On the development and application of the fast-response aerodynamic probe system in turbomachines - Part 2: Flow, surge, and stall in a centrifugal compressor

Citation
C. Roduner et al., On the development and application of the fast-response aerodynamic probe system in turbomachines - Part 2: Flow, surge, and stall in a centrifugal compressor, J TURBOMACH, 122(3), 2000, pp. 517-526
Citations number
29
Categorie Soggetti
Mechanical Engineering
Journal title
JOURNAL OF TURBOMACHINERY-TRANSACTIONS OF THE ASME
ISSN journal
0889504X → ACNP
Volume
122
Issue
3
Year of publication
2000
Pages
517 - 526
Database
ISI
SICI code
0889-504X(200007)122:3<517:OTDAAO>2.0.ZU;2-C
Abstract
The present paper, Part 2 of a trilogy, is primarily focussed on demonstrat ing the capabilities of a fast-response aerodynamic probe system configurat ion based on the simplest type of fast-response probe. A single cylindrical probe equipped with a single pressure sensor is used to measure absolute p ressure and both velocity components in an essentially two-dimensional flow field. The probe is used in the pseudo-three-sensor mode (see Part 1). It is demonstrated that such a one-sensor probe is able to measure high-freque ncy rotor-governed systematic fluctuations (like blade-to-blade phenomena) alone or in combination with flow-governed flow-frequency fluctuation sas r otating stall (RS) and mild surge (MS). However, three-sensor probes would be needed to measure stochastic (turbulence-related) or other aperiodic vel ocity transients. The data shown refer to the impeller exit and the vaned d iffuser of a single-stage high-subsonic centrifugal compressor. Wall-to-wal l probe traverses were performed at the impeller exit and different positio ns along the vaned diffuser for different running conditions. the centrifug al compressor was operated under stable as well as unstable (pulsating or s talled) running conditions. The turbomachinery-oriented interpretation of t hese unsteady flow data is a second focus of the paper. A refined analysis of the time-resolved data will be performed in Part 3, where different spat ial/temporal averaging methods are compared. Two different averaging method s were used for the data evaluation: impeller-based ensemble-averaging for blade-to-blade systematic fluctuations (with constant period length at a co nstant shaft speed), and flow-based class averaging for the relatively slow MS and RS with slightly variable period length. due to the ability of fast -response probes to simultaneously measure velocity components and total an d static pressure, interesting insights can be obtained into impeller and d iffuser channel flow structures as well as into the time behavior of such l arge-domain phenomena as RS and MS. [S0889-504X(00)01103-X].