Storable propellant engine start-up is a complex phase which involves non-stationary hydraulic effects, two-phase flow. Each element of the engine, oxidizer and fuel lines, combustion chamber, has to be verified and qualified. For this purpose test facilities are an indispensable element for development and acceptance for space system, subsystem and components.
The partial failure of Ariane 5 flight No 510 on December 7 2001 called for the attention of liquid rocket propulsion engineers on the importance of the simulation of high frequency (HF) instabilities in propellant feed lines. The disturbances observed during the start-up of the flying engine were not triggered in any of the tests performed. As a consequence of the difference in the engine and test bench lines, transient phenomena result not predicable from a test campaign and so highly dangerous for engines performances and operation.
The knowledge of the flow characteristics in both the test benches and flying rocket stages is essential for future hardware design. Efforts have to be made for advancing the understanding of the transient flow behaviour in pipes for safe operation of the engine and for reducing the high costs and risks associated with tests. In this context, the development of tools for simulating the behaviour of flight-like feed systems by using non-flight-like test bench equipment are widespread.
In the present work a numerical investigation and evaluation of the critical fluid-system parameters is performed by means of the simulation and modeling software EcosimPro 4.4, based on C++ programming language. A hydraulic model of the test bench P2 in Lampoldhausen up to the test facility main valve is built up, as well as a simplified model of the Aestus engine. To validate and qualify the combined test bench-engine model, real on-ground test transients and steady-state results are compared with numerical results.
The numerical model has been successfully modified and adjusted and simulation results matched the measured values within acceptable ranges. Good agreement with steady-state-pressure, propellant mass flow rates behaviour and transient start-up in terms of water hammer peaks have been obtained. However, the water hammer frequencies have been not matched accurately due to 1D-restrictions of the code and mainly to different characteristics of the simulated fluids.
Future efforts have to be done to improve the implementation of fluid property data bases implementation, especially for NTO (equilibrium condition between NO2 and N2O4 depending on pressure and temperature) and for deeper understanding and investigation of possible influences of the content of pressurization gas content in the liquid fluid on the transient behaviour of the propellant feed system.