Development of a HS-LIF-system for Lagrangian correlation measurement

• Published in: Experiments in fluids 2009-04, Vol.46 (4), p.607-616
• Main Authors: Winkler, Anton, Wäsle, Johann, Sattelmayer, Thomas
• Format: Article
• Language: English
• Subjects: Applied sciences; Combustion; Combustion. Flame; Computational fluid dynamics; Energy; Energy. Thermal use of fuels; Engineering; Engineering Fluid Dynamics; Engineering Thermodynamics; Exact sciences and technology; Fluid flow; Fluid- and Aerodynamics; Heat and Mass Transfer; Noise; Research Article; Spectra; Theoretical studies; Theoretical studies. Data and constants. Metering; Turbulence; Turbulent flow; Wavenumber
• ISSN: 0723-4864; 1432-1114
• DOI: 10.1007/s00348-008-0585-2

Within the presented work, a key assumption for a combustion noise model is validated. Heat release fluctuations are the main reason for the noise emission of turbulent premixed flames. Within the combustion noise model of Hirsch et al. [31st Symposium (Int.) on Combustion, pp 1435–1441, 2006], the heat release is computed in the wavenumber domain and transferred into the frequency domain, subsequently. The transformation of the spectra requires a power law dependence of the scalar spectra upon the wavenumber proportional to and upon the frequency proportional to f −2 in the inertial subrange. The validation of the latter assumption requires a measurement system, which allows time dependent recording of fluid properties, e.g. the progress variable. These are provided by a HS-LIF-system, which supports a repetition rate of 1 kHz with sufficient energy to detect OH-radicals. From the high speed video data, the motion of the flame front is reconstructed. The presented study shows the set up of the HS-LIF-system as well as the various image post processing steps, including data binarization, flame front tracking and finally, computation of the lagrangian correlation for the progress variable. It can be shown that the spectral distribution of the progress variable in the Lagrangian frame is as assumed by the above mentioned combustion noise model.