Evaluating the Euler-Euler approach for predicting a strongly 3D bubble-induced recirculatory flow with OpenFOAM. (16th January 2021)

• Record Type: Journal Article
• Authors: Duguay, Jason; Lacey, Jay; Massé, Alexandre
• Is Part Of: Chemical engineering science. Volume 229(2021)
• Publication Date: 2021-01-16
• Subjects: Multiphase flow -- OpenFOAM -- TwoPhaseEulerFoam -- Euler-Euler -- Stereoscopic particle image velocimetry -- Computational fluid dynamics -- Grit chamber -- Wastewater treatment; Chemical engineering -- Periodicals; Génie chimique -- Périodiques; Chemical engineering; Periodicals; Electronic journals; 660
• DOI: 10.1016/j.ces.2020.115982
• Languages: English
• ISSNs: 0009-2509

Abstract

Highlights: Including lateral lift considerably improved recirculatory flux predictions. Improvements in recirculatory flux seen with increasing lateral flux coefficients. All models produced simplified predictions of the measured 3D flow field. Near-bed velocities were underpredicted which would affect sediment transport. Rise velocity, void fraction and free-surface bulking profiles were reasonable. Abstract: Wastewater treatment plants use bubbly flows to promote grit removal from the plant's influent. Computational fluid dynamics is an attractive tool to virtually prototype grit-chamber designs. This study assesses a two-phase unsteady Reynolds-averaged Euler-Euler (uRANS-EE) approach to predict bubble induced recirculatory flow in a laboratory tank of similar complexity to that expected in a full-scale grit-chamber. Numerical results are validated against a comprehensive experimental data set consisting of a 3D reconstruction of the flow field using a multi-plane stereoscopic particle image velocimetry technique, void fraction profiles using an optical probe and bubble rise-velocities using high-speed video. Depending on the combination of modeling parameters the predicted volumetric fluid flux returning towards the diffuser varied between 60 and 97% of the experimental flux. Predicted rise velocities exceeded measured values by a factor between 1.15 to 1.75 and near bed velocities were underpredicted by upwards of 25%. The principal 3D flow field features were predicted to a sufficient accuracy to support the use of the uRANS-EE approach for numerical grit-chamber prototyping.