This work proposes an automated strategy for the preliminary design of turbomachinery, based on the application of a throughflow code and of a highly flexible and efficient optimization strategy. The code solves for the circumferentially-averaged flow equations, including the effects of aerodynamic and friction forces and of blade thickness; the outcome of the code is the flow distribution on the meridional surface. The fluid-dynamic solver is coupled with the optimization tool in order to determine the "optimal" mean flow surface, as a result of a multiobjective optimization procedure, in which nonconcurrent goals are simultaneously considered. A global optimization strategy is applied, based on the combination of a Genetic Algorithm with a metamodel to tackle the computational cost of the process. The optimization method is applied to a low speed axial compressor, for which the optimization goals are the minimization of aerodynamic loss and discharge kinetic energy at the exit of the stage, as well as the uniformity of work exchange along the blade span. The method proves to match all the objectives, providing a clear improvement with respect to classical and well-established design methods. The optimization provided by the automated design is finally assessed by high-fidelity calculations performed with a fully three-dimensional CFD code on both the baseline and optimized machine configurations. Improvements are confirmed for all the goals specified in the optimization strategy, resulting in a more efficient machine.
Optimization of Turbomachinery Flow Surfaces Applying a CFD-based Throughflow Method
PASQUALE, David;REBAY, Stefano
2013-01-01
Abstract
This work proposes an automated strategy for the preliminary design of turbomachinery, based on the application of a throughflow code and of a highly flexible and efficient optimization strategy. The code solves for the circumferentially-averaged flow equations, including the effects of aerodynamic and friction forces and of blade thickness; the outcome of the code is the flow distribution on the meridional surface. The fluid-dynamic solver is coupled with the optimization tool in order to determine the "optimal" mean flow surface, as a result of a multiobjective optimization procedure, in which nonconcurrent goals are simultaneously considered. A global optimization strategy is applied, based on the combination of a Genetic Algorithm with a metamodel to tackle the computational cost of the process. The optimization method is applied to a low speed axial compressor, for which the optimization goals are the minimization of aerodynamic loss and discharge kinetic energy at the exit of the stage, as well as the uniformity of work exchange along the blade span. The method proves to match all the objectives, providing a clear improvement with respect to classical and well-established design methods. The optimization provided by the automated design is finally assessed by high-fidelity calculations performed with a fully three-dimensional CFD code on both the baseline and optimized machine configurations. Improvements are confirmed for all the goals specified in the optimization strategy, resulting in a more efficient machine.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.