Numerical Simulation and Advanced Modelling of Hydrogen-Fuelled Propulsion Systems for Reusable Launch Vehicles

点击翻译按钮获取中文摘要

1 OverviewReaction Engines Ltd (REL) is a UK-based company formed in 1989 to design and develop technologies for a new class of hypersonic propulsion system - the Synergetic Air-Breathing Rocket Engine (SABRE). REL's breakthrough technology is an ultra-lightweight heat exchanger that enables the SABRE engine to use conventional aerotechnology at unconventionally high speeds by preventing engine components overheating. SABRE technology will enable aircraft to fly over five times the speed of sound in the atmosphere, and realise single-stage-to-orbit space launch vehicles, reducing costs and improving access to space.Key to the success of the SABRE engine is using hydrogen as fuel, burning under unconventional conditions that are far from well-understood. Combustion at these conditions includes inherent instabilities (thermodiffusive, hydrodynamic and thermoacoustic) that pose signicant challenges to combustor design. Prototyping such world-leading technology is extremely expensive, and numerical simulation can provide unparalleled insight into the fundamental physical processes involved in such novel combustors, thereby aiding the design and development process. However, existing numerical simulations have demonstrated a sensitivity to models used to describe turbulent mixing and flame physics; there is a need for high-resolution numerical simulations to develop and validate reliable turbulent combustion models that can be used for combustor design and development.2 MethodologyA database of high-fidelity three-dimensional Direct Numerical Simulations (DNS) will be constructed that consider both the turbulent mixing of direct injection of hydrogen fuel into a sub-sonic cross-flow, and the subsequent downstream lean combustion at conditions relevant for SABRE. NASA's low-emission lean-hydrogen burner will be used as a target test cases, and Reaction Engines will provide further necessary conditions that can be used for initialisation.The DNS calculations will be carried out using PeleLM (and PeleC, if necessary), developed at the Lawrence Berkeley National Laboratory in California, and built on a framework with which the PI has over fteen years' experience. Key advantages include exploitation of the low Mach number approximation and adaptive mesh renement, both of which signicantly reduce computational expense, and is optimised for use on massively-parallel supercomputers, making it a world-leading computational tool for turbulent combustion. Simulations increasing in size will be carried out in stages, starting from a high-specication workstation, through the NU HPC facility "Rocket", up to the national supercomputer "Archer".The DNS data will be analysed in detail to assess existing turbulent flame models used for Reynolds Averaged Navier Stokes (RANS) and Large Eddy Simulation (LES) approaches, and to guide development of new models as appropriate. The models will be implemented and tested in an engineering CFD software framework, and validated on the same cases as conducted for the DNS simulations; this will allow direct comparisons to be made and to iterate on the model development, thereby establishing the condence in the approach so that it can be applied more generally to aid the design and development of SABRE combustor technology.Suggested research areas: Continuum Mechanics, Combustion Engineering, Fluid Dynamics and aerodynamics, Numerical Analysis, Hydrogen and alternative energy vectors