Characterization of Flow Freezing in Small Channels for Ice Valve Applications
Freezing of water flowing through a small channel can be used as an efficient and cost-effective flow control mechanism for microfluidic platforms. Ice valves provide a leak-proof, non-invasive and high-pressure-tolerant method of flow control compared to their conventional micromechanical counterparts. To develop, design and implement ice valves an understanding of the processes and parameters that govern ice formation in small channels is required. The aim of this dissertation is to understand the freezing process of water flowing in small diameter channels and the factors affecting the same, so as to develop a simple physical model that predicts the ice growth process and channel closing time. Further, a stand-alone ice valve formation device is developed that is suitable for implementation in high-pressure microfluidic applications.
While ice valves have various advantages over conventional microvalves, their successful implementation is in part hindered due to their long response time. An understanding of the factors that affect the ice growth process during flow freezing in small diameter channels (commonly encountered in microfluidic devices) would allow reduction and control of the response time of ice valves. In this dissertation, freezing in a pressure-driven water flow through a channel is investigated using measurements of external channel wall temperature and flow rate synchronized with high-speed visualization. A test setup is designed and demonstrated to control the external cooling boundary conditions during visualization of the ice formation modes in a small channel; the external wall thermocouple and the water flow rate readings are synchronized with the high-speed images. Firstly, the effect of water flow rate on the freezing process is investigated in a glass channel of 500 m inner diameter in terms of the external wall temperature, the growth duration of different ice modes, and the channel closing time. Freezing initiates as a thin layer of ice dendrites that grows along the inner wall and partially blocks the channel, followed by the formation and inward growth of a solid annular ice layer that leads to complete flow blockage and ultimate channel closure. A simplified analytical model is developed to determine the factors that govern the annular ice growth, and hence the channel closing time. The model identifies the water flow rate and the channel diameter as the two key parameters that govern the channel closing time. For a given channel, the model predicts that the annular ice growth is driven purely by conduction due to the temperature difference between the outer channel wall and the equilibrium ice-water interface. The flow rate affects the initial temperature difference, and thereby has an indirect effect on the annular ice growth. Higher flow rates require a lower wall temperature to initiate ice nucleation and result in faster annular ice growth (and shorter closing times) than at lower flow rates.