Experimental Characterization of Sample Tubing Dynamics for the Improvement of Droplet Microfluidic Feedback Control Systems

Abstract

This thesis presents the development of an experimental methodology to characterize pressure transient dynamics across liquid sample tubing in the plant of a pressure-driven droplet microfluidic feedback control system (PDMFCS). To progress the PDMFCS towards being a widely adopted fluidic analysis tool for non-expert end-users in various biochemical fields, i.e. progressing the PDMFCS towards modularity, will require utilization of an accurate plant model to establish informed and robust control system design procedures. Increased accuracy in the plant model can be obtained through development of experimental methods to characterize the dynamics associated with the plant components. Previous PDMFCS implementations have approximated the sample tubing dynamics using a hydrodynamic equivalent circuit model (HECM), but did not experimentally validate this model. As well, pressure transient studies performed for other fluid applications do not model a flow scenario physically similar to that occurring through the sample tubing during PDMFCS operation, further justifying the need for this present study.

Pressure transient dynamics across the sample tubing of the PDMFCS plant were found to be characterized as an approximately linear first-order system with transport lag through estimation of a transfer function (TF), with 95% confidence in uncertainty in the estimated parameters, from an experimental frequency response obtained by simultaneously measuring pressure waves at the inlet and outlet of the sample tubing. Decreasing the average inlet pressure, increasing the tubing length to inner diameter ratio, or increasing the fluid viscosity, resulted in a decrease of the corner frequency (an increase in the time constant) of the frequency (step) response of the experimentally estimated TF. Comparing experimentally estimated TF dynamics to those predicted by the HECM showed that, due to the assumption of hydrodynamic steady-state flow inherent to its derivation, the HECM fails to quantitatively approximate the pressure transient dynamics across the sample tubing. The primary conclusion of this study is that the experimentally estimated TF should be used, instead of the HECM, to approximate the sample tubing dynamics within the PDMFCS plant model. Using the experimentally estimated TF to approximate the pressure transient dynamics across the sample tubing should improve the plant model accuracy, such that informed and robust control system design methodologies can be developed for the PDMFCS, which will enhance the modular potential of the system.