A. Introduction

Microcooling elements in micro-thermal-mechanical systems (MTMS) are a modern high-end technology. Microchannel heat sinks are extensively used in high-heat flux systems, e.g., microelectronic, optical and microfluidic devices, high-performance supercomputers, electric vehicles, advanced military avionics, power and laser devices, very-high temperature gas-cooled reactors and liquid metal fast reactors, radiator panels of spacecraft, thermal control of satellites, etc. In pharmaceutical and chemical industries, by integrating heat exchangers, mixers, and reactors as a block of microcomponents, it is possible to achieve a significant decrease in the system size. Heat exchangers based on microscale channels are used in automotive fuel cells where the heat dissipation system should be compact and contain a minimum number of components. The heat dissipation rate is extremely high in these devices. Phase-change processes such as flow boiling are very important in such devices due to possible higher heat transfer coefficients and much lower temperature gradients along the heat exchanger. In addition, two-phase microchannel heat sinks are one of the strongest candidates for heat removal devices in high-heat flux environments by virtue of their large surface area-to-volume ratio, compact dimensions, and low flow rate requirements. Flow boiling in microchannels has drawn considerable attention from researchers worldwide. Microchannels are used in MEMS and high-end microprocessor cooling applications. Microchannels are used in applications involving high operational pressures (up to 600 bar) and temperatures (1,000 °C). There is a scale effect that becomes evident when the hydraulic diameter is less than the capillary length, i.e., [(σ/g)(ρ_l - ρ_v)]^0.5.

In microchannels, the hydraulic diameter of the channels is smaller than the capillary length. Identifying the threshold, or the transition band, beyond which a two-phase flow may be considered "micro" is an open issue. The channel classification has been given by Kandlikar and Grande (2003). This classification should be used as a mere guide indicating the size range, rather than rigid demarcations based on specific criteria. Relevant literature in this direction would be Kawaji and Chung (2003). The confinement number (Co), as the distinguishing parameter, gives only a rough idea,

\[\label{eq1} \mbox{Co}=\frac 1D\sqrt{\dfrac{\sigma}{(\rho_l-\rho_v)\,g}} \tag{1}\]