Numerical modelling
NCCAT leverages cutting-edge numerical modelling to unlock deeper insights into combustion systems—pushing beyond the limits of experimental methods and driving next-generation technology.
Numerical modelling work is undertaken at NCCAT to gain further understanding of the combustion system and processes than is possible using experimental methods alone.
The modelling tools developed can also be used directly in the development of next generation combustion systems. The following sections give an idea of some of the areas in which NCCAT develops and applies numerical models.
Combustion CFD
Simulations using both commercial and in-house CFD codes, using both RANS and LES methods, are performed on combustion systems. Research into the models themselves is undertaken, for example, considering the coupling of simulations of the combustor with those of the up and downstream turbomachinery, or how modelling strategies for hydrocarbon fuels can be adapted for Hydrogen combustion. The simulations are also used to understand and develop experimental methods.
Combustor aerodynamics
NCCAT has a long history of investigating and improving the aerodynamic performance of combustion systems and CFD methodologies are integral to this. RANS and LES approaches are used as necessary and the close collaboration with experimental work ensures that these simulations are well validated.
Combustor network models
Full LES simulations of combustion systems can be computationally expensive which limits their utility as a design tool, particularly when complex chemical reaction processes need to be included. Combustor network models represent a potentially powerful way of including these complex reaction processes in a low cost simulation tool. Work at NCCAT looks at how these models can be built, used and integrated with other simulation and experimental work.
Fuel spray modelling
Predicting and understanding fuel atomization is central to the development of any liquid fuelled aircraft engine. However, this process is challenging to measure. Work at NCCAT uses high fidelity simulations to look at the atomization process in detail. These simulations can offer more than experiments alone as the full topology of the spray, as well as velocity and pressure fields, are known at each instant. These computationally expensive simulations cannot be used as routine design tools and work at NCCAT also takes place to develop Lagrangian spray models within practical combustion simulations.
Machine learning
Machine learning (ML) techniques are also being developed and applied to help understand and make use of the large quantities of data that can be produced using modern experimental and simulation techniques. For example, ML image processing techniques can be used to correlate flame appearance to emissions performance.