Open Access

ARTICLE

Ray-Tracing Based Image Correction of Optical Distortion for PIV Measurements in Packed Beds

Fabio J. W. A. Martins^a,*, Clauson Carvalho da Silva^b, Christian Lessig^b, Katharina Zähringe^a

a Laboratory of Fluid Dynamics and Technical Flows (ISUT), Otto-von-Guericke-Universität Magdeburg, Universitätsplatz 2, D-39106 Magdeburg, Germany
b Institute for Simulation and Graphics, Otto-von-Guericke-Universität Magdeburg, Germany

* Corresponding Author:* Correspondence Author, E-mail:

Journal of Advanced Optics and Photonics 2018, 1(2), 71-94. https://doi.org/10.32604/jaop.2018.03870

Issue published 22 April 2021

Abstract

Packed beds are employed in a wide range of industrial processes, making their optimization an important objective. This requires a thorough understanding of the flow characteristics in the interstices of the bed and the influence of parameters such as packing material, fluid type and flow rate. Next to numerical simulations that have been performed, liquid flows in packed beds have been experimentally investigated by Particle Image Velocimetry (PIV), with optical accessibility obtained using refractive index matching and transparent geometries. For gaseous flows through packed beds, where refractive index matching is not available, there is, in contrast, still a lack of experimental work, because the different optical properties of gas and transparent packing geometries generate optical aberrations. The approximation of gaseous flows with results from dynamically similar, liquid flows provide thereby no reliable alternative because they might be inaccurate and lead to incorrect conclusions about the actual gas flow. The high optical aberrations created by transparent packed beds in PIV experiments can be corrected using ray tracing simulations of a geometric-optical model of the viewing media. The present work proposes a novel image correction technique based on this approach that overcomes the considerable optical distortions caused by arrangements of transparent solid spheres. We analyze the influence of spheres along the camera view in three test cases: the first studies optical distortions of target images, the second PIV computations of a known displacement of a printed particle field, and the third PIV measurements of an air-jet flow. The image correction is qualitatively and quantitatively evaluated by computing errors in target images and in velocity fields. We show that the high optical aberrations that are caused by the spheres can mostly be corrected using ray tracing, allowing for accurate optical measurements in complex geometries, such as packed beds. The approach proposed in this work might be extended to even more challenging geometries in order to solve limitations of existing techniques in the literature.

---

Keywords

---