Thickened flame models are prolific in the literature and offer an effective method of resolving flame dynamics on coarse LES meshes. The current state of the art relies heavily on the use of efficiency functions to compensate for impaired wrinkling of the thickened flame. However in practice these functions can involve parameters that are difficult to determine, perform poorly outside of certain ranges or require \emph{a posteriori} analysis to evaluate performance. An alternative based on a generalised thickening is evaluated across a range of canonical configurations. The approach is demonstrated to perform well across a large range of thickening factors in capturing phenomena such as localised quenching and pinch off as well as generation of flame surface. Including good performance even in the case of large flame dynamics under acoustic forcing where the model has a clear advantage over DNS in achieving grid independence. Finally the approach is unified into an Large Eddy Simulation/Adaptive Mesh Refinement framework and applied to a turbulent Bunsen flame. The results show that even if the internal flame structure is poorly resolved on the original mesh, the global system behaviour is well predicted and compares favourably with other approaches.

Thickened flame models are prolific in the literature and offer an effective method of resolving flame dynamics on coarse LES meshes. The current state of the art relies heavily on the use of efficiency functions to compensate for impaired wrinkling of the thickened flame. However in practice these functions can involve parameters that are difficult to determine, perform poorly outside of certain ranges or require a posteriori analysis to evaluate performance. An alternative based on a generalised thickening is evaluated across a range of canonical configurations. The approach is demonstrated to perform well across a large range of thickening factors in capturing phenomena such as localised quenching and pinch off as well as generation of flame surface. Including good performance even in the case of large flame dynamics under acoustic forcing where the model has a clear advantage over DNS in achieving grid independence. Finally the approach is unified into an Large Eddy Simulation/Adaptive Mesh Refinement framework and applied to a turbulent Bunsen flame. The results show that even if the internal flame structure is poorly resolved on the original mesh, the global system behaviour is well predicted and compares favourably with other approaches.