(88b) Numerical Study of Gas-Liquid Flow in Trickle Bed Using MPS Method
World Congress on Particle Technology
2018
8th World Congress on Particle Technology
Fluidization & Multiphase Flow
Novel and Non-Conventional Reactors and Multiphase Flow Systems I
Wednesday, April 25, 2018 - 8:45am to 9:05am
Kei Sakakura1* ,Ryou Tanabe1 and Yasunobu Kaneko1
Technology & Engineering Center, Idemitsu Kosan Co., Ltd Chiba, Japan.
*E-mail: kei.sakakura@idemitsu.com
Abstract
1.Introduction
Trickle-bed reactors are widely used in chemical and energy fields due to simple geometry and easy reactors operation. However, the overall performance of reactors are deeply affected by catalyst bed (packing configuration, porosity, particle shape, size, coating), flow maldistribution, mass and heat transfer rate and reaction kinetics. Some of the important issues in microscopic modeling of gas-liquid flow through packed beds are closure models for interphase drag, wetting, and capillary forces. It is important to understand how liquid-gas interacts with solid surfaces with complex shapes. To clarify that, Liquid spreading in gas-liquid concurrent trickle beds is investigated using MPS approach considering the wall contact angle. Liquid phase is simulated by MPS as lagurangian particles, while gas phase is solved as eulerian method using mesh. To evaluate the MPS model, complex geometry of packed beds with sphere, quadrillobe and quadrihollow types of catalyst are obtained by DEM simulation. The MPS model is shown to achieve good agreement with experimental data and to predict accurately, tendencies of liquid spreading and tracer ink residence time through the complex geometry. .
2.Experimental setup
The experimental were conducted using 0.15m width,0.5m tall,0.015m thickness rectangle cold-model packed actual three demetallization catalysts whose shapes are sphere, quadrilobe, quadrihollow type, and water and air connected flow meter were supplied concurrently from top of packed bed. Each flow rate was adjusted as same superficial velocity as that of actual trickle bed reactor. After the liquid and gas flow was fully developed in the bed, black ink was injected in liquid as tracer for visualization of the liquid flow. Liquid dispersion rate and black ink passage time thorough the bed were measured by video camera for CFD validation.
3.Modeling description and simulation condition
Packing geometries are set up from catalyst packing simulation using DEM (Discrete Element Method). Each catalyst is composed by small DEM particles which are fixed in a catalyst. When catalysts contact each other in packed bed, small particles are used to judge the collision and predict the collision force and angular moment to solve the motion of the catalysts. After that, packing geometries are converted to CAD data(stl file) for MPS simulation.
MPS method is one of the particle methods which can be applied to fluid flow. Since the mesh is not required, large deformation of liquid free surfaces can be calculated without mesh distortion even in complex geometry. The governing equations are represented by lagrangian description and it is not necessary to discretize the convection terms. In this study, liquid movement is simulated by MPS with consideration of surface tension force and gas phase is solved by conventional fundamental equation using mesh. Therefore, lagragian and eularian method are coupled.
4.CFD Results
In packed bed composed of sphere type catalyst, the liquid flow though the bed does not dispersed well as compared with quadrilobe and quadrihollow type. On the other hand, quadrilobe type shows the good dispersibility to the wall direction due to high packing density and high fluidity caused by stick- like shape of catalyst. And CFD results show reasonably good agreement with the experiment results regarding the liquid dispersion and liquid passage time through the beds.In all the packed beds the gas velocity thorough the bed is almost uniformed and shows higher velocity at the wall due to higher void rate.
Keywords
DEM MPS trickle flow