Optimization of an axial fan for air cooled condensers

We report on the low noise optimization of an axial fan specifically designed for the cooling of CSP power plants. The duty point presents an uncommon combination of a load coefficient of 0.11, a flow coefficient of 0.23 and a static efficiency ηstat > 0.6. Calculated fan Reynolds number is equal to Re = 2.85 x 107. Here we present a process used to optimize and numerically verify the fan performance. The optimization of the blade was carried out with a Python code through a brute-force-search algorithm. Using this approach the chord and pitch distributions of the original blade are varied under geometrical constraints, generating a population of over 24000 different possible individuals. Each individual was then tested using an axisymmetric Python code. The software is based on a blade element axisymmetric principle whereby the rotor blade is divided into a number of streamlines. For each of these streamlines, relationships for velocity and pressure are derived from conservation laws for mass, tangential momentum and energy of incompressible flows. The final geometry was eventually chosen among the individuals with the maximum efficiency. The final design performance was then validated through with a CFD simulation. The simulation was carried out using a RANS approach, with the cubic k -  low Reynolds turbulence closure of Lien et al. The numerical simulation was able to verify the air performance of the fan and was used to derive blade-to-blade distributions of design parameters such as flow deviation, velocity components, specific work and diffusion factor of the optimized blade. All the computations were performed in OpenFoam, an open source C++- based CFD library. This work was carried out under MinWaterCSP project, funded by EU H2020 programme.