Numerical Investigation of First Stage Base Heat Flux of Secondary Nozzles in Bleed Engine and Gas Generator Cycles

Kurzfassung

The correct estimation and prediction of thermal loads acting on a launch vehicle plays a vital role in characterizing the requirements on the thermal protection system (TPS) of the spacecraft in order to avoid catastrophic failure. With the increasing demand for reusable launch vehicles (RLVs), which represent a resource- and cost-efficient alternative to conventional space transport systems, additional challenges in the assessment of the thermal loads arise due to the changing flow field during the return phase of the vehicle and especially the retro-propulsion and landing phase. While the flight trajectory and conditions are the same for RLVs and conventional space transport systems until MECO and stage separation, the counterflowing supersonic exhaust jets during the retro-burn have a strong effect on the thermal loads on the rocket structures and dominate the flowfield characteristics during this flight phase. While some studies on the base heating have been conducted for conventional launch vehicles [28] and RLVs [23], [44], [11], no published research has been devoted to the topic of side-effects due to the outflow of gas generators, bleed nozzles and air vents of cryogenic fuel tanks for both kinds of vehicles. Due to these secondary exhaust jets, unburned hydrogen, which is used to power the oxidizer and fuel pumps for the main combustion chamber, is ejected near the high temperature outflow of the main engines. In order to fill this knowledge gap and provide an assessment of the additional influence on the thermal loads acting on the vehicle baseplate due to these secondary nozzles, steady state CFD simulations were carried out for a first stage without aerodynamic control surfaces and landing legs. For all simulations the second-order finite-volume DLR TAU code [34] was applied using the Reynolds-Averaged Navier-Stokes (RANS) equations together with a one-equation Spalart-Allmaras turbulence model. A reduced Jachmimowski mechanism is applied for the chemistry modelling in order to capture the potenital effects of post-combustion. A detailed description of the numerical setup is given, including the free stream conditions for the simulated trajectory points during the ascent and retro-burn phase, the geometrical configuration of the bleed nozzle arrangement and aft-bay design, as well as the procedure for local refinement methods. A main aspect of the numerical setup was the derivation of the secondary nozzle outflow. For the simulation of these secondary nozzles, two main engine cycles, represented by an expander bleed engine cycle and a gas generator cycle, were identified as suitable modelling options, since they represent two state-of the art restartable engine cycles. The functionality of these engine cycles was described and the procedure for deriving and rescaling the secondary nozzle chamber conditions from already existing configurations to match the chosen engine characteristics is discussed in detail. The chamber conditions are then used to conduct separate 2D axisymmetric nozzle simulations, to reproduce the outflow conditions of the secondary nozzles and interpolate them on the 3D mesh used in the flowfield simulations. Additionally a grid convergence study was conducted to confirm the validity of the numerical mesh and provide an estimate on the numerical uncertainties. The influence of the additional bleed nozzle outflow is analyzed by comparing simulations for the ascent and retro-burn phase with activated and deactivated bleed nozzles. For the case of activated bleed nozzles a significant change in flowfield and thermal loads can be observed. Despite the fact, that the massflow of the secondary nozzles makes up only about 2% of the main engine mass flow, the recirculation and entrainment of hot main engine exhaust observed in the case of deactivated secondary nozzles is prevented, leading to a shielding effect of the baseplate and a reduction of thermal loads. While the mean heat flux along the baseplate for the case of deactivated secondary nozzles is given by 48.2 kW/m 2 and 20.9 kW/m 2 for the ascent and retro-burn phase, these values are reduced to -18.9 kW/m 2 and -2.8 kW/m 2 for the configuration with activated bleed nozzles. The wall temperature in these simulations equals 600 K. The influence of the applied main engine cycle was investigated, which only had a small influence on the base heat flux. Representative points along the ascent and descent trajectory were investigated and the observed flow patterns were described. For the ascent phase a converging behaviour of the base heat flux can be seen until the distribution of the thermal loads along the baseplate becomes independent of the increasing ambient pressure. During the retro-propulsion phase only three of the main engines are active, leading to a shielding effect against the incoming free stream. In case of activated bleed nozzles the observed heat flux values are in the negative range, indicating a reduction of the thermal loads compared to simulations with deactivated secondary nozzles. The presented results clarify the necessity of including the outflow from secondary nozzles in heat flux predictions of launcher configurations, both for conventional space transport systems and RLVs, to provide accurate assessments of the aerothermal loads on the vehicle base.