Faithin Omaha eBook Collection

Download or read book Kinematic Study of the Evolution and Properties of Flame Surfaces in Turbulent Nonpremixed Combustion with Local Extinction and Reignition written by Weirong Wang and published by . This book was released on 2013 with total page 206 pages. Available in PDF, EPUB and Kindle. Book excerpt: Turbulent nonpremixed combustion with local extinction and reignition is an important fundamental phenomenon that has many practical implications. The objective of the research reported in this thesis is to identify and understand the mechanisms that govern extinction and reignition by studying the kinematics of the stoichiometric mixture fraction isosurface (i.e., the flame surface). The configuration, area and the behavior of the flame surface can quantitatively describe the interaction of the flame with turbulent flow and thus the flame surface kinematics has a strong influence on the properties of the reactive scalars. One particular property of importance is the mean flame surface density [signma]. This study contributes to the determination of [sigma] by proposing three approaches for its measurement through the use of direct numerical simulations (DNS) data for the case of incompressible and isotropic turbulence. These direct and indirect approaches proposed for the computation and modeling of the time evolution of [sigma] are 1) direct numerical measurement, 2) theoretical prediction from the concept of level crossings with the application of Rice's theorem\cite{S.O.Rice1}, which leads to 3) two separate modeling approaches using statistical definitions and by balance equation of isoscalar surface area density. These approaches allow us to follow the growth of the surface due to local surface stretching by turbulence and its ultimate decrease due to molecular destruction. The statistical model fairly accurately predicts the evolution of [sigma]. We are also able to determine values of the principal terms in the evolution equation for [sigma], including the surface stretching term and the molecular destruction term. We find that the stretching and destruction term are approximately statistically independent of the isoscalar value of the surface. The difficulties in modeling of local flame extinction and reignition are demonstrated by evaluating the application of the steady-state flamelet (SFL) model. The transient effects (i.e., the lag of flame quenching behind fluctuations in the stoichiometric scalar dissipation rate X[subscript st] ) and the time lag of reignition relative to the relaxation of X[subscript st] are not considered by the SFL model. These are demonstrated by showing the mixing and reactive quantities along the stoichiometric mixture fraction contour lines. With the knowledge of the flame surface evolution and its properties, the mechanisms of local extinction and reignition can be directly investigated by the dynamics of flame hole (i.e., an extinguished flame surface) expansion and collapse. The flame surface is first resolved as a mesh of triangular surfaces within the three-dimensional DNS volume that represents the stoichiometric surface. A criterion is developed on this flame surface mesh to identify instantaneously local quenched regions and distinguish these from burning surface regions. Next, a new methodology is developed to numerically identify contiguous extinguished stoichiometric surface elements such that the size of a single flame hole can be identified. The new methodology is tested and found to be accurate. This numerical algorithm can provide rich geometric information on local flame hole structure and along with knowledge of the local reacting properties for each individual flame hole. The statistics of the geometric properties of the flame holes are then obtained. During the reignition process, the area size distribution of flame holes was found to be roughly invariant with time which results in an approximate time invariant average hole area. This appears to be due to the loss of small flame holes with time while larger holes become smaller. To first order, this process appears to approximate a self-preserving size distribution.