Research Article

CFD Simulation of Steady Blood Flow through Left

Subclavian Stenosed Artery, Turbulence Modeling
and Critical Narrowing
Fatima Moumen 1, Abed-El-Farid Djemaï 2
Abstract
In this work, we first propose a precise geometry for any arterial stenosis. Then we use the CFX-ICEM-ANSYS
15.0 code to perform a computational simulation of blood flow in the aortic arch, both in the healthy
case and in the case of presence of stenosis in the left sub-clavian artery with various stenosis degrees.
Then, we discuss our results obtained for the main hemodynamic parameters (velocity, pressure, wall
shear stress) and compare them with those obtained from medical imaging and numerical studies. We
also study the turbulent effects, via a modeling of the Reynolds number in function of stenosis degree,
the length of recirculation zone and the critical narrowing.

Keywords: Hemodynamic parameters, left sub-clavian artery (LSCA), geometry and degree of stenosis,
computational simulation, Reynolds number, turbulent effects
Nomenclature
SAS Subclavian Artery Stenosis Q Blood Flow
SAA Subclavian Artery Aneurysm ρ Blood density
SSS Subclavian Steal Syndrom ν Blood kinematic viscosity
CFD Computational Fluid Dynamics μ Blood dynamic viscosity
CVD CardioVascular Disease μ0 Zero shear rate viscosity
AVD Artherosclerotic Vascular Disease μ∞ Infinite shear rate viscosity
TIA Transcient Ischemic Attack λ Time constant
ADAN Anatomically Detailed Arterial Network α Stenosis degree
DUS Duplex UltraSound LR Recirculation zone length
WSS Wall Shear Stress u, V Blood velocity
OSI Oscillatory Shear Index Δp Pressure drop
MRI Magnetic Resonance Imaging Strain rate
Re Reynolds number n Power law index
De Dean number a Shape parameter
Introduction
Actually, CVD remains the main cause of death in the Western societies as well as the other societies. In fact, AVD is the

1
Department of Physics, Faculty of Exact and Applied Sciences, University of Oran 1-Ahmed Benbella, Oran, 31100, Algeria.
2
Department of Living and Environment, Faculty of Nature and Life, University of Sciences and Technologies USTO-Mohamed
Boudiaf, Oran, 31100, Algeria.

Correspondence: Abed-El-Farid Djemaï, Department of Living and Environment, Faculty of Nature and Life, University of Sciences
and Technologies USTO-Mohamed Boudiaf, Oran, 31100, Algeria.

E-mail Id: aef.djemai@gmail.com

Orcid Id: http://orcid.org/0000-0001-8983-8878

How to cite this article: Moumen F, Djemaï A. CFD Simulation of Steady Blood Flow through Left Subclavian Stenosed Artery,
Turbulence Modeling and Critical Narrowing. J Adv Res Appl Mech Compu Fluid Dyna. 2017; 4(3): 15-29.

Digital Object Identifier (DOI): https://doi.org/10.24321/2349.7661.201701

E ISSN: 2349-7661 l P ISSN: 2394-7055

© ADR Journals 2017. All Rights Reserved.

Moumen F et al.
principal cause of myocardial infarction (heart attack), stoke,
ischemic heart pain and sudden cardiac death, Lusis (2000).
These diseases account collectively for the first cause of
death in world. In addition, the international epidemic
of obesity and type 2 diabetes, which both represent
potent risk factors for atherosclerosis, contribute to sensibly
increase this ranking, World Health Organization (2014).
Many reviews have been realized, Lusis (2000), Glass and
Witztum (2001) and Libby (2003), to summarize the recent
findings concerning the description of the biological and
genetic bases of atherosclerosis. Atherosclerosis, disease of
the larger arteries, which also literally means “hardening of
the arteries”, is essentially caused by plaque deposits over
the artery walls, and that may induce a severe reduction
of blood flow through the arteries. Hence, this progressive
disease initiated by localized fatty streak lesions within the
arteries, occurs as early as the infancy and, over decades,
can develop into more complex plaques sufficiently large
to cause an important blockage to blood flow. This local
narrowing of the artery defines what it is commonly called
arterial stenosis, whose severity is generally measured by
the percentage reduction of the arterial diameter, called
the stenosis degree. Generally, the stenoses are considered
as the characteristic aspect of the atherosclerosis and
become clinically significant when their degree is greater
than 75%, Young (1979) and Ku (1997).
Artherogenesis refers to the growth of atheromatous
plaques in the inner lining of the arteries. The hemodynamic
conditions associated to artherogenesis are generally
observed where arteries change significantly their geometry
(bifurcation, junctions, branches, ect…). Moreover, it is
generally believed that low-shear stress regions are the
most vulnerable to artherogenesis, Ku and al (1985).
Stenosis is certainly the most frequent vascular complication,
Donnelly and London Nick (2009). In fact, arterial Stenosis
is abnormal narrowing or restriction present in the inner
wall of blood vessel due to the deposition of cholesterol,
fatty materials, cellular waste etc. Then, it is often described
only by its percentage in both clinical and scientific studies.
We found in the literature many attempts of mathematical
modeling of the stenosis geometry with various degrees
and eccentricities, Deshpande (1977), Ahmed and Giddens
(1983), Ahmed and Giddens (1984), Varghese and al (2007a-
b), Sadeghi and al (2011), Akbar and Nadeem (2014), Kartik
(2016), to name but some. These models have been used
to study hemodynamic parameters of blood flow (Velocity,
pressure and Wall shear stress). In fact, the mostly of these
models showed, for instance, that the low wall shear
stress appears to be closely related to the development
of arteriosclerosis, Grigioni and al (2002), and that the
turbulent aspect of the flow (via its Reynolds number) may
further the plaque rupture, Loree and al (1991).
E ISSN: 2349-7661 l P ISSN: 2394-7055
J. Adv. Res. Appl. Mech. Compu. Fluid Dyna. 2017; 4(3)
In this context, many CFD simulations have been performed
and contributed enormously to identify artherogenic
arterial sites, Berger and Jou (2000), Schächinger and
Zeiher (2002), Sloop and al (2015), to name but some. It
is shown that plaque build-up over the artery walls is most
common in the aorta and its main branches, especially
at the level of carotid arteries. In fact, the carotid artery
is particularly subject to deposition at its bifurcation and
ruptured plaque pieces may be deported to the brain,
where they may cause neurological symptoms or a stroke,
Ku (1997), Berger and Jou (2000). Nevertheless, locating the
probable sites of plaque rupture has been more difficult
and therefore, the attention was focused on the search of
the hemodynamic conditions that lead to plaque rupture
and then to an AVD. Furthermore, SAS is a significant form
of peripheral artery disease, which may be a marker of
diffuse atherosclerosis and increased risk for cardiovascular
events. It represents approximatively 2% of the free-living
population and 7% of the clinical population, Shadman
and al (2004). Hemodynamically significant stenosis of
the subclavian artery usually presents with symptoms of
upper limb ischemia on the ipsilateral side as the lesion. It
may also present as SSS with symptoms of vertebro-basilar
insufficiency as a result of retrograde flow in the ipsilateral
vertebral artery, Potter and Pinto (2014). The subclavian
steal phenomenon is certainly more prevalent in the left
subclavian artery and its symptoms, including diplopia,
dizziness, vertigo, syncope and dysarthria, are indications
of vertebrobasilar ischemia caused by diminished blood
pressure due to a proximal stenosis or occlusion of the
left subclavian artery.
Numerical and experimental studies may help to elucidate
in vivo data obtained by using DUS or MRI. Therefore, a
computational simulation based on a realistic geometric
model of stenosis can lead to an accurate understanding of
hemodynamic characteristics of blood flow in these stenotic
sites, and so it will permit an improvement of medical
diagnostics and surgical decisions made by clinicians.
In a previous work, Moumen and Djemaï (2016), and
using CFX-ICEM-ANSYS code, we have investigated the
hemodynamic parameters in the case of a normal aortic arch
anatomy and in the case of aortic coarctation pathology,
and compared the obtained results with those obtained
for a proposed model without and with coarctation, while
introducing a real tridimendional magnetic resonance
imaging geometry in the simulation process. We then
concluded that our proposed model reproduces, with a
high agreement, the real case obtained from imaging data.
The main aim of our work is to apply our realistic stenosis
geometry for the case of LSCA with stenosis, in the case
of commonly called SSS, and to perform a computational
simulation of blood flow in the healthy case and the stenotic
case in view to study the change in the hemodynamic
16
J. Adv. Res. Appl. Mech. Compu. Fluid Dyna. 2017; 4(3)
parameters, and comparing them with those obtained from
medical imaging and numerical studies. This paper deals
with the case of a steady flow through the left subclavian
stenosed artery with fixed boundary conditions, while a
second paper in preparation will treat the case of pulsative
flow, where physiological boundary conditions extracted
from imaging data will be adopted. Finally, we will focus
our attention on turbulent aspects.
Our Geometrical Model
Hereunder, we expose our geometrical model of the healthy
aorta and our model of the geometry of an arterial stenosis.
Figure 1.Geometry and meshing of healthy aorta model
This configuration only models and approximately
represents the real aortic arch. Although this geometry
appears to be somewhat simple, however it is fairly well
representative of the statistically real anatomical case.
Moreover, many authors have used geometries quite similar
to ours, for instance, the relatively recent articles, Sultanov
and al (2008), Sultanov and Guster (2008), Hiratzka and
al (2010), Benim and al (2011), Gao and al (2011), Wittek
and al (2013).
Figure 2.STL file, geometry and meshing of a real healthy aorta
Our stenosis geometry model
In the literature, there are many various geometries that
are proposed to represent the arterial stenosis. In the
following, we present our own geometrical model of a
axisymmetric arterial stenosis, which is not considered in
17
Moumen F et al.
Geometry and meshing of our healthy aorta model
Firstly, in the proposed model, Moumen (2010), we
represent the aorta arch as a 180° curved circular tube
with an interior radius of 0.37 cm and an external radius
of 2.57 cm, to be statistically in the average of real aorta
sizes. We link to this tube a rectilinear cylindrical tube of
diameter of 2.20 cm and length of 4 cm to represent a part
of the descending aorta. The three principal aortic arteries
are also represented by three cylinders with diameters of
respectively 0.94 cm, 0.60 cm and 0.66 cm. Using ICEM-
CFX code, we obtain the following meshing (262410 nodes
and 1547453 tetrahedral elements) and geometry (Fig. 1).
Secondly, this model of aorta was validated in a previous
article, Moumen and Djemaï (2016). In fact, this validation
was done through the realization of the geometry and
meshing (with 154,331 nodes and 901,105 tetrahedral
elements) of a real aorta (MRI data furnished by LIB-UPMC-
Paris under STL format), (Fig. 2), and then a comparison
(with great success) of our aorta model with the real aorta.
the literature at all.
Firstly, we consider that the healthy arteries of medium or
large caliber are of circular cylindrical geometry, (Fig. 3):
E ISSN: 2349-7661 l P ISSN: 2394-7055
Moumen F et al. J. Adv. Res. Appl. Mech. Compu. Fluid Dyna. 2017; 4(3)

Figure 3.Circular cylindrical geometry of healthy large arterial segment of length L

Then, we consider that the process of plaque build-up over equal to h ( ).
the inner arterial wall will create a particular geometrical
form for the incipient stenosis, under the direct influence Among various possible geometrical forms that stenosis
of the variations of the hemodynamical parameters may take, we have opted for the choice of a hyperboloid
characterizing the pulsatile blood flow at some artherogenic of one sheet, (Fig. 4).
arterial sites. Here, we consider the length of this site is

Figure 4.One sheet-Hyperboloid form of the arterial stenosis of length h

This choice has been tested in Moumen and Djemaï (2016), while is the volume of the non-stenosed cylinder
to represent the geometry of an aortic coarctation and of total height h.
gives very hopeful results.
In this work, we adopt for the following definition of
Let’s recall some properties of the one-sheeted hyperboloid. stenosis degree:
It is a doubly ruled surface and when it is oriented along
the z-axis, the one-sheeted hyperboloid with skirt radius
is described by the following Cartesian equation:
where is the unobstructed area and
is the throat area.

where represents exactly the radius in the plan In this case, knowing the radius R of the considered artery,
, i.e at the peak of the stenosis, (Fig. 4), and its the stenosis is then parameterized by its length h, its degree
parametric equations are given by: and its skirt radius :

, ,

where . and we have:

The radius r(z) for some is given by:

Computational Simulations Of Blood Flow

so that the radius at the top cross section ( ) is General assumptions and considerations
given by:
Firstly, the blood is considered as an incompressible
non-Newtonian fluid, although the Newtonian feature
is generally accepted as a reasonable first approximation
The volume of this one-sheeted hyperboloid with total to the non-Newtonian behavior of blood at shear rates
height h is given by: observed in large arteries, Perktold and al (1991a). The non-
Newtonian viscosity characteristic of blood is incorporated
.
by using the Carreau model, Perktold and al (1991b), Akbar
and Nadeem (2014):

E ISSN: 2349-7661 l P ISSN: 2394-7055 18

J. Adv. Res. Appl. Mech. Compu. Fluid Dyna. 2017; 4(3)
where , n=0.3568, a=2, ,
. Furthermore, the blood fluid is assumed
to have a density of 1.0595 gr/cm3.
Secondly, we consider the blood flow as steady fully
developed turbulent flow. Thirdly, we ignore the
distensibility of the vessel walls, such that they are assumed
to be smooth rigid (non-elastic) walls. This assumption
appears to be more realistic especially in the case of plaque
build-up which makes the inner artery walls more rigid
with no slip conditions.
We also adopt as an inlet boundary condition the normal
speed of 150 cm/s, and as an outlet boundary condition
the relative pressure of 60 mmHg. Several recent numerical
studies have treated the stenotic steady flows as we do in
this paper, Varghese and al (2007a), Sherwin and Blackburn
(2005), Griffith and al (2008), Javadzadegan and al (2013),
Hazarika and Sarmah (2014), Gaur and Gupta (2014), to
name but some. We motivate our choice of these boundary
conditions as follows. Firstly, since we are dealing with the
steady flow case, the choice of the normal averaged velocity
value (150cm/s) as an inlet boundary condition appears
to be the more indicated choice, since it corresponds to
the usual pressure inlet condition (80mmHg) commonly
used in diverse steady state simulations. Secondly, since in
the presence of a narrowing the flow exhibits a resistance
and consequently an increase of both velocity and shear
stress and also a pressure drop (All these are quantities
of physiological relevance), we have assigned the value of
60mmHg (≈8kPa) as an outlet boundary condition in view
to study the variations of these hemodynamic parameters,
and in particular the pressure drop Δp since this latter is a
very important clinical index (See for instance, Effat and al
(2016)). We have tried other values of this outlet pressure,
for instance the value 100mmHg (≈13.33kPa), and find that
the measured drop pressure is exactly the same, while the
profiles of velocity and WSS remain exactly the same as for
the first outlet condition. Obviously, our outlet condition
is applied to all outlets.
Our computational simulation results using ANSYS-ICEM-
CFX 15.0 concerns the profiles of pressure, velocity and WSS
19
Moumen F et al.
in all the cases and, in order to ensure the convergence of
each time step, we set the convergence criterion for the
relative residual for all dependent variables equal to 10-4.
In all our simulations, we apply the high Reynolds number
k–ε turbulence standard model, because we are interested
to highlight the fully developed turbulent aspect of the
flow. The k-ω turbulence model is rather more adapted for
the low Reynolds pattern. In fact, as shown in Bluestein and
al (1997), and Bluestein and al (1999), the development of
vortical structures in the flow field near the walls for high
Reynolds numbers leads to enhanced platelet deposition.
Henceforth, in this work, the Reynolds number is equal
to Re=10812 for the case of healthy aortic arch, and to
Re=10869 for the case of a stenotic left sub-clavian artery
with α=75%. With these values of Reynolds number greater
than 10000, and according to Incropera and al (2007), we
are effectively in presence of a fully developed turbulent
flow in the LSCA.
In another hand, it is actually well admitted that the most
important factors that characterize the artherogenesis
and plaque rupture are: Low WSS, high OSI, He  and
Ku (1996), and secondary flow. Since, in this paper we
deal with a steady flow, we can only discuss the WSS and
the recirculation flow.
Hereunder, we present our computational simulations for
the case of a healthy aortic arch and also for the only case
of stenotic left sub-clavian artery with stenosis degree
α=75%. We have also realized computational
simulations for the cases α=25%, 50%, 80%, 85%, 90%,
92%, 95%, 97% and 99% whose results will serve us to treat
the turbulent via a modeling of the Reynolds number in
function of the stenosis degree. We reserve the section
6 for an overall discussion about all the results obtained
in this work.
Case of our healthy aortic arch model
Using our geometrical model of healthy aorta described in
section 2.1., we realize computational simulations of the
blood flow through the aorta and its main branches and
we obtain the following profiles of pressure, velocity and
wall shear stress, (Fig. 5-7).
E ISSN: 2349-7661 l P ISSN: 2394-7055
Moumen F et al. J. Adv. Res. Appl. Mech. Compu. Fluid Dyna. 2017; 4(3)

Figure 5.Pressure profile in the case of a healthy aortic arch

Figure 6.Velocity profile (streamline) in the case of a healthy aortic arch

Figure 7.Wall shear stress profile in the case of a healthy aortic arch
Case of a stenotic left sub-clavian artery
In this section, we have used our geometrical model of
stenosis introduced in section 2.2., in view to study the left
sub-clavian artery stenosis. We adopted a realistic position
of the stenosis at the beginning of the left sub-clavian artery
(at 0.5cm from the beginning of this aortic branch) as shown
in Figures below corresponding to the case of stenosis
degree α=75%. In all the case, the length h of the stenosis is
considered equal to 1cm, and the obtained computational
meshing for the three cases are respectively: (261929 nodes
and 1544468 tetrahedral elements), (261549 nodes and
1542179 tetrahedral elements), and (261320 nodes and
1540893 tetrahedral elements), that provide a satisfactory
spatial resolution.
Inserting our geometrical model of stenosis into our
geometrical model of aorta, at the level of the left subclavian
artery, we realize the following computational simulation of
the blood flow through the stenosis and we get the profiles
of pressure, velocity and wall shear stress, for the cases
α=25%, 50%, 80%, 85%, 90%, 92%, 95%, 97% and 99%.
Hereunder, we present only the figures corresponding to
the case of α=75%, (Fig. 8-10).

E ISSN: 2349-7661 l P ISSN: 2394-7055 20

J. Adv. Res. Appl. Mech. Compu. Fluid Dyna. 2017; 4(3) Moumen F et al.

Figure 8.Pressure profile in the case of our stenosis model at

Figure 9.Velocity profile (streamline) in the case of our stenosis model at

Figure 10.Wall shear stress profile in the case of our stenosis model at
Turbulent Effects. Reynolds Number Vs Stenosis that may at its turn cause plaque rupture. For instance, for
Degree symmetric stenoses at carotid artery flow rates, Grigioni
and al (2002), peak pressure fluctuations were observed
It is well known that turbulence develops when the blood 1-1.5 upstream diameters distal to the stenosis, but there
passes through an arterial stenotic location. Many models were no significant turbulent pressure fluctuations at
have been studied to study the hypothesis that turbulence coronary artery flow rates. However, stenosis asymmetry
near a deposed plaque can give rise to pressure fluctuations strongly increased the intensity of turbulent pressure

21 E ISSN: 2349-7661 l P ISSN: 2394-7055

Moumen F et al. J. Adv. Res. Appl. Mech. Compu. Fluid Dyna. 2017; 4(3)

fluctuations at flows simulating carotid flow and resulted where δ is the radius of curvature (in our aortic arch model,
in significant pressure fluctuations for coronary flow it is equal to 1.47cm).
conditions. These studies predict peak to peak pressure
fluctuations of 15 mmHg in a 90% asymmetric coronary When the artery is obstructed by means of a stenosis whose
stenosis. In conclusion, it is possible that turbulence may degree is , then the diameter reduces to ,
play a role in acute damage of atherosclerotic plaques, where, (See section 2.2):
particularly in asymmetric stenoses.

The Reynolds number is a dimensionless quantity that is

used essentially to help distinguish between flow patterns On the other hand, the velocity of the blood flow at stenosis
in various fluid flow situations. It is defined as the ratio of site also varies in function of . Then, we can deduce
inertial forces to viscous forces, and is written as: mathematically that the velocity of the blood flow at
stenosis site will vary in function of as follows:

where is the fluid density of the blood, V its velocity,
its dynamic viscosity, its kinematic viscosity and D=2R
is the (stenosis degree-dependent) artery diameter. In
fact, to study blood flow through curved arteries (like the
aortic arch), it is more indicated to use the Dean number
defined as follows:
.
In the table 1, we resume the essential data obtained not
only for Reynold numbers, but also for the averaged values
of pressure, of velocity and of WSS at the stenosis peak
by, issued from our computational simulations for various
stenosis degrees.

Table 1.Obtained data from our computational simulations

Α (%) Re Meshing Averaged value of Averaged value of Averaged value of
Number Number of Pressure at the peak Velocity at the peak WSS at the peak
of nodes elements (x103 Pa) (m/s) (Pa)
0 10812 262410 1547453 7.95 2.40 36.80
25 10813 262293 1546706 7.78 2.44 37.48
50 10832 261929 1544468 5.30 2.61 48.30
75 10869 261549 1542179 1.48 3.65 76.45
80 10876 261520 1542084 1.08 3.72 80.85
85 10918 261505 1541984 1.52 3.61 73.70
90 10926 261320 1540893 2.45 3.25 62.81
92 10918 261333 1540329 2.91 3.21 59.86
95 10906 261188 1540145 4.24 2.61 38.02
97 10877 261224 1540307 4.44 2.57 35.72
99 10804 261216 1540329 4.98  1.29  13.38

Now, using these data, we have realized the following 10 and d = 4/3, such that we get :
modeling of the behavior of the Reynolds number in
function of the stenosis degree by the function:
It follows that:

and using Mathematica, we found the following graph,

(Fig. 11), with a = 10812 = Re(0), b = 7208=2 Re(0)/3, c =

E ISSN: 2349-7661 l P ISSN: 2394-7055 22

J. Adv. Res. Appl. Mech. Compu. Fluid Dyna. 2017; 4(3) Moumen F et al.

Figure 11.Mathematical modeling of Reynolds number in function of stenosis degree

Using Origin software, we also obtain the following For discussions concerning all these results, we refer to
graphs for pressure, velocity and WSS at the stenosis the section 6.
peak in function of the stenosis degree α, (Fig. 12-14).

7
Pressure at the peak
Pressure at the peak (103 Pa)

0 10 20 30 40 50 60 70 80 90 100
stenosis degree (%)

Figure 12.Plotting pressure at the peak vs stenosis degree

4.00
3.75
3.50 Velocity at stenosis peak
Velocity at stenosis peak (m/s)

3.25
3.00
2.75
2.50
2.25
2.00
1.75
1.50
1.25
1.00
0 10 20 30 40 50 60 70 80 90 100
stenosis degree (%)

Figure 13.Plotting velocity at the peak vs stenosis degree

90
85
80
75 WSS at the peak
70
65
60
WSS at the peak (Pa)

55
50
45
40
35
30
25
20
15
10
5
0
0 10 20 30 40 50 60 70 80 90 100
stenosis degree (%)

Figure 14.Plotting WSS at the peak vs stenosis degree

Duplex Ultrasonography, Left Subclavian Stenosis Let’s recall that Duplex ultrasonography imaging, Huda
And Subclavian Steal and Slone (2003), Gerhard-Herman and al (2006), Mrdjen
(2013), Haidekker (2013), is an imaging technique which is
Vascular ultrasonography is the main branch of radiology a non-invasive evaluation of blood flow through arteries
that uses duplex. It is an inexpensive, non-invasive way to and veins.
determine pathology. In fact, it helps identify the location
and the severity of vascular stenosis  (narrowing) or Before presenting the recent results concerning sub-clavian
occlusion (complete blockage) within an artery. Its results steal on DUS, as a consequence of a left subclavian stenosis,
may be very useful for therapeutic decisions. which is our main aim in this paper, let’s first recall some

23 E ISSN: 2349-7661 l P ISSN: 2394-7055

Moumen F et al.
basic physical notions. In fact, these can offer some insight
into the manifestation of sub-clavian stenosis. When fluid
travels through a smaller area A and flow Q is constant,
then its velocity V increases:
Hence, blood traveling through an area of stenosis will
travel more quickly than in the normal case. Moreover,
Bernoulli principle states that an increase in the speed of
the fluid occurs simultaneously with a decrease in pressure
or a decrease in the fluid’s potential energy. So according
to this principle, when blood moves at higher velocity, it
exerts lower pressure. Thus, the fast moving blood distal
to a stenosis is at lower pressure than usual for that area.
During systole, flow is greatest, therefore velocity is greatest
and then pressure is lowest distal to the stenosis. Since
blood flows from areas of high to low pressure, blood flow in
the left vertebral artery may favor transient reversal during
systole, Potter and Pinto (2014). When superimposed on
otherwise forward flowing blood, this change in pressure
gradient manifests as a transient slowdown in blood flow
velocity.
During diastole, when sub-clavian pressure normalizes, left
vertebral artery returns to normal full velocity antegrade
flow. These transient slowdowns in flow, caused by transient
drops in subclavian artery pressure, manifest on spectral
DUS as a notch in the normal monophasic waveform, that
is the first type of vertebral artery waveform that was
identified in Kliewer and al (2000).
With even greater stenosis, the pressure differential can
be so great as to cause a transient stoppage of blood flow
during diastole, as regular forward flow is matched by a
gradient favoring reversed flow. This manifests on spectral
DUS as a deep notch reaching the x axis (the point of zero
velocity). As even greater stenosis causes a greater pressure
differential, flow transiently reverses in the vertebral artery.
This is represented on the spectral DUS by regions of
negative velocity during systole.
Furthermore, a recent review, Ochoa and Yeghiazarians
(2010), reveals that few practitioners consider the
importance of upper extremity arterial disease, like the left
subclavian stenosis. As a matter of fact, the left subclavian
artery is four times more likely to be affected than the right
or innominate arteries, Schillinger and al (2002). However,
once obstructive disease affects other aortic arch vessels,
namely the carotid or vertebral arteries, the likelihood of
steal or ischemic symptoms drastically will increase.
While the definitive diagnosis usually rests on medical
imaging, the importance of a thorough physical exam must
be underscored. As such, a systolic cuff pressure difference
or gradient (in the case of invasive assessment) of > 10
mmHg can be used to delineate hemodynamic significance
E ISSN: 2349-7661 l P ISSN: 2394-7055
J. Adv. Res. Appl. Mech. Compu. Fluid Dyna. 2017; 4(3)
and justify further investigation. Once a significant blood
pressure difference is found, the next step should involve
diagnostic imaging. In this case, DUS with color flow is
the most indicated non-invasive exam. It can detect a
significant stenosis if its findings include some indications
about waveform dampening or monophasic changes, color
aliasing suggestive of turbulent flow, and increased blood
flow velocities at the suspected site(s) of stenosis.
Although more expensive, other complementary invasive
imaging exams could be used to confirm the diagnosis,
such as CT angiography (CTA) and/or magnetic resonance
angiography, or contrast angiography, Schillinger and al
(2002). The first provide an excellent anatomical resolution
with potential assessment of not only the subclavian artery
but other supra-aortic vessels as well, and the second is
very helpful with its invasive hemodynamic measurement
of subclavian lesions.
Discussions, Conclusions, Limitations And
Perspectives
Relevance of the subject
One of the main goals for numerical investigations of
the blood flow in the thoracic aorta, and in particular
in the aortic arch, is to well understand changes in flow
patterns and associated pressures in the presence of aortic
pathologies, such as sub-clavian steal disease. Indeed,
this pathology induces an internal geometry deformation
of the left subclavian artery, which induces at its turn
significant changes in the hemodynamic characteristics of
the flow. These phenomena would affect the normal blood
circulation, and strong turbulence would occur, and then
affect the arterial walls, leading ultimately to the need of
surgical intervention to avoid fatal events.
The pathogenesis is mainly embolic sometimes
hemodynamic (stenosis or occlusion by thrombosis
superimposed in the absence of sufficient substitution).
Hemodynamic mechanism is well aware of some transient
ischemic attacks (TIAs) arising on the occasion of a
temporary drop in blood pressure (orthostatic hypotension,
antihypertensive medication, shock, ...) such as the
subclavian steal, which is characterized by the vertebral
artery flow inversion, due to a stenotic lesion in the origin
of the subclavian artery. Subclavian steal syndrome is a
constellation of signs and symptoms caused by retrograde
flow of blood in the vertebral artery due to ipsilateral arm
exercise or exertion. In general, the so-called SAS refers to
narrowing of one or both of the subclavian arteries. The
risk of stroke is correlated to the degree of stenosis, the
irregularity of the plate, the hypoechoic appearance and
the worsening of stenosis on successive scans.
Although SAS is not very common, it is estimated in the
general population as a whole to be around 2-4%, while in
24
J. Adv. Res. Appl. Mech. Compu. Fluid Dyna. 2017; 4(3)
those with peripheral vascular disease, it can be as high as
18%, (See the excellent review, Ochoa and Yeghiazarians
(2010)).  It can arise from a number of pathological
conditions, with atherosclerosis being the most common
by far. It is considered as a disease of smokers and of
diabetics, and it was observed that the LSCA is involved
more often than the right (Statistically, 4 times more
commonly affected).
SAS can be suspected when some significant blood pressure
difference is found between the arms. The presence of
subclavian stenosis is suspected when this difference is
greater than 20 mmHg. In these patients, routine blood
pressure should be measured in the higher side. It is very
important to know that up to 50% of patients may have
concurrent coronary artery disease. Thus, it is expressely
indicated to complete the exam with radiographic imaging
tests, such as Ultrasound technique, that may show
evidence of subclavian steal on the ipsilateral side, and
CT/MR angiography, that will allow direct visualization of
degree and extent of stenosis.
It could happen that SAS is complicated with SAA, Wu J.,
and al (2016). Nowadays, clinically it is well established
that angioplasty and stenting are the only good option to
treat SAS and SAA and then the Coronary-subclavian Steal
Syndrome, with high level of technical and clinical success.
As CFD is a technique for cardiovascular medicine and a
very strong tool for clinician practionners, this work has as
objective to provide them with some precious indications
concerning the variation of the various hemodynamic
parameters, and in particular arterial pressure and turbulent
effects, in the case of left SAS for different stenosis degrees.
To our knowledge, this subject has not been previously
studied, in this precise context. We have planned to treat
this subject in two phases. First, we treat the limit case of
steady flow in view to get some useful indications. Then,
we will treat the case of a pulsative flow in a future work
(in preparation).
Choice of geometries and limitations
Firstly, our models of aorta and stenosis geometries have
proved their efficiency in the case of aortic coartation,
Moumen and Djemaï (2016). Nevertheless, we think
that there is a lot of to do to make these models more
statistically representative of the real physiological aorta
geometry and the real stenosis shape. This limitation
could be overcome by using CT image based aorta models.
Another limitation was the assumption of rigid artery wall
and the no-slip condition. Aorta walls are thick and elastic
in nature. The fluid-structure interaction has to be taken
into account to get findings from the CFD simulation very
close to the reality.
25
Moumen F et al.
In this work, we have also considered the blood fluid as an
incompressible non-Newtonian fluid, and a steady fully
developed turbulent flow, using Carreau model for the
blood viscosity and k–ε turbulence standard model for
turbulence effects. As a logical continuation of this work,
we plan to use the pulsative aspect of the flow and use
the adapted inlet/outlet boundary conditions extracted
from original imaging data. We could also choose other
viscosity models (For instance, Casson model) and also
another turbulence model (The k-ω turbulence model is
more suitable to treat the pulsatile flow through stenosis,
as we have already noticed it before). We have also to
study other parameters tied to this case, like the OSI index
especially. These are other limitations that will be resolved
in our future work.
Finally, from Table1, it is deserving to notice that the meshing
and geometry used in our computational simulations are
very satisfactory and provide us a good spatial resolution
since the number of nodes is superior to 260000 and the
number of tetrahedral elements is superior to 1500000.
CFD simulations, hemodynamic parameters profiles,
recirculation zone and indications
The profiles obtained from numerical simulation of blood
flow through our healthy aorta model, (Fig. 5-7), show
that the velocity of flow is growing in the aortic arch and
reaches its maximum at the bifurcation points where vortex
loops form, that the pressure increases as we move away
from the center of curvature of the aorta (This increase is
also visible in the vicinity of the branches of the arc), and
that shear forces are highly dynamic, but are generally
high along the outer wall near the branches and low along
the inner wall.
In the stenotic cases with increasing α, we observe that the
blood passes to high speeds and produces high gradient
velocity field, (See Fig. 8). Two vortices are developed at
both proximal regions of the stenosis, and a recirculation
zone, which is associated with the separation of the
boundary layer from the right internal wall, appears in
the post-stenotic region. We also observe that these
vortices and recirculation flow become more important
when α increases. The formation and propagation of the
recirculation zone is one of the most important phenomena
in the post-stenotic region, Mittal and al (2003).
Thus, in our work, we observe that, at a point on the
proximal downstream region of the stenosis, the flow
separates and a shear layer and recirculation zone are
formed. The length LR of the recirculation zone behind
the stenosis is used to better characterize the blood flow,
(See for instance, Griffith (2007), Chapter 5: “Steady Flow
Through an Axisymmetric Stenotic Geometry”, page101).
In fact, the recirculation length can be easily
E ISSN: 2349-7661 l P ISSN: 2394-7055
Moumen F et al.
determined by tracking the wall shear stress and
marking the z-location at which it changes sign. We explain
the fact that the recirculation length LR increases with
stenosis degree α by the following relation, Mittal and al
(2003):
where K is a constant (to be adjusted), D=2R is the non-
stenosed left subclavian artery diameter and Re = Re(α) is
the Reynolds number for some degree stenosis α, whose
modeling is given by, (See section 4):
Another important point of interest in this study is the
observation of a pressure drop in the stenotic left subclavian
artery which increases with the stenosis degree α, (See Fig.
9). From these profiles, we observe that the blood flow
first enters in the constriction and the increase in velocity
results in a steep drop in the pressure coefficient which
reaches its minimum at the stenosis site. Beyond this point,
the flow enters the expansion and the pressure coefficient
exhibits a steep increase.
However, because of this strong adverse pressure gradient,
the boundary layer on the wall separates just behind the
stenosis. Pressure losses generally become significant
only for stenoses whose degree is greater than 50% (See
Table1). Our findings show that the pressure drop at the
stenosis peak can reach nearly 50mmHg at the famous
critical value of stenosis degree.
Separation of flow can also contribute to pressure loss and
is a major factor at lower stenosis degrees. Thus, separation
will occur in most arterial stenoses. A strong shear layer
at the interface of the separation region and the central
jet creates additional viscous losses.
Figure 10 shows that the wall shear stress is maximal
at the throat of the stenosis. This will also contribute to
viscous losses. We deduce that, for low-degree stenoses,
the wall and separation shear layers produce most of the
losses. For high-degree stenoses, turbulence is the major
loss mechanism.
Reynolds number, behaviour of pressure, velocity and
WSS at the stenosis site
From Table1, we have tried to study the evolution of
the Reynolds number as well as the other hemodynamic
parameters (Pressure, velocity and WSS) at the center of
stenosis in function of the its severity measured through
its degree α.
Since we have focused our attention on the fully developed
turbulent behavior of the flow, we have been naturally
interested to find a modeling of the Reynolds number
E ISSN: 2349-7661 l P ISSN: 2394-7055
J. Adv. Res. Appl. Mech. Compu. Fluid Dyna. 2017; 4(3)
in function of α. It is given by Fig. 11. We find that Re
increases with α until nearly the value α = 90%, and then it
decreases. This means that the turbulent effects (creation
of vortex and formation of a recirculation zone) become
more and more important in both sides of the stenosis
(See Fig. 10). In this work, we have proposed a modeling of
this turbulence evolution in function of the severity of the
stenosis. This finding could be directly tied to the possible
rupture of plaque and then to an Atherosclerotic Vascular
Disease (AVD) which is the principal cause of myocardial
infarction (heart attack), stoke, ischemic heart pain and
sudden cardiac death.
It is well established actually from statistical results that
four primary regions (of low WSS) in the human body are
the most common sites of plaque formation: The coronary
arteries, the major branches of the aortic arch such as the
carotid and subclavian arteries, the visceral branches of
the abdominal aorta such as the renal arteries, and the
terminal branches of the abdominal aorta such as the
femoral arteries.
Such sites are of particular clinical interest since these
arteries supply the heart, the brain, and other organs
essential for homeostasis. Any anomaly that could disrupt
blood flow to such organs could cause significant morbidity
or mortality.
In our work, we have chosen the left subclavian artery
stenosis as an example, and we adopted for it the closest
position to the real case.
From Fig. 14, we see that the WSS increases with the
stenosis degree α until a critical value of α slightly greater
than 80%, and then it decreases. The same remark could
be made for the velocity and this could be easily explained
by Bernoulli equation, (See Fig. 13). In this situation, the
external pressure may be greater than the internal fluid
pressure, and the artery could really collapse if obviously
we consider that the walls are no longer rigid. Indeed, in
a physiological blood flow, artery walls are compliant and
respond to the fluid pressure and wall shear stresses.
Nevertheless, we note strangely that the velocity decreases
from this critical value of α. We consider this critical value
as a strong signal of an almost complete occlusion of the
artery from which the SSS begins.
Conversely, as expected, the pressure decreases until
the same critical value and then increases. Since for non-
vanishing values of α, the pressure is lower than the outlet
pressure, we conclude that we are we are in presence of
more and more important reverse flows.
Other findings in literature, conclusions and perspectives
From our extensive bibliographic research concerning
SAS or SSS caused by a SAS, we have found only articles
26
J. Adv. Res. Appl. Mech. Compu. Fluid Dyna. 2017; 4(3)
reporting some particular clinical cases described by means
of invasive or non-invasive medical tests (Ultrasound,
Angiography), or clinical treatment of SAS by means of
stenting or angioplasty and some articles dealing with CFD
simulations, Päivänsalo and al (1998), Pollard and al (1998),
Kliewer and al (2000), Schillinger and al (2002), Shadman
and al (2004), Ochoa and Yeghiazarians (2010), Sakima and
al (2011), Viduetsky (2011), Pandit and Lierbermann (2013),
Potter and Pinto (2014), Reyna and al (2014), Blanco and
al (2016), Salman and al (2016), Tascanov M.B. (2016), Wu
and al (2016), to name but some, and references therein].
In the other hand, the majority of references using CFD
simulation concern the stenosis present at some types of
arteries (coronary, carotid, renal etc…) other than the left
subclavian artery or simply a stenosed arterial segment,
generally considered to be cylindrical and straight. The
most recent article, Blanco and al (2016), uses the ADAN
model with pulsatile flow to simulate subclavian steal
and coronary-subclavian steal phenomena. The goal of
this work was to perform a computational study to gain
insight into the hemodynamic forces that drive blood flow
steal mechanisms caused by subclavian artery stenosis.
The authors presented a comparison between the ADAN
model predictions for different degrees of stenosis (80%,
90%, 95% and 99%) and velocity measurements (mean and
standard deviation, measured in cm/s) obtained from DUS
in the ipsilateral (left) and contralateral (right) vertebral
artery of controls, and of patients with subclavian steal,
Päivänsalo and al (1998).
The only comparison that can be made between their
results and ours is the fact that we agree with them in their
conclusions concerning the observation that, as the degree
of stenosis is increased, they first match the partial steal
condition and consistently approach the complete steal
condition as it is defined in Päivänsalo and al (1998). This
coincides with our finding of the famous critical value of
stenosis degree located between 80% and 90%.
Furthermore, comparing the range of values of the mean
velocity of blood flow through the left subclavian stenosis
obtained by DUS imaging, Pollard and al (1998), Sakima
and al (2011), Viduetsky (2011), Pandit and Lierbermann
(2013), with those obtained in our work (See Table 1), we
find a good agreement.
All these ascertainment lead us to conclude that our CFD
simulation of steady blood flow using our original aorta and
left subclavian stenosis geometries give excellent results.
We expect to improve our results by applying CFD simulation
with our geometries to study pulsatile blood flow (work
in progress) with well adapted physiological inlet/outlet
boundary conditions and taking into account the elasticity
of the walls. We also plan to deepen our study concerning
turbulence modeling, the critical value of stenosis degree
27
Moumen F et al.
and the behavior of hemodynamic parameters.
Currently there is no standardized procedure to grade the
physical severity of the stenosis. We hope that our work
will contribute to a good knowledge of hemodynamic
parameters, such as velocity, flow rate, wall shear stress
and pressure drop that will permit a good diagnosis and
surgical decision-making.
References
1. Ahmed S.A. and Giddens D.P. (1983), Velocity
measurements in steady flow through axisymmetric
stenoses at moderate Reynolds numbers, J. Biomech.
Vol.16, p505-516.
2. Ahmed S.A. and Giddens D.P. (1984), Pulsatile
poststenotic flow studies with laser Doppler
anemometry, J. Biomech. Vol.17, p695-705.
3. Akbar N.S. and Nadeem S (2014), Carreau fluid model
for blood flow through a tapered artery with a stenosis,
Ain Shems Engineering Journal 5, p1307–1316.
4. Berger S.A. and Jou L-D. (2000), Flows in stenotic
vessels, in Annual Review of Fluid Mechanics, Annual
Reviews, Palo Alto.
5. Blanco P.J. and al (2016), Computational modeling of
blood flow steal phenomena caused by subclavian
stenoses, J. Biomech. 49 (9), p1593–1600.
6. Bluestein D. and al (1997), Fluid Mechanics of Arterial
Stenosis: Relationship to the Development of Mural
Thrombus, Ann. Biomed. Eng. 25, p344–356.
7. Bluestein D and al (1999), Vortex Shedding in Steady
Flow Through a Model of an Arterial Stenosis and Its
Relevance to Mural Platelet Deposition, Ann. Biomed.
Eng. 27, p763–773.
8. Deshpande Mohan D. (1977), Steady Laminar and
turbulent flow through vascular stenosis models, PhD
Thesis, Georgia Institute of Technology.
9. Donnelly R. and London Nick J.M. (2009), Eds of ABC
of Arterial and venous disease, 2nd Edition, Wiley-
Blackwell, BMJ Books.
10. Effat M.A.  and al (2016), Clinical outcomes of
combined flow-pressure drop measurements using
newly developed diagnostic endpoint: Pressure drop
coefficient in patients with coronary artery dysfunction,
World J. Cardiol. 8(3), p283–292.
11. Gao F. and al (2011), Numerical Simulation in Aortic
Arch Aneurysm , DOI: 10.5772/18566, In book, chap.
12: Etiology, Pathogenesis and Pathophysiology of
Aortic Aneurysms and Aneurysm Rupture, edited by
Reinhart Grundmann.
12. Gaur M. and Gupta M. K. (2014), Steady slip blood
flow through a stenosed porous artery, Advances in
Applied Science Research 5(5), p249-259.
13. Gerhard-Herman M. and al (2006), Guidelines for
Noninvasive Vascular Laboratory Testing: A Report
from the American Society of Echocardiography and
E ISSN: 2349-7661 l P ISSN: 2394-7055
Moumen F et al.
the Society of Vascular Medicine and Biology, J. Am.
Soc. Echocardiogr. 19, p955-972.
14. Glass C.K. and Witztum J.L. (2001): Atherosclerosis:
The road ahead, Cell 104, p503-516.
15. Griffith M.D. (2007), The stability and behaviour of flows
in stenotic geometries, Doctorate Thesis, University of
Provence Aix-Marseille I, (2007), HAL Id: tel-00437443.
16. Griffith M.D. and al (2008), Steady Inlet Flow in Stenotic
Geometries: Convective and Absolute Instabilities,
Journal of Fluid Mechanics 616, p111-133.
17. Grigioni M. and al (2002), The role of wall shear stress
in unsteady vascular dynamics, Progress in Biomedical
Research Vol7, p204-212.
18. Haidekker M.A. (2013), Medical Imaging Technology,
Springer.
19. Hazarika G. C. and Sarmah A. (2014), Blood Flow
through a Stenosed Artery with the Effect of
Transverse Magnetic Field using a Non-Newtonian
Model, International Journal of Computer Applications
Vol.101– No.1, p5-8.
20. He X. and Ku D. N. (1996): Pulsatile flow in the human
left coronary artery bifurcation: average conditions, J.
Biomech. Eng. 118, p74.
21. Hiratzka L.F. and al (2010), Guidelines for the Diagnosis
and Management of Patients With Thoracic Aortic
Disease, Circulation 121, e266-e369, (see Figure 12).
22. Huda W. and Slone R. (2003), Review of Radiologic
Physics, 2nd edition. Lippincott Williams & Wilkins,
Chapter 11 – Ultrasound pp. 173-191.
23. Incropera F.P. and al (2007): Fundamentals of Heat
and Mass Transfer, John Wiley and sons, sixth edition.
24. Javadzadegan A. and al (2013), Flow recirculation
zone length and shear rate are differentially affected
by stenosis severity in human coronary arteries, Am.
.J Physiol. Heart Circ. Physiol. 304, H559–H566.
25. Jianjun Wu and al (2016), Endovascular treatment of
subclavian artery stenosis complicated with subclavian
artery aneurysms: a case report, Int. J. Clin. Exp. Med.
9(10), (p20401-20405.
26. Kartik J. (2016), Direct numerical simulation of
transitional pulsatile stenotic flow using Lattice
Boltzmann Method, Peer PrePrints, https://doi.
org/10.7287/peerj.preprints.1548v3, CC-BY 4.0 Open
Access.
27. Kliewer M.A. and al (2000), Vertebral artery Doppler
waveform changes indicating subclavian steal
physiology, American Journal of Roentgenology 174,
p815-819. Erratum in Am. J. Roentgenology 174(5),
p1464.
28. Ku, D.N. and al (1985), Pulsatile flow and artherosclerosis
in the human carotid bifurcation, Artherosclerosis 5,
p293-302.
29. Ku, D.N. (1997), Blood flow in arteries, Ann. Rev. Fluid
Mech. 29, p399-434.
30. Libby, P. (2003), Vascular biology of atherosclerosis:
overview and state of the art, Am. J. Cardiol. 91, p3-6.
E ISSN: 2349-7661 l P ISSN: 2394-7055
J. Adv. Res. Appl. Mech. Compu. Fluid Dyna. 2017; 4(3)
31. Loree H.M. and al (1991), Turbulent pressure
fluctuations on surface of model vascular stenoses,
American Journal of Physiology – Heart and Circulatory
Physiology Vol. 261 no.3, H644-650.
32. Lusis, A.J. (2000), Atherosclerosis, Nature 407,p233-
241, http://dx.doi.org/10.1038/35025203.
33. Mittal R. and al (2003), Numerical study of pulsatile
flow in a constricted channel. J. Fluid Mech. vol 485,
p337-378.
34. Moumen F. (2010), Etudes des effets turbulents de
l’écoulement sanguin dans la crosse aortique. Théorie
et simulation, Magister thesis, Univ. of Oran Es-sénia.
35. Moumen F. and Djemaï A.E.F. (2016), Computational
Fluid Dynamics Analysis of the Aortic Coarctation,
Natural Science 8, p271-283.
36. Mrdjen S. (2013), Basic Principles of Doppler
ultrasonography, European Congress of Radiology,
DOI: 10.1594/ecr2013/C-0363.
37. Ochoa V.M. and Yeghiazarians Y. (2010), Subclavian
artery stenosis: A review for the vascular medicine
practitioner, Vascular Medicine 16(1), p29–34.
38. Päivänsalo M. and al (1998), Duplex ultrasound in the
subclavian steal syndrome, Acta Radiol. 39, p183–188.
39. Pandit S. and Lierbermann G. (2013, October): To
Catch a Thief: Imaging of Subclavian Steal, Beth Israel
Deaconess Medical center, http://eradiology.bidmc.
harvard.edu/LearningLab/cardio/Pandit.pdf.
40. Perktold K. and al (1991a), Pulsatile non-Newtonian
blood flow in three-dimensional carotid bifurcation
models: a numerical study of flow phenomena under
different bifurcation angles, J. Biomech. Eng. 13(6).
41. Perktold K. and al (1991b), Pulsatile non-newtonian
flow characteristics in a three-dimensional human
carotid bifurcation model, Transactions of the ASME
113, p464–475.
42. Pollard H. and al (1998), Subclavian Steal Syndrome.
A Review, Australasian chiropractic and osteopathy
(ACO) Vol. 7(1), p20-28.
43. Potter B.J. and Pinto D.S. (2014), Subclavian Steal
Syndrome, Circulation 129, p2320-2323.
44. Reyna J. and al (2014), Subclavian Artery Stenosis:
A Case Series and Review of the Literature, Rev.
Cardiovasc. Med. 15(2), p189-195.
45. Sadeghi M.R. and al (2011), The effects of stenosis
severity on the hemodynamic parameters-assessment
of the correlation between stress phase angle and
wall shear stress, Journal of Biomechanics Vol.44,
p2614-2626.
46. Sakima H. and al (2011), Correlation between the
Degree of Left Subclavian Artery Stenosis and the
Left Vertebral Artery Waveform by Pulse Doppler
Ultrasonography, Cerebrovascular Diseases 31(1),
p64-67.
47. Salman R. and al (2016), Treatment of subclavian artery
stenosis: A case series, International Journal of Surgery
Case Reports 19, p69–74.
28
J. Adv. Res. Appl. Mech. Compu. Fluid Dyna. 2017; 4(3)
48. Schächinger V. and Zeiher A.M. (2002), Atherogenesis-
recent insights into basic mechanisms and their clinical
impact, Nephrol Dial Transplant 17, p2055–2064.
49. Schillinger M. and al (2002), Outcome of conservative
versus interventional treatment of subclavian artery
stenosis. J. Endovasc. Ther. 9, p139–146.
50. Shadman R. and al (2004), Subclavian Artery Stenosis:
Prevalence, Risk Factors, and Association With
Cardiovascular Diseases, Journal of the American
College of Cardiology 44 (3), p618–623.
51. Sherwin S.J. and Blackburn H.M. (2005), Three-
dimensional instabilities and transition of steady and
pulsatile axisymmetric stenotic flows, Journal of Fluid
Mechanics 533, p297-327.
52. Sloop G.D. and al (2015), The Interplay of Aging, Aortic
Stiffness and Blood Viscosity in Atherogenesis, Journal
of Cardiology and Therapy 2(4), p350-354.
53. Sultanov R.A. and al (2008), 3D Computer Simulations
of Pulsatile Human Blood Flows in Vessels and in
the Aortic Arch: Investigation of Non-Newtonian
Characteristics of Human Blood, arXiv:0802.2362v1.
54. Sultanov R.A. and Guster D. (2008), Computer
Simulations of Pulsatile Human Blood Flow Through
3D-Models of the Human Aortic Arch, Vessels of Simple
Geometry and a Bifurcated Artery: Investigation of Blood
Viscosity and Turbulent Effects, arXiv:0811.1363v1.
29
Moumen F et al.
55. Tascanov M.B. and al (2016), Cardiac Arrest in a Patient
with Critical Left Subclavian Artery Stenosis, Clinics in
Surgery - Cardiovascular Surgery Vol.1, Article 1165, p1.
56. Varghese S.S. and al (2007a), Direct numerical
simulation of stenotic flows. Part1. Steady flow, Journal
of Fluid Mechanics Vol.582, p253-280.
57. Varghese S.S. and al (2007b), Direct numerical
simulation of stenotic flows. Part2. Pulsatile flow,
Journal of Fluid Mechanics Vol.582, p281-318.
58. Viduetsky A. (2011), Subclavian Steal Syndrome:
Diagnostic Imaging Correlations and Review
of Literature, Journal of Diagnostic Medical
Sonography Vol.14.
59. Wittek A. and al (2013), Ed. Computational Biomechanics
for Medicine: Models, Algorithms and Implementation,
Springer.
60. World Health Organization, (2014), The top 10 causes
of death, (Accessed March 26, 2015), http://www.who.
int/mediacentre/factsheets/fs310/en/
61. Wu J. and al (2016), Endovascular treatment of
subclavian artery stenosis complicated with subclavian
artery aneurysms: a case report, Int. J. Clin. Exp. Med.
9(10), p20401-20405.
62. Young, D.F. (1979), Fluid mechanics of arterial stenosis,
J Biomech. Eng. 101, p157-173.
E ISSN: 2349-7661 l P ISSN: 2394-7055