Impact of modeling a Heat Exchanger as a porous medium on the noise source of an industrial engine cooling fan

This thesis investigates the influence of modeling a heat exchanger as a porous medium on the aerodynamic and aeroacoustic behavior of an industrial engine cooling fan with varying radiator-fan axial spacing. A combined experimental and numerical approach is adopted to evaluate how inflow turbulence, generated by the heat exchanger, affects fan broadband noise. Experimental measurements were conducted using hot-wire anemometry, while microphones were used in a semi-anechoic chamber. Experiments were compared against numerical simulations performed with a lattice-Boltzmann method framework in which the radiator is modeled as a porous medium. Velocity statistics and spectral analyses indicate general agreement in trends between experiments and simulations, although the numerical model underestimates velocity fluctuations at high frequency and slighly overpredicts acoustics. Nevertheless, experimental and numerical acoustic signals show a fairly acceptable match up to 5 kHz. An analytical assessment using the 1D Amiet model for leading edge noise shows that the turbulence interaction noise contribution at the leading edge is 15–20 dB lower than the total measured sound pressure level. This confirms that, in the current configuration, turbulence interaction noise is negligible compared to other acoustic sources. Therefore, modeling the heat exchanger as a porous medium provides a reliable approximation for predicting the fan noise in configurations with limited radiator-fan axial spacing.