Experimental investigation of vibrational and acoustic phenomena for detecting the stall and surge of a multistage compressor

Nowadays, the operative range limit of compressors is still a key aspect of the research into turbomachinery. In particular, the study of the mass flow rate lower limit represents a significant factor in order to predict and avoid the inception of critical working conditions and instabilities such as stall and surge. The importance of predicting and preventing these dangerous phenomena is vital since they lead to a loss of performance and severe damage to the compression system and the compressor components. The identification of the typical precursors of these two types of compressor unstable behaviors can imply many advantages, in both stationary and aeronautic applications, such as i) avoiding the loss of production (in industry) and efficiency of systems and ii) reducing the cost of maintenance and repairing. Many approaches can be adopted to achieve this target, but one of the most fascinating is the vibroacoustic analysis of the compressor response during operation. At the Engineering Department of the University of Ferrara, a test bench, dedicated to the study of the performance of an aeronautic turboshaft engine multistage compressor, has been equipped with a high frequency data acquisition system. A set of triaxle accelerometers and microphones, suitable for capturing broad-band vibration and acoustic phenomena, were installed in strategic positions along the compressor and the test rig. Tests were carried out at different rotational speeds, and with two different piping system layouts, by varying the discharge volume and the position of the electric control valve. Moreover, two different methodologies were adopted to lead the compressor towards instability. This experimental campaign allowed the inception of compressor stall and surge phenomena and the acquisition of a great amount of vibro-acoustic data which were firstly processed through an innovative data analysis technique, and then correlated to the thermodynamic data recorded. Subsequently, the precursor signals of stall and surge were detected and identified demonstrating the reliability of the methodology used for the study of compressor instability. The results of this paper can provide a significant contribution to the knowledge of the inception mechanisms of these instabilities. In particular, the experimental data can offer a valid support to the improvement of surge and stall avoidance (or control) techniques since it presents an alternative way of analyzing and detecting unstable compressor behavior characteristics by means of non-intrusive measurements.