Outdoor unit split fan noise prediction using a hybrid lighthill analogy method

Requirements of outdoor units and split HVAC systems have been increasing with system performance over the years. Since fans are the primary source of noise in these systems, selecting a fan that can meet the thermal requirements and at the same time produce lower level of noise is the most effective way of designing a quieter system. Fan suppliers normally provide fan sound characteristics at a certain distance from the fan at a rated RPM. This information is not sufficient to paint a holistic picture of the aero-acoustic phenomena that can produce unwanted tones during operation. Thus, a Model-Based Design (MBD) approach is applied in the design process during which different concepts are tried in modeling before being tested. This paper discusses a new approach in implementing Lighthill analogy method to capture the tonal and broadband noise characteristics of an outdoor split unit. This method allows for more accurate results compared to FW-H and faster results compared to a direct aero-acoustic CFD simulation. This paper shows a novel signal processing method for fan noise. Tonal and broadband results are separated and processed separately. This allows to capture accurately the tones while smoothing out unrealistic fluctuations in Sound Pressure Level (SPL) at high frequencies occurring because of the limited sampling time.