Studies published over the past decade have established the importance of sliding bubbles in enhancing the heat transfer in subcooled two-phase flow in channels small enough to confine the bubbles. Recent work in this laboratory (Ozer et al., 2011, 2012) proposed that the primary enhancement mechanism is a single-phase convective mechanism: the transport of cold fluid nearer the wall due to the mixing behind the bubble. This is in contrast to two phase-change mechanisms: distributed bubble nucleation and the evaporation of the liquid microlayer between a sliding bubble and the surface. The work reported here explores this hypothesis by comparing the heat transfer enhancement produced by injected air bubbles to Ozer’s measurements obtained with naturally nucleated vapor bubbles. Data were collected under similar conditions in a highly subcooled laminar flow of Novec 649 in a horizontal rectangular minichannel of 1.21 to 1.484 mm channel spacing. The channel was formed by an electrically heated metallic upper wall and an unheated transparent lower wall. For the air/liquid flow, bubbles were injected at either a single point on the lower wall or through a sintered metal plug. The latter system produced a more channel-filling distribution of bubbles. A high-speed imaging system recorded the bubble motion and liquid crystal thermography recorded time-averaged surface temperature data. The comparison is presented in the form of the streamwise evolution of surface temperatures and the enhancement in time-averaged Nusselt number. Also, results for the passage of a single air bubble are presented. The air/liquid flow produced a Nu enhancement of between 120–350% compared to a single-phase flow at the same conditions. The passage of the single gas bubble produced a decrease in the wall temperature directly behind the bubble of 2–5 °C. The Nu enhancement produced by the air/liquid data and the nucleated vapor data is well correlated to appropriate dimensionless groups involving bubble diameter and frequency. The results from both data sets support the contention that a transient transport/mixing model developed previously for the vapor/liquid case captures the dominant single-phase convective mechanism in sliding bubble flows in highly confined channels.

No abstract