Validation Experiments

The validation of the experimental technique, data analysis procedure and computer implementation has been carried out in two ways: (i) Comparison between different measurement techniques, (ii) Comparison of the standard calculated quantities with that in the literature. The validation results of mean velocity, vorticity patterns, Strouhal number and drag coefficient are discussed in the following sections.

Comparison of hotwire and PIV data

Benchmark experiments were conducted at selected planes in the wake of the square cylinder under identical conditions. Both hotwire and PIV measurements were recorded. A cross-check in terms of the two components of time-averaged velocity is presented in Figures 3.37 to 3.41 at two different Reynolds number (Re=410 and 610). At =5 the u-velocity profiles recorded by the two techniques show significant difference in the velocity deficit, Figure 3.37. Since the recirculation zone extends upto =6, the time-averaged velocity is negative at all earlier -locations. The hotwire probe cannot resolve the negative stream-wise velocity component, while PIV does not suffer from this drawback. Hence, adjacent to the cylinder, the PIV data can be assumed to be more accurate. As one moves downstream, the match between the two measurements is better. At =15 units, the stream-wise velocity profiles from hotwire and PIV measurements merge with each other (figure 3.39).

Figure 3.38 to 3.40 compare the transverse -velocity for cylinder at different orientations and Re=410 and 610. The overall trends in the v-velocity are similar. At =5 location, the maximum deviation of -velocity between hotwire and PIV measurements is within the uncertainty of the velocity measurements. This is in contrast to the high deviation in the u-velocity measurements in Figure 3.37. The hotwire measurement error in the recirculation zone is higher for the -velocity than the -velocity. The -velocity trend at =10 and 15 are similar between hotwire and PIV measurements. The higher percentage deviation of the v-velocity at these locations compared to the earlier x-location can be be attributed to the lower absolute magnitude of the v-velocity. Figure 3.43 shows the comparison of velocity at two locations. At =5 the PIV data shows a slightly higher peak RMS value compared to hotwire measurements. At =10, both PIV and hotwire results coincide.