THERMODYNAMIC CONTROL SYSTEMS FOR CRYOGENIC PROPELLANT STORAGE DURING LONG MISSIONS

摘要

One of the key problems for space launcher, in the context of long duration missions using cryotechnic propulsion, is the pressure rise of the propellant tank due to residual thermal inputs (solar fluxes, mechanical connexions). In order to control self-pressurization and as an alternative to simple but propellant-consuming direct venting (DV) systems, a concept of thermodynamic venting systems (TVS) is considered. The main objective of the TVS concept is to control pressurization with a minimum penalty in mass and propellant losses. The present work combines 0-dimensional modelling and laboratory experiments to quantify the performance of the proposed control cycles. The TVS system relies on the following key ideas: some liquid propellant is removed from the tank, cooled down thanks to a pump and a heat exchanger and re-injected inside the tank as a spray in both vapour and liquid phases. This cold spray makes the vapour condense and homogenizes the liquid phase temperature by mixing. The real propellant (i.e. oxygen or hydrogen) is replaced in the experiment, close to room temperature and pressure, by a simulation fluid with a behaviour similar to that of real fluids in cryogenic conditions. This desirable feature is achieved in the experiments using a fluoroketone Novec by 3M. The geometry of the experiment is mostly a cylindrical tank containing the volatile fluid as a liquid and its vapour close to equilibrium. A regulated electrical heating simulates the thermal input. The tank wall temperature is carefully imposed by a constant temperature water loop such that the temperature control set point equals the inside fluid temperature. Consequently there is almost no heat exchange at the wall. Taking advantage of the easy-to-manage experimental conditions (compared to cryogenic ones), various operating conditions of the simulation tank can be imposed: constant volume, constant pressure, constant mass. The evolution of the control performance is analysed when changing control parameters such as: injection pressure and temperature, injection flow-rates. The 0-dimensional model can be applied with the same choice of parameters, which allows a detailed comparison between model predictions and measurements. A series of about 100 experiments on transient cooling and pressure control has been performed. The cross-comparison of 0-dimensional modelling simulations with measurements demonstrates the validity of the model but also evidences its limits.