vapor compression system, transient cycles, heat exchangers modeling
The aeronautical industry is developing new environmental control system architectures, based on the more electrical aircraft approach, with the aim of improving the overall performance of aircrafts. The analysis of theses architectures is very complex as they include different thermal systems interacting between each other. Their design is commonly carried out with the support of numerical simulations. The aim of this work is to present a Modelica library to simulate vapor compression cycles at both steady and transient conditions. The study of these specific cycles is particularly challenging from both the phenomena and the numerical resolution point of views. The main goal of the library is to provide, not only accuracy and robustness, but also very low coputational time. Most of the components developed for the library are simulated by means of look-up tables or efficiencies for the sake of accuracy and quickness. Therefore the efforts were focused on two critical aspects of the system. On the one hand, the heat exchangers were developed based on a switching moving boundary approach in order to have a relatively high ratio between accuracy and resolution time. This method takes into account the distribution of phases along heat exchangers (this is a crucial thermal aspect for the accurate simulation of evaporators and condensers). The time consumption is very low compared to distributed models as only one control volumen per phase is used. On the other hand, several features of the whole vapor compression cycle resolution were tackled with meticulosity, namely, the refrigerant amount management, the initilization procedure, the thermostatic control loop, and both the steady and transient simulation aspects. The results shown in this work are devoted to highlight three main characteristics of the simulations carried out from the developed library. 1) The numerical robustness of the system and its components. In particular, the heat exchanger switching moving boundary model was tested at all posible transitions, while the vapor compression cycle was also subjected to a wide range of boundary conditions. 2) The low computational time required for simulations. The system resolution at steady state conditions needs very few seconds to converge. 3) The accuracy of predictions. The simulations at both component and system levels showed good agreement with experimental data.