Simulation and prediction of gas turbine performance is a very important issue in design process or in actual behavior analysis. In these models physical behavior of components such as compressors, combustion chambers and turbines are simulated related to each other. The compressor is the most important part of simulation. This paper presents a model for simulating a compressor using stage stacking procedure with the aid of a genetic algorithm. The most important feature of the proposed method is that qualitative and quantitative rules based on turbo-machinery knowledge of compressors are used as constraints to the genetic algorithm to find the corrected situations of design. This knowledge is evaluated with both industrial and aero gas turbine engines (501F & CF6 (LM2500)). The model is based on an analytical solution and provides an insight into the effects of choices made during the compressor design process on performance and off-design stage matching. The results of the model highlight the capability of the method which accurately reproduces the available data. In addition to obtaining design conditions, this model can find and calculate stages that are highly loaded and this information is vital to control the compressor.

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