Investigating direct-contact heat transfer of a spherical-cap liquid/vapour two-phase bubble evaporating in immiscible liquid: An experimental study

Abstract

For the first time, this study presents the experimental investigation into the direct-contact heat transfer of a single spherical-cap liquid/vapour two-phase bubble evaporating in an immiscible liquid medium. The experiments were conducted using a rectangular Perspex column with dimensions of 600 mm × 100 mm × 100 mm. N-pentane droplets, with a diameter of 3.5 mm at saturation temperature (∼36 °C), served as the dispersed phase, while distilled water at three different initial temperatures, creating three temperature differences (ΔT), which is equivalent to Jacobs number (Ja = 18, 30 and 45), was used as the continuous phase. The experiments were performed at an active height in the continuous phase (H) of 500 mm. A high-speed camera operating at 250 frames per second was utilised to record the evaporation process of the droplets along the continuous phase active height. The captured images were analysed using FASTCAM (PFV-4) and AutoCAD (3D) software for precise data evaluation. The formation of the spherical-cap two-phase bubble was observed under various operating conditions. Based on the experimental output, the thickness of the yet-to-be vaporised liquid averaged over the area of the flat base of the two-phase spherical cap bubble (δ), the size of the evaporating spherical-cap two-phase bubble $\left(D_e/D_o\right)$, and the convective heat transfer coefficient, expressed in terms of the Nusselt number (Nu), were determined and analysed. As anticipated, the experimental results demonstrated that Nu increased with a rise in the Reynolds number (Re) and $\left(D_e/D_o\right)$ with a significant inverse effect from the Jacobs number (Ja). Additionally, δ was found to decrease exponentially over time, and Nu increased as δ decreased. An empirical correlation for Nu as a function of Pecklet number (Pe) and Ja was developed and compared successfully with the experimental data with a maximum error of about 8 % at Ja = 30.