Two-phase reservoir: development of a transient thermo-hydraulic model based on bond graph approach with experimental validation

The main purpose of the project FUI THERMOFLUID is to study the feasibility of a new electronic cooling system embedded on flying objects (missile, satellite, and airplane). The technology chosen consists of a pumped two-phase flow cooling loop (PTPFL). It is an innovative technology with a transport capacity of the thermal power up to 10 MW.m, exceeding in this way the performance of all other technologies. A PTPFL is a cooling loop based on the exploitation of the latent heat properties of the fluid trapped inside the loop, and moved by a pump. The components constituting a PTPFL are: a two-phase reservoir (TP-R), a mini-channels evaporator, a brazed plate condenser, a pump, and pipes. The global research work is devoted to propose a dynamic model and experimental validation of the PTPFL. The present article is exclusively dedicated to the TP-R. Indeed, this element plays a key role in the functioning of PTPFL. Historically, the TP-R did not equip the first cooling loop. However, due to its advantages, its introduction was essential. The developed dynamic model will be used in another work to predict the thermal hydraulic efficiency of the PTPFL from its mechanical and fluidic parameters, to conduct the study of transitional regimes and instability problems, and provides an original tool dedicated to design the TP-R in function of the thermal power levels to be evacuated and the selected refrigerant. The bond graph methodology is adopted for modelling works because of its energetic approach and multi-physics character of the studied system. The new model proposed in this article has many originalities: First, it is based on bond graph approach. Nowadays, the open literature shows that no bond graph model has been developed for such thermo-fluid system. Second, the dynamic model of TP-R pays great attention to phenomena that have never been taken into account in works cited in the present article, such as evaporation and condensation. Third, different conducto-convective heat exchanges are modelled without any experimental recalibration of the thermal exchange coefficients, unlike models proposed in the literature. In fact, all coefficients are systematically calculated using adequate correlations.