Conference Year



turbocharger compressor performances, centrifugal compressors


Most centrifugal compressors operate in conditions with negligible heat transfer (adiabatic compression). Their plant tests conditions are similar or close to adiabatic conditions. Test regulations establish measures to diminish influence of a heat transfer “compressor body – atmospheric air” to an exit temperature. Therefore a temperature rise in a compressor is used to calculate a work input coefficient and efficiency. Unlike it high pressure centrifugal compressors of gas turbines and superchargers operate in conditions with very active heat transfer with ambience, lubricant and hot turbine parts. Non adiabatic compression process evidently influence on temperatures inside a flow path, gas density, velocity triangles. But this aspect of a problem is out of the discussed problem. This problem is how to define gas dynamic performance of a compressor. The author has at his disposal hot test data of a small turbocharger compressor with the impeller diameter 48mm. Data were provided by the colleague Prof. J. Seume (Institute of Turbo machines, Hanover University, Germany): mass flow rate, total pressures and four total temperatures: directly at compressor borders and on a distance of them. The difference of values demonstrates strong heat transfer in inlet and exit pipes. The detailed study in the Institute of Turbo machines, Hanover University has shown that compression process is sufficiently non adiabatic. Unrealistic influence of rotation speed on efficiency points on it indirectly. The author applied the Universal modeling method of Prof. Y. Galerkin to reduce test data firstly. The 5-th generation computer programs were developed recently and successfully applied to model gas dynamic performances of subsonic compressors. The German colleagues made a supposition that measured temperature difference is very close to an adiabatic process at design RPM 202000. This performance was modeled with the standard complex of empirical coefficients. The roughness of cast surfaces was taken into account. Test data of TU SPb show that work input coefficient is linear function of a flow coefficient at an impeller exit independent of Mach number in subsonic area. The linear supposition was applied for transonic and supersonic flows as well. This procedure was applied with 6-th generation of computer programs. The 6-th generation program takes into account shocks and calculates losses in 3-D impellers in quasi-3-D mode. In result the modeling of performances in range of RPM 104000-202000 is more satisfactory. The set of empirical coefficients for calculation of head losses and work input coefficient can be applied for test data reducing of other small turbocharger compressor performances.

1123_presentation.pdf (497 kB)