Identification and characterization of flow structures in chemical process equipment using multiresolution techniques

Abstract

Planar information of velocity from 2D particle image velocimetry (PIV) and large eddy simulation (LES) data have been studied using multiresolution wavelet transform (WT) formalisms, i.e., discrete and continuous WT. Identification of dominant energy containing structures with their characterization in terms of fractal spectra have been carried out for industrially important equipment exhibiting turbulent behavior. These include annular centrifugal contactor, jet loop reactor, ultrasound reactor, channel flow, stirred tank and bubble column reactor. The characterization of their dynamics based on denoising the data and studying the local energy along the WT scales show sensitive variation and this helps in identifying the size and shape of structures. A dependency is seen between mixing time and the higher order moments of length scale distribution, viz., skewness and kurtosis and a generalized correlation has been built up for important types of equipment and associated flow parameters. The correlation is not only based on the knowledge of reactor geometry and operating conditions but also on the flow structures via their statistical parameters. Wavelet transform modulus maxima (WTMM) methodology has been used to study the evolution of structures and their interaction in a reduced dimensionality by evaluating the fractal spectra. Classification studies have been carried out using principal component analysis (PCA) of the fractal spectra. The results obtained show clear classes for the six types of equipments and delineate regimes to obtain benchmark patterns of flow hydrodynamics based on PCA co-ordinates. This methodology offers a generalized way for the optimal design and operation of different types of reactors.