Using chaos analysis for evaluating the quality of F-PIV measurements in a bubble column

- Estimation of chaos invariants from 2D-2C PIV velocity fluctuation data.
- Proposal of correlation integral as an indicator of F-PIV quality.
- Evaluation of PIV post-processing effect on the velocity module and its components.
- Proposal of using quality indicators for optimizing PIV setup and processing.
- Relationship between the flow complexity and quality of the F-PIV information.

A new approach is proposed for evaluating the quality of fluorescent particle image velocimetry (F-PIV) measurements for different superficial gas velocities (UG) in a bubble column (1.0 m long×0.145 m i.d.) using chaos analysis. The 2D velocity fields of a F-PIV system were obtained in a 0.145×0.10 m area located 0.552 m above the gas distributor. Four thousand pairs of images were obtained at a frequency of 4.2 Hz at nine radial positions for superficial gas velocities (UG) of 0.158, 0.528, 2.106 and 4.212×10−2 m/s. Using the standard cross-correlation (SCC) approach, PIV quality was analyzed for the velocity module and its components by investigating the cross-correlation coefficient (CCC), the signal-to-noise ratio (SNR) and the uncertainty (CS method) estimated by correlation statistics. Moreover, it was possible to extract the velocity fluctuation series for chaos analysis (correlation dimension, DML, and standard deviation of the correlation integral, DevML). The effect of the quality of F-PIV measurements evaluated by chaos analysis is based on the noise limitation in amplitude and normalization of the integral correlation distances. The results presented similarities in the behavior of the PIV uncertainty and chaotic parameters along the radius for all UG investigated. For this reason, chaos analysis can be used in other PIV processing settings where quality indicators, such as uncertainty, SNR and CCC, do not exist. Thus, it was possible to use this analysis to gain a better understanding of the nonlinear interactions in a bubble column as well as the performance of a PIV post-processing filter in order to reduce the effect of noise.

296