Skip to main content
PGT Hub

Research with impact in Aerospace Engineering

Queen Mary's General Engineering research publications were ranked 4th in the UK in the 2014 REF. Find out more and read about how we are developing robust and effective methods to reduce jet engine noise at source. 

GPU LES of flow around a jet engine

The School of Engineering and Materials Science provides internationally-leading research activity represented by five academic divisions. Research is supported by industrial, clinical and academic partnerships as well as a thriving Joint Research Institute with Northwestern Polytechnical University in China. We have a long track record of worldwide research impact.

Our research has impact on industrial practice. In addition, we work to ensure that our innovations are understood and appreciated by the broader public.

We are developing computational models to reduce the noise produced by jet engines. Aircraft noise is known to be responsible for many physiological and psychological effects. It is, therefore, of great importance for major aviation industries, such as Airbus and Embraer, to better understand the aerodynamic noise generation mechanisms and develop more robust and effective methods to reduce the noise at source.

This work is supported by the EPSRC through the JINA project, which involves Queen Mary University of London and the University of Bristol.

The JINA team, led by Professor Sergey Karabasov, works on the development of next-generation high-fidelity computational aeroacoustics methods tailored for Graphics Processing Unit (GPU) computations. Our recent studies demonstrated that advanced computer architectures (GPU) in combination with modern computer algorithms (high-resolution schemes, compact stencil, asynchronous time stepping, low memory footprint based on single precision arithmetic), which exploit existing open source grid generation and wall modelling libraries (OpenFOAM) in combination with the standard acoustic integral methods (FW-H) make first-principle methods such as Large Eddy Simulations (LES) feasible in ‘compute under the desk’ mode. In comparison with conventional LES implementations the current approach offers a reduction in time/cost more than 10 times. The significant acceleration of high-fidelity simulations achieved makes it feasible to use such tools inside a Multi-Disciplinary-Optimisation loop where one of the goals is noise reduction. The application of high-fidelity computational methods such as LES in the design optimisation process at low cost may revolutionise the design process of new aero engines where the aeroacoustic performance, which requires high-fidelity modelling, is increasingly important.

More research stories are available on our website.

Back to top