Modeling of bubble coalescence and break-up using the Log-normal Method of Moments. (18th May 2022)

• Record Type: Journal Article
• Authors: Habiyaremye, V.; Komen, E.M.J.; Kuerten, J.G.M.; Frederix, E.M.A.
• Is Part Of: Chemical engineering science. Volume 253(2022)
• Publication Date: 2022-05-18
• Subjects: Bubbly flow -- Population balance -- Moment method -- Log-normal -- Coalescence -- Break-up; Chemical engineering -- Periodicals; Génie chimique -- Périodiques; Chemical engineering; Periodicals; Electronic journals; 660
• DOI: 10.1016/j.ces.2022.117577
• Languages: English
• ISSNs: 0009-2509

Abstract

Highlights: The Log-normal Method of Moments is extended to capture coalescence and break-up. These phenomena are modeled using source terms in the moment transport equations. Gaussian quadrature is used to approximate the integrals in these source terms. Good agreement is found between our model and an accurate class method. A significant gain in computational time is achieved compared to the class method. Abstract: This work focuses on sub-grid modeling of bubbly flows using the two-fluid model, coupled with the Log-normal Method of Moments (LogMoM) as a population balance model in order to efficiently capture bubble poly-dispersity. In this work, LogMoM is extended to account for bubble coalescence and break-up by formulating source terms for the Moment Transport Equations (MTEs). Gauss-Hermite and Gauss–Legendre quadrature are used to accurately evaluate the integrals in these source terms. The model is implemented into OpenFOAM, and is verified against analytical solutions for simple polynomial coalescence and break-up kernels in a zero-dimensional setting. Subsequently, LogMoM is successfully validated against a discrete class method already available in OpenFOAM as well as reference measurements from the TOPFLOW and METERO experiments using well established physical kernels for coalescence and break-up. Good agreement is found between LogMoM and the reference class method calculations, with a reduction in computational time of 31% to 62%.