In wholly single-phase flows, or two- or multiphase flows where the phases are homogeneously mixed, fluid motion is invisible to the eye. The introduction of tracer particles into the flow enables flow patterns and behavior to be visualized. Furthermore, tracking the motion and behavior or these particles, either individually of as groups, allows the fluid motion to measured. So-called tracer methods currently provide the most accurate, means by which to monitor and measure fluid velocity.
All tracer methods have light scattering as a basis. They require some form of illumination, followed by collection of light scattered from the particles, and analysis of this scattered light to provide a measure of the particle motion from which the fluid velocity is inferred. They have many advantages over conventional probe methods such as pilot tubes and hot wires, including accuracy, reliability and high spatial and temporal resolution. Their nonintrusive nature, whereby nothing needs to be physically inserted into the flow, enables their use in hostile environments. They additionally offer the potential for instantaneous global mapping of flow fields. However, suitable optical access is required but which is not always available.
Tracer particles can be solid, liquid or gaseous and are usually injected directly into the fluid. Ideally they should be spherical and monosized. Since the host fluid velocity is inferred from their motion, selected particles must be capable of following the fluid motion with accuracy and precision. This requires that they are small but also neutrally buoyant. A simple but valuable measure is a particle's aerodynamic diameter Da = Δρ1/2Do, where Do is the particle's physical diameter and Δρ, the difference in density between it and the host fluid in gm cm−3. The smaller Da, the more responsive the particle is to flow fluctuations. In addition, tracer particles should exhibit good light scattering ability, facilitated by larger size and a refractive index which differs significantly from that of the fluid. An adequate number of particles must be present, as required by the experiment, but not so many as might affect the fluid behavior. Invariably, choice of particle is a compromise.
There are two generic types of tracer method. The first, laser Doppler velocimetry (LDV), measures the Doppler shift which results when light is scattered from a moving particle Durst et al. 1976). It is a point measurement method and yields accurate, time-averaged, mean velocity and turbulence information. (See also Anemometers, Laser Doppler.) Velocity maps are built up by traversing the flow. The second, particle image velocimetry (PIV), has each particle velocity (and often its direction) encoded in the recorded image(s) of the flow. Instantaneous velocity maps are captured within a plane or volume of the flow and traversing is usually unnecessary.
In the most common (two-beam) form of LDV (Figure 1), a laser beam is split into two beams which are each focused and made to intersect measurement volume (mv), typically 0.1 mm3. When a particle crosses this volume, light from the two beams is scattered, with scattering from each undergoing a phase shift due to the Doppler effect. A photodetector receives scattered light as beat signals (Doppler bursts) which are usually monitored using an oscilloscope and are passed to a signal analyzer for processing. The burst frequency f is measured and is directly proportional to the particle velocity. The signal has the same form as if the two beams had created an interference pattern within the mv with fringe spacing λf = λo/2sinθ, where λo is the laser wavelength and θ the bisector angle of the two beams. The measured velocity component is normal to these fringes, and is found as u = λff.
Each particle crossing the mv generates its own Doppler burst. Usually a probability distribution function (pdf) is built up from several thousand individual measurements, from which an average velocity is computed along with a root mean square (rms) value. Modern signal analysis is by digital autocorrelation or Fourier transforms and in excess of 104 signals per second may be processed. Multiple velocity components can be measured by arranging different color interference patterns at other orientations.
Another LDV method requires only a single beam to be focused into the flow. The direct Doppler shift is then measured using a Michelson interferometer, (see Interferometry.) Doppler global velocimetry (DGV), using a laser sheet, extends this approach to two-dimensional velocity mapping, wherein velocity magnitude is encoded in the recorded intensity of scattering. Such methods are for high speed applications.
PIV is, essentially, a modern extension of so-called chronophotographic methods Somerscales (1981), where particles are recorded as streaks, encoded tracks, or multiple images captured at known times. Particle velocity is measured as the length of a streak, or the displacement between multiple images. A pulsed or chopped/modulated laser beam is generally used, and is typically introduced as a light sheet which defines a precise plane of interest, Figure 2. Photography is gradually being replaced by electronically gated high resolution charge coupled device (ccd) cameras as the means of recording images.
PIV data reduction requires sophisticated image processing techniques. Particle tracking methods are used where there are relatively few recorded particles or where high velocity gradients exist; particle tracks are then individually identified and measured. The alternative, a correlation method, is used when there are many particles, or where images are particularly noisy. With these, a grid is applied to the recorded image which is analyzed segment-by-segment for average particle displacements.
Backlighting and holographic methods offer alternatives to using laser sheets; holography, specifically, enabling an instantaneous velocity map to be captured throughout a volume. In addition, acquisition of image sequences allows the temporal development of flows to be captured in a similar manner to photochromic dye tracing. PIV methods are particularly applicable to transient flows. (See also Holograms, Holographic Interferometry and Photochromic Dye Tracing)
LDV and PIV each measure different aspects of behavior within flows. In general, they are applicable to flows of all types, as evidenced by the general literature [Adrian et al. 1991], The two approaches complement each other and are often used in combination when studying fluid behavior.
Adrian, R. I., Durao, D. F. G., Durst, F., Maeda, M. and Whitelaw, J. H. (Eds.) (1991) Applications of Laser Techniques to Fluid Mechanics, Springer-Verlag, NY.
Durst, F., Melling, A. and Whitelaw, J. H. (1976) Principles and Practice of Laser-Doppler Anemometry, Academic Press, NY. DOI: 10.1016/0030-3992(76)90012-8
Somerscales, E. F. C. (1981) Measurement of velocity, Methods of Experimental Physics: Fluid Dynamics Part A, Emrich, R. J. Edn., Academic Press, NY.