Eventos Anais de eventos
ENCIT 2020
18th Brazilian Congress of Thermal Sciences and Engineering
MULTIDIMENSIONAL ANALYSIS OF OPTIMIZED FLAMELET GENERATED MANIFOLD PROGRESS VARIABLES
Submission Author:
Renan Balbinotti Kops , RS , Brazil
Co-Authors:
Renan Balbinotti Kops, Lisandro Maders, Fernando Pereira
Presenter: Renan Balbinotti Kops
doi://10.26678/ABCM.ENCIT2020.CIT20-0516
Abstract
The numerical study of diffusive flames using techniques that rely on laminar flamelet models require the definition of a control variable capable of describing the progress of the combustion from unburned to burned states. Such variable is called a Progress Variable, and its definition significantly affects the result obtained. Usually, the Progress Variable is defined by the user, which may require a trial-and-error approach. In this study, a set of Progress Variables for pure methane is obtained through an optimization algorithm that tries to create a perfectly monotonic manifold, while reducing gradients of stored parameters in the manifold. This set of Progress Variables is then used to build the manifolds used to perform multi-dimensional simulations using the Flamelet Generated Manifold technique. Species and source term gradients are analyzed. The goal is to find which controlling parameters are important to obtain a well-behaved manifold, and evaluate if the different objective functions proposed in the literature are suitable for multi-dimensional simulations. A correlation between the optimized parameters and the solutions accuracy was observed.
Keywords
Flamelet-Generated-Manifold, Optimization, progress variable, Combustion, Genetic Algorithms
DOWNLOAD PDF VIEW PRESENTATION
