With the space industry growing, environmental considerations become increasingly important, especially with respect to the propulsion systems used to launch satellites into space and control their position in orbit. Since traditional satellite propellants are highly toxic, there is an increased demand for green, i.e., environmentally friendly, substitutes like hydrogen peroxide.
This work explored the modelling of hypergolic bi-liquid thrusters in the framework of the Greenlam project, which aims to develop a 100N hydrogen peroxide kerosene thruster. While previous works were either experimental or focused on staged H2O2–RP-1 engines with a catalyst bed, this thesis investigated a numerical approach and focused on unstaged engines, aiming to identify and validate models viable to simulate the decomposition of hydrogen peroxide and subsequent combustion with kerosene with the aid of a catalyst.
Transient three-dimensional simulations were performed. k-ω SST, the Peng Robinson real-gas equation of state and Species Transport with Finite Rate chemistry were employed to model turbulence, gas properties and reactions, respectively. The effect of the catalyst was represented by adapting the Arrhenius rate parameters. Propellants were injected using the Discrete Phase Model. The Eulerian model was shown not to be suitable to simulate the propellant injection and atomisation.
A coaxial, an impinging-jet and a pintle injector were considered. Simulations with the coaxial injector showed good agreement with data obtained from CEA and with other rocket engines. Simulations with the impinging-jet and pintle injector failed to capture droplet impingement and consequent atomisation and thus could not be validated.
Both stoichiometric and fuel-rich propellant mixtures and H2O2 concentrations of 95% and 98% were simulated. Thrust was between 62 and 63N under sea-level conditions, equivalent to 103 to 105N in vacuum and hence approximately 3 − 5% higher than anticipated. Chamber temperature reached up to 2763K. Chamber pressure was 7.6bar. The stoichiometric mixtures showed higher thrust output, higher chamber temperature and higher wall temperature than the fuel-rich mixtures. The higher concentrations led to higher chamber and wall temperatures. Analysing the kerosene mass fraction in the exhaust showed that in any case at least 9% of the injected kerosene was ejected unburnt due to a lack of mixing, and most of the additional kerosene in the fuel-rich mixtures was also simply ejected. The chamber walls reached temperatures of up to 3271K, about 500K higher than bearable by the material. While the coaxial injector was shown to be a cause for the high wall temperatures due to unfavourable propellant distribution, an adiabatic wall boundary condition was assumed which likely also led to an overestimation of the temperature.
A set of models applicable for simulating hypergolic bi-liquid rocket engines was found and validated. More work is required in terms of injector design and modelling, confirmation of reaction rate parameters and wall modelling.