Prof. N. Swaminathan

Cambridge University, Engineering Department Cambridge, UK CB2 1PZ

Gas turbine OEMS are constantly looking for new methods and concepts to develop “green” combustors to meet the energy demand and tight emission levels together. A robust and precise predictive capability is indispensable to achieve this. To develop such capabilities a thorough understanding of fundamental processes involved is required. It is well known that direct numerical simulation (DNS) is a powerful tool to gain these knowledge. However, they are usually limited to simple geometries and low Reynolds number flows. The use of this methodology for flow at high Reynolds number in complex geometries of engineering interest is impractical because of the computational demand. However, the fundamental knowledge gained from simple canonical problems can be leveraged to build relevant models for practical use. A case study to highlight this will be discussed and it will cover (1) DNS and its data analysis, (2) the new fundamental knowledge gained, (3) translation of this knowledge into a simple model for turbulent combustion of contemporary interest, and (4) sample results from industry standard simulations of a gas turbine combustor operating at 1 and 3 bar conditions, using these combustion models. Appropriate comparison to experimental measurements will also be made.