The interaction between rotor and stator airfoils in a multistage turbomachine causes an inherently unsteady flow field. In addition, different relative circumferential positions of several stator rows and rotor rows, respectively, have an influence on the flow behaviour in terms of loss generation, energy transport and secondary flow. The objective of the presented study is to investigate the effects of stator airfoil clocking on the performance of an 1-1/2 stage axial cold air turbine. The investigated axial turbine consists of two identical stators. The low aspect ratio of the blades and their prismatic design leads to a three-dimensional outlet flow with a high degree of secondary flow phenomena. Nevertheless, the small axial gaps between the blade rows are responsible for strong potential flow interaction with the radial wake regions in the measurement planes. Consequently, parts of the wakes of the first stator are clearly detected in the rotor outlet flow. To give an overview of the time-averaged flow field, measurements with pneumatic probes are conducted behind each blade row at ten different clocking-positions of the second stator. Further, an optimised clocking position was found due to a minimum in pressure loss behind the 2nd stator. The unsteady measurements are carried out with hot-wire probes for three selected stator-stator positions. Animations of selected flow properties show the influence of different circumferential positions of the second stator on the unsteady flow behaviour and secondary flow field. In addition and compared with experimental results three-dimensional unsteady viscous flow computations are performed.

Endwall losses contribute significantly to the overall losses in modern turbomachinery, especially when aerodynamic airfoil load and pressure ratios are increased. Hence, reducing the extend and intensity of the secondary flow structures helps to enhance overall efficiency. This work will focus on secondary flow reduction in typical aero engine low pressure turbines. From the large range of viable approaches, a promising combination of axis symmetric endwall contouring and 3D airfoil thickening was chosen. Aerodynamic design, experimental verification and further analysis based on numerical simulation are described in a two part paper. In the second part the implications of the 3D modifications on the flow structure are analyzed by employing a 3D Navier-Stokes simulation based on the experimental data reported in part one. For obtaining reliable flow simulations at typical LP turbine conditions, it is important to apply a 3D Navier-Stokes solver with proven turbulence and transition modeling to the three-stage LP turbine of the Institute of Aeronautical Propulsion at Stuttgart University. Numerical and experimental results exhibit regions, where the modified design leads to a change in flow pattern in accordance with the design intent, as well as regions with an actual increase in loss production. The flow changes in both regions are evaluated and discussed. It is found that a certain local loss increase phenomenon can also be found in other LP turbine rigs. The reasons for this behavior are analyzed by a comparison with data from other turbine rigs and by an additional variation of the 3D design of the first stage of the investigated turbine.

Experimental investigations of flow mixing due to film cooling of turbine blades have been performed. In a 1½-stage axial air turbine cooling gas (cool nitrogen down to −130 °C) was blown directly onto the leading edge of the first stator by special gas injector devices. In order to provide a database for the verification of numerical codes and to give an impression of the mixing process the gas has been injected at different radial positions. Furthermore the cooling massflow and cooling temperature were varied. The measuring data were obtained using pneumatic 5-hole probes with temperature sensors. The presented experimental data were simultaneous acquired in the planes behind both stators and the rotor. The results are compared and, discussed with reference measurements without cooling gas injection. It is shown that the effect of cooling gas injection is apparent in the wake of the first stator where it causes a small decrease in the pressure distribution as a result of increased flow mixing. Behind the first stator differences in the circumferentially averaged pitchwise flow angle due to the injected gas were not measured. Furthermore, temperature measurements clearly show the effect of the cooling gas injection in all planes. Even behind the second stator the different magnitudes of the temperature distribution are caused by the various injection of cooling gas.