Abstract:
Designing a gas turbine meter (GTM) to maintain certain levels of performance features is usually done by manufacturing and testing prototypes in experimental setups. However an experimentally supported Computational Fluid Dynamics (CFD) analysis may drastically reduce the cost and time in the design process of GTMs. One of the most important design parameters effecting the performance of GTMs is the blade geometry. Many variables may be defined in the geometry of the blades. Blade pitch angle, number of blades and the tip clearance are the three essential design variables studied in this thesis considering the results from CFD tools such as Fluent. Firstly blade pitch angle and number of blades are taken as design variables for different design approaches. Then the chosen blade shape is improved regarding the tip clearance and trailing edge shape; afterwards the final GTM is linearized according to the different Reynolds numbers. OptGTM3D, which is a user-interface Matlab computer program capable of commanding Gambit and Fluent for GTM blade design, is written. A methodology which is also quite demanding in problems involving rotating parts and flow induced torque, is adapted into this program for externally converging the moment coefficient in Fluent to find the angular speed for specified flow rates. Producing geometry and mesh, running it in Fluent is automatically done through OptGTM3D for the specified geometries. The geometry is generated as a volume around one blade in Gambit and run in Fluent by making use of Multiple Rotating Reference Frame to save computational time.
