By enhancing the premixing of fuel and air prior to combustion, recently developed lean-burn combustor systems have led to reduced NOx and particulate emissions in gas turbines. Lean-burn combustor exit flows are typically characterized by nonuniformities in total temperature, or so-called hot-streaks, swirling velocity profiles, and high turbulence intensity. While these systems improve combustor performance, the exiting flow-field presents significant challenges to the aerothermal performance of the downstream turbine. This paper presents the commissioning of a new fully annular lean-burn combustor simulator for use in the Oxford Turbine Research Facility (OTRF), a transonic rotating facility capable of matching nondimensional engine conditions. The combustor simulator can deliver engine-representative turbine inlet conditions featuring swirl and hot-streaks either separately or simultaneously. To the best of our knowledge, this simulator is the first of its kind to be implemented in a rotating turbine test facility.The combustor simulator was experimentally commissioned in two stages. The first stage of commissioning experiments was conducted using a bespoke facility exhausting to atmospheric conditions (Hall and Povey, 2015, “Experimental Study of Non-Reacting Low NOx Combustor Simulator for Scaled Turbine Experiments,” ASME Paper No. GT2015-43530.) and included area surveys of the generated temperature and swirl profiles. The survey data confirmed that the simulator performed as designed, reproducing the key features of a lean-burn combustor. However, due to the hot and cold air mixing process occurring at lower Reynolds number in the facility, there was uncertainty concerning the degree to which the measured temperature profile represented that in OTRF. The second stage of commissioning experiments was conducted with the simulator installed in the OTRF. Measurements of the total temperature field at turbine inlet and of the high-pressure (HP) nozzle guide vane (NGV) loading distributions were obtained and compared to measurements with uniform inlet conditions. The experimental survey results were compared to unsteady numerical predictions of the simulator at both atmospheric and OTRF conditions. A high level of agreement was demonstrated, indicating that the Reynolds number effects associated with the change to OTRF conditions were small. Finally, data from the atmospheric test facility and the OTRF were combined with the numerical predictions to provide an inlet boundary condition for numerical simulation of the test turbine stage. The NGV loading measurements show good agreement with the numerical predictions, providing validation of the stage inlet boundary condition imposed. The successful commissioning of the simulator in the OTRF will enable future experimental studies of lean-burn combustor–turbine interaction.

Recently developed lean-burn combustors offer reduced NOx emissions for gas turbines. The flow at exit of lean-burn combustors is dominated by hot-streaks and residual swirl, which have been shown–individually–to impact turbine aerodynamic performance. Studies have shown that residual swirl at inlet to the high-pressure (HP) stage predominantly affects the vane aerodynamics, while hot-streaks affect the rotor aerodynamics. Studies have also shown that these changes to the HP stage aerodynamics can affect the downstream intermediate-pressure (IP) vane aerodynamics. Yet, to date, there have been no published studies presenting experimental turbine test data with both swirl and hot-streaks simultaneously present at inlet. This paper presents the first experimental and computational investigation into the effects of combined hot-streaks and swirl on turbine aerodynamics. Measurements were conducted in the Oxford Turbine Research Facility, a short-duration rotating transonic facility that matches non-dimensional engine conditions. Two turbine inlet flows are considered. The first is uniform in total pressure, total temperature, and flow angle. The second features a non-uniform total temperature (hot-streak) profile featuring strong radial and weak circumferential variation superimposed on a swirling velocity profile. Area surveys of the flow were conducted throughout the turbine. Measurements and URANS predictions suggest that the inlet temperature non-uniformity was relatively well preserved upon being convected through the turbine, and relatively poor comparisons between URANS and experiment highlight the challenge of accurately predicting the complex IP vane flow.

Under the EU LEMOCTEC programme, the Oxford Turbine Research Facility (OTRF) was upgraded to include a modern 1½ stage, high-pressure turbine with film cooled highpressure guide vanes (HPVs) and low-turning intermediate pressure vanes (IPVs).The facility has also been upgraded to include a third-generation engine-representative combustor temperature and swirl simulator at inlet, allowing the study of turbine interactions with inlet conditions representative of a modern lean burn combustor.This paper presents the aerodynamic and mechanical design of the LEMCOTEC highpressure turbine and its integration and commissioning in the OTRF. Test data with uniform inlet flow is presented, acting as a baseline to assess the performance benefit of optimising the turbine design for a non-uniform combustor exit flow-field. Measurement techniques are discussed, and experimental data is compared to pre-test design CFD results from both the Rolls-Royce HYDRA code and the commercial CFX code.

It is known that a secondary effect of rotor-casing effusion cooling is to modify and potentially spoil the rotor over-tip leakage flow. Studies have shown both positive and negative impacts on high-pressure stage aerodynamic performance and heat transfer, although there remains no consensus on whether the net effect is beneficial when both aerodynamic and thermal effects are accounted for simultaneously. An effect that has not been extensively discussed in the literature is the change in stage operating point that arises due to mass introduction midway through the machine. This effect complicates the analysis of the true performance impact on a turbine and must be accounted for in an assessment of the overall benefit of such a system. In this paper, we develop a low-order (“mean-line”) analysis in an attempt to bring clarity to this issue. We then present results from experiments conducted in the Oxford Turbine Research Facility, a 1.5-stage transonic rotating facility capable of matching non-dimensional engine conditions, with effusion cooling implemented over a sector of the rotor casing. Finally, we present steady and unsteady Reynolds-Averaged Navier-Stokes (RANS) simulations both with and without cooling. The results are used show that with cooling, there are flow changes both locally to the cooled casing (changes to the tip-leakage and secondary flow structures) and globally (changes to the bulk-flow velocity triangles). For the present configuration, both changes contribute positively to stage efficiency.