Multi-Objective Optimization of Aero Engine Combustor Adopting an Integrated Procedure for Aero-Thermal Preliminary Design

Abstract

The preliminary design of an aero-engine combustor is a multidisciplinary process that involves an extensive and systematic analysis of the design space. Simulation-driven approaches, in which several design configurations are numerically analyzed, may lead to heterogeneous models interacting with each other, sharing miscellaneous information within the process. Iterative and user-defined approaches, moreover, are inefficient when multiple and conflicting requirements are in place. To rely on integrated design methodologies has been demonstrated to be beneficial, especially if adopted in a structured approach to design optimization. In this paper, the application of the Combustor Design System Integration (DSI) to the definition of an optimal combustor preliminary configuration will be presented. Given a combustor baseline design, the multi-objective optimization problem has been defined by targeting an optimal distribution for temperature profiles and patterns at the combustor’s exit. Dilution port characteristics, such as hole number and dimension as well as the axial position of the row have been selected as design variables. To guarantee a water-tight design process while minimizing the user effort, the DSI tools were included in a dedicated framework for driving the optimization tasks. Here, a proper CFD domain for RANS, constituted by the flame tube region extended to the dilution port feeds, was adopted for imposing the air split designed for the combustor. Concerning a “complete” combustor sector, this allows a reduction in the computational effort while still being representative for its aero-thermal behavior. The optimization task was performed using a Response Surface Method (RSM), in which multiple, specific combustor configurations were simulated and the CFD result elaborated to build a meta-model of the combustor itself. Finally, the suitability of the resulting optimized configuration has been evaluated through an “a posteriori” analysis for thermal conditions and emission levels (NOx and CO). A lean combustion concept developed by Avio Aero with the aim of the homonymous EU research project, the NEWAC combustor, has been considered as test case.

Design methodology for combustor in advanced gas turbine engines: a review

“…The ideal method for gas turbine combustor design optimization, which is difficult to evaluate performance and has many design targets, is to use an optimization tool that combines 3D performance prediction for multiple design targets. Recently, Elmi et al (2021) proposed one way to solve this problem by using a response surface method (RSM) that simulates multiple specific combustor configurations and elaborates the CFD results to build a meta-model of the combustor itself. Sampling methods and response surfaces evaluated for the specific optimization problem are shown in Figure 12.…”