Gas Turbine Combustor Flow Structure Control Through Modification of the Chamber Geometry

As combustors are put in service, problems such as erosion, hot spots, and liner oxidation occur, and a solution based on lessons learned is essential to avoid similar problems in future combustor generations. In the present paper, a combustor flow structure control via combustor geometry alteration is investigated using laser Doppler velocimetry. Mainly, three configurations are studied. The first configuration is that of a swirl cup feeding a dump (rectangular cross section) combustor. The rectangular chamber is configured with a width to breadth (w/b) ratio of 85%. The second configuration is similar to the first one, but a combustion dome is installed. The dome is configured with a 9 deg difference in the expansion angle on both sides (asymmetric dome). The third configuration is that of a swirl cup and a combustion dome installed in a prototype combustor (single annular combustor (SAC) sector), with both primary and secondary dilution jets blocked. The SAC is configured with a cross sectional area that decreases toward the exit through the tilting of the inner combustor liner. The results show that the combustion dome eliminates the corner recirculation zone and the low velocity region close to the combustor walls. The combustion dome asymmetry results in a significant asymmetry in the velocity magnitude, as well as the turbulence activities and the tilting of the central recirculation zone (CRZ) toward the surface with the higher expansion angle. The liner tilting results in a 40% reduction in the length of the CRZ. However, once the primary jets are open, they define the termination point of the CRZ. The chamber w/b ratio results in a CRZ with the same diameter ratio (85%) in all configurations. Interestingly, the maximum reverse flow velocity is roughly constant in all measurement plans and configurations up to a downstream distance of 1R (R is the flare radius). However, with open primary jets, the CRZ strength increases appreciably. It appears that the confinement dictates both the flow field outside the CRZ and the size of the CRZ, while the swirl cup configuration mainly influences the strength of the CRZ. Regarding turbulence activities, the presence of the dome damps the fluctuations in the expanding swirling jet region. On the other hand, the primary jets increase the turbulence activities appreciably in the jet impingement region, as well as the upper portion of the CRZ (60% increase).