| Popis: |
Non-intrusive velocity measurements in turbulent, stratified flows are hampered by index of refraction fluctuations. These fluctuations can reduce the data rate and quality in laser Doppler velocimetry by displacing the beams. In an experiment with particle image velocimetry (PIV), the displacement of light reflected from particles can produce spurious particle positions and velocities. Although matching the indices of refraction of the stratifying agents can eliminate these problems, the costs associated with stratifying a large facility can be high. Thus, we seek to quantify the errors incurred in PIV measurements as a function of the intensity of the turbulence. We present results of a laboratory experiment in which images of a stationary target mimicking typical PIV images are acquired under different conditions. Because the target is fixed, any displacement registered by the PIV system is an error. The experimental facility consists of a salt-stratified tank stirred with arrays of vertical rods oscillating horizontally. Direct force measurements allow the work done by the rods on the water, and thus the average dissipation ea, to be estimated. Concurrent density measurements during the stirring are used to compute the length and time scales of the largest eddies. PIV errors are then expressed as a function of the turbulence intensity and the path length LP. The former quantity is measured by ea/νN 2 , where N is the buoyancy frequency. |
