University of Leicester
Browse

Scalar Mixing and Coherent Structures in Simulations of the Plane Turbulent Mixing Layer

Download (94.4 MB)
thesis
posted on 2018-05-16, 10:05 authored by Stephan Nicholas Hug
For more than half a century turbulent mixing layers have been the subject of intense experimental and numerical investigation. With the discovery of primary, spanwise aligned and secondary, streamwise oriented vortices the interest in low and high Reynolds number mixing layers has been invigorated. The immense increase of computational capabilities in recent years has lead to an ever growing number of numerical simulations of mixing layers. However, numerical simulations have had great difficulties in reproductions the structure dynamics and entrainment mechanisms observed in the experiments. In this study Large Eddy Simulations of the low and high Reynolds number spatially developing, three-dimensional mixing layer are performed. At the heart of the presented studies lies the focus on the inlet conditions of the simulations. The effects of spatial and temporal correlation of the inlet conditions are studied for the low and high Reynolds number planemixing layer. It is shown that physically correlated inlet fluctuations lead to the development of the spatially stationary, streamwise oriented vortices observed in experiments. The effects of the presence of the streamwise vortices on the momentumand passive scalar fields are investigated in detail. In the latter parts of this work, the effects of varying the inlet fluctuation levels are reported. By altering the inlet fluctuation magnitudes the number and strength of the spatially stationary streamwise vortices can be controlled. The implications of this on the dynamics of the primary, spanwise aligned vortices are discussed. A change in the number and strength of the spatially stationary streamwise vortices is shown to be critical for the shape of the obtained probability density functions. If spatially stationary streamwise vortices are present, the obtained probability density functions are of the non-marching type. A lack of spatially stationary streamwise vortices produces marching probability density functions.

History

Supervisor(s)

McMullan, Andrew; Gao, Shian

Date of award

2018-03-13

Author affiliation

Department of Engineering

Awarding institution

University of Leicester

Qualification level

  • Doctoral

Qualification name

  • PhD

Language

en

Usage metrics

    University of Leicester Theses

    Categories

    No categories selected

    Keywords

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC