In order to meet the increasingly strict emissions targets required in modern civil
aviation, lean burn combustors are being pursued as a means to reduce the environmental
impact of gas-turbine engines. By adopting a lean air/fuel mixture NOx production may be
reduced. The increase in proportional amount of high pressure air entering directly into the
combustor reduces the amount available for cooling of the combustor liner tiles. A reduced
mass of air places restrictions on the porosity of cooling arrays, requiring a departure from
applications of pedestal and slotted film cooling typically used to cool double skin combustor
liners. An alternative approach applied to lean burn combustors places impingement and
effusion arrays on the cold and hot skins respectively for cooling of both sides of the hot liner
skin. Although impingement cooling is well established as a means of promoting forced
convection cooling, there are many areas on a liner tile where cooling behaviour is not well
characterised. Additionally, film cooling reduces combustive efficiency and increases the
production of NOx and CO, prompting interest in reducing its use in combustor cooling.
The research for this thesis has focussed on investigations into current and proposed
geometries to identify methods to enhance cold side cooling in lean burn applications. A
fully modelled combustor liner tile has been used for investigation into the impact of
structural and pressure blockages on cold side cooling performance of an impingementeffusion
array using a transient liquid crystal technique to measure heat transfer performance.
Research has found structural blockages can reduce heat transfer performance to ~60% of
typical values, with crossflow development due to pressure blockage producing similar
reductions in Nusselt values to ~70% of typical.
A second investigation explored enhanced cooling geometries combining a distributed
impingement feed over roughened channels of pedestals at variable height (H/D) and pitch
(P/D). A newly proposed ‘Shielded Impingement’ concept combines full height pedestals, to
protect impingement jets from developing crossflow, with quarter height pedestals for
turbulence enhancement of crossflow cooling. The research has found that Shielded
Impingement geometries displayed the strongest cooling performance of all tested designs
due primarily to increased downstream Nusselt numbers. Pressure losses were comparable to
short pedestal geometries, with little apparent effect of full height pedestals. Low pressure
losses mean that application to extended channels in line with the full tile geometry is
A Doctoral Thesis. Submitted in partial fulfilment of the requirements for the award of Doctor of Philosophy of Loughborough University.