Reynolds-Averaged Navier-Stokes (RANS) calculations and Large-Eddy Simulations (LES) of the flow in two asymmetric three-dimensional diffusers were performed. The numerical setup was chosen to be in compliance with previous experiments. The aim of the present study is to find the least expensive method to compute reliably and accurately the impact of geometric sensitivity on the flow. RANS calculations fail to predict both the extent and location of the three-dimensional separation bubble. In contrast, LES is able to determine the amount of reverse flow and the pressure coefficient within the accuracy of experimental data.

The current study investigates the flow conditions of a twin scroll asymmetric turbine. This is motivated by the operating conditions of the turbine at a heavy-duty reciprocating internal combustion engine with exhaust gas recirculation. The flow conditions of the turbine at the engine can be described best with the turbine scroll interaction map. Standard hot gas measurements of a turbocharger turbine are presented and discussed. Due to the strong interaction of the turbine scrolls, further hot gas measurements are performed at partial admission conditions. The turbine inlet conditions are analysed experimentally, in order to characterize the turbine performance. The turbine scroll pressure ratio is varied, leading to unequal twin turbine admission conditions. The flow behaviour is analysed regarding its ability for further extrapolation. Beyond scroll pressure ratio variations, unequal temperature admission conditions were studied. A way of characterizing the representative turbine inlet temperature, regarding the reduced turbine speed, is presented. The different scroll parameter ratios are evaluated regarding their capability of describing flow similarity under different unequal turbine admission conditions. In this content, turbine scroll Mach number ratio, velocity ratio and mass flow ratio are assessed. Furthermore, a generic representation of the turbine flow conditions at the engine is presented, based on standard turbine performance maps.

Heat transfer measurements on a highly loaded low-pressure turbine airfoil with a separation bubble on the pressure surface are presented. The experiments were conducted in a linear cascade at various free-stream turbulence intensities (Tu1 = 1.6% to 10%) and Reynolds numbers of the inflow. The effect of both quantities on heat transfer, separation and laminar-turbulent transition is quantified. Particle-Image-Velocimetry has been performed to study the characteristics of the separation bubble. The results reveal a considerable influence of the boundary layer separation on the local heat transfer. The size of the separation region is strongly influenced by free-stream turbulence level and Reynolds number.

An experimental study on film cooling performance of laterally inclined diffuser shaped cooling holes is presented. The measurements have been conducted on a flat plate with coolant ejected from a plenum. The film cooling effectiveness downstream of a row of four laidback fanshaped holes with sharp edged diffusers has been determined by means of infrared (IR) thermography. A variety of geometric parameters has been tested, including the inclination angle, the compound angle, the area ratio, and the pitch to diameter ratio. All tests have been performed over a wide range of engine typical blowing ratios (M=0.5–3.0). The hot gas Reynolds number and the coolant to hot gas density ratio have been kept constant close to engine realistic conditions. The results, presented in terms of contour plots of related adiabatic film cooling effectiveness as well as laterally averaged related values, clearly show the influences of the cooling hole geometry. Increasing the area ratio and the compound angle, in general, leads to higher values of the effectiveness, whereas steeper injection causes a reduction of the effectiveness.

This manuscript presents the application of an improved CFD methodology to simulate the scavenge system film flow phenomena in a real aero engine bearing chamber environment i.e. influence of seals and rotational shaft. Near the scavenge off-take, the usual thin film approach is not valid due to the occurrence of relative thick films (up to 5mm, comp. [1]) where film internal dynamics become very important. Therefore, other multiphase modelling techniques need to be explored. Young and Chew suggest in [2] that the Volume Of Fluid (VOF) method is the most suitable technique for air/oil system applications. Hashmi et al. reported in [3] that this free surface method for shear driven thick wall films in the bearing chamber environment needs additional provisions for turbulence modelling. Accordingly, a simple correction is made to the k–ε RNG turbulence model to improve the simulation results. The improved CFD methodology is applied to an engine representative geometry and proves to be robust and computationally efficient. The test conditions in the simulation was chosen in a way to avoid any droplet stripping from the film surface. It is shown that the applied methodology together with the correction in the turbulence modelling prove to play a vital role for a good comparison with experimental data. After validation the simulation results are used to describe the flow phenomena which occur in the bearing chamber for the investigated condition. The introduced CFD modelling technique shows large potential for the development and trouble shooting purposes in the industrial environment.