Numerical investigation of regional particle deposition in the upper airway of a standing male mannequin in calm air surroundings

Computers in Biology and Medicine 52 (2014) 73–81

Contents lists available at ScienceDirect

Computers in Biology and Medicine journal homepage: www.elsevier.com/locate/cbm

Numerical investigation of regional particle deposition in the upper airway of a standing male mannequin in calm air surroundings Arash Naseri a, Omid Abouali a,n, Pejman Farhadi Ghalati a, Goodarz Ahmadi b a b

School of Mechanical Engineering, Shiraz University, Shiraz, Iran Department of Mechanical and Aeronautical Engineering, Clarkson University, Potsdam, NY, USA

art ic l e i nf o

a b s t r a c t

Article history: Received 25 January 2014 Accepted 12 June 2014

A 3-D realistic computational model of the airway system integrated into a standing male mannequin was developed. The computational domain includes the regions around the mannequin and the inside of the airway passages. The simulation was performed for low activity breathing rates with calm air around the mannequin. The flowfield of the inhaled air was first obtained from solving the Navier–Stokes and continuity equations. Then the particles were released in the domain around the mannequin and their trajectories were evaluated by using the Lagrangian approach for solving the particle equation of motion. The regional aerosols deposition was evaluated for different parts of the human airway system and the results were compared with those obtained from the separate modeling of the airway system without the interaction of the airflow with the mannequin external face. The results showed when the upper airway is integrated into the mannequin, the regional deposition of inhaled particles mainly changes in the airway system. & 2014 Elsevier Ltd. All rights reserved.

Keywords: Calm air Regional deposition CFD Mannequin Aerosol Outdoor and indoor condition Human upper airway Particle

1. Introduction Understanding the process of inhalation of aerosols has attracted considerable attention because of its significance to human exposure to particulate pollutants, and transmission of infectious diseases. In this connection, the nasal airway plays an important role in the human respiratory system. It heats and humidifies the inhaled air, and protects the lungs by capturing suspended particulate matters. The effective filtering of the nose becomes more critical in atmospheres with carcinogenic particles which could lead to serious health effects [1,2]. In the past few years, inhalation drug delivery via pharmaceutical aerosols through breathing has been developed. In order to maximize the absorption of pharmaceutical aerosols in the upper airways (nasopharynx and larynx), particles larger than 20 μm have been used to ensure their deposition and to prevent particles from entering the lungs [3]. In other cases, it is desirable to have the drug delivered deep into the lungs. Therefore, submicrometer aerosols are used to decrease the undesirable deposition in upper airways [4]. Estimating the regional deposition fraction of pharmaceutical aerosols is of interest for treating localized respiratory tract infection.

n

Corresponding author. Tel.: þ 98 711613304. E-mail address: [email protected] (O. Abouali).

http://dx.doi.org/10.1016/j.compbiomed.2014.06.007 0010-4825/& 2014 Elsevier Ltd. All rights reserved.

For an acceptable average drug dose, it is still important to be aware of the regional deposition rates. Since, a high local drug dose may cause tissue injuries or initiate a new disease. Thus, understanding the regional particle deposition in human upper airway and the fraction of particles entering the lungs are very important for controlled pulmonary drug delivery. Experimental and numerical study of the flow dynamic and particle deposition in the nasal cavity has been the subject of many researches. The investigations of Zachow et al. [5] and Wen et al. [6] are examples of some recent numerical studies for the airflow in the nasal cavity. Earlier Zamankhan et al. [7], Xi and Longest [8], Wang et al. [9], Inthavong et al. [10] and Moghadas et al. [11] presented the numerical results for the micro/nanoparticle deposition in the nasal cavity. Hahn et al. [12], Kelly et al. [13] and Doorly et al. [14] performed a series of experimental investigations of airflows in the nasal cavity. Kelly et al. [15,16] measured the particle deposition in a nasal cavity replica model with different surface qualities. These included a Stereo-lithography (SLA) nasal replica model with greater surface roughness and a Viper nasal replica model (manufactured with a Viper Si2 machine) with a smooth surface. More recently considerable attention was given to the regional deposition of particles in the respiratory system (Abouali et al. [17], Farhadi Ghalati et al.[18], Tavakoli et al.[19] and Jayaraju et al. [20], Wang et al. [21], Mylavarapu et al. [22], Xi et al. [23], Ghahremani et al. [24], Dastan et al. [25]). In these earlier

74

A. Naseri et al. / Computers in Biology and Medicine 52 (2014) 73–81

investigations, a separate airway system with either a uniform pressure or uniform inlet airflow velocity was considered and the external body was not included in the computational model. It is, however, generally known that many parameters such as particle size, air ventilation conditions, ambient airflow speed and temperature, as well as even people positions significantly affect the particle inhalation and particle deposition in the respiratory track. Recently, understanding the effect of external body feature on human exposure to particulate matter has attracted some attention. Kennedy and Hinds [26] used a full-sized and torso mannequin and studied particle inhalation through nose or mouth during breathing. Se et al. [27] studied the effect of human facial feature on particle inhalation and the total deposition in human airways using computational fluid dynamic (CFD) for a crude mannequin model. In their experimental studies, Berry and Froude [28] and Baldwin and Maynard [29] paid more attention to the calm air environment. They defined the calm air as a condition where the wind speed rarely exceeds 0.3 m/s. Using a mannequin, Der-Jen Hsu et al. [30] measured the inhalablility of ultra large aerosol in a calm air environment under different breathing conditions. They showed that the breathing condition and age do not significantly affect the aspiration ratio and suggested an empirical correlation for inhalability for nasal breathing. i.e.,

study, Taylor et al. [34] investigated the effect of the inflow geometry on flow predictions for the numerical simulation of steady nasal inspiration. They compared the cases of: blunt velocity profile applied to the nares, a tapered pipe inflow and a model that fully replicate the internal and external nasal airways. Investigating the airflow inside the nasal airway, they found that a tapered pipe inflow provides a better approximation for the natural inflow than a blunt velocity profile applied to the nares. Recently, Inthavong et al. [35] investigated the source and trajectories of deposited particles in the upper respiratory system using a computational model that combined the mannequin and respiratory airway in an indoor environment for free stream airflow velocities of 0.05, 0.2 and 0.35 m/s, which is common in an indoor environment. The presented brief literature review shows that the study of regional deposition in the upper human airway was reported only for a detached model of the airway. In this paper the regional particle deposition in the upper airway in a realistic mannequin situated in a calm air environment was studied. The computational model included the surrounding of the mannequin, as well as the upper airways, so that effects of the external body on regional particle deposition could be assessed.

2. Model description E ¼ 3:01 þ 0:64ðLog dae Þ2  2:78 Log dae 2.1. Geometry Here dae is the particle aerodynamic diameter. In their work, torso mannequins of an adult and a child were exposed to non-spherical airborne particles. Several volume flow rates corresponding to human activity condition (rest, moderate and heavy exercise) were studied. Yu-Tung Daia et al. [31] measured the nasal inhalability of aerosol particles under a calm air condition for ten subjects exposed to different inspiratory flow rates. Comparison of in vivo and mannequin measurements suggested that the natural convection from the human body could affect the particle motions in calm air and cause slightly higher inhalability for particles smaller than 50 mm. Li et al. [32] simulated simplified model of humanoid and then investigated effect of facial features on nostril velocity profile and then conducted resulted velocity profile as an inlet boundary condition of nasal cavity. They stated facial features effect only on a small region in front of the face (10–20 mm) and lead into non-uniform velocity profiles at the inlet of the nostril that it may affect particle inhalation. Inthavong et al. [33] integrated the human upper airway into a mannequin and placed inside a room, facing different airflow speeds. This study revealed better understanding about the air and particle flow patterns near mannequin exposed to different indoor conditions. In a detail

The mannequin used in the study is shown in Fig. 1. The mannequin resembles a typical American male [36,37]. The external computational domain is a 4 m side cube with the mannequin standing at its center as shown in Fig. 2. This computational domain is expected to be sufficiently large to eliminate the boundary effect. The grid sensitivity study was performed and the grid independency was achieved for a mesh of about 5.5 million cells. In order to investigate the importance of including the mannequin external body in computational study of deposition of particles in the human airway, a 3-D model of the airway is constructed and is connected to the external computational domain. Fig. 3 shows some samples of the coronal and sagittal cross sections of the upper airways of a healthy adult male. These CT scans were used for construction of the nasal and the rest of upper airway passages. The CT scans were provided by TABA imaging center (Shiraz, Iran). A CTI whole body scanner was obtained using a GE medical imaging system with the following parameters: 0.625 mm slices increment, 24.2 cm field of view, 120 kV peak and 99 MA. The boundary between the airway

Fig. 1. Views of the mannequin and details of facial features used in the study. All lengths are in centimeters.

A. Naseri et al. / Computers in Biology and Medicine 52 (2014) 73–81

mucosa and air in the upper respiratory tract was identified in CT scans slices. To construct the 3D model of nasal cavity, pharynx, larynx and trachea, the images were imported into ICEM, which is a CFD pre-processor software. The boundaries are connected to each other and a smooth surface was created to form the airway volume shown in Fig. 4. The same airway was studied before by Farhadi Ghalati et al. [18] for the internal airflow field. The attachment of the 3D airway model to the mannequin is shown in Fig. 5. Using Delaney method, an unstructured tri/quadrilateral volume was generated for the airway model. Three prism layers were created for boundary layer near the walls. The airway passage has a grid size of approximately 3,200,000 cells, which was selected after the grid independency tests.

the airflow is laminar and no slip boundary condition was assumed for the airway and mannequin’s surfaces. To discretize the governing equations, the finite volume scheme was employed and the SIMPLE algorithm was used for coupling between the pressure and velocity field. The convective and gradient terms of transport equations were discretized using second-order-upwind scheme and least squares cell based, respectively. The effect of particle motion on the flow was neglected. These assumptions are reasonable for the dilute particle concentrations in airflow. Therefore, the airflow field was first simulated and then the trajectories of particles were evaluated. The Lagrangian method was used to track the micro-particles. The particles are initially released with a uniform distribution into the volume that surrounds the mannequin and their initial velocity was set equal to settling velocity. This volume is a rectangular cube with the height of 2.3 m and width and depth of 1 m. The mannequin was set in the center of this rectangular cube. Three to six million

2.2. Flow solver and boundary conditions The flow regime was assumed to be steady inhalation and the periodic nature of breathing was not considered. Particular attention was given to the regional particle deposition morphology. Steady inhalation rates of 7, 10, 15 l/min, which are typical for rest or low activities were used in these simulations. To achieve these breathing rates, sub-ambient pressures were imposed on the outlet of trachea for which a pressure outlet boundary condition was used. Zero gauge pressure boundary condition was set on the boundaries of the computational domain to simulate the atmospheric pressure surrounding the mannequin. The sub-ambient pressures were needed for the trachea outlet to generate the appropriate flow rate are listed in Table 1. At these breathing rates

Fig. 2. Schematic of the computational domain.

Coronal plane

75

Fig. 4. The developed 3D model of the upper airways.

Axial plane Fig. 3. CT scans cross sections of the upper airways.

Sagittal plane

76

A. Naseri et al. / Computers in Biology and Medicine 52 (2014) 73–81

and C c is the Cunningham slip correction factor given as    2λ dp 1:257 þ0:4exp  1:1 Cc ¼ 1 þ dp 2λ

ð5Þ

In Eq. (5) λ is the air mean free path. The ANSYS-Fluent 6.3 software was used to solve the governing equations. The lift force also was included in the equation of motion for particles for some sample cases and no noticeable effect was observed, hence only drag and gravity were included in the equation.

4. Results and discussion

Fig. 5. Model of the mannequin with the attached upper airway.

Table 1 Boundary condition for the studied cases. Breathing condition

Rest condition

Low activity 1

Low activity 2

Volume flow rate (l/min) Gage pressure at trachea (Pa)

7  13.5

10  23.3

15  45

particles were used to generate deposition fractions that are independent of the number of particle used. As noted before, to represent the outdoor for a calm wind condition, the zero gage pressure boundary condition for all surfaces surrounding the computational domain except the floor was imposed.

AR ¼

3. Governing equation The steady incompressible Navier–Stokes and continuity equations are solved for evaluating the airflow field. The corresponding continuity, momentum equations are given respectively as ! ∇ u ¼0

ð1Þ

∇P ! ! ! u ∇u ¼  þ ν∇2 u

ρ

The breathing rates of 7, 10, and 15 l/min were studied to cover the range of rest to low activity conditions. The airflow in the nasal cavity is expected to remain in laminar regime for these breathing rates and up to about 15 l/min [13]. For a breathing rate of 7 l/min, Fig. 6 shows the path lines of inhaled air from the environment around the mannequin. The path lines are colored by velocity magnitude. It is seen that the peak velocity occurs in the pharynx region where the cross section area is the minimum. Sharp changes in the path lines occur in the nasal cavity and in the oro-pharynx region before the trachea. The detail of the flowfield for internal upper airway can be found in the work of Farhadi Ghalati et al. [18] The present computational model is validated by comparing the model predictions for inhalation of large aerosols in calm air with the experimental data of Hsu and Swift [30] for a breathing mannequin. Hsu and Swift [30] placed a torso mannequin in a modified shower stall and exposed it to (non-spherical) Al2O3 airborne particles with a specific gravity of 3940 kg/m3. The particle generation system was on the top of the chamber and the breathing machine was attached to the mannequin to provide the desired volume flow rate corresponding to rest and moderate exercise. They measured the aspiration ratio defined as the ratio of the concentration at the nostril to that at the upper surface of the domain. That is, Nn Q 1 N1 Q n

where Nn and N1 are, respectively, the number of particles entering the nasal cavity and entering the upper surface of the mannequin environment, Q n is the nasal breathing flow rate in m3 =s and Q 1 is defined as Q 1 ¼ ðV set þ V 1 ÞAsurf ace , where V set is settling terminal velocity of particle which is given as 2 V setteling ¼ ρp dp gC c dp =18μf and V1 is the average airflow velocity

ð2Þ

! where u , P, and ρ are respectively, the air velocity vector, pressure and density. Here, ν is the kinematic viscosity. The micro-particle transport and deposition were evaluated by the Lagrangian trajectory analysis approach. The corresponding particle equation of motion is given as !p du 3μC D Rep ! !p ! ¼ ð u  u Þþ g 2 dt 4ρp dp C c

ð3Þ

!p In Eq. (3), u is the particle velocity vector, dp is the particle diameter, ρp is the particle density, μ is the dynamic fluid viscosity, g is the acceleration gravity vector, and Rep ðRep ¼ ρjuj  upj jd=μÞ is the particle Reynolds number. Here C D is the particle drag coefficient CD ¼

24 ð1 þ 0:15Re0:687 Þ p Rep

ð4Þ

Fig. 6. Path lines of inhaled air colored with velocity magnitude for the breathing ate of 7 l/min.

A. Naseri et al. / Computers in Biology and Medicine 52 (2014) 73–81

at the top surface, which is very small. Here ρp, dp are, respectively, the particle density and diameter, and Cc, g, μ are, respectively, Cunningham correction factor, acceleration of gravity and air dynamic viscosity. To validate the present computational model, a set of simulations for conditions identical to those of the experimental study of Hsu and Swift [30] was performed. In particular, the wall boundary condition for the side walls of the chamber as well floor was imposed. The pressure boundary condition for the top surface remained unchanged. The newly simulated airflow in the environment was used to analyze a series of computer trajectory for different particle sizes. In these simulations, particle density was identical to those in the experiment. The model predictions for the aspiration ratio are shown in Fig. 7 and are compared with the experimental data of Hsu and Swift [30]. It is seen that the numerical model predictions for the aspiration ratio are in good agreement with the experimental data. Fig. 7 shows that the aspiration ratio is close to 95% for 5 μm particles and drops to about 65% for 20 μm particles. For particles larger than 50 μm particles, the aspiration ratio becomes negligibly small. Simulation results for the particle deposition in the studied model of the airway connected to a mannequin are presented in this section. A density of 1000 kg/m3 was used for the particles. Fig. 8 shows the total particle deposition in the upper airways from nostril to the trachea, defined as TDF ¼Nd/Ntotal, where N d is the number of particles that is deposited in the airway and N total is total number of particles that entered the upper airway at the nostril. Fig. 8 compares the simulation results for the case of a separate airway with the case of an airway connected to the mannequin. For the case of separate airway, the particles are injected into nostril in a uniform distribution. For the case of airway of a mannequin, the particles are distributed uniformly around the mannequin in the computational domain. The effect of using the external projection of human face and body depends on

77

both the breathing rate and particle diameter. For breathing rate of 7 l/min, Fig. 8 shows that including the external body of mannequin in the model reduces the total deposition in the upper airway compared with the separated airway. But for higher flow rates (10 and 15 l/min) and large particles the effect reverses and it increases the total deposition fraction for some specific sizes of particles. For flow rate of 10 l/min for particles with diameter of 10 mm or smaller, releasing the particles around the mannequin decreases the deposition fraction. For higher flow rate of 15 l/min this phenomenon is observed for particles with diameter of 5 mm or smaller. Fig. 9 shows the distribution of 2 mm particles at the nostrils when they are released uniformly into the volume surrounding the mannequin. It is seen that particle distribution is non-uniform at the nostrils, and also the particle distributions at the two nostrils are different. This particle distribution difference is perhaps due to the differences in the left and right nasal passages leading to different flow rates for the two sides of the nose. Table 2 shows the flow rate of each nostril for various total nasal flow rates. It is seen that the airflow rate thought the right nasal passage is about 73% of that in the left passage, and the percentage drops to below 70% for low breathing rate of 7 l/min. It appears that the flow resistance of the left nasal passage is lower than that for the right one and this leads to a higher flow rate through the left nostril. Another issue that should be noticed is that the inhaled particles from the surrounding are located in the high velocity regions at the nares. In fact due to the effect of external face and

Aspiration ratio

1 Present model (Q=7 lit/min)

0.8 0.6

Experimental data [30]

0.4 0.2 0 0

50

100

Particle diameter (µm) Fig. 7. Comparison of the predicted aspiration ratio versus particle diameter with the experimental data of Hsu and Swift [30] for the breathing rate of 7 l/min.

Depostion fraction (percent)

TOTAL DEPOSITION FRACTION 100 Mannequin(Q=7 Lit/min)

80

Separate airway (Q=7 Lit/min)

60

Mannequin(Q=10 Lit/min) Separate airway (Q=10 Lit/min)

40

Fig. 9. Distribution pattern of 2 mm particles released uniformly around the mannequin at the nostrils for 7 l/min breathing rate. The contours show the velocity magnitude.

Mannequin (Q=15Lit/min ) Separate airway (Q=15 Lit/min)

20

Table 2 The flow rate of each nostril for various total nasal flow rates.

0 0

5

10

15

20

25

30

35

7 l/min

10 l/min

15 l/min

2.85 4.15

4.22 5.78

6.42 8.58

Particle diameter (µm) Fig. 8. Comparison of the total deposition fractions in the detached upper airway with the airway in the mannequin.

Right nostril Left nostril

A. Naseri et al. / Computers in Biology and Medicine 52 (2014) 73–81

Fig. 10 shows that the general trend of sharp increase of the deposition fraction with particle size is captured by all four models. The real situation is case 1 that accounts for the nonuniformity of the particle and airflow distributions at the nostrils. It is seen that the effect of non-uniform particle distribution is more important than the airflow field at the nostril. When the particle injection at the nostril for the detached airway model is the same as that for the real case in the presence of the mannequin, the results for the particle deposition fraction are almost the same as for case one with slight overestimation for large particle and slight underestimation for small particles. However, when the particles are released uniformly at the inlet of the nostrils, the deposition fraction was noticeably underestimated for large particles and overestimated for small particles. This trend is the same for the detached airways with a uniform inlet pressure as well as for the attached airways when the airflow velocity at the nostril is computed properly. As mentioned before, Taylor et al. [34] presented a similar comparison for the flow filed inside the nasal airway comparing the cases of: truncated uniform velocity profile applied to the inlet of nostrils, a connected pipe to the nares and a model that includes the external nose too. The models of internal airway connected to a pipe and also the airway integrated to a sphere covering the nose and a part of face are also studied in present work (not shown for sake of brevity) and the

TOTAL DEPOSITION FRACTION Depostion fraction (percent)

(1) Airway attached to the mannequin: Particles are initially released in the environment surrounding the mannequin. (Particle distribution, as well as, airflow velocity at the nostril is computed as part of the solution.) (2) Airway attached to the mannequin: Particles are initially released with a uniform distribution in the mannequin nostrils. (Airflow field at the nostril is computed as part of the solution.) (3) Detached airways: Particles are initially released with a nonuniform distribution in the nostrils. The particle distribution is evaluated for the attached airways of case 1. (Airflow field at the nostril is uniform.) (4) Detached airways: Particles are initially released with a uniform distribution in the nostrils. (The inlet pressure at the nostril is uniform.)

results showed that although a pipe inflow covers both simplification and good approximation for the flowfield (as it was shown before by Taylor et al. [34]) but it does not provide a good approximation for the total and regional particle deposition in the nasal airway. The effects of releasing the particles surrounding the human body on regional deposition in human airways are studied in this section and the results are presented in Figs. 11–15. The regional deposition is defined as RD ¼ N zone =N total where N zone is the number of particles deposited on the special zone of airway. These figures clearly show the effect of external body due to presence of the mannequin markedly affects the regional deposition pattern. For vestibule and main airway regions of the airway model in the mannequin, Fig. 11 shows a sharper increase in

100 REALISTIC FLOW FIELD AND PARTICLE INJECTION

80

REALISTIC FLOW FIELD BUT UNREALISTIC PARTICLE INJECTION

60

SEPERATED AIRWAY FLOW FIELD BUT REALISTIC PARTICLE INJECTION SEPERATED AIRWAY FLOW FIELD AND UNREALISTIC PARTICLE INJECTION

40 20 0 0

5

10

15

20

25

30

35

Particle diameter (µm) Fig. 10. Comparison of deposition fraction for different models.

VESTIBULE AND MAINAIRWAY

100 Depostion fraction (percent)

body a non-symmetrical fully developed flow forms at the nostrils. This shear flow approaching the nose focuses the particles near the high velocity regions at the nares. When the particles are being released uniformly at the inlet of the nostril for a separate airway, some of them are located in low velocity regions near the nostril wall. Hence, for the integrated airway model, larger particles have more chance to be collected in the airway walls because of a higher initial inertia. On the other hand, for small particles or even the large ones in lower breathing rates when the initial inertia for particles at the center of nares are not enough for a high deviation from the streamlines, their deposition would be even lower compared with those for the uniform distribution at the inlet of the separate airway model. This is because the particles located near the boundary have more chance to be collected with a small deviation from the main stream or with interception effect. The non-uniform particle distributions at the nostrils for the upper airway that is connected to the mannequin lead to different total and regional particle deposition patterns compared with those for the separated upper airway with uniform particle concentration at the nostril. To provide a better understanding of the importance of the real (non-uniform) pattern of particle distribution at the nostril, a series of simulations for various cases for separated and attached upper airways are performed and the results are compared in Fig. 10. Particular attention is given to the following cases:

90 80

Mannequin (Q=7 Lit/min)

70

Separate airway (Q=7 Lit/min)

60 50

Mannequin (Q=10 Lit/min)

40

Separate airway(Q=10 Lit/min)

30 20

Mannequin (Q=15 Lit/min)

10

Separate airway (Q=15 Lit/min)

0 0

10

20

30

40

50

Particle diameter (µm) Fig. 11. Comparison of the regional deposition fractions in the vestibule and main airway of the detached upper airway with the airway in the mannequin for different breathing rates.

Depostion fraction (percent)

78

NASOPHARYNX

4.5 4

Mannequin(Q=7 Lit/min)

3.5

Separate airway (Q=7 Lit/min)

3

Mannequin(Q=10 Lit/min)

2.5

Separate airway (Q=10 Lit/min)

2

Mannequin (Q=15 Lit/min)

1.5

Separate airway (Q=15 Lit/min)

1 0.5 0 0

10

20

30

40

50

60

Particle diameter (µm) Fig. 12. Comparison of the regional deposition fractions in the nasopharynx of the detached upper airway with the airway in mannequin.

Depostion fraction (percent)

A. Naseri et al. / Computers in Biology and Medicine 52 (2014) 73–81

2

OROPHARYNX Mannequin(Q=7 Lit/min)

1.5

Separate airway (Q=7 Lit/min) Mannequin(Q=10 Lit/min)

1

Separate airway(Q=10 Lit/min) Mannequin (Q=15 Lit/min)

0.5

Separate airway (Q=15 Lit/min)

0 0

10

20

30

40

50

Particle diameter (µm)

Depostion fraction (percent)

Fig. 13. Comparison of the regional deposition fractions in the oropharynx of the detached upper airway with the airway in the mannequin.

25

LARYNX Mannequin(Q=7 Lit/min)

20

Separate airway (Q=7 Lit/min)

15

Mannequin(Q=10 Lit/min) Separate airway (Q=10 Lit/min)

10

Mannequin (Q=15 Lit/min) Separate airway (Q=15 Lit/min)

5 0 0

10

20

30

40

50

Particle diameter (µm)

Depostion fraction (percent)

Fig. 14. Comparison of the regional deposition fractions in the Larynx of the detached upper airway with the airway in the mannequin.

TRACHEA

2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0

Mannequin(Q=7 Lit/min) Separate airway (Q=7 Lit/min) Mannequin(Q=10 Lit/min) Separate airway (Q=10 Lit/min) Mannequin(Q=15 Lit/min) Separate airway (Q=15 Lit/min)

0

10

20

30

40

50

Particle diameter (µm) Fig. 15. Comparison of the regional deposition fractions in the trachea of the separate upper airway with the airway in the mannequin.

regional deposition fraction with particle size compared to that for the detached airway. The deposition fraction for the attached model is also higher for larger particles and lower for small particles compared with those for the detached airway model. This trend is perhaps due to the focusing of the particles at the inlet of the nostril due to the effect of mannequin face and body on the inhaled flowfield and particles. The concentration of large particles in the high velocity central regions increases the chance of deposition by inertia impaction mechanism. For small particles, however, being away from the walls at the nostril decreases the chance for deposition by the interception, and their overall deposition decreases, compared to the case with uniform concentration at the nostril. Fig. 12 compares the regional deposition fractions in the nasopharynx airway region as predicted by the detached airway model and the airway in the mannequin model. It is seen that the regional depositions predicted by the two models are comparable with the detached airway model predicting higher level of deposition compared to the airway model in the mannequin.

79

This is particularly the case for larger particles. The regional deposition fraction also decreases with increase of the inhalation flow rate. For the detached airway model, the particles that pass through the nasal airway and reach naso-pharynx region are in the range of 1–30 mm. For the airway model in the mannequin, however, this range is 1–15 mm as most of the larger particles are captured in the earlier part of the nasal passage. For the separated airway model, the peak deposition occurs for 15 mm particles for all breathing rate considered, while for the airway model in the mannequin the size for peak deposition varies with the breathing rate. For the inhalation rate of 7 l/min, the peak deposition occurs for 15 mm particles, but for higher flow rates the peak shifts to smaller 10 mm particles. The regional deposition fractions in oro-pharynx are shown in Fig. 13. This figure shows that the deposition fraction in the oropharynx region does not change monotonically with air flow rate, and in fact the fraction of deposited particles is the highest for the intermediate inhalation rate of 10 l/min. The deposition fraction in the oro-pharynx region predicted by the separated airway model is generally higher than those predicted by the complete model of airway in the mannequin. Fig. 13 also shows that the deposition fraction in the oro-pharynx region is quite low and is less than 2 percent. The regional deposition fraction in the Larynx is shown in Fig. 14. It is seen that large amount of deposition occurs for some particle sizes especially for lower breathing rates. Furthermore, the variations of deposition fraction predicted by both airway models are roughly similar. In particular, for the inhalation rate of 7 l/min, the deposition fraction predicted by the detached airway model and the airway model in the mannequin are almost the same with slightly higher predictions of the separated airway model for 15 mm and 20 mm. For the flow rate of 10 l/min, the separated model of the nasal upper airway leads to a higher deposition fraction in the larynx for 20 mm particles. For the flow rate of 15 l/min, the separated airway model predicts that the deposition fraction is highest for particles with diameter of 15 mm. However, the peak deposition shifts to 10 mm particles for the airway in the mannequin. It is also noticed that the deposition fraction predicted by the airway in the mannequin model is lower than that for the detached airway model. For the trachea region, the deposition fraction is generally low as shown in Fig. 15. For the airway in the mannequin model, the effects of external face on the particle path shift the peak deposition fraction in the trachea region to smaller particles. When the simulation is performed for the separated airway model, the peak deposition fraction for all three studied inhalation rates occurs for 15 mm particles. The airway model in the mannequin, however, predicts the peak deposition occurring for 10 mm particles. Also generally the deposition fraction decreases with increase in the inhalation rate. This might seems counter-intuitive as the inertia deposition is the main mechanism for these particle sizes. However, it should be remembered that for higher flow rates the particles are filtered out in earlier sections of the nasal airway and a smaller number of particles have the chance to reach the trachea region. This observation is consistent with the very low deposition fraction for large particles in the trachea region. As the calm air is more typical of the indoor environment, to test the validation of the present simulation results for closed indoor environments, the computational analyses were repeated for a cubical room with 4 m sides and one opening in one of the side walls. For this opening a 0.4 by 0.4 m square plane was placed in the wall fronting the mannequin and pressure inlet was set for its boundary conditions. Here the no-slip boundary condition was used on all side walls. The airflow condition and the associated particle depositions in the airway in the mannequin were evaluated. Fig. 16 compares the total deposition fraction and regional

80

A. Naseri et al. / Computers in Biology and Medicine 52 (2014) 73–81

VESTIBULE AND MAIN AIRWAY

TOTAL DEPOSITION FRACTION 100

80 60 40

OUTDOOR (Q=7 Lit/min)

20

INDOOR CONDITION (Q=7 Lit/min)

0 0

20

Depostion fraction (percent)

Depostion fraction (percent)

100

80 60 40

OUTDOOR (Q=7 Lit/min) INDOOR CONDITION (Q=7 Lit/min)

20 0 0

40

Particle diameter (µ µm)

20

40

Particle diameter (µm)

Fig. 16. Comparison of the deposition fractions in the airway in the mannequin for indoor and outdoor conditions.

2 µm

20 µm

5 µm

10 µm

30 µm

50 µm

Fig. 17. Particle deposition patterns in different regions of human upper airways for the breathing rate of 7 l/min.

deposition fraction in the vestibule and main airway as predicted by the model with no-slip wall boundary conditions (indoor environment) and zero gauge pressure boundary conditions (outdoor environment). It is seen that the corresponding total and regional deposition fractions are not different from those obtained for computational model with zero gauge pressure boundary conditions at the side boundaries. Therefore, it may be concluded that the presented results are valid for calm air environments in both indoor and outdoor. Fig. 17 shows the spatial deposition pattern for different particle sizes (5, 10, 20, 30, 50 mm) in the human upper airways for a breathing rate of 7 l/min. As expected, smaller particles penetrate deeper into the upper airway and reach the larynx and the trachea regions. All particles with sizes equal or larger than

30 mm are filtered out in the vestibule. Particles with diameter of 10 mm or smaller could reach into the trachea for low breathing rates. Larynx because of its rather complex geometry is a hot spot for deposition of particles passing through the upper airways.

5. Conclusions Aerosol deposition in the human upper airway inhaled from a calm air environment around a mannequin was studied numerically and the results were compared with those obtained for the detached airway model. The results showed that integrating the airway model with the mannequin significantly affected the airflow velocity and particle distribution at the nostrils. In particular, the particle distribution at the inlet of the nasal airway was non-uniform. As a

A. Naseri et al. / Computers in Biology and Medicine 52 (2014) 73–81

result, the regional depositions were significantly different compared to those of the commonly used separated airway with uniform inlet particle concentration. The differences in the total deposition predicted by these models, however, were moderate. The regional deposition patterns were also affected by the flow around the mannequin. The significant differences in the model predictions for the regional deposition fraction confirm the importance of the including the mannequin and the surrounding environment in the computational model for studying the regional particle deposition in the human upper airway. It should be pointed out that the present study focuses on inhalation in the completely calm air under isothermal condition and the thermal effect including the thermal plume around the mannequin was not considered. It is however known that the thermal plume generated by the air free convection around the human body can affect the particle transport and bring the particles to the breathing zone [38]. As a result, the particle inhalation into the respiratory system could be significantly affected. The inclusion of thermal effects and in particular thermal plume on particle deposition fraction in the upper airways, however, is left for a future study. Conflict of interest statement All the authors declare that there is no potential conflict of interest including any financial, personal or other relationships with other people or organizations within that could inappropriately influence (bias) this work. References [1] I. Bala´sha´zy, W. Hofmann, T. Heistracher, Local particle deposition patterns may play a key role in the development of lung cancer, J. Appl. Physiol. 94 (2003) 1719–1725. [2] M. Kreuzer, K.M. Muller, A. Brachner, M. Gerken, B. Grosche, T. Wiethege, H.E. Wichmann., Histopathologic findings of lung carcinoma in German uranium miners, Cancer 89 (2000) 2613–2621. [3] J.D. Suman, B.L. Laube, R. Dalby, Validity of in vitro tests on aqueous spray pumps as surrogates for nasal deposition, absorption, and biologic response, J. Aerosol Med. 19 (4) (2006) 510–521. [4] P.W. Longest, J. Xi, Computational investigation of particle inertia effects on submicron aerosol deposition in the respiratory tract, J. Aerosol Sci. 38 (1) (2007) 111–130. [5] S. Zachow, A. Steinmann, T. Hildebrandt, R. Weber, W. Heppt, CFD simulation of nasal airflow: towards treatment planning for functional rhinosurgery, Int. Congr. Ser. (2006) 165–167. [6] J. Wen, K. Inthavong, J.Y. Tu, S. Wang, Numerical simulations for detailed airflow dynamics in a human nasal cavity, Respir. Physiol. Neurobiol. (2008) 125–135. [7] P. Zamankhan, G. Ahmadi, Z. Wang, P.K. Hopke, Y.S. Cheng, W.C. Su, D. Leonard, Airflow and deposition of nano-particles in a human nasal cavity, Aerosol Sci. Technol. 40 (2006) 463–476. [8] J. Xi, P.W. Longest, Numerical predictions of submicronmeter aerosol deposition in the nasal cavity using a novel drift flux approach, Int. J. Heat Mass Transf. 51 (2008) 5562–5577. [9] S. Wang, K. Inthavong, J. Wen, J.Y. Tu, C.L. Xue, Comparison of micron and nano particle deposition patterns in a realistic human nasal cavity, Respir. Physiol. Neurobiol. 166 (2009) 142–151. [10] C.M. Se, K. Inthavong, J.Y. Tu, Inhalability of micron particles through the nose and mouth, Inhal. Toxicol. 22 (4) (2010) 287–300. [11] H. Moghadas, O. Abouali, A. Faramarzi, G. Ahmadi, Numerical investigation of septal deviation effect on deposition of nano/micro particles in human nasal passage, Respir. Physiol. Neurobiol. 177 (2011) 9–18.

81

[12] I. Hahn, P.W. Scherer, M.M. Mozell, Velocity profiles measured for airflow through a large-scale model of the human nasal cavity, J. Appl. Physiol. 75 (1993) 2273–2287. [13] J.T. Kelly, A.K. Prasad, A.S. Wexler, Detailed flow patterns in the nasal cavity, J. Appl. Physiol. 89 (2000) 323–337. [14] D.J. Doorly, D.J. Taylor, R.C. Schroter, Mechanics of airflow in the human nasal airways, Respir. Physiol. Neurobiol. 163 (2008) 100–110. [15] J.T. Kelly, B. Asgharian, J.S. Kimbell, B.A. Wong, Particle deposition in human nasal airway replicas manufactured by different methods, part I: inertial regime particle, Aerosol Sci. Technol. 38 (2004) 1063–1071. [16] J.T. Kelly, B. Asgharian, J.S. Kimbell, B.A. Wong, Particle deposition in human nasal airway replicas manufactured by different methods, part II: ultrafine particles, Aerosol Sci. Technol. 38 (2004) 1072–1079. [17] O. Abouali, E. Keshavarzian, P. Farhadi Ghalati, A. Faramarzi, G. Ahmadi, M.H. Bagheri., Micro/nano-particle deposition in maxillary sinus pre and post virtual uncienctomy, Respir. Physiol. Neurobiol. 181 (2012) 335–345. [18] P. Farhadi Ghalati, E. Keshavarzian, O. Abouali, A. Faramarzi, J. Tu, A. Shakibafard, Numerical analysis of micro- and nano-particle deposition in a realistic human upper airway, Comput. Biol. Med. 42 (2012) 39–49. [19] B. Tavakoli, O. Abouali, M.H. Bagheri, M. Yazdi, G. Ahmadi, Micro particles transport and deposition in realistic geometry of human upper airways, Int. J. Eng. 25 (4) (2012) 315–322. [20] S.T. Jayaraju, M. Brouns, S. Verbanck, C. Lacor, Fluid flow and particle deposition analysis in a realistic extrathoracic airway model using unstructured grids, J. Aerosols Sci. 38 (2007) 494–508. [21] Y. Wang, Y. Liu, X. Sun, Sh. Yu, F. Gao, Numerical analysis of respiratory flow patterns within human upper airway, Acta Mech. Sin. 25 (2009) 737–746. [22] G. Mylavarapu, Sh. Murugappan, M. Mihaescu, M. Kalra, S. Khosla, E. Gutmark, Validation of computational fluid dynamics methodology used for human upper airway flow simulations, J. Biomech. 42 (2009) 1553–1559. [23] J. Xi, X. Si, J. WonKim, A. Berlinski, Simulation of airflow and aerosol deposition in the nasal cavity of a 5-year-old child, J. Aerosol Sci. 42 (2011) 156–173. [24] E. Ghahremani, O. Abouali, H. Emdad, G. Ahmadi, Numerical analysis of stochastic dispersion of micro-particles in turbulent flows in a realistic model of human nasal/upper airway, J. Aerosols Sci. 67 (2014) 188–206. [25] A. Dastan, O. Abouali, G. Ahmadi, CFD simulation of total and regional fiber deposition in human nasal cavities, J. Aerosols Sci. 69 (2014) 132–149. [26] N.J. Kennedy, W.C. Hinds, Inhalability of large solid particle, J. Aerosols Sci. 33 (2002) 237–255. [27] C.M.K. Se, K. Inthavong, J. Tu, Inhalability of micron particles through the nose and mouth, Inhal. Toxicol. 22 (2010) 287–300. [28] R.D. Berry, S. Froude, An Investigation of Wind Conditions in the Workplace to Assess Their Effect on the Quantity of Dust Inhaled, HSE Internal Report IR/L/ DS/89/3, Health and Safety Executive, London, 1989. [29] P.E.L Baldwin, A.D. Maynard, A survey of wind speeds in indoor workplaces, Ann. Occup. Hyg. 42 (1998) 303–313. [30] D.H. Hsu, D.L Swift, The measurements of human inhalability of ultralarge aerosols in calm air using mannequin, J. Aerosol Sci. 30 (1999) 1331–1343. [31] Y.T. Daia, Y.J. Juang, Y.Y Wu, P.N. Breysse, D.J Hsu, in vivo measurements of inhalability of ultralarge aerosol particles in calm air by humans, J. Aerosol Sci. 37 (2006) 967–973. [32] X. Li, K. Inthavong, J. Tu, Particle inhalation and deposition in a human nasal cavity from the external surrounding environment, Build. Environ. 47 (2012) 32–39. [33] K. Inthavong, Q.J. Ge, X.D. Li, J.Y. Tu, Detailed predictions of particle aspiration affected by respiratory inhalation and airflow, Atomos. Environ. 62 (2012) 107–117. [34] D.J. Taylor, D.J. Doorly, R.C. Schroter, Inflow boundary profile prescription for numerical simulation of nasal airflow, J. R. Soc. Interface 7 (2010) 515–527. [35] K. Inthavong, Q.J. Ge, X. Li, J.Y Tu, Source and trajectories of inhaled particles from a surrounding environment and its deposition in the respiratory airway, Inhal. Toxicol. 25 (2013) 280–291. [36] R.L. Huston, Principles of Biomechanics, Taylor and Francis Group, Boca Raton, FL 33487-2742, 2009. [37] Z. Zhuang, D. Landsittel, S. Benson, R. Roberge, R. Shaffer, Facial anthropometric differences among gender, ethnicity, and age groups, Ann. Occup. Hyg. 54 (2010) 391–402. [38] M. Salmanzadeh, Gh. Zahedi, G. Ahmadi, D. Marr, M. Glauser, Effect of thermal plume adjacent to the body on the movement of indoor air aerosol particles, J. Aerosol Sci. 53 (2012) 29–39.