Detection of turbulent/non-turbulent interface for an axisymmetric turbulent jet: evaluation of known criteria and proposal of a new criterion

• Published in: Experiments in fluids Vol. 47; no. 6; pp. 995 - 1007
• Main Authors: Anand, Ravi Kumar, Boersma, B. J., Agrawal, Amit
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer-Verlag 01.12.2009; Springer
• Subjects: Axisymmetric; Computational fluid dynamics; Criteria; Engineering; Engineering Fluid Dynamics; Engineering Thermodynamics; Exact sciences and technology; Fluid dynamics; Fluid flow; Fluid- and Aerodynamics; Fundamental areas of phenomenology (including applications); Heat and Mass Transfer; Instrumentation for fluid dynamics; Jets; Physics; Research Article; Shear stress; Turbulence; Turbulent flow; Turbulent flows, convection, and heat transfer; Vorticity; Velocity Criterion; Root Mean Square Difference; Vorticity; Particle Image Velocimetry; Direct Numerical Simulation; Data analysis; Turbulent flow; Circular jet; Digital simulation; Velocity distribution; Experimental study; Interfaces; Algorithms; Criterion; Probability density function; Particle image velocimetry; Detection; Velocity measurement
• ISSN: 0723-4864; 1432-1114
• DOI: 10.1007/s00348-009-0695-5

In this paper, we evaluate several criteria for the detection of turbulent/non-turbulent interface using direct numerical simulation and particle image velocimetry data of an axisymmetric turbulent jet. The possibility of identifying the interface from information available in wholefield velocity data alone is also explored. The present results using a Concentration thresholding technique compare well against available results obtained using a similar detection criterion. It is noted that Concentration and Vorticity criteria are difficult to apply with standard PIV data and therefore a new criterion based on azimuthal vorticity and streamwise velocity—quantities available from such data, is proposed. The proposed criterion scores over previously employed criteria in terms of its simplicity of evaluation, and can possibly be applied to other flows not tested here. The instantaneous location of the interface as detected from the different criteria differs substantially. However, the conditionally averaged streamwise velocity, azimuthal vorticity, and Reynolds shear stress across the interface obtained from the new criterion, as well as from the previous criteria, agree reasonably well against available results. The present work further suggests that different criteria, even with slightly sub-optimal threshold value, can provide quantitatively similar ensemble-averaged results.