Basic System Components for Traditional PIV Experiments

In order to determine the displacement of seed particles, several basic hardware components are required to conduct a traditional PIV experiment. As shown in Figure 12.4, the fluid must be “seeded” with particles; therefore, a means to both generate particles and evenly disperse them in the flow is needed. Once the particles are added to the fluid, they must be visible to a camera. Therefore, the particles must reflect light. In the most basic PIV experiments, two-dimensional velocity fields are obtained on a single plane. A light source must be chosen to produce a planar light sheet with a minimal thickness. A suitable camera is required to capture the light reflected by the particles, and to ensure the camera captures images while the particles are illuminated, a synchronizer unit is required to properly time the light source with the camera. Finally, a computer is needed to save the images and process the images to yield velocity maps representative of the flow field. The basic requirements for each of these hardware components are discussed in the following sections.

Seed Particles

As mentioned earlier in the chapter, the selection of proper seed particles is critical to ensure an accurate representation of the fluid velocity. The displacement of a group of particles is measured over a given time, and the measured velocity of these

Basic setup for a traditional PIV experiment

FIGURE 12.4 Basic setup for a traditional PIV experiment.

particles is assumed equal to the velocity of the surrounding fluid. Therefore, the seed particles must follow the fluid without disrupting its natural behavior.

When considering the selection of the ideal seed material, it is necessary to consider the Stokes number. The Stokes number represents how a foreign particle behaves when immersed in fluid. For example, when a large sphere is placed in moving fluid, the motion of the fluid changes to move around the sphere, while the “particle” (large sphere) remains relatively unchanged. Depending on the Reynolds number of the fluid (relative to the sphere size), the flow will exhibit the traditional separation and turbulent vortex street forming downstream of the sphere. The relatively large sphere clearly alters the fluid motion around it. For PIV applications, the foreign particle must not alter the flow field, and it must naturally follow the fluid.

The Stokes number represents the interaction between a foreign particle and its surrounding fluid. Non-dimensionally, the ratio of the particle’s inertial force to the fluid’s inertial force is represented with the Stokes number. More commonly, the Stokes number is presented as the ratio of the relaxation time of the particle compared to the characteristic time of the surrounding fluid. If the particle motion is independent of the fluid motion (the movement of the particle is not influenced by the fluid), both the relaxation time of the particle and the inertial force of the particle are large. In the extreme case of fixed foreign particles, the Stokes number approaches infinity. On the other hand, if the particle is identically following the motion of the fluid, it is not exerting unnatural force on the fluid, and its relaxation time approaches zero. Therefore, the Stokes number approaches zero. For PIV applications, it is necessary to select the seed particles so their Stokes number approaches zero, and they are neutrally buoyant in the surrounding fluid. Generally, a particle is acceptable if the Stokes number is less than 0.1; under such a condition, the error associated with deviation of the particle from the fluid is less than 1% [5].

Due to the relatively high density of liquids, the Stokes number criteria is easily achieved. A variety of solid particles can be dispersed in the fluid, and the density of the particles is comparable to that of the fluid. Table 12.1 provides several solid particles commonly seen in liquid PIV experiments. For liquid experiments, the size of the particle can be several orders of magnitude larger than typically used in gas flow

TABLE 12.1

Common Seed Particles for Liquid Flows

Particle Material

Approximate Density Range (g/cm1)

Approximate Particle Diameter Range

Polymer particles


10-100 pm

Coated, hollow glass beads


10-100 pm

Fluorescent particles


200-1000 nm

experiments. The size of the particle effects how much light is reflected by the particle; therefore, having the larger diameter particles is an advantage of liquid flows.

The density of gases is several orders of magnitude less than the density of liquids. Likewise, the density of solid- or oil-based particles is also much larger than the gas density. With the large variation of density between the potential seed particles and the surrounding fluid, the particle diameter must be sufficiently small to meet the Stokes number criteria. Comparing the seed particles listed in Tables 12.1 and

12.2 shows the particles dispersed in gas flows are much smaller than those used in liquid flows. With gas flows that might be open to the atmosphere, caution must be taken for the dispersion of particles into a contained laboratory space. DEHS oil is commonly used; it is a non-toxic, lightweight oil that will evaporate from surfaces. Most cooking oils are also suitable, as they are non-toxic, but a longer lasting residue can be left on exposed surfaces. The Silica and Titanium powders are desirable due to their white, reflective surface. However, these powders can cause irritation to the respiratory system, so proper precautions should be taken to minimize the exposure to the airborne particles.

When adding the seed particles to the fluid, it is important that particles remain distinctly separate from one another. Not only does this control the size of the particles (affecting their ability to follow the flow), but it also aids in the analysis of the PIV images. The seed density of the particles should be controlled in conjunction with sizing the interrogation regions for the image. The dispersion of too many or too few particles will lead to increased uncertainty in the velocity calculation. General guidelines point to approximately 10 distinct particles in each interrogation region. In both extreme cases (too many or too few particles), it is difficult to use the correlation functions to match the particles between the two images.

TABLE 12.2

Common Seed Particles for Gas Flows

Particle Material

Approximate Density Range (g/cm3)

Approximate Particle Diameter Range (pm)

Di-ethyl-hexyl-sebacat (DEHS) oil



Vegetable oil



Silica dioxide (SiO,) powder



Titanium dioxide (TiO,) powder



< Prev   CONTENTS   Source   Next >