Aeroacoustic Simulation of an Axial Fan Including the Full Test Rig by Using the Lattice Boltzmann Method

F1 - Lattice Boltzmann Methods (i)

The measurement of the sound emission of axial fans is standardized by internationally accepted guidelines. These guidelines require an undisturbed cubic volume upstream of the fan with the minimum length of the edge of two rotor diameters. In recent experimental studies it has been found by the authors that these requirements may be insufficient for acoustic fan tests. Even if the fan intake is placed in a vented empty room with a volume more than 1000 times larger a low-motion recirculating flow can develop which eventually causes a disturbed inflow to the fan rotor. Tonal "interaction" sound is a result which one must not attribute to the fan itself. In principle the recirculating flow field in large rooms is difficult to quantify. Unless expensive field measurement methods like particle image velocimetry are used, high-resolution probes (e.g. a hot wire or hot film probe) provide data only at a very limited number of data points in space. On the other hand Navier-Stokes based computational fluid dynamics simulations are limited to small computational domains when one needs to resolve the acoustically relevant spatial and temporal flow. It is the objective of this study to apply the numerical Lattice-Boltzmann method (LBM) to the flow in a computational domain comprising two very disparate sub-domains, a large inflow region and a comparably small fan. Not only the unsteady flow field is of interest, LBM promises the direct computation of the acoustics (here the inflow induced interaction tones) from the unsteady flow field data.
The flow in the entire aeroacoustic test rig for axial fans of the University of Siegen is modeled with the Lattice-Boltzmann Solver PowerFlow 5.0. The large room the fan takes air from is an anechoic chamber. The fan is the generic rotor-only axial fan "USI7" with five cambered and swept blades. Its aerodynamic and acoustic characteristics have been studied experimentally in numerous recent studies. The large scale environment is varied in terms of two different arbitrarily chosen variations to study the impact of the flow conditions far upstream of the fan section on the tonal sound emitted by the fan.
It turns out that a certain characteristic simulation time is needed to account for the impact of the large scale environment and eventually predict the tonal sound accurately. Geometrical variations of the large scale environment far upstream of the fan section can finally have a distinctive impact on the tonal sound emitted by the fan. A Fourier analysis of the inflow and the visualization of the blade pressure fluctuations reveal that the tonal sound emerge in the tip region of the blades and that the spectral distribution of the tonal sound depends on the inflow conditions and eventually on the flow conditions even far away from the fan.