Graph-Based Dynamic Modeling of Two-Phase Heat Exchangers in Vapor Compression Systems

Abstract

Vapor compression systems are a popular thermal management solution for high heat load applications because they benefit from the high heat transfer rate of two-phase fluid flow. However, model-based design and control of these systems proves challenging because of the complex and coupled hydro-thermal dynamics as the working fluid changes phase in the heat exchangers. When modeling these heat exchangers, there exists a strong tradeoff between model accuracy and computational complexity. While different techniques have aimed to reduce computational complexity of the model, they come at the cost of a decrease in accuracy. On the other hand, high fidelity models are computationally expensive to run. To bridge the gap between computational efficiency and accuracy, this work develops a novel multi-state graph-based dynamic modeling approach. The multi-state graph modeling approach is used to develop a dynamic model of a two-phase heat exchanger. The heat exchanger model is integrated into a vapor compression system model and the model behavior is verified against other models present in the literature. The results demonstrate that the multi-state graph model can accurately capture system dynamics within 1% of the current state of the art modeling approaches while providing valuable modularity for system level modeling. Additionally, the computational load can be made comparable to less accurate approaches.