Experimental and scale resolving numerical investigations of the physical performance of circumferential grooves in subsonic axial compressor stages

点击翻译按钮获取中文摘要

Axial compressors, used in gas turbine installations and process engineering are highly important economically, with constantly increasing demands for efficiency, pressure ratio and flexibility of use. Operation range and efficiency heavily depend on the radial gaps between rotating and stationary parts, which vary with operation conditions. These gaps cannot be scaled in the same way as the blades, so that the current trends in the design of compressors lead to larger relative gap widths. Circumferential grooves in the casing above the rotors can alleviate the negative impact of the radial gaps, but in particular for large gaps only insufficient experimental data are available. To numerically design casing treatments with circumferential grooves requires reliable simulation tools. State of the art, however, are RANS methods which are deficient in this respect. Here, it is mandatory to employ unsteady turbulence-resolving approaches to determine efficiency and stability limit. These approaches, however, still require improvement to reliably capture the highly complex physical mechanisms with the available computational resources. The project will address the open questions related to the functioning of circumferential grooves by detailed investigations. It will particularly address the impact of the grooves on efficiency and flow modification near the stability limit, as well as the interaction between different stages. To reach its goals, the project combines experiments in a low-speed compressor with turbulence-resolving simulations. This approach generates substantial added value due to the complementary of the data being generated. The experiments provide global data over a wide range of conditions and, with more involved techniques, reliable information in selected planes. The costly unsteady simulations deliver detailed insight into the interaction of the various effects all over the flow, e.g. also inside the grooves, and can be validated with the experiments. These simulation methods, however, are presently not sufficiently reliable for the complex flows to be addressed. If the large computational burden is reduced by employing a coarser grid or by stronger modelling assumptions the quality of the results can deteriorate and even go below the quality of standard methods. Hence, in this proposal both applicants join forces to create a practically applicable hybrid LES/RANS procedure for the situations investigated here. This will be done by advancing present approaches as well as by intense comparative studies and formulation of best practice guidelines.