Gas turbine is a turbomachine in which energy is transferred to, or from continuously flowing fluid by the dynamic action of one or more moving blade rows. The design of level 2 is ‘mean line design’ which contains a large proportion of empirical fluid dynamic models and loss correlations. The general purpose of this procedure is to predict the component discharge conditions from the known inlet conditions and component geometry. In order to design, it is necessary to have separate submodels with which to calculate values of the discharge flow area, the relative flow angle and the entropy gain in the duct. The decision of relative angle requires the knowledge of the slip factor.
By calculating the location and size of the relative eddy formed in the rotating impellers with the logarithmic spiral vanes, a new simple but accurate slip factor is analytically derived. The proposed slip factor depends on only one parameter that is a function of the number of vanes and the vane exit angle. Predicted slip factor for various cases are compared with those estimated by a number of previous slip factors as well as a recent theoretical calculation by Visser et al.. It is found that the present slip factor yields almost similar results to Wiesner`s which has been empirically formulated based on the theoretical calculation of Busemann.
The fluid discharges from the impeller at high velocity, and consequently it is essential to convert the kinetic energy efficiently into static pressure. A wide variety of diffuser designs are applied for this purpose. Also the diffuser is responsible for flow stability. The diffuser system of a centrifugal compressor is usually constructed of either a vaneless diffuser or a combination of a vaneless diffuser followed by a vaned diffuser.
The design procedure of the diffuser is considered, and the loss model for the diffuser is introduced. The test data of the centrifugal compressor for the APU engine are compared with the prediction. On the design rpm the prediction is agreeable to the test data. But off the design rpm the discrepancy is large. Further study is needed to establish the good prediction model.
