| description abstract | Lowspeed model testing has advantages such as great accuracy and low cost and risk, so it is widely used in the design procedure of the high pressure compressor (HPC) exit stage. The lowspeed model testing project is conducted in Nanjing University of Aeronautics and Astronautics (NUAA) to represent aerodynamic load and flow field structure of the seventh stage of a highperformance tenstage highpressure compressor. This paper outlines the design work of the low speed fourstage axial compressor, the third stage of which is the testing stage. The first two stages and the last stage provide the compressor with entrance and exit conditions, respectively. The hightolow speed transformation process involves both geometric and aerodynamic considerations. Accurate similarities demand the same Mach number and Reynolds number, which will not be maintained due to motor power/size and its lowspeed feature. Compromises of constraints are obvious. Modeling principles are presented in hightolow speed transformation. Design work was carried out based on these principles. Four main procedures were conducted successively in the general design, including establishment of lowspeed modeling target, global parameter design of modeling stage, throughflow aerodynamic design, and blading design. In global parameter design procedure, rotational speed, shroud diameter, hubtip ratio, midspan chord, and axial spacing between stages were determined by geometrical modeling principles. During the throughflow design process, radial distributions of aerodynamic parameters such as Dfactor, pressurerise coefficient, loss coefficients, stage reaction, and other parameters were obtained by determined aerodynamic modeling principles. Finally, rotor and stator blade profiles of the low speed research compressor (LSRC) at seven span locations were adjusted to make sure that blade surface pressure coefficients agree well with that of the HPC. Threedimensional flow calculations were performed on the lowspeed fourstage axial compressor, and the resultant flow field structures agree well with that of the HPC. It is worth noting that a large separation zone appears in both suction surfaces of LSRC and HPC. How to diminish it through 3D blading design in the LSRC test rig is our further work. | |
