A Fully Coupled Approach for the Integration of 3D-CFD Component Simulation in Overall Engine Performance Analysis

C. Klein,S. Reitenbach,D. Schoenweitz,F. Wolters

DOI: https://doi.org/10.1115/gt2017-63591

2017-06-26

Abstract:Due to a high degree of complexity and computational effort, overall system simulations of jet engines are typically performed as 0-dimensional thermodynamic performance analysis. Within these simulations and especially in the early cycle design phase, the usage of generic component characteristics is common practice. Of course these characteristics often cannot account for true engine component geometries and operating characteristics which may cause serious deviations between simulated and actual component and overall system performance. This leads to the approach of multi-fidelity simulation, often referred to as zooming, where single components of the thermodynamic cycle model are replaced by higher-order procedures. Hereby the consideration of actual component geometries and performance in an overall system context is enabled and global optimization goals may be considered in the engine design process. The purpose of this study is to present a fully automated approach for the integration of a 3D-CFD component simulation into a thermodynamic overall system simulation. As a use case, a 0D-performance model of the IAE-V2527 engine is combined with a CFD model of the appropriate fan component. The methodology is based on the DLR in-house performance synthesis and preliminary design environment GTlab combined with the DLR in-house CFD solver TRACE. Both, the performance calculation as well as the CFD simulation are part of a fully automated process chain within the GTlab environment. The exchange of boundary conditions between the different fidelity levels is accomplished by operating both simulation procedures on a central data model which is one of the essential parts of GTlab. Furthermore iteration management, progress monitoring as well as error handling are part of the GTlab process control environment. Based on the CFD results comprising fan efficiency, pressure ratio and mass flow, a map scaling methodology as it is commonly used for engine condition monitoring purposes is applied within the performance simulation. Hereby the operating behavior of the CFD fan model can be easily transferred into the overall system simulation which consequently leads to a divergent operating characteristic of the fan module. For this reason, all other engine components will see a shift in their operating conditions even in case of otherwise constant boundary conditions. The described simulation procedure is carried out for characteristic operating conditions of the engine.