Study Of Multi-dimensional Flamelet Generated Manifolds Of Aminar Flame

Modeling of combustion systems often requires adequate dimension reduction procedures applied, since detailed chemistry calculation of chemical reactions will lead to unbearable computational burden for reactions with a number of species. Methods as intrinsic-low dimensional manifolds(ILDM) and computational singular perturbation(CSP) have been introduced into combustion simulation, and further approaches like flame prolongation of ILDM(FPI) and flamelet generated manifolds(FGM) have been developed in recent years, which include diffusion effects for regions where flame structure is influenced by transport phenomenon. In the FGM approach, single-dimensional flamelet equations are calculated priori to create a series of low-dimensional manifolds, which will capture the main characteristic of flame front structure in actual flame. Via decoupling both flow and chemical reactions, numerical computational efforts can be drastically reduced. FGM approach has been applied to both laminar and turbulent flame simulation successfully, with premixed or non-premixed generated manifolds. When it comes to flame controlled by multi-regime characteristic, existing flamelet models cannot reproduce local flame structure accurately. For instance, non-premixed FGM cannot capture propagation characteristics of multi-regime flame, while the non-premixed characteristic between iso-surfaces of mixture fraction cannot be captured by premixed FGM.To overcome these deficiencies, multi-dimensional flamelet generated manifolds have been proposed by Nguyen et al. Multi-dimensional flamelet equation set in composition space is deducted from the basic equations of chemical reaction flow, by choosing specific flamelet coordinate. Flamelet solutions based on solving multi-dimensional flamelet equations can be tabulated to form the (MFM). For example, multi-dimensional flamelet generated manifolds for multi-regime flame structure simulation, as in partially premixed flames, can be deducted through designating mixture fraction Z and flamelet progress variable Yc as components of composition space coordinate system. Partial premixed flame structure can be captured by solving both non-premixed and premixed contribution terms in multi-dimension flamelet equation set, which will improve the modeling capability of FGM approach by adding additional dimension.Single-dimensional flamelet models as steady laminar flamelet model(SLFM) and flamelet generated manifolds based on solving premixed flamelets(premixed FGM) have been discussed in present paper, using flamelet generated transformation, to examine the capability and limitations during multi-regime flame structure modeling. Further analysis of the influence on simulated results using different flamelet parameterization forms in SLFM has been performed, meanwhile study of redefining flamelet progress variable with principle component analysis (PCA) has been carried out. Based on research of single-dimensional flamelet models, two new multi-dimensional flamelet modeling approaches on the foundation of multi-dimensional flamelet generated manifolds have been proposed in present paper:an interactive multi-dimensional flamelet model based on interactive correction of scalar dissipation rate in flamelet equations, and a Z-c multi-dimensional flamelet generated manifolds approach with normalized flamelet progress variable c replacing progress variable Yc as flamelet coordinate component. Numerical simulations for a laminar lifted flame and a series of counter-flow laminar partially premixed flames have been performed, to verify modeling accuracy for these new approaches, with detailed chemistry solutions as references. Numerical simulation results indicate that multi-dimensional flamelet model can capture both premixed and non-premixed characteristic of multi-regime flame structure, which leads to better approximation accuracy by comparing with single-dimensional premixed FGM approach. Application of these new multi-dimensional flamelet generated manifolds approaches can lead to significant improved solutions close to detailed chemistry references, with greatly reduced computation efforts.