How does primary liquid break-up determine the downstream spray characteristics of airblast atomisers?

Better fuel sprays {e.g. smaller drop sizes to improve evaporation, better 'patterns' - and low droplet 'clustering' in time - to avoid fuel rich zones for soot and NOx production, maintaining these qualities between idle to full power } are known to be a key to the design of greener aeroengines. Although fuel atomization is an old technology, remarkably little fundamental knowledge exists which would reliably answer the simple question: given this shape of atomizer, what kind of fuel spray would I produce? The answer must be able to explain how and why modifications of the order of a fraction of a millimeter to the atomiser radically changes the spray. To formulate the answer, we need better and more extensive measurement of the 'fundamental' processes than hitherto: and we must advance our ability to calculate the flow (waves, instabilities, ligament formation, 'pinch-off', etc.) from first principles - and check the advance against fundamental, simple yet representative sprays.The overall aim of the proposal is to generate new computational fluid dynamic (CFD) modelling strategies for the atomisation process of liquid jets in airblast atomisers with an emphasis to geometries used in combustion systems and aero-engines. Our approach will be to make novel time- and spatially- resolved measurements of the liquid, spray and gas motions in the primary breakup regions of co-flowing and cross-flow atomiser configurations using optical instrumentation. We will use these to evaluate the results of Large eddy simulations (LES) of two atomiser geometries, based on an open source CFD code, to predict the temporal development of the atomisation process, in terms of internal liquid flow and surrounding gas flow, liquid breakup and spray characteristics downstream of liquid jet intact length. We intend to establish the extent to which CFD can predict the three transitions - waves to ligaments to droplets - in primary atomization in these geometries, which is as yet completely untested and unknown. We will use the evaluated code to perform 'numerical experiments' with emphasis on the dense liquid core, where measurements are hard to make, to further elucidate the physics.