The study of flame dynamics and structures in an industrial-scale gas turbine combustor using digital data processing and computer vision techniques

In this paper, a combined effort has been made to study the flame dynamics and structures in a gas turbine combustor using a range of imaging and digital data processing techniques. The acoustic characteristics of the combustor have been investigated extensively. It is found that there is no straightforward way to alter the peak frequency of one of the peculiar combustion modes of the rig. High speed imaging is applied to investigate the flame dynamics and quantitative analysis of the image database has been demonstrated. The results show that the frequency spectrum of the mean pixel image intensity of seeded flame is in good agreement with the acoustic spectrum. To recover the loss in depth information present in conventional imaging technique, both the optical and digital stereo imaging techniques have been applied. The important flame position relative to the combustion chamber could be resolved.