Abstract
Augmented backside cooling refers to the enhancement of the backside convection of a combustor liner using extended heat transfer surfaces to fully utilise the cooling air by maximising the heat transfer to pumping ratio characteristic.
Although film cooling has and still is widely used in the gas turbine industry, augmented backside cooling has been in development for decades now. The reason for this, is to reduce the amount of air used for liner cooling and to also reduce the emissions caused by using film cooling in the primary zones.
In the case of micro gas turbines, emissions are of even greater importance, since the regulations for such engines will most likely become stricter in the following years due to a global effort to reduce emission.
Furthermore, the liners investigated in this paper are for a 10 kWe micro turbine, destine for various potential markets, such as combine heat and power for houses, EV hybrids and even small UAVs. The majority of these markets require long service intervals, which in turn requires the combustor liners to be under the least amount of thermal stress possible.
The desire to also increase combustor inlet temperatures with the use of recuperated exhaust gases, which in turn increase the overall system efficiency, limits the cooling effectiveness of the inlet air. Due to all these reasons, an advanced form of augmented backside cooling would be of substantial significance in such a system.
Currently some very simple designs are used in the form of straight plain fins, transverse strips or other similar geometries, but the creation of high heat transfer efficiency surfaces in such small sizes becomes very difficult with traditional subtractive manufacturing methods. When using additive manufacturing though these types of surfaces are not an issue.
This paper covers the comparison of experimental results with conjugate heat transfer CFD models and empirical heat balance models for two different AM liner cooling geometries and an AM blank liner. The two cooling fin geometries include a rotating plain fin and an offset strip fin. The liners were tested in an AM built reverse flow radial swirl stabilised combustion chamber at a variety of operating conditions. During the experiments the surfaces were compared using a thermal camera to record the outer liner temperature which was viewed through a quartz outer casing.
The experimental results showed that the cooling surfaces were effective at reducing the liner temperatures with minimal pressure losses for multiple operating points. Those results were then compared against the conjugate heat transfer CFD models and the empirical calculations used to design the surfaces initially.
From this comparison, it was noticed both the CFD and empirical calculations under predicted the wall temperatures. This is thought to be due to inaccuracies in the predicted flame temperatures and the assumed emissivity values used to calibrate the thermal imaging camera. Further uncertainties arise from the assumption of a constant air and hot gas temperature and mass flow along the cooling surfaces and the lack of data for the surface roughness of the parts.
The Application of Discontinuous Galerkin Methods in Conjugate Heat Transfer Simulations of Gas Turbines
