Recent trend toward lightweight, compact compression systems for advanced aircraft gas turbine engines has created a need for very high pressure ratio fan and compressor stages. One way of achieving pressure ratio in excess of 3:1 in an axial blade row is to introduce splitters (partial vanes) between the principal blades, a concept pioneered by Wennerstrom during early 70s for application in a 3:1 pressure ratio single axial stage. This paper presents an advanced methodology for high pressure ratio splittered rotor design. The methodology centers around combining a meridional flow calculation, an arbitrary meanline blade generation procedure, and 3-D inviscid and viscous analyses. Methods for specifying work distribution, solidity, loss, and deviation distributions, as well as the airfoil generation and splitter vane placement are discussed in detail. Importance of 3-D viscous effects along with results from a 3-D viscous calculation for Wennerstrom’s splittered rotor are also presented.

In this second part of a two-part paper on the application of the NASA multistage aerodynamic simulation CFD tool, the APNASA code, work is presented on how the code was used successfully in the design of a brand new, four-stage axial compressor aimed for a modern turbofan engine. In particular, the code was used to guide the blade geometry changes such that the right stage matching throughout the compressor was achieved. The overall performance is shown to be in good agreement with the map generated prior to the detailed design from a one-dimensional model, including the surge line. This success is attributed to the fact that the compressor has been successfully matched at its design intention. More detailed comparisons of the measurements with the prediction demonstrate capability of APNASA in capturing details of flow, such as the rotor exit pressure and temperature profiles and the stator exit flow angle. Largely as a result of the application of APNASA, the compressor described in this paper has been designed “right the first time”, resulting in significant cost savings in a new engine development program for Honeywell Engines and Systems.

The NASA Average-Passage multistage turbomachinery flow analysis code “APNASA” by J.J. Adamczyk (1985) has been validated, calibrated, and demonstrated at Honeywell Engines and Systems for the design of multistage axial compressors. APNASA was first calibrated against test data of two existing compressors and then used as a design tool in the design of a new modern multistage axial compressor. The results of the calibration, design effort, and the data measurements are presented in this two-part paper. In the present paper (Part 1) the results of the calibration for two multistage axial compressors are presented. The first compressor consists of four axial stages that were designed in the mid 1980s. The second compressor consists of three axial stages and was designed in the mid 1990s using viscous, three-dimensional CFD code, with airfoil optimization performed in single blade row fashion. The calibration work was aimed at developing meshing and modeling best practices and validating the code capability to simulate flow behavior in a multistage environment. Predictions are compared with test data for the axial compressor overall performance, individual stage performance, and detailed radial profiles at the stator vanes leading edge planes, throughout the compressor. Results show good agreement between APNASA predictions and measurement data. In particular, the results clearly demonstrate the ability of APNASA to capture the stage matching of multistage machines. As a result of this calibration/validation work, a new multistage axial compressor was subsequently designed, by using APNASA as the primary source of information for airfoil optimization (presented as Part 2 of this paper). Test results for the new compressor reveal that the design achieved its performance and operability goals in its first build. Details of the compressor design philosophy using APNASA and the comparison between APNASA simulation results and test data are presented in Part 2.