A methodology for coupled CFD-DEM modeling of particulate suspension rheology

The flow properties, or rheology, of particulate suspensions are highly dependent on the properties of the particles suspended within the base fluid (e.g. size, shape and surface properties). An understanding of the suspension rheology can help in the prediction of its behaviour under various flow conditions. Many studies focus on the experimental measurement of suspension properties, commonly employing devices such as rheometers to measure fluid properties under different conditions. A numerical model that is able to simulate the real-world interactions that determine particulate suspension rheology would complement those experimental studies. Accordingly, this work outlines a methodology for the development of such a model. Due to the differences between the two phases in a suspension, two different numerical methods were used, namely Computational Fluid Dynamics (CFD) and the Discrete Element Method (DEM). CFD uses a continuum approach to model the fluid component, while DEM resolves the behaviour of each individual particle. Two separate software programmes were used. For CFD, Open FOAM® was chosen, and for DEM, a programme called LIGGGHTS was used. These two different codes were coupled together with another programme called CFDEM. All three packages are open source software. To measure the rheology of the mixture, it was decided to simulate a rheometer. In particular, a rate-controlled, concentric-cylinder arrangement was chosen. Flow would be driven by a moving inner wall. Particle surface charge was accounted for by including both the van der Waals and electrostatic long-range forces between particles. This combination is known as the DLVO force. Plain particles, with no DLVO forces, were also considered. To the author's knowledge, using a coupled CFD-DEM approach to model suspension rheology had never been attempted before. Therefore, it was decided the development of the model would be done in stages, adding more complexity as each stage proved successful. The first step was to model a reduced rheometer geometry using CFD. Both a Newtonian and a non-Newtonian single-phase fluid were tested. Water and a Herschel-Bulkley mineral slurry were used respectively. Different rheometer geometries were tested. Results from these models correlated well with experimental values. The single-gap rheometer geometry with a 500μm gap between the inner and outer walls was found to perform the best. Final CFD model parameters used in these simulations were used as the basis for the coupled model. To reduce computational complexity, the model size and shape had to be reduced from a full-sized rheometer to that of a small rectangular box, with opposing flat walls acting as inner and outer cylinders of a rheometer. This improved computational efficiency. CFD tests conducted on the new box geometry showed that a box with sides of length 50μmproduced results equivalent to larger, full-sized, single-gap rheometer geometries with curved walls.

Grantor dc:publisher.institution

Department of Chemical Engineering

Year dc:date.issued

2015

Author dc:creator

• Smuts, Evan Matthew

Advisors dc:contributor.advisor

• Deglon, Dave
• Meyer Chris

Language dc:language.iso

eng

Handle dc:identifier.uri

http://hdl.handle.net/11427/16782

OAI identifier oai:identifier

oai:open.uct.ac.za:11427/16782