A novel integrated PDB-FDB model for the prediction of flow boiling heat transfer under high sub-cooling and very high heat flux conditions

•The heat-transfer characteristics of heated circular channels were examined.•The effect of control parameters on flow boiling heat transfer was analyzed.•The predictability of existing heat transfer correlations was evaluated.•The effect of bubble agitation on sub-cooled flow boiling heat transfer was analyzed.•A novel integrated PDB-FDB heat-transfer correlation was developed. Although numerous studies on subcooled flow boiling heat transfer have been conducted, those evaluating very high heat fluxes of several MW/m2 under one-side heating conditions are limited. In this study, the heat transfer characteristics of a circular smooth channel are evaluated under one-sided heating conditions with a high heat flux of up to 11.82 MW/m2. An experiment to measure the heat transfer (including boiling performance) from single-phase to critical heat flux (CHF) under an inlet pressure of 0.1–1 MPa, mass flux of 2000–6000 kg/m2s, and inlet bulk temperature conditions of 30–140 °C is conducted. The boiling regime is analyzed by dividing it into a partially developed nucleate boiling (PDB) regime and fully developed nucleate boiling (FDB) regime, using the onset of the nucleate boiling (ONB) point and onset of the fully developed nucleate boiling (OFDB) point. A total of 15 existing subcooled flow boiling heat transfer correlations for the PDB and FDB regimes are evaluated with the experimental results. In the PDB regime, the Liu and Winterton correlation reveal good predictability within 11%. However, all the evaluated existing correlations of the FDB regime exhibit high error rates. In this study, a new FDB correlation is developed by considering the agitation heat flux (an expected heat transfer characteristic in the FDB regime) in the Liu and Winterton correlation with Python code and the AI regression method. An evaluation of the developed correlation with the data of other previous studies, such as those of Araki et al. and Jianguo Yan et al. reveals that the newly proposed correlation can predict the FDB heat transfer under conditions of higher pressure (approximately 5 MPa) and higher flow rate (approximately 10,000 kg/m2s) with a low error rate of 17%.