Simultaneous Measurement of Turbulence and Particle Kinematics Using Flow Imaging Techniques

Numerous fields in both engineering and natural science involve problems associated with fluid particle interaction. This method provides a relatively low-cost, non-intrusive simultaneous optical measurement of both particle trajectories and flow velocities. Here we measure the settling velocities of sediment particles in a turbulent flow, enabling detailed characterization of the particle trajectories while simultaneously measuring turbulent velocities in the same location.

To begin the particle image velocimetry setup, fix a dual head high intensity pulse laser horizontally level on an optical plate, in line with the flow facility. Place a cylindrical lens in line with the laser to produce a 2D plane of light that will be below the oscillating grid. Then mount a spherical lens after the cylindrical lens at a distance that will generate a light sheet that is 5 to one millimeter thick.

Next, place a double exposure CCD camera perpendicular to the light sheet to record images for PIV. Attach a lens to the camera, turn it on, and set it to free and continuous mode. Coarsely focus the PIV camera on the turbulent flow facility.

Adjust the aperture and the camera position until the image is smaller than or close to the desired light sheet boundaries. Then turn off the camera and turn on the laser at low intensity. Confirm that the light sheet is perpendicular to the floor, and then place a calibration target marked with a grid precisely at the center of the light sheet.

It is essential to ensure that the PIV camera is perpendicular to the light sheet and that the light sheet is perpendicular to the floor or the facility bottom. Misalignment result in incorrect velocity projections, and hence fluid velocity error. Turn off the laser and turn the camera back on.

Focus the camera on the calibration target and capture a single image. Open the image in image processing software, and confirm that the row, height, and column spacing are consistent across the target. The corner marker sizes should differ by no more than one pixel and ideally, they should be identical.

If the image meets these criteria, remove the calibration target. Install the grid, and run the facility. Then introduce about a tablespoon-full of PIV tracer particles to the fluid.

Wait until the tracers and fluid are well mixed before continuing. Then turn on the laser and set it to external control and high power. Turn off the room lights and capture an image pair to evaluate the tracer density.

Gradually increase the tracer concentration by teaspoon-fulls to the desired visual density. Then set the PIV camera frame rate to the highest possible value and set the time between consecutive PIV images. Confirm that the laser is configured appropriately.

Then turn off the lights and collect data in free mode for a few seconds. Cross correlate the image pairs and confirm that the acquired data is of good quality. Stop the grid oscillation when finished.

To begin setting up 2D particle tracking, place a monochromatic LED line light under the oscillation grid facility so that the light sheet will be centered within the LED line. Turn on the LED line light and the laser on low power. Confirm that the light sheet and the line light are well aligned, and then turn them off.

Next attach a lens to a CMOS high speed camera to be used for particle tracking. Turn on the camera in free continuous or live mode, and coarsely focus it on the region of interest. Adjust the particle tracking camera aperture height and distance until the region of interest is with it's field of view and the camera is level and perpendicular to the line light.

Turn off the camera. Turn on the line light and place the calibration target at the center of the line light. Then turn off the line light, turn on the camera, and focus it on the target.

Capture an image of the calibration target and confirm that the particle tracking camera is level, perpendicular to the target and in focus with no image distortion at the edges. Remove the calibration target afterwards. Then set the number of high speed images to be collected.

Based on the expected particle velocity, set the frame rate and resolution to values that should achieve particle displacement of three to 10 pixels between images. Install the grid, turn on the LED line light, and darken the room. Start the grid oscillation and introduce a small portion of the particles of interest into the flow.

When the particles appear on the high speed camera, capture a few frames. It is important for the particle tracks to be clearly visible in the images, indicating that the particles remain in plane and do not frequently overlap. Failure to meet these criteria will result in an inability to accurately track the particles.

Confirm that there are no visible entrance effects, particle overlap is infrequent, and particle motion is primarily in plane. Stop the oscillation when finished. To begin the final calibration, with the lights dimmed place the calibration target within the LED and laser light sheets.

Turn off the laser and LED and turn on the room lights. Ensure that the calibration target is in focus within the camera FOVs, and has a unique mark visible to both cameras. Capture an image of the calibration target on both cameras.

Note the relevant placements of the unique mark, and confirm that the cameras are still level and show no distortion around the edges. Then remove the calibration target, install the grid, and start the oscillation. Let it run for at least 20 minutes to allow the flow to reach a steady state.

Then darken the room, turn on the LED line light, and introduce the particles into the flow. Simultaneously start the laser pulses and image acquisition for both systems when particles appear in the particle tracking camera FOV. When data acquisition has finished, save the images and stop the grid oscillation.

Analyze the flow velocity distribution and the particle trajectories. The PIV images can be processed into instantaneous fluid velocity and vorticity distributions. Here the fluid velocity vector distribution is overlaid on a vorticity color map.

With this setup, the magnitude of the spatial mean of root mean squared fluid velocity fluctuation over the PIV field of view should increase with oscillation frequency for both horizontal and vertical velocity components. Particle trajectories and velocities can be determined from the high speed particle tracking images. The distribution of particle velocities should be roughly gaussian.

Here larger irregularly shapes particles generally showed particle velocity distributions with larger standard deviations than those of the smaller, spherical particles. Although both sets of particles showed distributions with larger mean vertical velocities and larger standard deviations as the grid oscillation rate increased. The stagnant flow settling velocities of synthetic particles, industrial sand, and locally gathered sand determined from their particle trajectories all roughly agreed with the Dietrich curves.

The tendency of particles settling velocities to increase with grid oscillation frequency was explored further in subsequent analysis. Simultaneous optical measurement of both particle kinetics and fluid dynamics, specifically turbulence, is challenging because of the potential for interference between the two imaging techniques, resulting in measurement inaccuracies. Flows that are strongly three dimensional are not well suited for this technique, because out of plane motions will produce errors in both the 2D tracking and the particle velocimetry analysis.

The concentration of tracked particles must be relatively low to maximize confidence that the same particle is being tracked in consecutive images. Also, PIV tracers and the particles being tracked must be sufficiently different in size to distinguish them. The integration of the flow velocity information with the particle trajectory depends on what's being investigated.

For example, this method can also examine the flow velocities at specific instances in time along the trajectory of the particle. This technique was demonstrated with sediment transport, an application for motion sciences but it is relevant in many applications where fluid flow interacts with natural or manmade particulates.