Development of an optimization procedure to improve performance of turbomachinery. The methodology couples a genetic algorithm with a CFD code (zFlow, CFX, OpenFOAM) and an automatic mesh generator. The computational cost of the genetic algorithm is reduced exploiting a Kriging model. This approach has been used to optimize i) the vaned diffuser of an automotive turbocharger centrifugal compressor (to be used in a microturbine for combined heat and power application), ii) the stator of an existing one-stage impulse radial turbine for an Organic Rankine Cycle, and iii) a centrifugal pump.
CFDlab@uniBS