A validation study of the flamelet approach ’ s ability to predict flame structure when fluid mechanics are fully resolved

Modern combustion devices are designed to balance the need for efficient performance with the need for robust performance. Presently available design techniques are capable of meaningfully informing how this balance should be struck. To truly optimize the efficiency and robustness of a combustor, however, the predictive capabilities of numerical simulations will have to be more heavily relied upon. This is because the variables that describe the multi-dimensional combustor design space are related in highly non-linear ways. While empirical numerical models that depict transitions between burned and unburned states are readily available, these models do not provide the level of detail needed for optimizing a combustor’s efficiency. For example, the evolution of important quantities such as CO, NOX, and soot often depend on the details of the time and spatial scales that characterize the structure of the reaction zone. Flamelet approaches are a particular class of combustion model that do capture the time and spatial scales associated with a given asymptotic limit of combustion physics (Peters 2000; Pierce & Moin 2004; Pitsch 2006). This ability to capture the correct limiting physics has historically been viewed as the advantage of the flamelet approach. In modern combustors where mixed-regime behavior is expected, however, the applicability limits of the flamelet approach are not yet well bounded. Single regime non-premixed flamelet implementations, for instance, would not be expected to accurately describe partially premixed or unsteady combustion processes. Furthermore, because the sensitivity of CO, NOX, and soot to predicted reaction zone structure is not fully understood, the accuracy with which singleor multi-regime flamelet approaches predict pollutants in realistic combustors is not known. Therefore it is not currently possible to perform the kind of cost and benefit analysis that is important in selecting an appropriate combustion model for a given application. The objective of this study is to analyze how accurately current flamelet approaches describe complex reaction zone structure in a fully resolved, multi-regime, multi-dimensional laminar flame. Two questions will be used to target this objective. First, how well do single regime (non-premixed) flamelet approaches simulate the detailed flame structure and pollutant signature? Second, do recently developed implementations of multi-regime flamelet methods capture flame structures in any more detail? These questions will be answered by using the flamelet approach to model an unsteady laminar triple flame that has been simulated using a finite rate chemical mechanism for n-heptane.