Zhenyu Xu

My PhD thesis is mainly about the numerical simulation of the porous media’s drying. Drying of porous media is a process of significant scientific and commercial interest in a number of industrial areas including coating, food processing, production of building materials, pulp and paper industry, and pharmaceuticals.

There are two major categories of numerical simulation methods for porous media problems: the continuum and discrete methods. In the continuum methods, the porous medium is treated as a continuous material with volume-averaged macroscopic properties, while ignoring the porous media’s pore-scale structure and attributes. These methods are useful in large space-scale problems, but its accuracy heavily depends on the determination of macroscopic properties, which must come from experiments or pore-scale simulations. The discrete methods, instead, take the pore-scale structures and phenomena in the porous media into account. These methods are usually used to simulate small space-scale problems or to determine the macroscopic properties for continuum methods.

The pore network model, one of the discrete models, has been studied in many researches and still is a very active topic. However, the multiscale problems existing in the transportation between the micrometer level pore-scale pore network and the meter level lab-scale porous media has rarely been studied. So the first problem I addressed is to clarify the factors influencing the multiscale transport problem, including the spatial and temporal schemes, algorithms, and mesh. An efficient method couple-solving the invasion-percolation pore network model and outside flow is developed and verified.

Previous research proved that the numerical results of basic pore network models have good agreement with experiments in phase distributions, but large errors in drying rate and drying time. It is further proved that the film effect is critical in prediction of the drying rate and drying time. So the second problem I addressed is to incorporate the film effect into my coupled transport model. This unique model is verified to have a good agreement with experiments in all three aspects. This model is used to study the slow drying of thin porous media in flow.

The current film effect model treats the film thickness as constant in the whole pore network. This defect makes it unsuitable for studying multi-porosity porous media. My third effort is to modify the current film effect model and to use it in the simulation of some dual-porosity porous media with variable pore shapes. A software, the byproduct of my thesis, PORODRY, based on these methods is developed in C++ from scratch and tested running in multiple platforms, including laptops, workstations and the clusters of UWM high performance computing center.

Contact Information:  zhenyuxu@uwm.edu