This is the manuscript of my doctoral thesis. Even in the surprising case that you're not interested in the fascinating topic of turbocharger acoustic research, it may be possible that the overall style of the LaTeX document could prove useful for the typesetting of your own projects. So I switched this repo that I used for tracking changes to public; see the final PDF document in release v3.0 here.
Experiments on turbochager compressor acoustics
Jorge García Tíscar
Abstract
As turbocharging requirements raise to face an increasingly stricter environmental regulation of internal combustion engines, concerns regarding their acoustic emission become more pressing. Since downsized engines require higher boost pressures and lower mass flow rates, the turbocharger compressor is forced to work at more unstable regimes, where flow patterns become more complex and noise levels rise.
This thesis aims to thus to investigate these issues, proposing methodologies to characterize the acoustic emission of turbocharger compressors and implementing them in different experiments with a special focus on the link between acoustic emission and flow behaviour at unstable conditions.
A literature review is carried out in order to assess the state of the art principally regarding experimental techniques related to this issue but also including the latest developments in terms of understanding the flow characteristics through numerical simulations. Different methodologies are consequently proposed and implemented into a custom test rig inside an anechoic chamber as to experimentally measure and analyze the acoustic output of the compressor.
From this measurement campaign a characterization of the noise spectral content across the compressor map is obtained and described, identifying different acoustical phenomena such as blade passing tonal noise, low frequency content associated to deep surge, higher frequency broadband ascribed to tip clearance interaction, and broadband noise in the plane wave range known as whoosh in the literature, of special concern for automotive manufacturers. This particular phenomenon is detected even at more stable conditions at higher flow rate, and rising in level as flow rate is diminished to the point of being masked by lower frequency content.
After a validation of the selected procedure in realistic engine conditions, experimental data is compared against a numerical model of the compressor developed in a parallel work to assess its validity and propose different post-processing techniques to extract additional insights about the behaviour of the flow at different conditions, hinting at the main generation mechanism for whoosh being located at the compressor diffuser.
Since numerical simulations predict a reduced amount of reversed unstable flow at conditions where whoosh noise is still measured, an experimental campaign is performed where detailed measures of local temperature near the compressor wheel are used to determine the evolution of reversed flow, with supplemental readings through pressure probes in the inducer and diffuser being used to link this evolution with the fluctuation of spectral content. Temperature results are also correlated with measurements of the velocity field through particle imaging, demonstrating a clear link between the reversed flow field and temperature readings.
Different experimental campaigns are then described where modifications of the inlet geometry immediately upstream the compressor are performed in order to assess how the air flow presentation can influence acoustic performance. Geometries featuring reservoirs, nozzles and guide vanes are shown to potentially reduce noise levels. A 90º elbow parametric study is performed, demonstrating how the inner radius of the elbow influences circumferential temperature differences and noise levels, leading to the hypothesis of geometry influence on whoosh noise being related to air presentation changes that promote lower or higher growth of reversed backflow, which in turns carries upstream the spectral content being generated in the compressor diffuser.
Lastly, additional experimental and numerical studies are proposed to further explore these issues, in order to provide a better understanding of how inlet designs may delay and mitigate the appearance and transmission of these adverse acoustical phenomena.