The overall fraction of the power produced by renewable sources in the energy market has significantly increased in recent years. The power output of most of these clean sources is intrinsically variable. At present day and most likely in the upcoming future, due to the lack of inexpensive and reliable large energy storage systems, conventional power plants burning fossil fuels will still be part of the energy horizon. In particular, power generators able to promptly support the grid stability, such as gas turbines, will retain a strategic role. This new energy scenario is pushing gas turbine producers to improve the flexibility of their turbomachines, increasing the need for reliable numerical tools adopted to design and validate the new products also in operating conditions far from the nominal one.
Especially when dealing with axial compressors, i.e. machines experiencing intense adverse pressure gradients, complex flow structures and severe secondary flows, CFD modelling of offdesign operation can be a real challenge.
In this work, a state-of-the art CFD framework for RANS analysis of axial compressors is presented. The various aspects involved in the whole setup are discussed, including boundary conditions, meshing strategies, mixing planes modelling, tip clearance treatment, shroud leakages and turbulence modelling. Some experiences about the choice of these aspects are provided, derived from a long-date practice on this kind of turbomachines. Numerical results are reported for different full-scale compressors of the Ansaldo Energia fleet, covering a wide range of operating conditions. Furthermore, details about the capability of the setup to predict compressor performance and surge-margin have been added to the work.
In particular, the setup surge-margin prediction has been evaluated in an operating condition in which the turbomachine experiences experimental stall. Finally, thanks to several on-field data available at different corrected speeds for operating conditions ranging from minimum to full load, a comprehensive validation of the presented numerical framework is also included in the paper.