High-Resolution Experimental Measurements and Mechanistic Modelling of Saturated Cryogenic Pool Boiling Heat Transfer

Abstract

Refueling cryogenic rockets in low Earth orbits has the potential to significantly enhance the duration and the reach of future space missions. However, the development of such capabilities is not without challenges, which exacerbated by the complexity and the cost of testing such equipment in microgravity, e.g., a fuel depot placed in orbit. The low boiling point of cryogenic fuels (hydrogen, methane, oxygen) makes them prone to boil when transferred through superheated pipes, or simply stored in fuel tanks, resulting in the presence of two-phase mixture. Boil-off gas can lead to pressurization of components like fuel line and tanks, cause two-phase flow instabilities during fuel transfer, or significantly reduce the usable amount of cryogenic fuel. Progresses in multiphase computational fluid dynamics (mCFD) can be leveraged to predict the two-phase flow behavior in such cases. However current boiling models offer poor prediction accuracy of the boiling heat flux in most applications, but also miss critical boiling characteristics, e.g., the amount of vapor produced. This thesis proposes a fully closed formulation of a mechanistic boiling model adapted to saturated cryogenic pool boiling. The model leverages exhaustive measurements of boiling parameters. A new pool boiling setup was designed for that purpose, using pressurized nitrogen as a proxy for cryogenic fuel. The apparatus allows to measure the typical boiling curves, i.e., boiling heat flux and wall temperature, measurement of dry area fraction, triple contact line density as well as more fundamental parameters such as bubble lift diameter, nucleation frequency and nucleation site density among others. The heating surface inclination was varied between 0° (upward facing horizontal) and 179° (downward facing horizontal) to probe for the effect of buoyancy on the boiling parameters and overall heat transfer. The analysis of individuals bubble using both phase-detection and shadowgraphy showed the lack of microlayer. Instead, the large size of the bubble footprint observed experimentally strongly suggested an intense evaporation process at the triple contact line of the bubbles and occurring right after nucleation. Based on this observation, the evaporation rate at the triple contact line has been indirectly estimated on numerous bubbles and appeared consistent with analytical models describing such evaporation mechanism in other fluids. In the tested operating conditions, the linear evaporation rate was measured at around 5 W/m, accounting for 20% of the boiling heat flux. The amount of heat removed by quenching of the heating surface through transient conduction has also been evaluated using phase-detection recordings and was 3 shown account for an additional 20% of the boiling heat flux. The remaining portion of the boiling heat flux could be explained by neither triple contact evaporation nor transient conduction during quenching. The heat flux partitioning model proposed in this work allows to predict the quenching heat flux, the triple contact line evaporation as well as the observed dry area fraction, contact line density and bubble density during nitrogen boiling. Minimal effect of surface inclination has been observed on the nucleate boiling heat transfer, except at the departure from nucleate boiling.