Simulations of Transport Phenomena and Porous Structures Using Dissipative Particle Dynamics

Abstract: The topic of this work is simulations of porous materials and transport phenomena at a mesoscopic length scale, i.e., a length scale larger than an individual atom but smaller than the continuum. The porous material studied is a membrane used in fuel cells. Mass transfer has also been studied in the membrane. Heat and mass transfer has been studied in parallel plate channels, first in saturated channels, then the study is expanded to also cover droplet flow. In both cases, solidification, i.e. phase change from liquid to solid, has been studied. So what is the purpose of this study? Many of our merging technologies rely on physical processes that occur at small length scales. In a fuel cell membrane, protons are transported in pores with a diameter of only a few nanometres. The chemical reactions in the fuel cell are dependent on oxygen, protons and electrons reacting with each other on a nanometre-sized active surface. In both cases, fluid flows are of pivotal importance to the performance of the entire fuel cell, and these flows are creeping at a flow rate on the scale of nanoliters per second! A flow of this magnitude will behave quite differently than one in a larger application. To understand how these fluid processes work in detail and how the active porous materials are constructed can help us to construct cheaper and more effective batteries and fuel cells. This is the motivation for the work carried out in this thesis. The aim of this work is to achieve a deeper understanding of transport phenomena and porous materials at a mesoscopic length scale. To conduct experiments is costly, and therefore computer simulations have been chosen as the approach for this work. To be more specific, the aim has been to reconstruct the mesoscopic structure of a fuel cell membrane and evaluate its structural and transport properties, and to evaluate heat and mass transfer in a parallel plate channel where specific care has been taken to describe the heat transfer in the inlet of the channel and simultaneous fluid flow and solidification. This work is based on computer simulations using the technique dissipative particle dynamics (DPD). The study shows that the model of the porous structures agrees with previous studies, both regarding mass transport and structural parameters. The studies of mesoscopic flow in parallel plate channels predict a more effective heat transfer in the inlet of the channel as compared to a comparable macroscopic channel, and that solidification will depend on the temperature difference between the inlet and the cold walls and that the impact of flow velocity will be negligible.