Integrated Compressor Duct with Bleed: Experimental Validation of a Hybrid RANS/LES Approach

Other conference contribution, 2019

In this study, a comparison is performed between measurements done in an in-house experimental test rig at GKN Aerospace Engine Systems and simulations done using the commercial CFD solver ANSYS CFX. The aim is to compare a hybrid RANS/LES model to a more industrially applied RANS model in terms of capabilities of predicting the measurements.
The geometry represents an intermediate compressor duct from an aircraft engine, with an integrated bleed-pipe upstream of the duct. The bleed-pipe gives a better engine control when run at part speed by uncoupling the mass-flow through the upstream and downstream components. Previous studies indicated no major separations in the intermediate compressor duct and therefore, to save computational resources, wall-functions were employed for the hybrid simulations.
Overall, the CFD simulations compare well with measured data in terms of wall-pressure coefficient in the intermediate compressor duct as well as at upstream and downstream locations. In terms of normalized total pressure profiles, the CFD and the experiments agree upstream of the test section. However, at an evaluation surface downstream of the duct, only the hybrid model simulations are capable of predicting the measured total pressure. Furthermore, it has to be kept in mind that even though the SBES simulations where significantly more expensive compared to the RANS simulations, the RANS results only represent the averaged flow field, with minimum information about transient behaviour. On the contrary, the SBES simulations represent the transient flow field, where the results have to be time-averaged.