The wide range of applications for radial inflow turbines is the reason for an increasing interest in this type of machines in the recent years. An appropriate method to achieve good efficiencies under different operating conditions is the use of variable nozzle vanes. Because of the limited knowledge of the flow phenomena involved, simple models are used for aero-thermodynamical machine design. A research program was initiated to experimentally study the flow field in a transonic radial nozzle cascade. An air test rig with variable nozzle vanes, derived from an industrial radial inflow turbine, was used for the tests. In this rig no rotor was installed in order to study the flow in the nozzles without any influence of this element. Static pressure distributions have been measured on the vane suction and pressure surfaces and on the shroud endwall within the blading and downstream of the nozzles. Results are presented for different pressure ratios and nozzle setting angles. The experimental results are compared with numerical calculations for inviscid flow based on the two-dimensional Euler equations.

To evaluate the potential of the compressor of Sulzer’s Typ 3 gas turbine, a series of engine tests was analyzed with two computer codes. The comparison between measured and calculated performance map are given in the paper. The design goal was to find modifications, which can be applied easily to already operating engines. The simplest option-increase of shaft speed with the existing blades-would have caused high loss due to increased tip Mach number. The calculation revealed, that a newly designed first rotor blade is an appropriate modification to increase massflow and efficiency. No further change is required, because the calculations indicate, that all subsequent stages operate at near optimum incidence. The calculations were confirmed experimentally. The paper presents the new rotor blade and its influence on the compressor calculated and measured performance.