Session EL: Biofluids: Physiological Respiratory

Regional airflow and particle distribution in the lung with a 3D-1D coupled subject-specific boundary condition

Correct prediction of regional distribution of inhaled aerosol particles is vital to improve pulmonary medicine. Physiologically consistent regional ventilations of airflow and aerosol particles are simulated with a 3D-1D coupled subject-specific boundary condition (BC). In 3D CT-resolved 7-generation airways, large eddy simulations are performed to capture detailed airflow characteristics and Lagrangian particle simulations are carried to track the particle transport and deposition. Results are compared with two traditional outlet BCs: uniform velocity and uniform pressure. Proposed BC is eligible for physiologically consistent airflow distribution in the lung, while the others are not. The regional ventilation and deposition of particles reflect the regional ventilation of airflow. In this study, two traditional BCs yield up to 98{\%} (334{\%}) over-prediction in lobar particle ventilation (deposition) fraction. Upper to lower particle ventilation ratios of both left and right lungs read $\sim $0.4 with the proposed BC, while those for the other two BCs vary with the error up to 73{\%}.   

Newtonian to non-Newtonian flow transition in lung surfactants

The lining of normal lungs is covered by surfactants, because otherwise the surface tension of the aqueous layer would be too large to allow breathing. A lack of functioning surfactants can lead to respiratory distress syndrome, a potentially fatal condition in both premature infants and adults, and a major cause of death in the US and world-wide. We use a home-built Brewster angle microscope on an optically accessible deep channel viscometer to simultaneously observe the mesoscale structures of DPPC, the primary constituent of lung surfactant, on water surface and measure the interfacial velocity field. The measured interfacial velocity is compared to Navier-Stokes computations with the Boussinesq-Scriven surface model. Results show that DPPC monolayer behaves i) purely elastically at low surface pressures on water, ii) viscoelastically at modest surface pressures, exhibiting non-zero surface shear viscosity that is independent of the shear rate and flow inertia, and iii) at surface pressures approaching film collapse, DPPC loses its fluid characteristics, and a Newtonian surface model no longer captures its hydrodynamics.   

Stability analysis of the pulmonary liquid bilayer.

The lung consists of liquid-lined compliant airways that convey air to and from the alveoli where gas exchange takes place. Because the airways are coated with a bilayer consisting of a mucus layer on top of a periciliary fluid layer, a surface tension instability can generate flows within the bilayer and induce the formation of liquid plugs that block the passage of air. This is a problem for example with premature neonates whose lungs do not produce sufficient quantities of surfactant and suffer from respiratory distress syndrome. To study this instability a system of coupled nonlinear evolution equations are derived using lubrication theory for the thicknesses of the two liquid layers which are assumed to be Newtonian. A normal mode analysis is used to investigate the initial growth of the disturbances, and reveals how the grow rate is affected by the ratio of viscosities $\lambda $, film thicknesses $\eta $ and surface tensions $\Delta $ of the two layers which can change by disease. Numerical solutions of the evolution equations show that there is a critical bilayer thickness $\varepsilon _c $ above which closure occurs, and that a more viscous and thicker layer compared to the periciliary layer closes more slowly. However, $\varepsilon _c $is weakly dependent on $\lambda $, $\eta $ and $\Delta $. We also examine the potential impact of wall shear stress and normal stress on cell damage. This work is funded by NIH HL85156.   

Airflow Simulation and Particle Deposition in a 3D Rat Lung Model

Knowledge of the fate of aerosols in the lung is needed to understand the efficiency of inhaled drug therapy. Invasive animal experiments and imaging allows for detailed quantitative comparison with computational modeling. In this study we built a three-dimensional (3D) airway tree model using rat magnetic resonance images. A custom 3D finite element solver was used to obtain animal specific velocities and pressures. Inlet boundary conditions were chosen to match a previous rat ventilation experiment and resistance outlet boundary conditions were selected to match a desired airflow split based on uniform ventilation. The Maxey-Riley particle equations were solved using Lagrangian particle tracking methods with realistic aerosol particle dimensions and density. The particle dynamics were validated using analytical solutions in idealized geometries. The impact of the choice of outlet boundary conditions for airflow simulations is quantified and aerosol particle deposition and distribution within the lung lobes are explored.   

Numerical investigation of aerosolized drug delivery in the human lungs under mechanical ventilator conditions

Particle deposition for aerosolized drug delivery in the human airways is heavily dependent upon flow conditions. Numerical modeling techniques have proven valuable for determining particle deposition characteristics under steady flow conditions. For the case of patients under mechanical ventilation, however, flow conditions change drastically and there is an increased importance to understand particle deposition characteristics. This study focuses on mechanically ventilated conditions in the upper trachea-bronchial (TB) region of the human airways. Solution of the continuous phase flow is done under ventilator waveform conditions with a suitable turbulence model in conjunction with a realistic model of upper TB airways. A discrete phase Euler-Lagrange approach is applied to solve for particle deposition characteristics with a focus on the effect of the ventilator inlet waveform. The purpose of this study is to accurately model flow conditions in the upper TB airways under mechanically ventilated conditions with a focus on real-time patient specific targeted aerosolized drug delivery.   

Unsteady Simulation of a Human Respiratory System with Micron-Particles

Unsteady numerical simulations of air flow, mixed with micron particles, through a human lung conducting zone during inhalation/exhalation process have been performed. The process included importing images from a high resolution MRI into a CFD software, generation of the CFD model and then CFD simulation over a 4 seconds cycle. The inlet diameter was 16 mm and the flow rate was 7 liter/ min. The implicit-unsteady Reynolds Average Navier-Stokes equations with the Wilcox K-$\omega $ turbulence model were used for the simulation. The micron particles were solid round lead with 1000 Kg/m$^{3}$ density. Results indicate high correlation between regions of the secondary flows and particle deposits. This was mostly evident in the main bronchus. While most particles exit the lung during the exhalation process, however, areas of re-circulating flow and near the walls continue to have some particle deposits.   

Particle Deposition During Airway Closure

Inhaled aerosol particles deposit in the lung and may be from environmental, toxic, or medical therapy sources. While much research focuses on inspiratory deposition, primarily at airway bifurcations due to inertial impaction, there are other mechanisms that allow the particles to reach the airway surface, such as gravitational settling and diffusion depending on particle size. We introduce a new mechanism not previously studied, i.e. aerosol deposition from airway closure. The airways are lined with a liquid layer. Due to the surface tension driven instability, a liquid plug can form from this layer which blocks the airway. This process of airway closure tends to occur toward the end of expiration. In this study, the efficiency of the impaction of the particles during airway closure will be investigated. The particles will be released from the upstream of the airway and convected by the air flow and deposited onto the closing liquid layer. We solve the governing equations using a finite volume approach in conjunction with a sharp interface method for the interfaces. Once the velocity field of the gas flow is obtained, the path of the particles will be calculated and the efficiency of the deposition can be estimated.   

Pulsatile flow past a single oscillating cylinder

The potential for oscillating fibers to modify flow within a new artificial lung design is first examined in the present fundamental fluid mechanics study of flow past a single oscillating cylinder. This new design is intended to provide better gas exchange through vorticity enhancement by oscillating microfibers (cylinders) in a pulsatile flow environment. The Keulegan-Carpenter number (Kc=Uo/D$\omega $c) was used to describe the frequency of the oscillating cylinder ($\omega $c) while the pulsatile free stream velocity was fixed by imposing $\omega $/Kc=1 for all cases investigated. The parameters investigated in this study were amplitude of oscillation (0.5D$<$A$<$D), Kc corresponding to 1$<\omega $c$<$3 and Reynolds number (5$<$Re$<$20), all equivalent to operating conditions of the TAL. Vorticity was enhanced up to 246{\%} from the steady state condition for high amplitudes and low Kc for all Re. An opposite trend was observed for the drag coefficient. A ``lock-in'' phenomenon (cylinder oscillating frequency matching the vortex shedding frequency) was found when KC=1 for all cases. A jump in the drag coefficient was observed and attributed to this operating regime. These results suggest that this new design of the TAL could potentially enhance gas exchange through oscillation of the microfibers with a decrease in the drag coefficient if operating far from the lock-in regime. This work was supported by NIH grants R01HL69420 and R01HL089043.   

Coughing and sneezing

The emergence and explosive spread of virulent viral (e.g., H1N1, SARS) and bacterial (e.g., Tuberculosis) infections is a problem of global interest with enormous human and economic consequences. The nature of {\it contact} between infected and non-infected persons greatly influences the outcomes of the disease epidemic; nevertheless, the definition and mechanisms leading to contact remain nebulous. We here examine the manner in which fluid dynamics modeling can assist in our understanding of contact and transmission of respiratory diseases. Particular attention is given to modeling the effluent of discrete exhalation events (e.g., coughing, sneezing) as multiphase thermals, and to predicting the range of pathogen-bearing droplets.   

Numerical Simulations of the Propagation of a Liquid Plug through a 2D Airway Bifurcation

Numerous medical therapies require the instillation of liquids plugs and their delivery throughout the pulmonary airways. This process and the effect on the resulting liquid distribution is controlled by a number of parameters, including airway orientation with respect to gravity, initial plug volume, liquid physical properties, and the imposed airflow rate which drives the plug from behind. The airflow rate defines an operative Capillary number, Ca, and the influence of gravity appears as an effective Bond number, Bo, whose magnitude varies with orientation. In this study, we develop a numerical method for solving the propagation of a liquid plug into a two-dimensional airway bifurcation consisting of a parent channel branching into two daughter channels. We measure the splitting ratio, RS, which is defined as the ratio of the liquid plug volumes between the daughter branches. RS increases with Ca and asymptotes to 1 as Ca goes to infinity, which corresponds to an equal split, while increasing Bo requires a higher value of Ca for an equal split. We also examine the normal and shear stresses on the bifurcation walls and observe that the stresses on the upper walls increase as Bo increases while the stresses on the lower walls decrease as Bo increases.