Mean Flow Structures Inside the Human Upper Airway

Abstract

The lattice Boltzmann method (LBM) was used to conduct a direct numerical simulation study of the airflow inside an idealised human upper airway. Results from both a modest resolution (18 million control volumes, 320 Gb data set) and an extreme resolution (148 million control volumes, 800 Gb data set) LBM simulation were compared to those from experimental results (Johnstone, A.: Hot wire measurements in an oropharyngeal pathway. M.Sc. Thesis, Queen's University, Kingston, ON, Canada, 2002; Johnstone et al., Expt Fluids 37(5): 673-689, 2004). A coarse resolution simulation (2.4 million control volumes, 105 Gb data set) was used to record the entire time-varying flow field; the nature of the mean structures in the three-dimensional flow field was studied using this data set. For the mean statistics, the LBM calculations yield better results than do the Reynolds averaged NavierStokes methods (Ball et al., Comput Fluids, 2007); the LBM reproduces significant detail of experimentally observed flow features. The flow is three-dimensional, obviously, and the interrogation of the mean flow structure is found to be unsteady so that sagittal plane and time-integrated measurements alone are insufficient to verify the accuracy of computational predictions of this flow.

Large-scale CFD simulations of the transitional and turbulent regime for the large human airways during rapid inhalation

“…Other works, such as Jayaraju et al [6] and Ball et al [7], used the upper human airway with an idealized geometry and presented results concerning mean flow and flow structures. Contrary to Jayaraju et al…”

“…Many works have been performed with simplified models [11], idealized models [6,7], or using several realistic models [2,3], for example Choi et al [12] investigated the variabilities of intra-and inter-subject of airflow in the human lung.…”

On locating the obstruction in the upper airway via numerical simulation

“…It has been used in simulating biomedical flows, such as flows in the respiratory system (Ball et al, 2008; Finck et al, 2007; Hörschler et al, 2010; Eitel et al, 2010; Lintermann et al, 2012) and cardiovascular system (Munn and Dupin, 2008; Boyd and Buick, 2008; Kim et al, 2010). The published LBM studies related to the UA are mostly concerned with the laminar flow in the nasal cavity (Finck et al, 2007; Eitel et al, 2010).…”

“…These studies demonstrated the capability of the LBM for predicting the complex flow in the UA. Recently, the DNS-LBM has been used to simulate the laminar-transitional-turbulent flows in an idealized laboratory model of the airway (Ball et al, 2008). The results of Ball et al showed that the DNS-LBM was superior to RANS as it reproduced the critical flow features observed in the experiment.…”

Numerical investigation of the flow field in the upper human airways

“…Unlike former numerical and experimental investigations, in which a simplified geometry was used, the present method can be efficiently applied to variable, realistic airway geometries. For instance the flow field downstream of the laryngeal region has recently been investigated in [13] by an LBM. Since the numerical method is capable of reproducing small-scale features of lung flow, the results serve to fundamentally understand respiratory mechanisms.…”