Seltmann, H., Rajala, S. A. and Drake, T.G., 1997, Dual camera, digital particle velocimetry system studies of multiphase flow: abstract submitted to IEEE Image Processing conference
Understanding the processes that control sand movement in the nearshore marine environment is important for management of coastal resources, recreation and national defense, among other reasons. Waves and currents move sand continuously along and across the shoreline, but the physical processes that are involved are particularly difficult to observe, and consequently, our ability to predict sand motion given wave conditions is rather poorly developed. This work describes the design and configuration of laser and video instrumentation for observing, recording and analyzing the velocities of both water and sand very near the seabed, under the usual nearshore conditions of unsteady oscillatory fluid motion.
The focus of this study is on one particularly simple flow of considerable importance in studies of surf-zone sediment transport, namely, flow over a planar bed of similar sand-size particles [2]. Such transport, commonly called sheet flow, occurs under large waves or strong currents or both, and is thought to accompany rapid bathymetric evolution and shoreline change. Under a variety of conditions, the sea bed develops ripples and a variety of other non-planar geometric configurations, but the plane-bed, sheet flow configuration studied here is both highly relevant and relatively well studied from theoretical, experimental and computational perspectives.
Particle image velocimetry (PIV) is an optical, velocity-field measurement system [3]. The goal of this work is to set up a PIV system in an oscillatory flow tunnel to measure the water velocity field in a plane parallel to the direction of wave propagation, and perpendicular to a mobile, nominally planar bed of sand. The PIV system must be capable of obtaining simultaneous measurements of fluid and solid-particle fields with sufficient spatial and temporal resolution to test models for sediment transport in oscillatory flows.
One difficulty in making accurate fluid measurements is that probes inserted into the fluid flow can alter the flow itself. In a laboratory environment, PIV allows measurement of the velocity field using video cameras and lasers that can be located outside the flow, and are therefore non-invasive. A simple PIV system consists of the following two components: 1) a source of particle images that can be used for analysis, which can be as simple as a single video camera system or a more complex, multi-camera, frame-straddling system; and 2) a method for analyzing the images to obtain velocity data [1, 4, 5].
The unique contribution of this research was to develop a dual-camera extension of a commercial PIV system (TSI, Inc.) for use in an oscillatory flow tunnel to measure water flow velocities over a mobile bed of sand. Techniques used to measure velocity in the resulting images are described and data from the PIV system are compared to acoustic Doppler velocimeter (ADV) measurements to determine if the PIV system can acquire valid measurements in a complicated multiphase flow.
The research was done using a device known as the Oscillatory Flow Tunnel (OFT). The Oscillatory Flow Tunnel was developed at Scripps Institution of Oceanography[6]. A uniform layer of 6.5-mm-diameter cellulose acetate spheres are placed into the sample area and the OFT is sealed and filled with fresh water. Identical plastic spheres are used instead of sand for comparison to computer and theoretical models which assume spherical particles.
The design of the video system had two goals. First was collection of PIV data from experiments up to 2 hours long. Second, the system was designed to accommodate multiple cameras. The first requirement was met by using cross-correlation cameras and attaching them to SVHS video decks to record data. Laser synchronization, needed for frame straddling, is a simple matter of distributing the video signal from the video camera to the laser synchronizer. The necessity of multiple cameras added complexity to the system, and required a way to synchronize the cameras together with the laser system. The cross-correlation cameras (Pulnix TM-9700) did not have a standard genlock input, therefore requiring the addition of a sync separator to break the composite video signal into its components. Figure 1 shows the system that was developed. The analog video was then converted to a single digital video signal for further analysis.
Comparison of PIV data to independent measurements of fluid velocity in the OFT forms the basis of the experimental results. The correlation techniques work well with images in which objects of interest (e.g., sand particles) are easily distinguished from the background (Figure 2).
The system solves for the mean displacement between groups of particles captured in a sequence of image pairs. The particle displacements are found by first dividing each image into multiple regions of interest (ROIs). An ROI of 64x64 pixels is typical. Each of these ROIs are analyzed separately to find the mean velocity of the region. The particle pairs that straddle two regions are known as out-of-plane data and ROI selection is made to minimize such occurrences. [HS1]
Variations in seeding density result in cross-correlation errors, but in general, the PIV results agree within statistical error with independent ADV measurements of flow velocity (Figure 3). Additional results will be presented in the full paper, along with a study of potential errors.
The goal of this work was the setup and calibration of a dual-camera PIV system for laboratory use in the study of unsteady, multiphase oscillatory flow. For properly chosen ROIs, velocities measured using the PIV system correlate well with a reference free-stream velocity measured using an ADV system for each of two data sets. Vertical profiles show that the PIV system is capable of recording velocity information throughout the flow, in particular, near the bed in high-concentration, two-phase flows. Data from two synchronized cameras can be used to determine fluid velocity fields at two disparate length scales. This is particularly useful in flows that have large variation in velocity within the area of interest, such as occurs near boundaries.
1. Buchhave, P. "Particle Image Velocimetry," Optical Diagnostics for Flow Processes. p. 247-269, 1994
2. Drake, T.G., R.L. Shreve, W.E. Dietrich, P.J. Whiting, and L.B. Leopold. "Bedload transport of fine gravel observed by motion picture photography", J. Fluid Mechanics, v. 192, p. 193 - 218, 1988
3. Grant, I. "Particle image velocimetry: a review", Proc. Instn. Mech. Engrs., v. 211, p. 55-76, 1997
4. Grant, I. and X. Wang. "Directionally-unambiguous, digital particle image velocimetry studies using a image intensifier camera", Experiments in Fluids v. 18, p. 358-362, 1995
5. Keane, R. "Correlation methods of PIV analysis", Optical Diagnostics for Flow Processes, p. 271-289, 1994
6. King, D. B. Jr. Studies in oscillatory flow bedload sediment transport. Ph.D. dissertation, University of California, San Diego (Scripps Institution of Oceanography), 183 pp., 1991
7. Proakis, J. G., and D. G. Manolakis. Digital Signal Processing, 3rd ed., Prentice-Hall, Inc., 1996