Explicit numerical simulation of microfluidic liquid flows in micro-packed bed

Abstract

Microfluidic systems are of tremendous technological interest as demonstrated by their use in chemical analysis (so called ‘lab-on-a-chip’) and biochemical analysis (e.g. to detect biomarkers for disease), and in process intensification. Packed beds of micro-sized particles possibly utilized for enhancing heat and mass transfer in microfluidic devices, where the flow regime is normally laminar, as well as provide significant increases in surface area per unit volume for analytical chemistry and biochemistry, and for separation and purification. Whilst macro-scale packed beds have long been well understood, the same is not true of their microfluidic counterparts, which we term micro-packed beds or μPBs. Of particular concern is the effect that the small bed-to-particle diameter ratio has on the nature of the bed packing and the hydrodynamics of the flow within them. This lack of understanding stems in part from the challenges that are faced in experimentally assessing μPBs and the flow through to them. The study reported in this thesis addresses these concerns through a two developments. In the first body of work, a new method is proposed for the accurate reconstruction of the structure of a μPB from X-ray micro-computed tomography data for such beds. The porosity obtained from μPB was, within statistical uncertainty, the same as that determined via a direct method whilst use of a commonly used technique yielded a result that was nearly 10% adrift, well beyond the experimental uncertainty. This work particularly addresses the significant issues that arise from the limited spatial resolution of the tomography technique in this context. In the second part of the work reported here, a meshless computational fluid dynamics technique is used to study Newtonian fluid flow through μPBs, including determination of their permeability and the by-pass fraction due to wall effects, which are important in these beds. This use of a CFD allows determination of parameters that are difficult to determine experimentally because of the challenges faced in measuring the small pressure drops involved and the absence of the limited spatial and temporal resolutions of various imaging techniques. The meshless method used here also overcomes the challenges normally faced when seeking to discretise the complex three-dimensional pore space of the packed bed. The developments here open the way to studying more complex μPB configurations, and other processes within them such as non-Newtonian flows and mass and heat transfer.