Session: 12-06-02: Condensation and Phase Change Materials

Paper Number: 173749

A Simple Heat Transfer Model for Annular Flow Condensation 

A simple, physics-based model is proposed to describe the heat transfer in two-phase annular flow during condensation. Two-phase flow, and more specifically, annular flow, is a common phenomenon that takes place within a variety of engineering applications, such as nuclear reactors, HVAC, and electronics cooling. For these processes, the heat transfer coefficient needs to be accurately predicted as it cannot be directly measured. The key to accurate prediction is developing models which take an input of measurable parameters and output the heat transfer coefficient based on either previous data or physics-based modelling. Issues may arise with the development of certain models, however. Some models used to predict heat transfer coefficient heavily rely on empirical data, and use a two-phase multiplier on data from a single phase. This may allow for accurate prediction within the range of data given, but the model may deteriorate when given parameters outside of the range that was used to fit the model. Additionally, physics-based models may rely on solving a set of differential equations, which requires iteration and is not desirable. The model developed in this paper differs in both regards. The model is based on the analogy between single- and two-phase heat transfer, and it considers the thermal resistances for the boundary layers near the wall and the liquid-vapor interface, eliminating the need for either a two-phase multiplier or iteration. The model is not fitted to a specific set of data, remaining based in physics and adding flexibility to the range of parameters that it is able to predict. The model also yields an analytical expression for the two-phase multiplier, suggesting that it scales linearly with the inverse of the liquid fraction, which is confirmed by the dataset developed for testing the model. The scaling factor is linked to the dimensionless interfacial velocity (i.e., ratio of interface velocity to mean liquid velocity), which can be approximated as a universal constant. The model is compared with a curated database of heat transfer coefficient data from the literature, which has been digitized and collected into a large dataset for prediction and testing purposes, and it includes 27 sources with 26 refrigerants. The prediction error is 13.4% on average for data of small diameters (3.5 mm and smaller) and 17.1% for all data (including diameters from 0.76 to 14.4 mm). It also is cross validated against the same database, with the highest prediction error reaching only 14.5%. The model achieves prediction accuracy comparable to the existing empirical correlations, while offering the additional benefit of providing a physically grounded framework that can be further adapted and refined.

Presenting Author: Lingnan Lin University of Maryland, College park

Presenting Author Biography: Dr. Lingnan Lin is an Assistant Professor at the University of Maryland, College Park. Previously, he was a Research Scientist at the National Institute of Standards and Technology (NIST), where he initially trained as a visiting Ph.D. student and later as a Postdoctoral Research Associate. He earned his Ph.D. from Shanghai Jiao Tong University in 2018 and B.E. from the University of Science and Technology of China in 2013. Dr. Lin’s research is focused on fundamental phase-change heat and mass transfer, with applications to energy transport, conversion, and storage. He is a member of ASME and ASHRAE and serves on many committees within these organizations

Authors:

Chase Yankowski University of Maryland, College Park
Lingnan Lin University of Maryland, College park