Development and validation of a PIV processing routine for two-phase index-matched flows
- PDF / 1,635,948 Bytes
- 20 Pages / 595.276 x 790.866 pts Page_size
- 109 Downloads / 203 Views
R E G UL A R P A P E R
Andrew M. Bluestein • Douglas Bohl
Development and validation of a PIV processing routine for two-phase index-matched flows
Received: 14 January 2020 / Revised: 21 May 2020 / Accepted: 18 June 2020 Ó The Visualization Society of Japan 2020
Abstract Implementing particle image velocimetry (PIV) in particle-laden flows is a challenging task due to the optical effects of the added particles. The addition of particles reduces optical access in the measurement region, degrades image quality and increases the experimental uncertainty. One method for reducing this effect is by using refractive index matching (RIM) which matches the index of refraction of the fluid and added particles rendering the particles optically transparent. Exact matching of the refractive index is difficult to achieve in practice. Even slight mismatches in the refractive indices lead to degradation of the image quality which will impact the higher-order statistics. Comparisons of statistics between loading levels are difficult under these conditions. In this work, the effect of processing methodology on RIM PIV images is investigated. RIM is achieved through the use of superabsorbent hydrogel materials in water. Two processing methods are investigated. The baseline method is a basic PIV correlation process without any pre- or post-processing. The refined process utilizes image pre-processing, data post-processing, and experimental error compensation. The goal for the refined method was to optimize and implement a more sophisticated PIV processing routine using existing commercial software on a particle-laden flow to reduce progressive error artifacts. Images from a fully turbulent channel flow with particle loading levels of 0–15% are utilized. Past literature indicates that turbulence levels will be attenuated for the particles used. Turbulence levels calculated using the baseline method show a linear increase with increasing loading, which is a non-physical result. The refined processing method shows a slight attenuation of turbulence which is consistent with expectations. The study highlights the need for careful processing and interpretation of results in cases where changes in image quality can be expected between different experimental conditions. Keywords Index matching PIV Error analysis Turbulence modulation
1 Introduction Particle-laden turbulent flows are highly complex and are of interest in many areas including environmental, biological, and industrial settings. Acquiring experimental data in particle-laden flows has been mostly limited to flows with dilute loadings due to the effect that the added particles have on the experimental domain. Examples include Shokri et al. (2017), Li et al. (2012), Muste et al. (2009), and Kiger and Pan (2002) where volume loadings were between 10–3 and 10–6. This is especially true in optical experiments where the presence of even a dilute loading of particles can reduce optical clarity in the measurement region due to laser sheet attenuation, increased opacity of the f
Data Loading...
