Computational Analysis of the Dynamics Flow in the Duct System of an Continuous Positive Airways Pressure Device

Continuous positive air pressure (CPAP) device is used for Obstructive Sleep Apnoea (OSA) treatments. The airflow distribution system is the main system to provide pressurized airflow from the outside to patients during its treatment. The system consists of a duct flow path, a high-speed centrifugal compressor, and a humidification unit. During the operation, the airflow is drawn from the outside to the patient through a complicated flow path. 

In this paper, a 3-D duct model is developed to present the flow path inside the CPAP device. The numerical investigations are performed for inflow and outflow ducts in the CPAP device to identify potential noise source locations. First, the CFD simulation technique is used to generate noise sources. Second, the simulated results are inputted into acoustic models to calculate the noise level. The CAA with the hybrid approach is conducted in the ANSYS software environment. The CFD turbulent flows are modeled with the renormalizations group theory (RNG), k-epsilon model. The flowrates of 50 L/min to 150 L/min were used to apply at the inlet boundary condition. The aerodynamic characteristics of the internal duct flow are predicted and used to estimate the noise sources using broadband noise source models. 

The predicted results have shown the internal flow characteristics as circulation and separation behaviors. Those behaviors were found at several locations such as  90-degree corners, restricted cross-sectional areas, etc. High levels of turbulent energy are also found in those locations indicated those locations are likely to have a high level of noise as it shows in the predicted noise model. The experimental measurements were conducted to validate the predicted results. It was found the predicted models can be used to evaluate the flow behaviors inside the duct system. 

In conclusion, the CAA with a hybrid approach is used to investigate the potential noise source locations at the inlet and outlet ducts of CPAP devices in this research. The broadband noise source models were used as numerical tools to determine the loudness at the source location. The predicted results have explained and identified some potential noise source locations. These findings will be used as guidelines to optimize the internal flow and therefore improve the noise generated from ducts. The broadband noise source models have indicated as potential tools that can be used in the early product development process for reducing noise level and better sound quality as well as a cost-effective tool for noise prediction.