Transcript Slide 1

Polymer and Surface
Engineering Laboratory
(PolySEL)
Health Impacts of Micron-sized Particle Deposition in the Human Tracheobronchial Tree
Erick S.
1
Vasquez ,
Nicole
2
Stein ,
Keisha B.
1
Walters
and D. Keith
2
Walters
Dave C. Swalm School of Chemical Engineering1
Mechanical Engineering Department2
Mississippi State University, Mississippi State, MS 39762
http://www2.che.msstate.edu/research/PolySEL/
Methods (continued)
Motivation
Exposure to particulates impacts human health. Dust,
combustion exhaust, manufacturing toxins (e.g. dioxins),
and many other environmental contaminants exist in the
form of micron-sized particles. When inhaled, these
particles can create health issues such as emphysema
and cancer. Conversely, novel biomedical technologies
currently include micron-sized particles that are inhaled
as a drug delivery vehicle. Unintentional and intentional
particle inhalation can result in particles deposited on the
mucus lining; these particles can then be taken up and
enter the blood stream.
Computational modeling to predict the transport and
deposition of inhaled particles represents key enabling
technologies for improved drug delivery methods and for the
mitigation of detrimental health effects due to pollution.
Introduction
Applications and prior studies for particle fate/deposition in
human lungs include:
• drug delivery mechanisms 1
• prediction of dust, diesel or
coal particle deposition 2
• particle deposition in bronchial
airway bifurcations, which may
play a crucial role in lung cancer
induction 3
• relationships between atmospheric
pollutant impacts—both natural and
anthropogenic sources—and their health impacts 4-5
Methods
- Computational fluid dynamics (CFD) simulations using a
physiological realistic bifurcation (PRB) 6-7
- PRB geometry is based on a model by Weibel 8
- Parameters such as branching angle were modified to be
more physiologically accurate
- Different levels of refinement (cell number) and two
different mesh topologies, tetrahedral (TET) and hybrid
(HYB), were analyzed :
Physiological Realistic
Bifurcation (PRB)
Geometry
Tetrahedral (TET)
Mesh:
tetrahedral cells
throughout geometry
Hybrid (HYB) mesh:
hexahedral cells
near wall & center
tetrahedral cells
Results and Discussion (Continued)
Results and Discussion
Computational results are compared with experimental
data obtained by Oldham et al.9 In the experimental study,
rectangular microscopic fields (2.05 mm x 1.4 mm) were
used along with latex particles (~10 microns; density ~ 1
g/cc):
Lagrangian simulations were used to track 100K individual particles in
each mesh refinement and topology.
2nd order Eulerian
simulations using different
nominal particle sizes showed
larger particles result in higher
deposition rates in the PRB
geometry.
Particle diameter
(microns)
Cumulative
deposition (%)
TET Mesh
5
10
20
Example path lines for tracked
particles are shown.
Air flow was solved independently of particle phase (dilute
particle assumption).
Center plane air velocity magnitude
for coarse, tetrahedral mesh.
In the Eulerian simulations, particle phase velocity field differs from air
phase due to particle inertia effects, including wall normal component at
bifurcation locations. The particle concentration field is relatively complex.
Particle deposition simulations were performed using
(i) an Eulerian modeling approach in which the particles are
represented in terms of an average concentration at each
location in the lung airway, and
20 micron
10 micron
5 micron
Lagrangian and Eulerian simulations conducted for 1 micron
(1000 nm) diameter particles showed the mesh topology and
refinement greatly influenced the results obtained for small(er)
particles. This effect was not seen for the 10 micron (or larger)
particles.
For the 1 micron particles, Lagrangian simulations began to
approach the experimental data when using a Hybrid mesh and
high levels of refinement (8,645,000 cells).
(i) a Lagrangian modeling approach in which each particle
and its path is tracked within the lung model.
Center plane particle phase velocity
magnitude shown.
Center plane particle phase velocity
magnitude shown.
Illustration of qualitative similarity in airway wall deposition patterns
between CFD simulations and experimental data for 10 micron particles.
Simulations were completed using these two modeling
approaches for 1, 5, 10 and 20 micron particle diameters.
Experimental and computational results at 10 microns
were compared along with the impact of particle size.
Coarse HYB mesh
Computational Modeling
Computational fluid dynamics (CFD) simulations were
carried out using Fluent (version 12.0.16). Simulation
assumptions include:
- only drag and gravitational forces are considered
- a constant diffusivity value is used
- particle volume fraction is zero at the walls
- zero velocity at the walls (no slip boundary condition)
- second-order discretization for momentum equations;
SIMPLE pressure correction scheme; PRESTO! scheme
for pressure discretization
- Lagrangian: 100K randomly distributed particles injected
at inlet with velocity equal to air velocity
- Eulerian: particle volume fraction value of 1 E-6 % was
utilized as the initial inlet condition for the simulations;
first-order and second-order discretization methods are
studied for particles deposition analysis
31.6
77.3
98.9
Computational
Results (Eulerian)
•
•
•
•
Experimental
Results 5
Cumulative percent deposition was studied as a function of
distance into the airway through two bifurcations.
Significant differences were found between the Lagrangian
and Eulerian methods for all meshes studied.
Differences were observed due to mesh topology for smaller
particles (less than 10 microns).
Comparison with experiments (10 micron particle diameter)
does not clearly indicate that either method is superior.
Eulerian vs. Lagrangian
TET mesh
Number of Cells
% Particle Deposition
2,856,292 4,163,830 8,645,075
0.25
0.19
0.13
Experimental9
0.01
Conclusions and Future Work
• The level of refinement used for simulations with 1 micron
particle sizes was found to have a high impact on particle
deposition data for Lagrangian simulations. The effect of grid
refinement is still under investigation for the Eulerian methods.
• For 10 micron particles, the Eulerian method gave a better
value for percent total deposition, but worse agreement with
the experimental data through the first bifurcation.
• While the Lagrangian method appear to capture better the
qualitative deposition pattern, as compared with experimental
data, when a parabolic inlet velocity profile is used, the
cumulative deposition results are less accurate.
• A high number of cells is required to obtain close agreement
with experimental data for Lagrangian simulations.
• Future work will extend these analyses to more complex
geometries that better represent human lung airways.
References
Eulerian vs. Lagrangian
HYB mesh
1.Hofmann, W, Journal of Aerosol Medicine 1996, 9(3): 369-388.
2.Pope III, CA, et al., Journal of the American Medical Assoc. 2002, 287(9): 1132-1141.
3.Balásházy I, et al., Journal of Applied Physiology 2003, 94(5): 1719-25.
4.Hesterberg, TW, et al., Critical Reviews in Toxicology 2009, 39(9): 743-781.
5.Oldham, MJ, et al., Aerosol Sci. and Tech. 2000, 32(1): 61-71.
6.Heistracher, T and W Hofmann, Journal of Aerosol Science 1995. 26(3): 497-509.
7.Heistracher, T., and W Hofmann, Annals of Occup. Hygiene 1997, 41 Suppl. 1:537 542.
8.Weibel, ER, Morphometry of the Human Lung, 1963, Springer, Berlin.
9.Longest, P. W.; Oldham, M. J. Journal of Aerosol Science 2008, 39 (1), 48-70.
Acknowledgments
This work is supported by the National Science Foundation under Grant EPS-0903787 and the Mississippi State University Bagley College of
Engineering Ph.D. Fellowship.