Evaluation of Hybrid Aeroacoustics Models for HVAC Duct Noise Predictions

This study evaluates the effectiveness of two hybrid computational aeroacoustic methods—Lighthill wave model and perturbed convective wave model—in simulating HVAC duct noise in the automotive industry. Using component-level acoustic testing of a Ford HVAC duct, simulations were conducted at varying airflow rates to assess the accuracy of both models in predicting duct noise. The Lighthill wave model, suitable for noise analysis in regions outside turbulent flow areas, showed a good correlation with experimental data, especially in the frequency range of 100 Hz–5000 Hz, but sometimes struggled with pseudo-noise effects at low frequencies near turbulent regions. The perturbed convective wave model, which is suitable for noise analysis anywhere in the flow domain, underpredicted sound pressure levels at low frequencies as well. Both models underpredicted high-frequency noise (>5 kHz) due to insufficient mesh and time-step sizes. Despite these limitations, the Lighthill wave model provided better overall agreement with experimental results. The analysis also identified a resonance peak at 160 Hz, captured in both simulations and experimental data, offering valuable insights for further optimization. These findings underscore the potential of hybrid aeroacoustic models to improve the design of quieter HVAC systems, enhancing both vehicle comfort and customer satisfaction.