Initiation and growth of confined vapour bubbles in micro-channels

State-of-the-art mechanistic models for confined bubble flow boiling in microchannels, such as [1] by Thome et al., currently use sub-models for the initiation of bubble growth that are based on observations of conventional nucleate boiling. Reversed flow towards the micro-channel inlet has been attributed by Kandlikar [2] to a pressure spike at the location of bubble nucleation. Mukherjee and Kandlikar [3] have performed a fully 3-dimensional simulation of growth after nucleation through to the early stages of confined growth at constant saturation temperature, implying constant pressure in the bubble. Detailed observations of the inception of confined bubble growth during flow boiling of water in a rectangular 2 x 1 mm channel without artificial nucleation sites showed that nucleation did not lead immediately to continuous growth [4]. The observations are considered in more detail in this presentation. Inception may be separated from the originating nucleation event in time by up to 10 ms and in axial position by up to 10 mm. The stochastic processes of nucleation and inception are influenced by the local pressure variations and by the flow reversal associated with inlet compressibility. The pressure variations are caused by the confined growth of bubbles pushing liquid slugs along the channel, and not by bubble nucleation. A 1-dimensional simulation of experimental data for confined bubble growth in a capillary tube with no inlet flow and initially at uniform temperature, incorporating the feedback between the varying pressure and the superheat driving growth, has shown that confined growth is sensitive to the initial conditions [5]. Improvements to the model are presented, including an estimation of convective heat transfer to the nose of the bubble, and the difficulties of modelling the growth from nucleation to full confinement are discussed