A computational investigation of vane clocking effects on compressor forced response and performance

Jonathan Robert Salontay, Purdue University

Abstract

The objective of this research was to computationally investigate the effects of vane clocking, the circumferential indexing of adjacent vane rows with identical vane counts, on the first torsional mode forced response of an embedded compressor rotor row and on the performance of an embedded compressor stage. A reduced mesh with four Stator 1 vanes, three Rotor 2 blades, and four Stator 2 vanes was created to capture the effects of vane clocking on the Purdue 3-Stage Research Compressor, which has a vane and blade count of 44-33-44 for the modeled embedded stage. AU3D was used as the unsteady, aeromechanical, computational fluid dynamics code for the clocking forced response investigation, while JACC was used as the unsteady, aerodynamic, computational fluid dynamics code for the clocking performance investigation. The results from the investigations were compared to experimental tip timing and performance data previously acquired in the Purdue 3-Stage Research Compressor. The Rotor 2 first torsional mode forced response predicted by AU3D consistently over-predicted the measured response by 10-27%; however, the trends associated with vane clocking matched the trends observed experimentally. The minimum rotor response occurred when the maximum vortical gust from the upstream stator and the minimum potential gust from the downstream stator acted on the rotor blades at the same instant in time. The maximum response occurred when the maximum upstream vortical gust and maximum potential gust from the downstream stator acted on the rotor blades at the same instant in time. Therefore, vane clocking changed the phase of the upstream and downstream forcing functions, which resulted in a change in response of the embedded rotor blades. The predicted change in stage efficiency at design loading was 0.31 points between the minimum and maximum efficiency clocking configurations, which was comparable to the measured 0.27-point efficiency change. The leading edge impingement of the of the upstream stator wake on the downstream stator lead to the optimum stage efficiency while clocking configurations that located the S1 wake in the mid-passage of S2 resulted in the minimum stage efficiency. This agrees with the experimental observations. An investigation of the boundary layer on the suction side of the downstream stator showed that the upstream stator wake plays a role in shielding the downstream stator vanes’ boundary layer from the effects of the upstream rotor wakes. Both experimental and computational results showed a large change in spanwise efficiency at 70% span. Evidence of secondary flow effects suggest that 70% span is sensitive to vane clocking because there is little radial flow due to two passage vortices meeting around 70% span.

Degree

M.S.M.E.

Advisors

Key, Purdue University.

Subject Area

Mechanical engineering

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS