Journal of Biosciences and Medicines, 2013, 1, 23-27 JBM
http://dx.doi.org/10.4236/jbm.2013.12006 Published Online October 2013 (http://www.scirp.org/journal/jbm/)
Characteristic size research of human nasal cavity and the
respirato ry airflow CFD analysis*
Jun Zhang
Advanced Technology of Transportation Vehicle Key Laboratory of Liaoning Province, Dalian Jiaotong University, Dalian, China
Received 2013
ABSTRACT
To study the airflow distribution in human nasal cav-
ity during respiration and the characteristic parame-
ters for nasal structure, thirty three-dimensional,
anatomically accurate representations of adult nasal
cavity models were reconstructed based on processed
tomography images collected from normal people.
The airflow fields in nasal cavities were simulated
using the fluid dynamics with the finite element soft-
ware ANSYS. The results showed that the difference
of human nasal cavity structure led to varying airflow
distribution in the nasal cavities and the main airflow
passed through the common nasal meatus. The nasal
resistance in the regions of nasal valve and nasal ves-
tibule accounted for more than a half of overall resis-
tance. The characteristic model of nasal cavity was
extracted based on the characteristic points and di-
mensions deducted from the original models. It
showed that either the geometric structure or the air-
flow field of the two kinds of model was similar. The
characteristic dimensions were the characteristic pa-
rameters of nasal cavity that properly represented the
original model in research for nasal cavity.
Keywords: Nasal Cavity; Characteristic Dimension;
Three-Dimensional Reconstruction ; Numerical
Simula tion of Flow Field; Computational Fluid
Dyn a mic ; Finite Element Method
1. INTRODUCTION
Nose is the first barrier of defense to outer invasions in
the human respiratory system that is protective for life
long. It provides functions of filtering, warming, and
moistening inhaled air and protects the delicate structure
of the lower respiratory system. With the current devel-
opment of research towards the pathogenic mechanism
and the application of iatrical apparatus such as endos-
copes, it has been demonstrated that certain nasal diseas-
es are closely related to the abnormal structure of nasal
cavity [1]. Some researchers have investigated the air-
flow characters in nasal cavity to try to find the corre la-
tion between the nasal structure and the nasal disease [2].
The method of numerical simulation for airflow is help-
ful to this investigation. By simulating the structure and
function of the nasal cavity with three-dimensional re-
construction theory with a computer, we can profoundly
explore the outbreak, treatment and prevention of nasal
diseases. Keyhani [3] constructed a finite element mesh
of the human nasal cavity from the CAT scans. In his
work, the steady-state N avier-Stokes and continuity equ-
ations were solved numerically to determine the laminar
airflow patterns in the nasal cavity at quiet breathing
flow rates. The numerical results were validated by com-
parison with detailed experimental measurements from
Hahn’s [4] study. Martonen [5] et al. constructed a three-
dimensional computational model of the human upper-
respiratory tract that featured both sides of nasal cavity.
The model included airways of the head and the throat
based on a cast of a medical school teaching model. The
results showed the airflow patterns in different flow rate
values and the velocity profiles during inhalation and
exhalation. Subramaniam [6] et al. represented a three-
dimensional, computational model of an adult human’s
nasal cavity and nasopharynx, and solved the Navier-
Stokes and continuity equations for airflow using the
finite-element method under conditions of steady-state
inspiratory. The model was developed from magnetic
resonance imaging scans of a person’s nose. The nasal
cavity model was divided into several regions and the
flow apportionment among different regions of the nose
was detailed. Kim [7] investigated airflows in normal
and abnormal nasal cavities and surgically created mod-
els experimentally by Particle Image Velocimetry (PIV).
The average distributions of airflow in normal and ab-
normal nasal were obtained. In the case of simulation of
surgical operations, velocity distribution in coronal sec-
tion changed locally. Reimersdahl [8] and Hörschler [9]
presented the results of numerical simulation of the air-
flow in a model of the human nasal cavity which showed
*Project of Liaoning Province Education Department, LS2010030.

J. Zhang / Journal of Biosciences and Medicines 1 (2013) 23-27
24
a good agreement with the experimental findings. Till
now, little potent principle for describing the nasal cavi-
ties with characteristic parameters of nasal structure has
been put forward.
It is essential to build various numerical models to in-
vestigate airflow characters in different nasal structure
considering the individual difference of human nasal
cavity. In this paper, thirty finite element models of nasal
cavity of healthy volunteers were reconstructed. The
simulation results showed the distribution of airflow and
the relationship between the airflow distribution and the
nasal cavity structure. One of these models was com-
pared with its characteristic model in geometrical struc-
ture and airflow field to evaluate the feasibility of the
method for extracting characteristic dimensions of hu-
man nasal cavities.
2. METHODS
2.1. Reconstruction of Models
Thirty volunteers (18 males and 12 females aging from
25 to 55 years, median 30 years, Han nationality) from
Northeast China were randomly selected. They did not
have histories of nasal diseases or any other abnormity in
the nasal passages. Each volunteer was fully examined
by nasal anterior rhinoscopy and endoscopy which al-
lowed researchers to qualitatively designate his or her
septum as having no deviation. The nasal models were
developed from CT scans operated in the Second Affi-
liated Hospital of Dalian Medical University. The coron-
al images of nasal cavity at intervals of 3 mms wer e used
to complete the reconstruction since the coronal view
could best illustrate nasal structure. With the assistance
of a radiologist and a surgeon expertised at nasal CT
scans and anatomy, the interface between the nasal mu-
cosa and air in the nasal cavity was delineated from each
coronal image which would be linked together to form a
three-dimensional model. The models were constructed
and meshed automatically by the finite element software
of ANSYS after necessary artifact correction was carried
out.
The horizontal, sagittal and top views of a meshed
nasal model example are shown in Fi g u re 1 . The models
at the air outlet were lengthened artificially so that air-
flow could extend thoroughly there.
2.2. Numerical Simulation
The governing equations for the airflow through the up-
per airway are the conservation of mass (continuity) and
the Navier-Stokes equations, expressed as:
0
y
xz
u
uu
xyz
∂
∂∂
++=
∂∂∂
(1)
Figure 1. Three-dimensional reconstruction
model of the nasal cavity.
2
2
2
1
1
1
xxxx
x y zxx
y
x y zyy
zzzz
x y zzz
uuuup
uuufu
txyz x
u
uy uyuyp
uuu fu
txyz y
uuuu p
uuufu
txyz z
υ
ρ
υ
ρ
υ
ρ
∂∂∂∂∂
+++=−+ +∇
∂∂∂∂∂
∂
∂∂∂ ∂
+++=−+ +∇
∂∂∂∂ ∂
∂∂∂∂∂
+++=−++∇
∂∂∂∂∂
(2)
where
x
u
,
y
u
,
z
u
are the velocity component in the
Cartesian coordinates and p stands for the pressure. P is
the mass density of air and υ is its dynamic viscous coef-
ficient.
The nostril (Section Ω1 of Figure 1) directly opened
to the atmosphere with pressure boundary condition PΩ1
= 101,325 Pa. The interior wall (Section Ω of Figure 1)
of the nasal cavity was simplified as a rigid surface since
the deformation is minor and consequently weakly af-
fects the airflow field. The non-slip boundary condition,
uΩ = 0, was assigned to the inner wall. The regular in-
spiratory capacity for a relaxed, steady inhalation/exha-
lation is between 400mls and 600 mls per period [10]
with an inspiratory rate of 15 - 25 breath/minutes [11]
based on medical observations. The upper limit value of
600 mls was adopted in this paper. It was assumed that
the breathing period (the cycle of an inhalation and an
exhalation) is 3 seconds, and airflow velocity varies li-
nearly with time at the exit section, as shown in Figure 2.
The vertical axis and the horizontal axes showed airflow
flux and time, respectively. Point a showed the peak val-
ue of airflow flux in an inspiration period; point b
showed the peak value of airflow flux in an expiration
period. At the exit section, the peak velocity was calcu-
lated through uΩ2 = Q /0.75S, where Q was the tidal vo-
lume and S was the cross sectional area of the exit. The
velocity boundary condition was given at the top cross
section of the oropharynx (Section Ω2 in Figure 1)
based on the above assumptions.
Airflow through the nasal cavity was numerically si-
mulated over the entire breathing period after the model
was meshed with tetrahedron element. The airflow was
described as a transient-state turbulence flow with gas
parameters ρ = 1.25 kg/m3, υ = 1.7894 × 10−5 N·s/m2.

J. Zhang / Journal of Biosciences and Medicines 1 (2013) 23-27
25
Figure 2. The change of flow rate with time in a breath-
ing period.
The standard k-ε turbulent model was adopted in AN-
SYS.
2.3. Extraction of Characteristic Dimensions
The nasal cavity was divided into six main parts: nasal
vestibule, nasal valve, common nasal meatus, middle
nasal meatus, inferior nasal meatus and nasopharynx
region, which were defined as characteristic structures of
nasal cavity. The proper cross-section of nasal cavity
could be found out where the middle and inferior turbi-
nates just appeared or disappeared (blue lines in Figure
3(B)), and the juncture (red lines in Figure 3(B)) of ad-
jacent characteristic structures were. These were charac-
teristic sections of the nasal structure where vertexes
were defined as characteristic points. The sectional shape
of nasal vestibule, nasal valve and nasopharynx were
simplified as quadrilaterals, and the width and height
were defined as their characteristic dimensions. Anatomy
of nasal meatus was much more complex than the others.
Each meatus was simplified as a corner (as shown in
Figure 3(A)). Characteristic points in nasal meatus were
extracted as shown in Figure 4 and the widths of nasal
meatus were defined as characteristic dimensions. The
characteristic nasal cavity model of a volunteer was es-
tablished in ANSYS based on the coordinate data of the
person’s characteristic points and the airflow field was
numerically simulated. The comparison of geometry be-
tween the characteristic model and original one was
shown in Figure 3.
3. RESULTS
3.1. Airflow Distribution in Nasal Cavity
The pressure and velocity at any point in the nasal cavity
could be obtained after numerical simulations for thirty
nasal models were completed. The model shown in Fig-
ure 5 was a replic a t ion of a wom a n’s nasal cavity. A slice
at a proper position was selected to display the velocity
Figure 3. Extraction of characteristic points in nasal
meatus.
Figure 4. Velocity (left), pressure (middle) and vector (right)
plot at the moment b.
Figure 5. Distribution of airflow in the nasal passages at the
moment of point “a”.
distribution (Figure 5 left), velocity vectors (Figure 5
right) and pressure drops (Figure 5 middle) which pre-
sented airflow direction in th e nasal cav ity at the moment
b when the expiratory air flow flux and the pres sure drop
were on their peak values. The highest airflow velocity
appeared in the region of nasal valve. In the region of
nasal valve and nasal vestibule, the air pressure changed
sharply. By contrast, it changed slowly in the posterior
region of nasal proper cavity and the nasopharynx. In
these thirty examples, the airflow resistance in region of
3 cm distance from nostril accounted for from 50.5% to
77.8% of overall nasal airway resistances.
Several representative velocity distributions at the
moment were shown in Figure 6. These figures illu-
strated that airflow distribution in each model was a little

J. Zhang / Journal of Biosciences and Medicines 1 (2013) 23-27
26
Figure 6. Comparison of velocity (left) and pressure (right) distribution be-
tween characteristic model and origin one of nasal cavity.
different and the airflow flux on one side of the nasal
cavity was different from the other’s. The results indi-
cated that there were three airflow distribution modes in
the nasal airway:
1. The main stream passed through the common nasal
meatus and the residual part passed through the middle
and inferior nasal meatus (shown in Fig ure 6 left). In
this mode, the airflow flux through the common nasal
meatus accounted for 56.6% of overall flux. 2. The main
stream passed through the inferior nasal meatus and the
common nasal meatus (shown in Figure 5 middle). In
this mode, the airflow flux through inferior nasal meatus
and common nasal meatus accounted for 60.5% of over-
all flux. 3. The main airflow passed through the middle
nasal meatus and the common nasal meatus (shown in
Figure 5 right). In this mode, the airflow flux through
the middle nasal meatus and the common nasal meatus
accounte d for 77.0 % of o ve r all flux.
Among thirty examples, seven of them agreed with the
first mode; seven of them were categorized the second
mode. The other fourteen examples belonged to the third
mode.
3.2. Comparison of Airflow Distribution between
Characteristic Model and the Original One
By comparing the airflow distribution of velocity field
(Figure 6 left) and the pressure field (Figure 6 rig ht)
between the characteristic model and original one, it
showed that either the geometry structure or numerical
simulation was similar, and the numerical comparison
and difference was shown in Table 1.
4. DISCUSSIO N
The mechanism of airflow in human nose is important
for understanding many aspects of the biology and pa-
thology in the respiratory tract. The present investigation
showed that the airflow flux through left or right side of
nasal cavity lies on the airflow resistance or the cross-
sectional area. On each side, the airflow distribution de-
pends on the structure of airway as the main airflow
Tabl e 1. Difference of geometry dimensions and airflow cha-
racter between two models.
Original
model Characteristic model Difference
Cross dim. 0.036 m 0.046 m 12.20%
Vertical dim. 0.084 m 0.085 m 0.59%
Longitudinal dim. 0.136 m 0.136 m 0.00%
Pressure drop 94.8 Pa 79.8 Pa 8.59%
Maximal velocity 9.279 m/s 9.260 m/s 0.10%
passed through the route with wider airway. The resis-
tance is usually lower in the wider airway like the com-
mon nasal meatus or where it intersects with the middle
nasal meatus. So long as the presentative structure di-
mensions of nasal cavity are obtained, which were ex-
pressed as characteristic points and dimensions in this
paper, the airflow distribution in the real nasal cavity
could be well described. The models under the definition
of characteristic dimension can represent not only its
original model, but also the models with approximate
characteristic dimensions. Ulyanov [12] provided two
typical nasal models of the southern type and the north-
ern type. The characters of the northern type nasalcavity
were that the inferior turbinate was large in size and the
main airflow passed through the middle passage. The
characters of southern type nasal cavity were that the
inferior turbinate was small in size and the main airflow
passed through the inferior interior passage [6]. Because
of the large inferior turbinate, the inferior nasal meatus
was narrow and the resistance in this airway was high
which led to most of airflow passing through the middle
nasal meatus. The principle for the southern type nasal
cavity was the same as the northern type. This was a
good use of characteristic dimension for identifying hu-
mans with the structure character of nasal cavity.
5. CONCLUSION
A feasible method was developed to reconstruct the nu-
merical models of nasal cavities. Through numerical

J. Zhang / Journal of Biosciences and Medicines 1 (2013) 23-27
27
simulation results of thirty examples, the details of air-
flow distribution in the nasal cavity were illustrated. The
results showed that the wider the meatus were, the more
airflow would pass through. The distribution of the air-
flow would be changed on two sides of the nasal cavity if
any part of nasal structure varies. The numerical model
based on characteristic dimension was then reconstructed.
It showed that either the geometric structure or the air-
flow distribution in the characteristic model was similar
to the original one. The conclusion can be made that the
characteristic model can partly replace the original one
and even the models with approximate characteristic
dimensions during the model research towards nasal cav-
ity.
REFERENCES
[1] Uliyanov, Y.P. (1997) Surgical reconstruction of nasal
serodynamics. Proceedings of 16th World Congress of
Otolaryngology Head and Neck Surgery, XVI World
Congress of Otolaryngology Head and Neck Surgery,
Sydney, 1591-1595.
[2] Liu, Y.X., Yu, S., Sun, X.Z., et al. (2005) Structure of
nasal cavity and characters of airflow. Chinese Journal of
Otolaryngology Head and Neck Surgery, 40, 846-849.
[3] Keyhani, K., Scherer, P.W. and Mozell, M.M. (1995)
Numerical simulation of airflow in the human nasal cavi-
ty. Journal of Biomechanical Engineering, 117, 429-441.
http://dx.doi.org/10.1115/1.2794204
[4] Hahn, I., Scherer, P.W., Mozell, M.M., et al. (1993) Ve-
locity profiles measured for airflow through a large scale
model of he human nasal cavity. Journal of Applied Phy-
siology, 75, 2273-2287.
[5] Martonen, T.B., Quan, L., Zhang, Z., et al. (2002) Flow
simulation in the human upper respiratory tract. Cell Bi-
ochemistry and Biophysics, 37, 27-36.
http://dx.doi.org/10.1385/CBB:37:1:27
[6] Subramaniam, R.P., Richardson, R.B. and Morgan, K.T.
(1998) Computational fluid dynamics simulations of in-
spiratory airflow in the human nose and nasopharynx.
Inhalation Toxicology, 10, 91-120.
http://dx.doi.org/10.1080/089583798197772
[7] Kim, S.K. and Chung, S.K. (2004) An investigation on
airflow in disordered nasal cavity and its corrected mod-
els by tomographic PIV. Measurement Science and Tech-
nology, 15, 1090-1096.
http://dx.doi.org/10.1088/0957-0233/15/6/007
[8] Reimersdahl, Th., Hörschler, I. and Gerndt, A. (2001) Air-
flow simulation inside a model of the human nasal cavity
in a virtual reality based rhinological operation planning
system. International Congress Series, 1230, 87-92.
http://dx.doi.org/10.1016/S0531-5131(01)00021-8
[9] Hörschler, I., Meinke, M. and Schröder, W. (2003) Nu-
merical simulation of the flow field in a model of the
nasal cavity. Computers & Fluids, 32, 39-45.
http://dx.doi.org/10.1016/S0045-7930(01)00097-4
[10] Guitong, Y., Weiyi, C. and Jinbin, X. (1999) Biome-
chanics. Chongqing Press, Chongqing.
[11] Keyhani, K. a nd Scherer, P.W. (1995) Numerical simula-
tion of airflow in the human nasal cavity. Journal of Bio-
mechanical Engineering, 117, 429-441.
[12] Ulyanov, Y.P. (1998) Clinical manifestations the variants
of nasal aerodynamics. Otolaryngology Head and Neck
Surgery, 119, 152-153.