Study of the effects of walls on vortex formation and liquid maldistribution with two-phase flow around a spherical particle via numerical simulation

With the ever-increasing dependency on the activity of industrial catalysis in fixed bed reactors (FBRs), recent microscale studies based on liquid film spreading behavior on particle surfaces using numerical methods have attracted much attention. However, most studies have investigated one particle separately, and the effects of vessel walls or other particles are not considered. In our preceding study, we found that liquid spreading behavior and flow patterns are strongly related to the gas vortex and its shape under the particle. Therefore, in this work, the formation and scale of the gas vortex, as well as its generation and breakdown during the flow process, are considered as the parameters in the simulation. Two geometries named the near-wall configuration and the normal configuration were assumed and fully investigated using ANSYS FLUENT v15.0 with different operating parameters. As a result, both the size and the formation of the gas vortex without wall effects are larger and steadier than those with wall effects, and correspondingly, the liquid film is well distributed on the surface regardless of the gas and liquid (G/L) velocity because of the equivalent velocity of the gas trace near the particle. In contrast, in the near-wall configuration, when the flat wall approaches the particle, two vortexes under the particle probably break down and further merge into one larger vortex and cause liquid maldistribution on the particle surface during the development of the flow process. The velocities of the gas traces are inequivalent, even causing a reversed flow and thus trapping the liquid phase inside the interstices, which can be alleviated by increasing the rate of G/L velocity and setting a larger interstice between the particle and the flat wall. The newly uncovered information will enhance the understanding of liquid spreading behavior on particle surfaces and thus be applicable to the performance of FBRs. (C) 2019 Elsevier B.V. All rights reserved.