ASTOS – The Complete Toolchain for Launch Vehicle Trajectory Simulation, GNC Design and V&V

Abstract

The design cycle of a launch vehicle combines a variety of disciplines that each require expertise and dedicated tools in the development. A significant improvement in development efficiency can be achieved having one coherent toolchain that can cover all the different topics. The framework presented in this abstract comprises a set of interoperable, reconfigurable, and flexible tools based on ASTOS and MATLAB/Simulink. It allows for the rapid and agile high-fidelity simulation of launch vehicles trajectories and easy prototyping of GNC algorithms up to detailed design and V&V of GNC Flight Software. The toolchain is centred around the ASTOS scenarios that contain the definition of the vehicle characteristics (e.g., propulsion and structural design parameters), environmental (e.g., gravity and atmospheric models), and dynamic models (e.g., equations of motion). This allows the user to adapt the level of fidelity of their models depending on the task and development stage. With a few simple changes the user can modify their 3DoF scenario - used for trajectory and vehicle optimization as well as design of guidance and navigation algorithms - to a 6DoF scenario – used for detailed model-based control design and V&V of GNC algorithms. The simulator is automatically configured based on the underlying scenario guaranteeing consistency and data integrity. This enables users to focus their efforts on the actual development of GNC methods, saving precious time and resources developing and setting up simulators that need to be updated with every iteration and change of the launch vehicle. The simulations can be performed either within the ASTOS software itself or in MATLAB/Simulink making use of the ASTOS Simulink interface. The ASTOS toolchain proved its merits on a variety of activities ranging from early design to spacecraft checkout and operational support. While most use cases address expendable launcher configurations, also reusable concepts are strongly supported such as first stage return flights making use of retro engines and aerodynamic control flaps, towing of a fly-back booster as studied in the FALCon project, air-launch concepts or hypersonic flight vehicles. Flexible structures can be simulated in an ASTOS scenario via Linearized Flexible Dynamics (LFD) or Multi-Body Simulation (MBS). This allows to setup the flexible model in an agile way and capture the impact from the flexible deflection of launch vehicle structural components and sloshing on navigation and control algorithms. In addition, MBS allows for detailed analysis of stage and payload separation. ASTOS is well known for its trajectory optimization capabilities which is an important step in the GNC development. The optimized reference trajectory serves as a basis for both guidance and control development. Linearized models of the dynamics of the launch vehicle form the basis of all model-based control design methods and open the door to advanced (robust) analysis tools providing quick insight into the performance of a controller. The proposed toolchain provides a tool to linearise the user-defined scenario at chosen time instances along the optimized reference trajectory. The user is also free to select the range of physical effects to include in the linearised dynamics among launcher flexibility, sloshing, sensor dynamics, tail-wag-dog effect, local aerodynamics, jet damping and pitch-yaw coupling. This modularity allows the user to work with simpler linearised launcher dynamics in the initial analysis phases and progressively increase the complexity of the linear model, up to validation and verification. A Verification tool allows to validate the obtained linearized models against the nonlinear simulator, giving insights into their correctness and domain of validity about the chosen grid. To validate the performance and robustness of the developed GNC strategies in a high-fidelity non-linear simulation, a Monte-Carlo (MC) environment is included within the toolchain. This environment is responsible for configuring the simulator to perform the desired tests, either single runs or full MC campaigns, with the target launch scenario, GNC algorithms and the GNC parameterization. The MC environment allows to, not only disperse the “real-world” parameters according to the user inputs and scenario specification, but also change the configuration of the GNC programmatically to perform sensitivity and parametric analysis to aid the algorithms tuning. The MC environment employs several desired features, identified in previous activities. It separates the generation of the configuration files for the MC campaigns from the execution of the simulations. In this way, the execution of the simulation is not hindered by the configuration and the repeatability of the test is ensured. It allows the user to select which simulation outputs to store and at which rate. It is possible for the user to reconfigure the simulator with the parameters of a specific MC run to easily reproduced identified behaviours. It implements plotting and reporting utilities to expedite the post-processing and analysis of the results by the automated generation of plots and reports.