Circumferential groove (CG) casing treatment is known to be a good method to improve the stall margin of axial compressor. Recent studies on single groove indicate the stall margin improvement (SMI) is significantly affected by groove’s axial location, which means the effects of the groove on the passage flow should be further studied as the groove’s axial location is changed. In this paper the features of stall inception with and without the groove treatment are experimentally researched in a low-speed axial compressor. The contour of static pressure on casing is obtained by fast-response pressure transducers concentrative mounted as a line in chordwise direction, and the data acquired by transducers in circumferential direction is processed with the method of wavelet analysis to demonstrate the development of stall inception. It is found the stall inception is the typically spike-type for smooth casing (without treatment), which is manifested as short-scaled disturbance of only 2–4 blade passages when it appears. Two single grooves at different axial locations, which can be ineffective and effective in SMI, are further tested. The results indicate the stall inception is still spike-type for both of them. However, for the ineffective groove, the stall inception is found to be relatively longer-scaled disturbance. For the effective groove, it is still the typical short-scaled disturbance similar to the smooth casing. In the further study on double grooves treatment in which the ineffective groove is included, the stall inception is also found to be relatively longer-scaled type. According to these experimental results and previous CFD simulation, the groove with almost no SMI is believed to take different effects on the compressor which modifies the stall inception into the relatively longer-scaled type, and the corresponding tip–region flow structure deserves more future study in this regard.
Conclusions
The impacts of the circumferential groove treatment on stall inception are experimentally studied in this paper. Two single grooves, which are ineffective (CG5) and quite effective (CG7) in improving the stall margin, are selected to discuss whether the impacts on stall inception are different for the single grooves with quite different SMI. The unsteady measurements on the casing by fast-response pressure transducers of group A (concentrated mounted as a line) and group B (in circumferential direction) are performed for the study of stall inception.
The contours of instantaneous static pressure on casing when stall inception is first observed and the corresponding circumfer-ential detection by wavelet analysis are discussed respectively. The typical spike-type stall inception is found for SC (without treat-ment) as well as the groove CG7 with quite effective SMI, which is revealed as the short-scaled disturbance in the acquired data. However, as for the groove CG5 locating between the two peaks of SMI but providing almost no SMI, the stall inception is found to be relatively longer-scaled than SC and CG7. The further tests of the double grooves treatment CG59 (the combination of CG5 and CG9) indicate the stall inception is still longer-scaled, which is similar to the case of single groove CG5. According to these experimen-tal results, the groove CG5 is believed to take different effects on the compressor: it provides almost no SMI while modifies the stall inception into the relatively longer-scaled type in both the single and double grooves treatment. These results provide the validation about the change of stall inception in CFD simulation, which indi-cates the different effects of CG5 on the tip–region flow structure.
In this regard, more experimental work about how the tip–region flow is affected by groove CG5 and its relation with the relatively longer-scaled stall inception will be considered in future work.
The results have been published on Aerospace Science and Technology 59 (2016) 145–154.
