Mixing Time and Scale‐up Investigation of a Moving‐Baffle Oscillatory Baffled Column

Abstract

Mixing time and scale-up behavior were investigated for a moving-baffle oscillatory baffled column (MB-OBC). Computational fluid dynamics software was used to investigate the effects of the oscillation conditions and changes in the column scale on the dimensionless mixing time. The results indicate that small amplitudes combined with high frequencies are detrimental to the mixing efficiency and that the scale-up of an MB-OBC is linearly dependent on the column diameter. This work, in combination with previous studies, proves that the predictability of the hydrodynamic behavior during scale-up is inherent to oscillatory baffled columns. This trait is desirable for adapting these columns for industrial-scale applications.

Design, performance characterization and applications of continuous oscillatory baffled reactors

“…where 𝑢𝑢 𝑖𝑖,𝑟𝑟 = �𝑢𝑢 𝑥𝑥 2 + 𝑢𝑢 𝑦𝑦 2 Equation ( 8) is weighted by area, where J and 𝐼𝐼 are the total number grid data points in the axial and radial directions, and equation ( 9) is weighted by volume. A number of studies have used 𝑅𝑅 𝑉𝑉 to analyse the flow performance of OBRs with single orifice baffles (Jian and Ni, 2005;Manninen et al, 2013;Mortazavi and Pakzad, 2020;Ni et al, 2003a;Sutherland et al, 2021). They observed that with increasing 𝑅𝑅𝑅𝑅 𝑜𝑜 , transverse velocities increased, thereby decreasing 𝑅𝑅 𝑉𝑉 .…”

Revisiting the Effect of U-Bends, Flow Parameters, and Feasibility for Scale-Up on Residence Time Distribution Curves for a Continuous Bioprocessing Oscillatory Baffled Flow Reactor

“…Sutherland et al expected mixing time to decrease with tube diameter scale-up; however, when numerically modeled, a proportional increase in mixing time was observed using geometric similarity in a moving baffled column when scaling. 29 Ahmed et al found that when scaling a helical baffled column at three different scales by extending the column length and tube diameter, RTD curves are not affected, provided there is a geometric similarity and dimensionless numbers remain the same across scales. In their study, a model was also developed to predict the TiS value as a function of dimensionless numbers when scaling a single-column helical baffled system.…”

“…Design of experiments (DOE) to evaluate key operational parameters such as amplitude ( x 0 ), frequency ( f ), and their corresponding dimensionless numbers is used to maximize plug flow within the OBR system. ,,,, Other studies have investigated the impact of oscillatory parameters, tube diameter, geometric designs, and baffle types on both axial dispersion and mass transfer when scaling OBRs but usually have confined it to a single column identified in Table . Sutherland et al expected mixing time to decrease with tube diameter scale-up; however, when numerically modeled, a proportional increase in mixing time was observed using geometric similarity in a moving baffled column when scaling . Ahmed et al found that when scaling a helical baffled column at three different scales by extending the column length and tube diameter, RTD curves are not affected, provided there is a geometric similarity and dimensionless numbers remain the same across scales.…”

Power consumption in a moving baffle oscillatory baffled reactor: A CFD study

“…[ 20 ] In their latest work, they performed a scale‐up study on moving baffle OBR using the same CFD model and by considering power consumption along with velocity ratio (a metric to quantify mixing performance) by changing the column diameter and keeping and constant. [ 21 ]…”

“…[20] In their latest work, they performed a scale-up study on moving baffle OBR using the same CFD model and by considering power consumption along with velocity ratio (a metric to quantify mixing performance) by changing the column diameter and keeping Re o and St constant. [21] In our previous study, [10] we created an experimentally validated 2D axisymmetric environment model to study the mixing performance in a moving baffle OBR. Despite good agreement with experimental data, the 2D model could not fully replicate the asymmetrical and transient turbulent nature of the flow.…”