Sometimes two kinds of particles are used in the same flow: one is for velocity tracking and the other for temperature detection. A technique using thermo-sensitive micro-capsulated liquid crystal particles suspended in liquid was proposed by Kobayashi et al (1992). This technique was applied to a thermal buoyant water jet (Kobayashi, et al, 1995). Kimura et al (1997) reported that three-dimensional temperature distribution could be constructed by interpolating two-dimensional distributions, which adopted a color-to-temperature transformation algorithm using a multi-layer feed-forward neural network in investigating natural convection in a rotating cylindrical cell. An analysis of buoyancy and thermocapillary flow was made using PIV with liquid crystal tracers (Wozniak & Wozniak, 1994). PIV application to natural convection in water heat storage vessel was also reported (Dahl, et al, 1995).

3.3  Flow structure

In separation flows, particles must be carefully seeded in a shear layer so that a separation region may contain a satisfactory number of particles. Rearrangement of velocity vectors from particle points to grid points becomes necessary for revealing some flow phenomena because of the lack of effective particle images in these regions. An investigation into the separation flow around airfoil shows that velocity data may miss from some critical regions on the PIV pictures. But calculation of the out-of-plane vorticity contours from the acquired PIV velocity fields allows details of the structure and locations of the vortices involved in a thin reverse-flow layer on the airfoil upperside (Wernert, 1996). Backward-facing-step flows were successfully measured to determine vortex structure in the recirculating region with hydrogen bubbles and oxygen bubbles as tracers (Ma, et al, 1995). Recently microburst, which is modeled by releasing a small volume of heavy fluid into a large tank filled with a lighter fluid, was measured with PIV including the effect of the difference of refractive indices due to density differences up to 4%, and vortex structure was detected (Alahyari & Longmire, 1994).

Turbulence measurement is a challenge to PIV. Particle traceability and measurement accuracy are important problems to be solved in turbulence measurement. As one of the most fundamental flows, turbulent boundary layer was measured by the PIV technique. About one hundred frames were averaged to construct the velocity distribution in the logarithmic region. Plots of the fluctuation quantities and Reynolds stresses demand over one thousand of frames for reliable statistics (Willert, et al, 1996). Feasibility of a PTV technique reported by Malik et al (1993) permits a change of search radius in a target frame based on the turbulent velocity fluctuation in rms (root mean square). In the measurement of turbulence in a channel, accuracy of fluctuating velocity may reach 1% of the full-scale mean velocity, and the encouraging good agreement of the accuracy with LDV measurements and direct numerical simulations was reported (Adrian, 1991).

3.4  Atomization flow

Image measurement techniques also play an important role in the study of atomization flows. By photographing the prototype flow patterns of jet nappe atomization of a large hydropower station, images of flow patterns of atomization can be obtained. The atomization images were analyzed with digital image processing techniques. Assuming image brightness is proportional to atomization intensity or density of droplet number, one can evaluate distribution contours of atomization intensity by calculating brightness contour maps (Hu, 1994). It is also possible to evaluate distribution contours of atomization intensity in a three-dimensional space by processing flow images using the method of three-dimensional reconstruction (Zhou, et al, 1995). Instantaneous break-up process of a round water jet by a high-speed annular air jet was visualized to study the underlying physical mechanisms involved in the primary break-up of the water jet. Visualization revealed that the break-up process consists of the stripping of water sheets, or ligaments, which subsequently break into smaller lumps or drops (Lasheras, Villermaux and Hopfinger, 1998).

4  CONCLUSIONS

It is undoubted that PIV and image measurement techniques are extensively applied to fluid dynamics, and contribute to modern experimental fluid mechanics. Algorithms of PIV basically contains correlation method and particle tracking method. It is also important to develop the techniques for delecting spurious velocity vectors and extracting physical informations such as vorticity, pressure and stream function. The function of PIV has been rapidly extended to measure multiphase flows using image processing techniques in recent years. We can expect that PIV and image measurement method will play an important role in the future development of modern hydraulics.

References

Adrian RJ (1991): Particle-imaging techniques for experimental fluid mechanics. Annual Review of Fluid Mechanics, Vol. 23, 261-304.

Adrian RJ (1996): Bibliography of particle velocimetry using imaging methods: 1917-1995. TAM Report No. 817, University of Illinois at Urbana-Champaign.

Alahyari A; Longmire EK (1994): Particle image velocimetry in a variable density flow: application to a dynamically evolving microburst. Experiments in Fluids, Vol. 17, 434-440.

Baek SJ; Lee SJ (1996): A new two-frame particle tracking algorithm using match probability. Experiments in Fluids, Vol. 22, 23-32.

Barnhart DH; Adrian RJ; Meinhart CD; Papen GC (1995): Phase-conjugate system for holographic particle image velocimetry through thick curve windows. Proceedings of The International Workshop on PIV-Fukui'95, edited by T. Kobayashi and F. Yamamoto, VSJ, 1-6.

Canny J (1986): A computational approach to edge detection. IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. PAMI-8, No. 6, 679-698.

Chang KA; Liu PL (1999): Experimental Investigation of turbulence generated by breaking waves in water of intermediate depth. Physics of Fluids, Vol. 11, No.11, 3390-3400.

Crowe C; Sommerfeld M; Tsuji Y (1998): Multiphase flows with droplets and particles, CRC Press.

Dahl J; Hermansson R; Tiberg S-E; Veber P (1995): Use of video-based particle image velocimetry technique for studies of velocity fields in a water heat storage vessel. Experiments in Fluids, Vol. 18, 383-388.

Dimotakis PE; Debussy FD; Koochesfahani MM (1981): Particle streak velocity field measurements in a two-dimensional mixing layer. Physics of Fluids, Vol. 24, 995-999.

TH. Dracos, Three-dimensional velocity and vorticity measuring and image analysis techniques, Kluwer Academic Publishers, 1996.

Gui LC; Merzkirch W (1996): A method of tracking ensembles of particle images. Experiments in Fluids, Vol. 21, 465-468.

Hassan Y; Blanchat T; Seeley JrCH (1992): Simultaneous velocity measurements of both components of a two-phase flow using particle image velocimetry. Int. J. of Multiphase Flow, Vol. 18, pp. 371-395.

Hewitt GF; Delhaye JM; Zuber N (1990): Multiphase instrumentation and experimental techniques, in Multiphase Science and technology, Vol.5, pp.65-80, Hemisphere Publishing Corporation.

Hu ML (1994): Study on jet nappe atomization, Journal of Hydrodynamics, Ser. A, Vol. 9, No. 3, pp. 344-349.

Huang HT; Fiedler HE; Wang JJ (1993a): Limitation and improvement of PIV, Part I: Limitation of conventional techniques due to deformation of particle image patterns. Experiments in Fluids, Vol. 15, 168-174.

Huang HT; Fiedler HE; Wang JJ (1993b): Limitation and improvement of PIV, Part II: Particle image distortion, a novel technique. Experiments in Fluids, Vol. 15, 263-273.

Ido T; Ishikawa M; Murai Y; Yamamot F (1997): Development of the CFD presumption method to restore whole flow field from discrete velocity data. J. of the Visualization Society of Japan, Vol. 17, Suppl. No.1, 235-238 (In Japanese).

Ishikawa M; Yamamoto F; Murai Y; Iguchi M; Wada A (1997): A novel PIV algorithm using velocity gradient tensor. Proceedings of The Second International Workshop on PIV'97-Fukui, edited by T. Kobayashi and F. Yamamoto, VSJ, 51-56.

Kaga A; Inoue Y; Yamaguchi K (1994): Pattern tracking algorithms for airflow measurement through digital image processing of visualized images. J. of Visualization Society of Japan, Vol. 14, 38-45 (in Japanese).

Kawahashi M; Yamamoto K (1995): Speckle method using beam scanning techniques. Proceedings of The International Workshop on PIV-Fukui'95, edited by T. Kobayashi and F. Yamamoto, VSJ, 155-158.

Khalighi B (1989): Study of the intake swirl process in an engine using flow visualization and particle tracking velocimetry. ASME-FED 85, 37-47.

Kimura I; Hyodo T; Ozawa M (1997): Temperature and velocity measurement of a 3-D thermal flow field using thermo-sensitive liquid crystals. Proceedings of The International Workshop on PIV'97-Fukui, edited by T. Kobayashi and F. Yamamoto, VSJ, 35-42.

Kobayashi T; Saga T; Doh D (1995): A three-dimensional simultaneous scalar and vector tracking method. Proceedings of The International Workshop on PIV'95-Fukui, VSJ, 33-43.

Kobayashi T; Saga T; Segawa S (1992): Simultaneous measurement of temperature and velocity field in the thermal flow through a color image processing; the second report: Simultaneous measurement of temperature and velocity using capsuled thermo-sensing liquid crystals and nylon tracer particles. J. of the Visualization Society of Japan, Vol. 12, Suppl. No.1, 71-74 (in Japanese).

Lasheras JC; Villermaux E; Hopfinger EJ (1998): Break-up and atomization of a round water jet by a high-speed annular air jet. J. Fluid Mech., Vol. 357, 351-379.

Ma G; Shen G; Pan X (1995): Bubbles used as flow tracer in PIV in water tunnel. Proceedings of The International Workshop on PIV-Fukui'95, edited by T. Kobayashi and F. Yamamoto, VSJ, 205-216.

Malik NA; Dracos Th; Papantoniou DA (1993): Particle tracking velocimetry in three-dimensional flows, Part II: Particle tracking. Experiments in Fluids, Vol. 15, 279-294.

Meinhart CD; Prasad AK; Adrian RJ (1993): A parallel digital processor system for particle image velocimetry. Meas. Sci. Technol., Vol. 4, pp. 619-626.

Meng H; Estevadeordal J; Gogineni S; Goss L; Roquemore WM (1997): Holographic flow visualization as a tool for studying 3D coherent structures and instabilities. Proceedings of The Second International Workshop on PIV'97-Fukui, edited by T. Kobayashi and F. Yamamoto, VSJ, pp. 27-34.

Miyazaki K; Chen G; Yamamoto F; Horii K (1999): Gas-solid two-phase spiral flow investigation with PIV techniques. Transactions of the Japan Society for Aeronautical and Space Sciences, Vol.41, 163-167.

Miyazaki K; Chen G; Yamamoto F; Ohta J; Murai Y; Horii K (1999): PIV measurement of particle motion in spiral gas-solid two-phase flow, Experimental Thermal and Fluid Sciences,Vol. 19, No.4, 194-203.

Murai Y; Watanabe S; Yamamoto F; Matsumoto Y (1997): Three-dimensional measurement of bubble motions in bubble plume using stereo image processing. Proceedings of The Second International Workshop on PIV'97-Fukui, edited by T. Kobayashi and F. Yamamoto, VSJ, 13-18.

Nishino N; Kasagi N; Hirata M (1989): Three-dimensional particle tracking velocimetry based on automated digital image processing. J. of Fluids Engineering, ASME, Vol. 111, 384-391.

 HN (1998): The role of surface disturbances in the entrainment of bubbles by a liquid jet. J Fluid Mech., Vol. 372, 189-212.

Okamoto K (1995): Three-dimensional particle tracking algorithms: velocity vector histogram and spring model. Proceedings of The International Workshop on PIV-Fukui'95, edited by T. Kobayashi and F. Yamamoto, VSJ, 21-32.

Raffel M; Kompenhans J (1995): Theoretical and experimental aspects of image-shifting by means of a rotating mirror system for particle image velocimetry. Meas. Sci. Technol., Vol. 6, 795-808.

Raffel M; Willert CE; Kompenhans J (1998): Particle Image Velocimetry, Springer.

Reese J; Chen RC; Fan, LS (1995): Three-dimensional particle image velocimetry for use in three-phase fluidization systems, Experiments in Fluids, Vol. 19, 367-378.

Reinauer R; Hager WH (1996): Shockwave in air-water flows, Int. J. Multiphase Flow, Vol. 22, No. 6, 1255-1263.

Sheng J; Meng H (1997): A 3D velocity field extraction technique using genetic algorithm. Proceedings of The Second International Workshop on PIV'97-Fukui, VSJ, 43-50.

Simpkins PG; Dudderar TD (1978): Laser speckle measurement of transient Benard convection. J. Fluid Mech., Vol. 89, 665-71.

Song X; Yamamoto F; Iguchi M; Koketsu M; Chen G (1996): 3-D PTV measurement of bubble rising flow in cylindrical vessel. ISIJ Int., 36, S54-S57.

Song X; Yamamoto F; Murai Y; Iguchi M (1997): Cross-correlation algorithm for PIV by Delaunay tessellation. Proceedings of The Second International Workshop on PIV'97-Fukui, edited by T. Kobayashi and F. Yamamoto, VSJ, 109-115.

Storr GJ; Behnia M (1999): Experiments with large diameter gravity driven impacting liquid jets. Experiments in Fluids, Vol. 27, 60-69.

Uemura T; Yamamoto F; Koukawa M (1990): High-speed algorithm for particle tracking velocimetry using binary. J. of Visualization Society of Japan, Vol. 10, 196-202 (in Japanese).

Veber P; Dahl J; Hermansson R (1997): Study of the phenomena affecting the accuracy of a video-based particle tracking velocimetry technique. Experiments in Fluids, Vol. 22, 482-488.

Wernert P (1996): Experimental and numerical investigations of dynamic stall on a pitching airfoil. AIAA Journal, Vol. 34, 982-989.

Wernet MP; Pline A (1993): Particle displacement tracking technique and +Westerweel J (1994): Efficient detection of spurious vectors in particle image velocimetry data. Experiments in Fluids, Vol. 16, 236-247.

Willert C; Raffel M; Kompenhans J; Stasicki B; Khler C (1996): Recent applications of particle image velocimetry in aerodynamic research. Flow Meas. Instrum., Vol. 7, 247-256.

Wozniak G; Wozniak K (1994): Buoyancy and thermocapillary flow analysis by the combined use of liquid crystals and PIV. Experiments in Fluids, Vol. 17, 141-146.

Yamamoto F; Ruan X; Song X; Iguchi M (1997): A method for counting particle number in overlapped particle images. J. of the Visualization Society of Japan, Vol. 17, Suppl. No.2, 107-110.

Yamamoto F; Wada A; Iguchi M; Ishikawa M (1996a): Visualization and image processing of torque converter internal flow. J. of Flow Visualization and Image Processing, Vol. 3, 51-64.

Yamamoto F; Wada A; Iguchi M; Ishikawa M (1996b): Discussion of the cross-correlation methods for PIV. J. of Flow Visualization and Image Processing, Vol. 3, 65-78.

Zhou DR; CHEN WQ (1995): Study on three-dimensional reconstruction of atomization flows in high-speed discharge, Report of research, Wuhan University of Water Resources and Hydroelectric Power, 1-9.