Abstract
The multidisciplinary design optimization (MDO) of re-entry vehicles presents many challenges associated with the plurality of the domains that characterize the design problem and the multi-physics interactions. Aerodynamic and thermodynamic phenomena are strongly coupled and relate to the heat loads that affect the vehicle along the re-entry trajectory, which drive the design of the thermal protection system (TPS). The preliminary design and optimization of re-entry vehicles would benefit from accurate high-fidelity aerothermodynamic analysis, which are usually expensive computational fluid dynamic simulations. We propose an original formulation for multifidelity active learning that considers both the information extracted from data and domain-specific knowledge. Our scheme is developed for the design of re-entry vehicles and is demonstrated for the case of an Orion-like capsule entering the Earth atmosphere. The design process aims to minimize the mass of propellant burned during the entry maneuver, the mass of the TPS, and the temperature experienced by the TPS along the re-entry. The results demonstrate that our multifidelity strategy allows to achieve a sensitive improvement of the design solution with respect to the baseline. In particular, the outcomes of our method are superior to the design obtained through a single-fidelity framework, as a result of the principled selection of a limited number of high-fidelity evaluations.
Highlights
Modern space missions are increasingly supported by vehicles able to perform complex assignments and return safely to the Earth’s surface
Our framework aims at capturing the multi-physics nature of the mission and of the design of a re-entry vehicle, accounting for the contributions introduced by the propulsion system, the re-entry trajectory at a given entry point, the aerothermodynamic effect characterizing the re-entry path and the thermo-structural aspects associated with the sizing of the thermal protection system
Our multifidelity domain-aware active learning method is compared to a single-fidelity Bayesian framework based on the efficient global optimization formulation (Jones et al 1998) eliciting evaluations of the low-fidelity aerothermodynamic model only
Summary
Modern space missions are increasingly supported by vehicles able to perform complex assignments and return safely to the Earth’s surface. This work presents a computational framework for the multidisciplinary design optimization (MDO) of re-entry vehicles that aims to use at best a limited amount of interrogations of the high-fidelity aerothermodynamic model to sensitively improve the design solution. Minisci and Vasile (2013) addressed the design optimization of a manned vehicle reentering the Earth atmosphere, Vasile et al (2014) proposed an MDO framework for the design of a Mars entry micro probe Both these works leverage a similar surrogate-based optimization framework, where a neural network is used to approximate the aerodynamic loads. We introduce an original multifidelity strategy based on an active learning scheme to compute an aerothermodynamic surrogate model in the context of the multidisciplinary optimization of an atmospheric re-entry vehicle.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
