Numerical Study of High-Frequency Oscillatory Air Flow and Convective Mixing in a CT-Based Human Airway Model

Abstract

High frequency oscillatory ventilation (HFOV) is considered an efficient and safe respiratory technique to ventilate neonates and patients with acute respiratory distress syndrome. HFOV has very different characteristics from normal breathing physiology, with a much smaller tidal volume and a higher breathing frequency. In this work, the high frequency oscillatory flow is studied using a computational fluid dynamics (CFD) analysis in three different geometrical models with increasing complexity: a straight tube, a single-bifurcation tube model, and a computedtomography (CT)-based human airway model of up to seven generations. We aim to understand the counter-flow phenomenon at flow reversal and its role in convective mixing in these models using sinusoidal waveforms of different frequencies and Reynolds numbers. Mixing is quantified by the stretch rate analysis. In the straight-tube model, coaxial counter flow with opposing fluid streams is formed around flow reversal, agreeing with an analytical Womersley solution. However, counter flow yields no net convective mixing at end cycle. In the single-bifurcation model, counter flow at high Re is intervened with secondary vortices in the parent (child) branch at end expiration (inspiration), resulting in an irreversible mixing process. For the CT-based airway model three cases are considered, consisting of the normal breathing case, the highfrequency-normal-Re case, and the HFOV case. The counter-flow structure is more evident in the high-frequency-normal-Re case than the HFOV case. The instantaneous and time-averaged stretch rates at the end of two breathing cycles and in the vicinity of flow reversal are computed. It is found that counter flow contributes about 20% to mixing in HFOV.

Role of carrier gases in enhancement of gas exchange and aerosol–drug delivery under invasive high frequency oscillatory ventilation

“…This allows turbulent structures to establish before the flow reaches that lower region of trachea. The current LES settings and treatment are consistent with other existing studies of pulmonary CFD Banerjee, 2015a, 2015b;Alzahrany et al, 2014a;Choi et al, 2010;Gemci et al, 2008;Jayaraju et al, 2008;Nagels and Carter, 2009). …”

Airway Gas Flow

“…The peak Re of HFOV is about three times higher than that of NORM and HFNR; thus, the flow in HFOV is expected to be most turbulent even at phase change. Choi et al (17) noted that convective mixing is most effective at peak inspiration because of the aforementioned mode 4 "longitudinal augmented dispersion" at high Re. At end inspiration and expiration, Figure 9 shows the velocity fields in the bifurcation plane at the airway bifurcation between generations 2 and 3.…”

“…To map the flow regimes of all airway segments in the lung, Choi et al (17) used a multidetector row computed tomography (MDCT)-based 3D airway model (up to seven generations) coupled to a 1D 25-generation conducting airway tree of the same subject (see the insert in Fig. 1) generated by a volume filling technique (88,89) to examine three breathing conditions.…”

A biomechanical model of pendelluft induced lung injury

“…A range of 0.5oy þ o3 was also reported by Dehbi (2011). For realistic complex geometries like the present CT-scan based lung model geometry, a tetrahedral mesh style has been considered appropriate in many past studies (Alzahrany and Banerjee, 2015;Choi et al, 2010;Farkas and Balashazy, 2007;Gemci et al, 2008;Luo et al, 2014). The mesh sensitivity was examined and it was found that a mesh of size 2.42 Â 10 6 elements resolves 491% of turbulence kinetic energy for all considered HFOV frequency and is considered appropriate as it resolves 480% of the total energy (Pope, 2001).…”