Dynamics of large and inertial scale turbulent structures in an internal combustion engine

This project will combine Large-Eddy Simulations (LES) with the time-resolved high-speed Particle-Image Velocimetry (PIV) of turbulent flows in a reciprocating internal combustion engine. While it is still common that in the engineering practice, the turbulence is modelled with either two-equations models, e.g. ubiquitous k-epsilon model, or with the approaches based on transport equations for individual components of the Reynolds-stress tensor, LES methods, despite all their promise, have not yet found a wide-spread use. On the other hand, with the recent development of fast repetition rate lasers, the time resolved PIV in engines have become possible but such applications are few and, at the moment, restricted to a "demonstration of capability". This project aims at combination of the two best available approaches in an attempt to elucidate the properties of the large and inertial scales of turbulent flows in engines. The previous experimental work shows that the combustion rate is determined by these scales rather than universal small-scale turbulence.

One particular obstacle to the application of either PIV or LES is the significant amount of data produced by these methods; this large amount is due to the unsteady character of the flows. Because of this, the project will consider methods of representation of the velocity fields in terms of a small number of modes using either Principal Orthogonal Decomposition (POD) or another similar technique.

The Combustion Group in the School of Mechanical Engineering in Leeds developed quite an extensive experimental database of burning rates of high-pressure premixed flames. Mr. Khan's study will concentrate on assessment and selection of various LES models, starting with the basic Smagorinsky's sub-grid model, using this database. This work will necessitate development of a computer code; it is anticipated that this development will benefit from the modern computer programming practices. The main outcome of Miss Khan's work should be a clear recommendation on what particular LES sub-grid model offer the best capabilities for combustion modelling under conditions approaching those existing in a modern gas turbine or a spark-ignition engine.