Surrogate data-driven physics-aware model for the conceptual aerodynamic design of rotating detonation engines

Purpose This work presents a methodology for the conceptual aerodynamic design of rotating detonation engines (RDEs). The purpose of this paper is to assist feasibility studies and to help in the planning of experimental campaigns. Design/methodology/approach The methodology is based on a surrogate data-driven physics-aware model. Simplified CFD is used to generate a data-tensor-form surrogate model of the RDE. Then, the surrogate model feeds an algorithm that solves the direct problem (prediction) and the inverse problem (multi-parameter design) of the RDE. The direct problem inputs are both operational and geometric RDE parameters. Outputs are performance parameters. The inverse design problem computes operational and geometrical parameters from prescribed performance parameters. Findings The main practical finding is that the use of surrogate model-based data-driven techniques helps to increase the fidelity level of aerodynamic conceptual design tools for RDE. Originality/value Traditionally, conceptual design studies of RDE have been carried out using thermodynamic path analysis in different forms. In this work, a methodological step ahead, which incorporates CFD simulations from the onset, is proposed. This was achieved by making use of a surrogate model approach. The main value of the proposed methodology is that it allows for extensive parametric studies with limited computational cost.