On the Application of POD to the Study of Flames

This paper deals with the application of Proper Orthogonal Decomposition (POD) to the analysis of flame phenomena. POD is used in combustion both as a powerful means of extracting the main features of flame phenomena from massive experimental data consisting of high-definition, high sampling rate image sequences, and as a means of construction of reduced order models. Both applications are illustrated, with more emphasis on the model reduction procedure with reference to a one-dimensional laminar propagating gas flame; open issues are addressed, and results computed by means of reduced models are compared with full model simulations.