Effects of Physical Modeling on Combustion Instability Predictions in a Single-Element Lean Direct Injection Gas Turbine Combustor

Abstract

Abstract : Simulations of a model configuration of a lean direct injection (LDI) gas turbine combustor were performed to assess the effects of fuel spray modeling in predicting combustion instability. Concurrently, experimental data were obtained at the same conditions in a test combustor with the same configuration for comparison. To improve the simulation predictions, fuel spray modeling was tuned and assessed in two major steps. First, focus is put on the Weber number-adaptive hybrid model that describes the secondary atomization. Tuning the model constants over a finite range does not have a significant effect on the predictions. Second, specified fuel spray distributions are tested to replace the linear stability primary breakup model (LISA). Both uniform and log-normal distributions are used with different mean drop diameters. A better match between pressure amplitude could be achieved by specifying a spray drop distribution, but the predicted dominant modes (3L/4L and 1T) seemed to be very sensitive to the distributions applied.