Imaging system produces 3-D views of turbulent flow fields

July 1, 1999
Whorls and eddies created when constituents mix in a turbulent flow can markedly affect the properties of a product or application. To discover how ingredients mix, Phil Roberts of the Georgia Institute of Technology (Atlanta, GA) has developed a laser-induced fluorescence system to obtain three-dimensional (3-D) images of turbulent flow fields using an image acquisition and storage system developed by MicroDISC (Yardley, PA).

Imaging system produces 3-D views of turbulent flow fields

Whorls and eddies created when constituents mix in a turbulent flow can markedly affect the properties of a product or application. To discover how ingredients mix, Phil Roberts of the Georgia Institute of Technology (Atlanta, GA) has developed a laser-induced fluorescence system to obtain three-dimensional (3-D) images of turbulent flow fields using an image acquisition and storage system developed by MicroDISC (Yardley, PA).

Based on a MotionVision CA-D6-0512W progressive-scan area camera from Dalsa (Waterloo, Ontario, Canada) interfaced to a RoadRunner digital frame grabber from BitFlow (Woburn, MA), the system is controlled using a multiprocessor PCI-based Pentium II central processing unit (CPU). On-line memory of 2 Gbytes helps the system to store several seconds of digital video at 262 frames/s. For longer imaging durations, MicroDISC`s digital video recorder disk system sustains image acquisition and recording to a RAID disk system at up to 160 Mbytes/s.

To induce fluorescence, a laser beam sweeps through a flow field at high speed in a programmed pattern. The beam sweep, which is controlled through a scan-controller board inside the computer system, manipulates the x and y movement of the optical scanners relative to a set of preadjusted mirrors. This setup allows simultaneous beam movement and optical-scanner positioning.

The scan controller then outputs a trigger signal that starts the MotionVision CA-D6-0512W camera. Therefore, one image is obtained for each vertical sweep of the beam. The beam then shifts in the y direction slightly and the process is repeated to capture multiple slices of the flow field. Next, these slices are recombined to obtain a 3-D view of the flow.

To build a useful image, the application requires between 100 and 1000 vertical slices per second. When the camera shutter is opened, the beam makes one (or more) sweep in the y direction; then the shutter is closed. The beam moves slightly in the x direction, and the process repeats. After completing a preset number of vertical sweeps, the beam moves back to the beginning x location, and the process is repeated.

Voice Your Opinion

To join the conversation, and become an exclusive member of Vision Systems Design, create an account today!