Turbulent statistics and energy budgets were calculated for a swirling turbulent flow using Generalized Feed Forward Neural Network (GFFNN) in a dump combustor model. Knowledge of turbulent statistics and energy budgets of fluid flow inside a combustor model is very useful and essential for better and/or optimum designs of gas turbine combustors. Several experimental techniques utilizing two dimensional (2D) or three dimensional (3D) Laser Doppler Velocimetry (LDV) measurements provide only limited discrete information at given points; especially, for the cases of complex flows such as dump combustor swirling flows. For these flows, numerical interpolating schemes are unsuitable. Recently, neural networks proved to be viable means of expanding a finite set of experimental measurements in order to enhance the understanding of complex phenomenon. This investigation showed that artificial neural networks are suitable for the prediction of turbulent swirling flow characteristics in a model dump combustor. These techniques are proposed for better designs and/or optimum performance of dump combustors.

This content is only available via PDF.
You do not currently have access to this content.