Stall Inception in the Compressor System of a Turbofan Engine

Abstract

Compressor flow instabilities have been the subject of a great number of investigations during the past decade. While most of this research work was done on isolated test-rig compressors, this paper presents stall inception measurements in the compressor system of a two-spool turbofan engine at various power settings. Several analyzing techniques such as temporal low-pass and band-pass filtering, temporal and spatial Fourier transforms including power-spectral-density calculations of the spatial coefficients, and a wavelet analyzing technique are applied. For the low-pressure compressor three different types of stall inception processes were observed depending on the rotor speed. At low speed, stall originates from spike-type precursors, while long wavy pressure fluctuations corresponding to modal waves were observed prior to stall at midspeed for undistorted inlet flow. At high speed, the rotor shaft unbalancing dominates the stall inception process as an external forcing function. In the case of distorted inlet flow spike-type stall inception behavior dominates throughout the speed range. While filtering and the Fourier spectra give a good insight into the physical background of the stall inception process (but with a very short warning time), the wavelet transform indicates the approach of the stalling process a few hundred rotor revolutions in advance independently of the type of precursor. Setting up a reliable stall avoidance control based on this analysis scheme seems to be promising. [S0889-504X(00)00401-3]

Visualization and Time-Resolved Particle Image Velocimetry Measurements of the Flow in the Tip Region of a Subsonic Compressor Rotor

“…Although operation under stalled condition is possible [3,4], it should be avoided because of the associated adverse effects, such as diminished overall performance and efficiency [5][6][7], vibratory stresses [8,9], and noise [10]. Approximate prediction of surge line and stall dynamics can be based on theoretical analysis [11,12], correlations [3], CFD simulations [13,14], and experimental data [15][16][17]. Yet despite considerable effort, the specific flow phenomena which trigger and sustain stall remained debatable.…”

Combination resonance and primary resonance characteristics of a dual-rotor system under the condition of the synchronous impact of the inter-shaft bearing

“…Ishii et al 7 paid attention to the numerical model for multiblade-row compressors to obtain the influence of multiple factors on the surge and rotating stall inception and post-stall dynamics in multistage axial compressors. Hoss et al 8 observed technological tools, including periodic low-pass and band-pass filtering, spatial and temporal Fourier transforms, and a wavelet transform, to detect three different types of L.P. compressor stall inception processes. Wilson et al 9 detailed the progression of flow instability at varying speeds, out of a small stopped fluid zone, including simply just a few blades to the entire machine’s dramatic surge action.…”

Real-time prediction for the surge of turboshaft engine using multi-branch feature fusion neural network

“…At present, the main time-frequency analysis methods used in stall precursor detection are wavelet analysis and Hilbert Huang transform. Hoss et al 17 collected the wall static pressure data of a two-stage low-pressure compressor LARZAC-04 of a double shaft turbofan engine and successfully detected the stall signal by using the wavelet analysis method; Xiang et al 18 used Hilbert Huang transform (HHT) and wavelet analysis methods to analyze the frequency characteristics of the compressor pressure signal. The result shows that HHT can obtain stall precursor signal more effectively; Yue et al 19 proposed a combined empirical mode decomposition and local mean decomposition (EMD-LMD) method to solve the negative frequency problem in the conventional HHT method and applied it to the detection of compressor stall precursor.…”