Sharma, Sidharath (2019) Numerical characterisation of flow-induced noise in a small high-speed centrifugal compressor with casing treatment. Doctoral thesis, University of Huddersfield.
Abstract

Centrifugal turbomachines of smaller sizes operating at higher speeds have become pervasive due to the increased specific power and reliability achieved by improvements in manufacturing, materials and computational methods. The presence of these small turbomachines, specifically compressors, in helicopters, unmanned aerial vehicles (UAVs), auxiliary power units (APUs), turbochargers and micro gas turbines necessitates superior aerodynamic performance over a broad operational range which is widely achieved by ported shroud casing designs. In addition to aerodynamic performance, acoustic emissions have become a critical aspect of design for these small centrifugal compressors due to high operational speeds. Therefore, in this thesis, a high-speed turbocharger compressor with the ported shroud (PS) casing treatment is used as a subject to understand the flow-induced noise in high-speed centrifugal machines using high-fidelity numerical (CFD) methods. Furthermore, the impact of PS design and operating speed on the acoustic emission of the compressor is also established by comparing the operation of PS open and PS blocked compressor configurations at 99 krpm and 130 krpm speedlines.

The numerical model to predict the acoustic characteristics of the compressor is developed and validated by comparing the acoustic and performance results with the experimental values. The impact of various critical parameters on the performance and acoustic predictions is quantified by exploring a range of statistical and scale resolving methods of turbulence formulations along with their sensitivity to spatial and temporal resolution. The results demonstrate the need for higher spatial resolution for scale resolving models to yield credible acoustic predictions.

The results from the selected numerical configuration are analysed to establish the relationship between the flow field and the acoustic characteristics of the compressor. The acoustic spectra for the design point are seen to be dominated by a characteristic ‘buzz-saw’ or Rotating Order (RO) tonal noise. These ‘buzz-saw’ or RO tones are confirmed to be caused by the sonic conditions on the leading edges of the impeller blades. For the near surge operation, the low-frequency broadband features associated with near surge operation are alleviated by the PS casing treatment and are not observed in the corresponding spectra. Furthermore, the characteristic ‘whoosh’ noise is also not observed in the spectra of either design or near surge points.

The flow is further investigated by the modal decomposition of the dynamic pressure field using Proper Orthogonal Decomposition (POD) to compute the high-energy coherent structures and their corresponding spectral characteristics. For the design operation, the results showed the accumulation of higher energy content in the first two modes that are related to the RO and blade pass tones. The diffuser and the PS cavity are found to house the energetic sources of the oscillations. For the near surge point, the energy is seen to be distributed much more evenly among the modes. The diffuser and volute are observed to house the more energetic sources of broadband content without any significant contribution from the PS cavity.

The comparison of the open and blocked configurations operating at design conditions shows higher tonal content in the inlet duct spectrum of the open configuration. For the near surge operation, the broadband elevations in the lower and medium frequency regions are observed for the blocked configuration. Furthermore, characteristic ‘whoosh’ noise is also identified in the outlet duct spectrum of the blocked configuration. The increase in operational speed causes a general increase in the overall acoustic emission of the compressor for both configurations.

Information
Library
Documents
[img]
Preview
FINAL THESIS - SHARMA.pdf - Accepted Version
Available under License Creative Commons Attribution Non-commercial No Derivatives.

Download (207MB) | Preview
Statistics

Downloads

Downloads per month over past year

Add to AnyAdd to TwitterAdd to FacebookAdd to LinkedinAdd to PinterestAdd to Email