FLUIDS MODULE

BME 429

Alexander Cook
Rebecca Nagurney

Section 2 Group 8
I.

Introduction
Atherosclerosis is a cardiovascular disease which affects blood flow through the arteries. Over
time, artery walls get blocked with fatty substances called plaque which narrow the area (also
called a stenosis) which blood can move through. Because of this, people who suffer from
atherosclerosis can suffer from a heart attack, stroke, kidney failure, coronary heart disease, etc.
which can be causes of death.1
Studies of fluid mechanics and dynamics dictates that through a stenosis, pressure drops and
velocity increases.2 Blood flow through a stenosis will also create recirculation zones which can
lead to thrombosis and atherogenesis.3 Thrombosis is when a blood clot develops in the artery.
When the blood clot is released into the blood stream, it can eventually travel to the heart, lungs,
or the brain and result in life threatening conditions. Because the flow of blood through the
arteries is so important and can result in many life threatening conditions, scientists and engineers
around the world are interested in studying flow patterns. Studying blood flow patterns and
understanding the influences of shear force, velocity, and pressure on the blood flow can lead to
new health treatments, implants such as heart valves, and surgical procedures.
Reynolds number is a numerical dimensionless value that can be calculated to determine if flow is
laminar or turbulent. A Reynolds number less than 2000 would be considered a laminar flow field
and a Reynolds number greater than 4000 would be considered a turbulent flow field. When
calculating Reynolds number, it is important to remember that it is directly proportional to the flow
velocity. Therefore, the higher the degree of stenosis results in a higher flow velocity and a higher
Reynolds number.4
Particle image velocimetry (PIV) is a way to measure flow through an area. This technique uses a
fluid and particles (known as tracers) to image of the fluid through a flow loop. The fluid that is
being used in the flow loop must be seeded with the tracer particles just enough to see the flow
but not too much or too little that they block the flow images or are not enough to image the flow.
A laser beam is placed at the desired imaging area and shined into the flow. Proper eye safety
procedure must be followed to prevent damage to the eyes. Results of the PIV testing gives images
of instantaneous measured planes that have vectors showing the direction of flow whether the
flow is steady or pulsatile.5
In the fluids lab module, groups used PIV to study the flow of a 45% glycerin solution through a
hemoflow system. The glycerin solution simulated the blood flowing through the carotid artery.
The data collected using this system was obtained through the FlowEX software system. The
experimental data was compared to computer simulated data of normal and stenotic flow through
the carotid artery models developed using the given SolidWorks software files.6
The goals of this fluids laboratory module is to learn to use and build a flow loop, use particle image
velocimetry techniques, and obtain data on fluid flow through a stenosis. The flow loop built in the
first half of the fluids module was to learn how to calibrate a pressure transducer and obtain
pressure measurements at different flow rates at the upstream, middle, and downstream locations

Section 2 Group 8

on a stenosis. The second half of the fluids module experiments was dedicated to obtaining PIV
data through the hemoflow system. The obtained data was analyzed and compared to COMSOL
simulations to determine if the experimental results are reasonable and if they follow the laws of
fluid mechanics. Based on the experimental and computer simulated data results, it is expected
that groups will have a better understanding of stenotic flow and diseases that are associated with
the reduced blood flow through the arteries.
II.

Methods and Materials

a. Alex and Rebecca
Two main experiments were completed. In the first one, a pulsatile flow loop with a symmetric
stenosis was constructed. Pressure measurements were taken at three taps along the stenosis to
observe any changes that occurred. In the second experiment, a commercial hemoflow loop was
used to study both steady and pulsatile flow in a stenotic carotid artery. Particle image velocimetry
(PIV) was used to establish the flow field for each set of parameters. All experiments were
conducted in a standard laboratory at room temperature ( 25 C). Tap water was used as the fluid
in the stenotic model and a 45 % glycerin solution was used in the PIV experiments (due to its
mechanical similarity to blood). All experimental methods were based on the protocols by
Manning.6,7
For the pulsatile flow loop, a circuit was constructed to supply power to the pressure transducer
(Ashcroft, Stratford, CT) and output the data to the computer for analysis. A DC power supply of 10
V and a resistor of 200 were used. The output of the pressure transducer was connected to analog
ports on the USB-6008 data acquisition (DAQ) board (National Instruments, Austin, TX). LabVIEW
SignalExpress (National Instruments, Austin, TX) was used to plot the current output as a function of
time. 10 000 data points were collected at a sample rate of 500 Hz, giving 20 seconds of pulsatile
pressure data for analysis.
Using this setup, the pressure transducer was calibrated in order to convert the current
measurements to pressure readings. The output current across the resistor was measured for
various heights of water in a narrow tube. Because pressure in a fluid is directly proportional to the
depth (through P = gh, where P is the pressure in Pascals, is the density of the fluid in kg/m3, g is
the acceleration due to gravity, and h is the height of the fluid in m), pressure was determined as a
function of current. This calibration curve was used to convert the current versus time data in
LabVIEW to pressure versus time data.
A simple pulsatile flow loop and motor setup was constructed. A five-gallon reservoir tank was
connected to the input of the artificial heart via PVC tubing. The plastic stenosis was connected to
the output of the heart and back to the reservoir tank. A pulsatile motor was connected to a latex
diaphragm to create the flow of fluid in the loop.
This setup was tested under two total flow rates: 1 L/min and 4 L/min. The beats per minute (BPM)
and stroke volume (SV) of the pulsatile motor were adjusted using the computer. The volumetric
flow rates were confirmed by measuring the output into the reservoir during a certain time interval.
For each flow rate, pressure measurements were taken upstream, downstream, and in the middle of
the stenosis according to the dimensions in Figure 1. MATLAB (MathWorks, Natick, MA) was used to

Section 2 Group 8
calculate the maximum pressure, minimum pressure, and average pressure for a number of cardiac
cycles. The mean pressure values were then calculated for each flow rate and port.

Figure 1: Dimensions of the plastic stenosis used for measuring pressure upstream, downstream, and in the middle

For the second experiment, a hemoflow loop system (Interactive Flow Studies Corporation, Billings,
MT) was studied using particle image velocimetry (PIV). The reservoir tank was filled with a 45 %
glycerin solution to simulate the mechanical properties of blood. Polyamid seed was mixed into the
solution to be used as tracer particles during the PIV measurements. The flow loop included the
carotid artery with a 43 % stenosis, the primary focus of the analysis. Figure 2 depicts the geometry
of the loop.

Figure 2: The geometry of the carotid artery used for the particle image velocimetry (PIV) experiments

Section 2 Group 8

The computer, laser, and camera settings were adjusted to provide the optimal parameters for
collection and analysis of data. The camera was run at 25.5 frames per second to image the flow.
The delta T within TU was adjusted to 1/67.3 for steady flow (condition one) and to 1/25.5 for
pulsatile flow (conditions two and three). The laser power was set to 1000 mW, and was pulsed at a
rate of 67.3 Hz to illuminate the flow field.
Three flow conditions were tested. A steady-state model was constructed by setting the motor to 0
beats per minute (BPM). For this condition, the flow rate was 4 L/min, divided evenly between
ventricular and carotid flow. Two pulsatile models were run at 75 BPM. Each one had a total flow
rate of 4 L/min, but had different divisions of the flow. The first model had a 3.5 L/min ventricular
flow rate and a 0.5 L/min carotid flow rate. The second pulsatile model had a 2 L/min flow rate for
both the ventricular and carotid tubes.
The resulting video was analyzed by pairs of frames. The interrogation window size and
interrogation window shift were set to 50 and 25 respectively to allow sufficient analyzation of the
microflows without requiring too much computing power. A sigma multiplier value of 3 was used to
filter out interfering light and shadows. 2-D plots of the velocity vectors were generated for the
entire duration of the videos captured. For pulsatile flow, the peaks and valleys of the velocity flow
field were isolated to characterize systole and diastole respectively. Plots were constructed to
provide a visual representation of this data. Additionally, data from the hemoflow systems onboard
computer were collected.

b.

Alex

There were many important simplifications that went into constructing the COMSOL models from
their real-world counterparts. For example, the plastic stenosis was constructed through the use of
cylinders and cones of the appropriate lengths and heights. Two 10-inch long cylinders were placed
before and after the stenotic region to minimize entrance and exit effects of the fluid. Water, a
built-in material in COMSOL, was used as the liquid to mirror the experiments.
The Reynolds number of each flow condition was calculated to determine the appropriate physics
to apply. Table 1 shows the results. For an inlet flow rate of 1 L/min for the stenosis, the Reynolds
number (920) is well below the threshold when turbulence becomes a significant factor (R 3500).
Therefore, a laminar flow model was chosen. The main equations for laminar flow are conservation
of mass and momentum:

For a flow rate of 4 L/min into the stenosis, the Reynolds number was close enough to the threshold
(3600 3500) to warrant the use of a turbulent model. COMSOL offers several choices to model

Section 2 Group 8

turbulence: the Algebraic yPlus model was chosen because it is relatively simple. Since the Reynolds
number is on the borderline between laminar and turbulent, it was assumed that a straightforward
model of turbulence would capture the most important details of the pressure and velocity.
There were no initial conditions set for both the steady-state and pulsatile models. Since the
stationary models do not depend on time, this was a reasonable justification. For the pulsatile
models, the inlet velocity condition was determined by a sine wave as a function of time; therefore,
initial conditions were unnecessary. The boundary conditions were always set as an inlet velocity
(constant for stationary models and a sine wave for pulsatile models), no-slip wall, and a pressure
outlet of zero. The amplitude of the velocity waveforms was determined from the steady-state inlet
velocity. The sine wave was tailored to oscillate around this value, reaching zero once every cycle to
simulate pulsing. The frequency of the sine waves was determined by the beats per minute of the
motor for each experiment.
The meshes were generally chosen based on reasonable accuracy of the results without taking
too much time to solve. Because some of these models were fairly complicated, this often involved
using coarse to extremely coarse meshes. The time-dependent stenotic models were run for 3.3 s,
which gave data for approximately five cardiac cycles.

Table 1: Calculated values of the Reynold's numbers for the four different conditions studied to evaluate turbulence

Model and flow conditions

Stenosis, 1 L/min
Stenosis, 4 L/min
Hemoflow, 0.5 L/min
Hemoflow, 2 L/min

Reynolds number
920
3600
180
720

The hemoflow models utilized most of the same principles as the stenotic models. The geometry
was obtained from pre-packaged SolidWorks files to model both the normal carotid and the 43 %
stenotic version. A steady (for stationary models) or pulsatile (for time-dependent models) velocity
inlet was defined as a boundary condition. Again, the walls were no slip. Finally, a zero-pressure
outlet was defined (except in this case, there were two surfaces through which the fluid was exiting.
Instead of water, the material properties of blood were used to mimic the 45 % glycerin solution
used in the experiments.8,9

Section 2 Group 8

6
Table 2: Alex's steady and pulsatile stenosis models at a flow rate of 1 L/min

Define the space

dimension

Define the physics

Define the study

type

Start COMSOL
Multiphysics
Under space dimension,
select 3D

Under select physics,

double-click Fluid Flow
Double-click SinglePhase Flow
Select Laminar Flow
(spf) and click the blue
arrow to continue
For a flow rate of 1
L/min, the Reynolds
number is small enough
(R 920 < 3500) to justify
the use of a laminar flow
model

Steady-state flow
A steady-state flow has
no time-dependent
parameters, so a
stationary study was
selected
Under select study, click
Stationary
Click on the checkered
flag to finish
Pulsatile flow
For the pulsatile flow, a
time-dependent model
was used to simulate the
varying inlet velocity

Define the
geometry

The dimensions of the

plastic stenosis can be

Section 2 Group 8

Define the material

type

seen in the schematic

diagram
This geometry was
constructed using
cylinders and cones of
the appropriate lengths
and diameters
Two 10-in. long cylinders
were placed before and
after the stenosis to
mitigate entrance and
exit effects of the fluid
The finalized geometry,
including a detailed view
of the stenosis, can be
seen in the
accompanying images

Right-click on Materials
and select Open
Material Browser
Double-click Built-In
and select Water,
liquid
Select the entire
geometry to add the
material to the COMSOL
model

Section 2 Group 8
Physical settings:
initial conditions

Physical settings:
boundary
conditions

8
Steady-state flow
Since this model was
stationary, there were no
relevant initial values to
consider
The temperature was set
to 298 K under
FluidProperties1 to
model the conditions of
the laboratory during the
experiments (close to
room temperature)
Pulsatile flow
There were no additional
initial conditions to
consider for the pulsatile
model
The velocity waveform as
a function of time is
explained under
Physical settings:
boundary conditions
Steady-state flow
Click on Wall1 and
ensure the boundary
condition is set to No
slip
Right-click Laminar
Flow and select Inlet
Under Inlet1, select
Velocity under
Boundary Condition
Click the Normal inflow
velocity radio button
and enter U0 = 3.3 cm/s
Right-click Laminar
Flow and select Outlet
Under Outlet1, select
Pressure under
Boundary Condition and
enter p0 = 0 atm (this
measures the pressures
relative to zero)
Pulsatile flow
The data contained 31
pulses over 20 seconds:
therefore, this correlated
to 93 BPM, or 1.55 Hz
A sine wave was created
using this frequency and

Section 2 Group 8

Mesh

an amplitude of 0.033
m/s (average velocity)
Set the sine wave as the
inlet velocity field with an
offset of 0.033 m/s
The Wall1 and
Outlet1 boundary
conditions were the
same as those used for
the steady-state model

Steady-state flow
Click on Mesh 1
Under Element size,
select Coarser
Click Build All
Number of elements:
55362

Section 2 Group 8

10
Pulsatile flow
Click on Mesh 1
Under Element size,
select Extremely coarse
Click Build All
Number of elements:
10340

Study: solve

Compute

Steady-state flow
Click on Study 1
(Stationary) and hit
Compute
Pulsatile flow
Click on Study 1 (Time
Dependent)
Set the following
conditions under times:
range(0,0.1,3.3)
These time parameters
will give data for
approximately five
cardiac cycles
Steady-state flow
Solution time: 55 s
Number of degrees of
freedom solved for:
67864
Pulsatile flow
Solution time: 266 s
Number of degrees of
freedom solved for:
15576

Section 2 Group 8

11
Table 3: Alex's steady and pulsatile stenosis models at 4 L/min

Define the space

dimension

Define the physics

Define the study

type

Start COMSOL
Multiphysics
Under space dimension,
select 3D

Under select physics,

double-click Fluid Flow
Double-click SinglePhase Flow
Double-click Turbulent
Flow and select
Turbulent Flow,
Algebraic yPlus
For a flow rate of 4
L/min, the Reynolds
number is large enough
(R 3600) to justify the
use of a turbulent flow
model
Steady-state flow
A steady-state flow has
no time-dependent
parameters, so a
stationary study was
selected
Under select study, click
Stationary
Click on the checkered
flag to finish
Pulsatile flow
For the pulsatile flow, a
time-dependent model
was used to simulate the
varying inlet velocity

Define the
geometry

The dimensions of the

plastic stenosis can be

Section 2 Group 8

12

Define the material

type

seen in the schematic

diagram
This geometry was
constructed using
cylinders and cones of
the appropriate lengths
and diameters
Two 10-in. long cylinders
were placed before and
after the stenosis to
mitigate entrance and
exit effects of the fluid
The finalized geometry,
including a detailed view
of the stenosis, can be
seen in the
accompanying images

Right-click on Materials
and select Open
Material Browser
Double-click Built-In
and select Water,
liquid
Select the entire
geometry to add the
material to the COMSOL
model

Section 2 Group 8
Physical settings:
initial conditions

Physical settings:
boundary
conditions

13
Steady-state flow
Since this model was
stationary, there were no
relevant initial values to
consider
The temperature was set
to 298 K under
FluidProperties1 to
model the conditions of
the laboratory during the
experiments (close to
room temperature)
Pulsatile flow
There were no additional
initial conditions to
consider for the pulsatile
model
The velocity waveform as
a function of time is
explained under
Physical settings:
boundary conditions
Steady-state flow
Click on Wall1 and
ensure the boundary
condition is set to No
slip
Right-click Turbulent
Flow and select Inlet
Under Inlet1, select
Velocity under
Boundary Condition
Click the Normal inflow
velocity radio button
and enter U0 = 13 cm/s
Right-click Turbulent
Flow and select Outlet
Under Outlet1, select
Pressure under
Boundary Condition and
enter p0 = 0 atm (this
measures the pressure
relative to zero)
Pulsatile flow
The data contained 31
pulses over 20 seconds:
therefore, this correlated
to 93 BPM, or 1.55 Hz
A sine wave was created
using this frequency and

Section 2 Group 8

14

Mesh

an amplitude of 0.13 m/s

(average velocity)
Set the sine wave as the
inlet velocity field with an
offset of 0.13 m/s
The Wall1 and
Outlet1 boundary
conditions were the
same as those used for
the steady-state model

Steady-state flow
Click on Mesh 1
Under Element size,
select Extremely
Coarse
Click Build All
Number of elements:
44730
Pulsatile flow
The mesh from the
steady-state model was
also used for the pulsatile
simulation

Section 2 Group 8

15

Study: solve

Steady-state flow
Click on Study 1
(Stationary) and hit
Compute
Pulsatile flow
Click on Study 1 (Time
Dependent)
Set the following
conditions under times:
range(0,0.1,3.3)
These time parameters
will give data for
approximately five
cardiac cycles
Steady-state flow
Solution time: 12 s
Number of degrees of
freedom solved for:
105864

Compute

Pulsatile flow
Solution time: 356 s
Number of degrees of
freedom solved for:
105864

Table 4: Alex's normal and stenotic carotid artery in a steady flow of 2 L/min

Define the space

dimension

Define the physics

Start COMSOL
Multiphysics
Under space dimension,
select 3D

Under select physics,

double-click Fluid Flow
Double-click SinglePhase Flow
Select Laminar Flow
(spf) and click the blue
arrow to continue

Section 2 Group 8

Define the study

type

16

For a flow rate of 2

L/min, the Reynolds
number is small enough
(R 720 < 3500) to justify
the use of a laminar
incompressible flow
model

A steady-state flow has

no time-dependent
parameters, so a
stationary study was
selected
Under select study, click
Stationary
Click on the checkered
flag to finish

Define the
geometry

The geometry of the

carotid artery was
obtained from a
SolidWorks file
Two geometries were
utilized: one with a
normal carotid artery and
one with a 43 % stenosis
Click on Geometry1
and change the length
unit to inches
Right-click on
Geometry1 and select
Import
Ensure the length unit is
specified From
COMSOL
The normal and stenotic
geometries can be seen
in the xy-plane in the
accompanying images
(top: normal, bottom:
stenotic)

Section 2 Group 8

Define the material

type

17

Physical settings:
initial conditions

Right-click on Materials
and select Material
Under Material1,
define the dynamic
viscosity (3.5 mPas) and
density (1060 kg/m3) of
blood
Add this material to the
entire geometry of the
model
Since these models are
stationary, there were no
relevant initial values to
consider

Section 2 Group 8
Physical settings:
boundary
conditions

18

Click on Wall1 and set

the boundary condition
to No slip
Right-click Laminar
Flow and select Inlet
Under Inlet1, select
Velocity under
boundary condition
Click the Normal inflow
velocity radio button
and enter U0 = 13.6 cm/s
Right-click Laminar
Flow and select Outlet
Under Outlet1, select
Pressure under
boundary condition
Enter p0 = 0 atm (this
measures the pressure
relative to zero)

Section 2 Group 8
Mesh

19

Click on Mesh 1
Under Element size,
select Coarser
Click Build All
Number of elements for
normal carotid: 29966
Number of elements for
stenotic carotid: 29357

Study: solve

Compute

Normal Carotid
Solution time: 119 s
Number of degrees of
freedom solved for:
39552

Click on Study 1
(Stationary) and hit
Compute

Stenotic Carotid
Solution time: 106 s
Number of degrees of
freedom solved for:
39084

Section 2 Group 8

20

Table 5: Alex's normal and stenotic carotid artery at a pulsatile flow rate of 0.5 L/min

Define the space

dimension

Define the physics

Define the study

type

Define the
geometry

Start COMSOL
Multiphysics
Under space dimension,
select 3D

Under select physics,

double-click Fluid Flow
Double-click SinglePhase Flow
Select Laminar Flow
(spf) and click the blue
arrow to continue
For a flow rate of 2
L/min, the Reynolds
number is small enough
(R 180 < 3500) to justify
the use of a laminar
incompressible flow
model
A pulsatile flow varies
with time, so a timedependent study was
selected
Under select study, click
Time Dependent
Click on the checkered
flag to finish
The geometry of the
carotid artery was
obtained from a
SolidWorks file
Two geometries were
utilized: one with a
normal carotid artery and
one with a 43 % stenosis
Click on Geometry1
and change the length
unit to inches

Section 2 Group 8

21

Define the material

type

Physical settings:
initial conditions

Right-click on
Geometry1 and select
Import
Ensure the length unit is
specified From
COMSOL
The normal and stenotic
geometries can be seen
in the xy-plane in the
accompanying images
(top: normal, bottom:
stenotic)

Right-click on Materials
and select Material
Under Material1,
define the dynamic
viscosity (3.5 mPas)8 and
density (1060 kg/m3)9 of
blood
Add this material to the
entire geometry of the
model
A sine wave was used to
define the inlet velocity
as a function of time
Therefore, no additional
initial conditions had to
be set (they were left as
zero)
See Physical settings:
boundary conditions for
a complete description of
the function used to
define the inlet velocity

Section 2 Group 8
Physical settings:
boundary
conditions

22

A sine wave was used to

model the pulsatile inlet
velocity
Right-click Global
Definitions and choose
Functions >>
Waveform
The frequency of the
waveform was 1.25 Hz
(75 BPM), with an
amplitude of 0.034 m/s
(steady-state inlet
velocity)
Click on Wall1 and set
the boundary condition
to No slip
Right-click Laminar
Flow and select Inlet
Under Inlet1, select
Velocity under
boundary condition
Click the Normal inflow
velocity radio button
and input the previously
defined sine wave with
an offset of 0.034
Right-click Laminar
Flow and select Outlet
Under Outlet1, select
Pressure under
boundary condition
Enter p0 = 0 atm (this
measures the pressure
relative to zero)

Section 2 Group 8
Mesh

23

Study: solve

Compute

Click on Mesh 1
Under Element size,
select Coarser
Click Build All
Number of elements for
normal carotid: 30604
Number of elements for
stenotic carotid: 29357

Click on Study 1 (Time

Dependent)
Set the range of time
values to 04 s with a
step of 0.1 s (this
captures five cardiac
cycles)
Hit Compute

Normal Carotid
Solution time: 861 s
Number of degrees of
freedom solved for:
39552
Stenotic Carotid
Solution time: 978 s
Number of degrees of
freedom solved for:
39084

Section 2 Group 8

24
Table 6: Alex's normal and stenotic carotid artery at a pulsatile flow rate of 2 L/min

Define the space

dimension

Define the physics

Define the study

type

Define the
geometry

Start COMSOL
Multiphysics
Under space dimension,
select 3D

Under select physics,

double-click Fluid Flow
Double-click SinglePhase Flow
Select Laminar Flow
(spf) and click the blue
arrow to continue
For a flow rate of 2
L/min, the Reynolds
number is small enough
(R 720 < 3500) to justify
the use of a laminar
incompressible flow
model
A pulsatile flow varies
with time, so a timedependent study was
selected
Under select study, click
Time Dependent
Click on the checkered
flag to finish
The geometry of the
carotid artery was
obtained from a
SolidWorks file
Two geometries were
utilized: one with a
normal carotid artery and
one with a 43 % stenosis
Click on Geometry1
and change the length
unit to inches

Section 2 Group 8

25

Define the material

type

Physical settings:
initial conditions

Right-click on
Geometry1 and select
Import
Ensure the length unit is
specified From
COMSOL
The normal and stenotic
geometries can be seen
in the xy-plane in the
accompanying images
(top: normal, bottom:
stenotic)

Right-click on Materials
and select Material
Under Material1,
define the dynamic
viscosity (3.5 mPas) and
density (1060 kg/m3) of
blood
Add this material to the
entire geometry of the
model
A sine wave was used to
define the inlet velocity
as a function of time
Therefore, no additional
initial conditions had to
be set (they were left as
zero)
See Physical settings:
boundary conditions for
a complete description of
the function used to
define the inlet velocity

Section 2 Group 8
Physical settings:
boundary
conditions

26

A sine wave was used to

model the pulsatile inlet
velocity
Right-click Global
Definitions and choose
Functions >>
Waveform
The frequency of the
waveform was 1.25 Hz
(75 BPM), with an
amplitude of 0.136 m/s
(steady-state inlet
velocity)
Click on Wall1 and set
the boundary condition
to No slip
Right-click Laminar
Flow and select Inlet
Under Inlet1, select
Velocity under
boundary condition
Click the Normal inflow
velocity radio button
and input the previously
defined sine wave with
an offset of 0.136
Right-click Laminar
Flow and select Outlet
Under Outlet1, select
Pressure under
boundary condition
Enter p0 = 0 atm (this
measures the pressure
relative to zero)

Section 2 Group 8
Mesh

27

Study: solve

Compute

Click on Mesh 1
Under Element size,
select Coarser
Click Build All
Number of elements for
normal carotid: 29966
Number of elements for
stenotic carotid: 29357

Click on Study 1 (Time

Dependent)
Set the range of time
values to 04 s with a
step of 0.1 s (this
captures five cardiac
cycles)
Hit Compute

Normal Carotid
Solution time: 1605 s
Number of degrees of
freedom solved for:
39552
Stenotic Carotid
Solution time: 1954 s
Number of degrees of
freedom solved for:
39084

Section 2 Group 8

28

c. Rebecca
Table 17 shows the pressure measurements (mmHg) taken at the upstream, middle, and downstream
measurements on the stenosis in the flow loop for part 1 of the fluids module. As seen in the table, the
average pressure measurements go from high (upstream) to low (downstream) for the 1 LPM and 4 LPM
flow rates. Figure 3 shows the calibration curve for the pressure transducer that was being used in the
experimental flow loop to take the pressure measurements at the different stenotic pressure taps.
Each of the following simulations see in Table 8 through Table 17 were developed using SolidWorks and
COMSOL software. The tables provide an overview of the process of building the stenosis models, the
materials being simulated, and the conditions provided to make the simulations run and produce flow
field velocity (m/s) and pressure (mmHg) results. The initial and boundary conditions were obtained
from experimental data conditions provided in the Hemoflow Protocol provided on Angel. Because the
experimental models have different flow rates in liters per minute (LPM), the values needed to be
converted (m/s) and the material values for blood and water were used to calculate the Reynolds
number for each of the systems. For the models using 1 LPM flow of water, the calculated Reynolds
number was 942 which could be considered laminar flow (Table 8 and Table 10). The stenosis models
using 4 LPM flow of water had a Reynolds number of 3,710 which is high enough to be considered
turbulent (Table 9 and Table 11).
The second half of the fluid module focused on fluid flow through a normal and a 43% stenotic carotid
model (Table 12 through Table 17). The normal and stenotic carotid models used conditions 2, 3 and 4
in the hemoflow protocol and the simulations were developed using the material properties of blood.
Table 11 and Table 14 used condition 2 with a 0.5 LPM flow rate. The Reynolds number for those
models was calculated to be 70 which is considered a laminar flow field (Table 12 and Table 15). The
remaining tables for the normal and stenotic models for conditions 2 and 3 used a flow rate or 2 LPM.
The Reynolds number for those models was calculated to be 276 which is also a laminar flow field (Table
13, Table 14, Table 16, and Table 17).
160
140

y = 49705x - 199.57

Pressure (mmHg)

120
100
80
60
40
20
0
0

0.001

0.002

0.003

0.004

0.005

0.006

Current (A)
Figure 3: Pressure transducer calibration curve

0.007

0.008

Section 2 Group 8

29

Table 7: Rebecca's Steady Stenosis Model 1 LPM

Space
Dimension

Open COMSOL
program, select model
wizard and then select
3D.

Physics

Use the down arrow to

open the options for
Fluid Flow physics.
Select the drop down
arrow for Single-Phase
flow, click Laminar
Flow, click add and
then study.

Section 2 Group 8

30

Study Type

Use the drop down

arrow to expand the
Preset Studies section.
Select Stationary and
then click Done.

Geometry

The geometry was

created using multiple
planes in solid works
and lofting to each
plane. The geometry
for the stenosis was
given to us. All of the
dimensions for the
stenosis were given in
inches. Each end of the
stenosis was extended
8 inches to prevent
entrance and exit
effects. The
SolidWorks file was
saved as Parasolid and
imported into
COMSOL. Finally, the
Build Selected was

Section 2 Group 8

31
used to build the
model in COMSOL.

Material Type

The entire model was

selected as the fluid
(Water) to simulate the
stenosis.

Section 2 Group 8

32

Section 2 Group 8
Physical
The inlet and outlet of
Settings/Initial the stenosis was set on
Conditions
the model. In the inlet,
the flow rate was set as
1 liter per minute with
a zero pressure and an
entrance length of 1
meter to offset any
entrance effects. The
outlet was set to
atmospheric pressure
and it also had an
entrance length set to
offset any exit effects.

33
Inlet:

Outlet:

Section 2 Group 8

34

Section 2 Group 8

35

Mesh

An extra coarse mesh

was used for the
model. Using anything
smaller than extra
course caused the
simulation to be too
much for the computer
to handle.

Study: Solve

This model was solved

using a stationary study
of laminar flow.

Compute

Click the compute

button under the study
section

Table 8: Rebeccas Steady Stenosis Model 4 LPM

Space
Dimension

Open COMSOL
program, select model
wizard and then select
3D.

Section 2 Group 8
Physics

Study Type

Use the down arrow to

open the options for
Fluid Flow physics.
Select the drop down
arrow for Single-Phase
flow, click Turbulent
Flow, Turbulent Flow,
Algebraic yPlus (spf)
click add and then
study.
Use the drop down
arrow to expand the
Preset Studies section.
Select Stationary and
then click Done.

36

Section 2 Group 8
Geometry

The geometry was

created using multiple
planes in solid works
and lofting to each
plane. The geometry
for the stenosis was
given to us. All of the
dimensions for the
stenosis were given in
inches. Each end of the
stenosis was extended 8
inches to prevent
entrance and exit
effects. The SolidWorks
file was saved as
Parasolid and imported
into COMSOL. Finally,
the Build Selected was
used to build the model
in COMSOL.

37

Section 2 Group 8

Material Type

The entire model was

selected as the fluid
(water) to simulate the
stenosis.

38

Section 2 Group 8

Physical
Settings/Initi
al Conditions

The inlet and the outlet

of the model was
selected. The inlet was
set to mass flow rate
and the outlet was set

39

Inlet:

Section 2 Group 8

40

to atmospheric
pressure.

Outlet:

Section 2 Group 8

Mesh

An extra coarse mesh

was used for the model.
Using anything smaller
than extra course
caused the simulation to
be too much for the
computer to handle.

Study: Solve

This model was solved

using stationary study
with a turbulent flow.
Click the compute
button under the study
section

Compute

Table 9: Rebecca's Pulsatile Stenosis Model 1 LPM

Space
Dimension

Open COMSOL
program, select model
wizard and then select
3D.

41

Section 2 Group 8
Physics

Use the down arrow to

open the options for
Fluid Flow physics.
Select the drop down
arrow for Single-Phase
flow, click Laminar Flow,
click add and then
study.

42

Section 2 Group 8
Study Type

Use the drop down

arrow to expand the
Preset Studies section.
Select Time Dependent
and then click Done.

Geometry

The geometry was

created using multiple
planes in solid works
and lofting to each
plane. The geometry
for the stenosis was
given to us. Each end of
the stenosis was
extended 8 inches to
prevent entrance and
exit effects. All of the
dimensions for the
stenosis were given in
inches. The SolidWorks
file was saved as
Parasolid and imported
into COMSOL. Finally,
the Build Selected was
used to build the model
in COMSOL.

43

Section 2 Group 8

Material Type

The entire model was

selected as the fluid
(water) to simulate the
stenosis.

44

Section 2 Group 8

45

Section 2 Group 8
Physical
Settings/Initi
al Conditions

The inlet and the outlet

of the model was
selected. The inlet was
set to a velocity field in
the z direction and the
outlet was set to
atmospheric pressure.

46
Inlet:

Outlet:

Section 2 Group 8

Mesh

An extra coarse mesh was

used for the model.
Using anything smaller
than extra course caused
the simulation to be too
much for the computer to
handle.

Study: Solve

This model was solved

using a time dependent
study with a range of 6
seconds and a laminar
flow.

47

Section 2 Group 8
Compute

Click the compute button

under the study section

Table 10: Rebecca's Pulsatile Stenosis Model 4 LPM

Space
Dimension

Open COMSOL program,

select model wizard and
then select 3D.

Physics

Use the down arrow to

open the options for
Fluid Flow physics. Select
the drop down arrow for
Single-Phase flow, click
Turbulent Flow,
Turbulent Flow, Algebraic
yPlus (spf), click add and
then study.

48

Section 2 Group 8
Study Type

Use the drop down arrow

to expand the Preset
Studies section. Select
Time Dependent and
then click Done.

Geometry

The geometry was

created using multiple
planes in solid works and
lofting to each plane.
The geometry for the
stenosis was given to us.
All of the dimensions for
the stenosis were given
in inches. Each end of
the stenosis was
extended 8 inches to
prevent entrance and
exit effects. The
SolidWorks file was saved
as Parasolid and
imported into COMSOL.
Finally, the Build Selected
was used to build the
model in COMSOL.

49

Section 2 Group 8

Material
Type

The entire model was

selected as the fluid
(water) to simulate the
stenosis.

50

Section 2 Group 8

51

Section 2 Group 8

Physical
Settings/Init
ial
Conditions

The inlet and outlet of

the model was selected.
The inlet was set to
BLANK and the outlet
was set to atmospheric
pressure.

52

Inlet:

Section 2 Group 8

53

Outlet:

Initial Values:

Section 2 Group 8

Mesh

An extra coarse mesh was

used for the model.
Using anything smaller
than extra course caused
the simulation to be too
much for the computer to
handle.

Study: Solve

This model was solved

using a time dependent
study with a range of 6
seconds and a turbulent
flow.
Click the compute button
under the study section

Compute

54

Section 2 Group 8
Table 11: Rebecca's Normal Carotid Condition 2

Space
Dimension

Open COMSOL program,

select model wizard and
then select 3D.

Physics

Use the down arrow to

open the options for
Fluid Flow physics. Select
the drop down arrow for
Single-Phase flow, click
Laminar Flow (spf), click
add and then study.
Use the drop down arrow
to expand the Preset
Studies section. Select
Time Dependent and
then click Done.

Study Type

Geometry

The geometry was set in

SolidWorks and the
model was imported into
COMSOL. The
SolidWorks file was saved
as Parasolid to open in
COMSOL.

55

Section 2 Group 8
Material
Type

Physical
Setting/Initi
al
Conditions

56

The material type was

added by right clicking
material, and selecting
blank material. Then, the
proper parameters were
input to simulate blood.
Inlet 1:

Outlet 1:

Section 2 Group 8

57

Outlet 2:

Initial Values:

Section 2 Group 8

Mesh

An extra coarse mesh

was used for the model.
Using anything smaller
than extra course caused
the simulation to be too
much for the computer
to handle.

Study/Solve

This model was solved

using a time dependent
study (0:0.1:6) seconds
with a laminar physics.
Click the compute button
under the study section

Compute

Table 12: Rebecca's Normal Carotid Condition 3

Space
Dimension

Open COMSOL program,

select model wizard and
then select 3D.

58

Section 2 Group 8
Physics

Study Type

Use the down arrow to

open the options for
Fluid Flow physics. Select
the drop down arrow for
Single-Phase flow, click
Laminar Flow (spf), click
add and then study.
Use the drop down arrow
to expand the Preset
Studies section. Select
Time Dependent and
then click Done.

Geometry

The geometry was set in

SolidWorks and the
model was imported into
COMSOL. The
SolidWorks file was saved
as Parasolid to open in
COMSOL.

Material
Type

The material type was

added by right clicking
material, and selecting
blank material. Then, the
proper parameters were
input to simulate blood.

Physical
Setting/Initi
al
Conditions

59

Inlet 1:

Section 2 Group 8

60

Outlet 1:

Section 2 Group 8

61

Outlet 2:

Initial Values:

Section 2 Group 8
Mesh

An extra coarse mesh

was used for the model.
Using anything smaller
than extra course caused
the simulation to be too
much for the computer
to handle.

Study/Solve

This model was solved

using a time dependent
study (0:0.1:6) seconds
with a laminar physics.
Click the compute button
under the study section

Compute

Table 13: Rebecca's Normal Carotid Condition 4

Space
Dimension

Open COMSOL program,

select model wizard and
then select 3D.

Physics

Use the down arrow to

open the options for
Fluid Flow physics. Select
the drop down arrow for
Single-Phase flow, click
Laminar Flow (spf), click
add and then study.
Use the drop down arrow
to expand the Preset
Studies section. Select
Stationary and then click
Done.

Study Type

62

Section 2 Group 8
Geometry

The geometry was set in

SolidWorks and the
model was imported into
COMSOL. The
SolidWorks file was saved
as Parasolid to open in
COMSOL.

Material
Type

The material type was

added by right clicking
material, and selecting
blank material. Then, the
proper parameters were
input to simulate blood.

Physical
Setting/Initi
al
Conditions

63

Inlet 1:

Section 2 Group 8

64

Outlet 1:

Outlet 2:

Section 2 Group 8

Mesh

An extra coarse mesh

was used for the model.
Using anything smaller
than extra course caused
the simulation to be too
much for the computer
to handle.

Study/Solve

This model was solved

using a stationary study
with a laminar physics.
Click the compute button
under the study section

Compute

65

Section 2 Group 8
Table 14: Rebecca's Stenotic Carotid Condition 2

Space
Dimension

Open COMSOL program,

select model wizard and
then select 3D.

Physics

Use the down arrow to

open the options for
Fluid Flow physics. Select
the drop down arrow for
Single-Phase flow, click
Laminar Flow (spf), click
add and then study.
Use the drop down arrow
to expand the Preset
Studies section. Select
Time Dependent and
then click Done.

Study Type

Geometry

The geometry was set in

SolidWorks and the
model was imported into
COMSOL. The
SolidWorks file was saved
as Parasolid to open in
COMSOL.

66

Section 2 Group 8
Material
Type

Physical
Setting/Initi
al
Conditions

67

The material type was

added by right clicking
material, and selecting
blank material. Then, the
proper parameters were
input to simulate blood.
Inlet 1:

Outlet 1:

Section 2 Group 8

68

Outlet 2:

Section 2 Group 8
Mesh

An extra coarse mesh

was used for the model.
Using anything smaller
than extra course caused
the simulation to be too
much for the computer
to handle.

Study/Solve

This model was solved

using a time dependent
study (0:0.1:6) seconds
with a laminar physics.
Click the compute button
under the study section

Compute

Table 15: Rebecca's Stenotic Carotid Condition 3

Space
Dimension

Open COMSOL program,

select model wizard and
then select 3D.

Physics

Use the down arrow to

open the options for
Fluid Flow physics. Select
the drop down arrow for
Single-Phase flow, click
Laminar Flow (spf), click
add and then study.

69

Section 2 Group 8
Study Type

Use the drop down arrow

to expand the Preset
Studies section. Select
Time Dependent and
then click Done.

Geometry

The geometry was set in

SolidWorks and the
model was imported into
COMSOL. The
SolidWorks file was saved
as Parasolid to open in
COMSOL.

Material
Type

The material type was

added by right clicking
material, and selecting
blank material. Then, the
proper parameters were
input to simulate blood.

Physical
Setting/Initi
al
Conditions

70

Inlet 1:

Section 2 Group 8

71

Outlet 1:

Outlet 2:

Section 2 Group 8

Mesh

An extra coarse mesh

was used for the model.
Using anything smaller
than extra course caused
the simulation to be too
much for the computer
to handle.

Study/Solve

This model was solved

using a time dependent
study (0:0.1:6) seconds
with a laminar physics.
Click the compute button
under the study section

Compute

72

Section 2 Group 8

73

Table 16: Rebecca's Stenotic Carotid Condition 4

Space
Dimension

Open COMSOL program,

select model wizard and
then select 3D.

Physics

Use the down arrow to

open the options for
Fluid Flow physics. Select
the drop down arrow for
Single-Phase flow, click
Laminar Flow (spf), click
add and then study.
Use the drop down arrow
to expand the Preset
Studies section. Select
Stationary and then click
Done.
The geometry was set in
SolidWorks and the
model was imported into
COMSOL. The
SolidWorks file was saved
as Parasolid to open in
COMSOL.

Study Type

Geometry

Material
Type

Physical
Setting/Initi

The material type was

added by right clicking
material, and selecting
blank material. Then, the
proper parameters were
input to simulate blood.
Inlet 1:

Section 2 Group 8

74

al
Conditions

Outlet 1:

Section 2 Group 8

75

Outlet 2:

Mesh

An extra coarse mesh

was used for the model.
Using anything smaller
than extra course caused
the simulation to be too
much for the computer
to handle.

Section 2 Group 8
Study/Solve

Compute

III.

76

This model was solved

using a stationary study
with a Laminar Flow
physics.
Click the compute button
under the study section

Results
a. Alex and Rebecca
Figures 5 through Figure 9 show the Pressure (mmHg) vs Time (Seconds) of each of the
cycles of data obtained in the stenotic flow loop in part 1 of the fluids module. Each of
these cycles were used to find the max, min, and average pressure values of each cycle.
Then, the values for each cycle were obtained using a MATLAB code written by Alex
Cook and shown in Table 17. Those values show the pressure drop that occurs in a
stenotic region at 1 and 4 LPM flow rates.

Table 17: Pressure measurements in the plastic stenosis as a function of flow conditions and location of the tap (mmHg)

1 L/min
upstream
1 L/min
middle
1 L/min
downstream
4 L/min
upstream
4 L/min
middle
4 L/min
downstream

Pmax

PmaxSTD

Pmin

PminSTD

Pavg

PavgSTD

30.7

4.15

-3.33

1.14

4.23

0.388

33.5

4.21

-2.82

0.839

4.35

0.244

31.3

1.18

-2.56

0.452

4.11

0.159

33.7

1.18

-12.7

0.328

6.54

0.0696

26.9

5.38

-12.7

0.344

4.50

0.0941

25.9

2.73

-11.6

0.310

4.95

0.071

Section 2 Group 8

77

45
40
35

Pressure (mmHg)

30
25
20
15
10
5
0
-5

-10

10

12

14

16

18

20

12

14

16

18

20

Time (Seconds)

Figure 4: 1 LPM Upstream Pressure vs. Time Stenosis Data

45
40
35

Pressure (mmHg)

30
25
20
15
10
5
0
0

10

-5
-10

Figure 5: 1 LPM Middle Pressure vs. Time Stenosis Data

Time (Seconds)

Section 2 Group 8

78

40
35
30

Pressure (mmHg)

25
20
15
10
5
0
0

10

12

14

16

18

20

12

14

16

18

20

-5
-10

Time (Seconds)

Figure 6: 1 LPM Downstream Pressure vs. Time Data

40

30

Pressure (mmHg)

20

10

0
0

10

-10

-20

Figure 7: 4 LPM Upstream Pressure vs. Time Data

Time (Seconds)

Section 2 Group 8

79

45

35

Pressure (mmHg)

25

15

10

-5

-15

Figure 8: 4 LPM Middle Pressure vs. Time Data

Time (Seconds)

12

14

16

18

20

Section 2 Group 8

80

35

Pressure (mmHg)

25

15

10

12

14

16

18

20

-5

-15

Time (Seconds)

Figure 9: 4 LPM Downstream Pressure vs. Time Data

Figure 10 through Figure 19 represent that experimental PIV data for hemoflow conditions 2, 3, and 4.
Figure 10, Figure 14, and Figure 18 show the graphs of velocity magnitude vs frame number. Figure 10
and Figure 14 show the pulsatility of 75 beats per minute (BPM) while Figure 18 shows the average
velocity magnitude vs. frame number at a steady state condition (no pulsatility). Figure 11 and Figure 13
show the peak systole obtained from Figure 10 for PIV condition 2. In the figures, one can see the
vectors pointing upwards indicating a pulse. Figure 12 shows peak diastole and the image shows the
inflow of fluid by the direction of the vectors. Figure 15 and Figure 17 show peak systole and Figure 16
shows peak diastole for Figure 14 which is PIV results for condition 3. Finally, Figure 18 and Figure 19
shows the average velocity magnitude vs. frame number for PIV condition 4.

Section 2 Group 8

Figure 10: Velocity magnitude (mm/s) vs. frame number at point (55.103, 5.8621) PIV condition 2 at frame number 8

Figure 11: PIV condition 2 peak systole

81

Section 2 Group 8

Figure 12: PIV condition 2 diastole at frame number 26

Figure 13: PIV condition 2 peak systole at frame 46

82

Section 2 Group 8

83

Figure 14: Velocity magnitude (mm/s) vs. frame number at point (49.828, 10.698) PIV condition 3

Figure 15: PIV condition 3 peak systole at frame 24

Section 2 Group 8

Figure 16: PIV condition 3 diastole at frame 30

Figure 17: PIV condition 3 peak systole at frame 33

84

Section 2 Group 8

Figure 18: Velocity magnitude (mm/s) vs. frame number at point (48.948, 7.0345) PIV condition 4

Figure 19: PIV condition 4 average flow (steady)

85

Section 2 Group 8

86

b. Alex
Stenosis Flow Loop Results
For the steady-state stenosis data, three 2-D cut planes were made before, after, and in the middle of
the occlusion according to Figure 1. The maximum velocities and pressures were computed on each
surface. The change in pressure and velocity between each port was calculated. Figures 20 and 21, as an
example, show the velocity and pressure maximums at each tap for the steady-state 4 L/min model.
These readings were repeated for the 1 L/min steady state and compiled into Table 19. The changes
between ports was calculated in Table 20.

Figure 20: Velocity maximums at each pressure tap for 4 L/min steady state

Figure 21: Pressure maximums at each tap for 4 L/min steady state

Section 2 Group 8

87
Table 18: Stenosis steady-state pressure and velocity data

Flow condition
1 L/min

4 L/min

Pressure tap
Upstream
Middle
Downstream
Upstream
Middle
Downstream

Pressure (mmHg)
0.013
-0.0038
0.0015
0.24
-0.033
-0.021

Velocity (cm/s)
4.9
8.1
7.4
18
33
30

Table 19: Change in pressure and velocity steady-state models between pressure taps

Flow condition
1 L/min
4 L/min

Pressure tap
Upstream to middle
Middle to downstream
Upstream to middle
Middle to downstream

P (mmHg)
-0.017
0.0053
-0.27
0.012

V (cm/s)
3.2
-0.70
15
-3.0

The pulsatile model pressures were calculated as functions of time and the average pressure differences
between taps were calculated. As an example, Figure 22 shows the pressure for the 1 L/min model.

Figure 22: Pressure as a function of time for the 1 L/min stenosis model. Blue is the upstream pressure, green is the stenotic
pressure, and red is the downstream pressure.

Table 21 shows the average change in pressure for between each of the pressure taps.

Section 2 Group 8

88
Table 20: Change in pressure and velocity pulsatile models between pressure taps

Flow condition
1 L/min
4 L/min**

Pressure tap
Upstream to middle
Middle to downstream
Upstream to middle
Middle to downstream

P (mmHg)
-0.073
0.058
-0.9**
0.4**

** The results of this turbulent simulation were difficult to attain and analyze, so these numbers of
approximate.

Hemoflow Carotid Artery Results

For the steady-state carotid artery data, two 2-D cut planes were made, one just before the bifurcation
and the other intersecting the stenotic or normal region (as well as the unaffected right-hand vessel).
The velocity and pressure were plotted on these surfaces and the maximum and minimum values were
found, as shown in Figures 23 and 24 for the stenotic model. The same measurements were repeated
with the normal artery. Table 22 summarizes these data for the steady-state carotid models.

Figure 23: Plot of the velocity across two surfaces within the carotid artery for the steady-state stenotic model: one before the
bifurcation and the other within the stenosis. The maximum velocity at the bifurcation is 1.1 m/s, while the stenotic region
experiences a maximum velocity of 2.5 m/s. The left scale bar corresponds to the stenotic region, while the right scale bar
corresponds to the bifurcation. The right-handed side after the bifurcation experienced an in-between velocity of 2.0 m/s.

Section 2 Group 8

89

Figure 24: Plot of the pressure across two surfaces within the carotid artery for the steady-state stenotic model: one before the
bifurcation and the other within the stenosis. The approximate pressure at the bifurcation is 13.9 mmHg, while the stenotic region
experiences a pressure of -6.3 mmHg. The left scale bar corresponds to the stenotic region, while the right scale bar corresponds
to the bifurcation. The right-handed side after the bifurcation experiences a pressure of about 3 mmHg.

Table 21: Summary of the velocity (m/s) and pressure (mmHg) values at the bifurcation and the stenotic/normal region of the
carotid loop

Steadystate
model
Normal
Stenotic

Velocity (m/s)

Pressure (mmHg)

Bifurcation

Stenosis/Normal

Bifurcation

Stenosis/Normal

1.0
1.1

1.1
2.5

0.1
1.4

3.0
13.9

0.85
-6.3

-2.2
-20.2

For the pulsatile carotid models, three 3-D cut points were added in the model: one just before the
bifurcation, one in the stenosis or normal region, and one in the alternate right-handed pathway for the
fluid. The pressure and velocity were plotted as functions of time to determine the cardiac cycle. Figures
25 and 26 show these velocities for the 0.5 L/min pulsatile flow rate normal and stenotic models
respectively. Figures 27 and 28 show the velocities for the 2 L/min models. Figures 2932 show the same
corresponding flow and geometric conditions, but for pressures instead of velocities.

Section 2 Group 8

90

Figure 25: Velocity magnitude versus time for the normal carotid artery with a 0.5 L/min pulsatile flow rate. Blue is the bifurcation,
green is the normal region, and red is the alternate pathway.

Figure 26: Velocity magnitude versus time for the stenotic carotid artery with a 0.5 L/min pulsatile flow rate. Blue is the bifurcation,
green is the stenotic region, and red is the alternate pathway.

Section 2 Group 8

91

Figure 27: Velocity magnitude versus time for the normal carotid artery with a 2 L/min pulsatile flow rate. Blue is the bifurcation,
green is the normal region, and red is the alternate pathway.

Figure 28: Velocity magnitude versus time for the stenotic carotid artery with a 2 L/min pulsatile flow rate. Blue is the bifurcation,
green is the stenotic region, and red is the alternate pathway.

Section 2 Group 8

92

Figure 29: Pressure versus time for the normal carotid artery with a 0.5 L/min pulsatile flow rate. Blue is the bifurcation, green is
the normal region, and red is the alternate pathway.

Figure 30: Pressure versus time for the stenotic carotid artery with a 0.5 L/min pulsatile flow rate. Blue is the bifurcation, green is
the stenotic region, and red is the alternate pathway.

Section 2 Group 8

93

Figure 31: Pressure versus time for the normal carotid artery with a 2 L/min pulsatile flow rate. Blue is the bifurcation, green is the
normal region, and red is the alternate pathway.

Figure 32: Pressure versus time for the stenotic carotid artery with a 2 L/min pulsatile flow rate. Blue is the bifurcation, green is
the stenotic region, and red is the alternate pathway.

Table 23 summarizes the pulsatile velocity trends during systole for the PIV data.
Table 22: Pulsatile velocity trends during systole for the carotid artery models

Flow
condition
(L/min)
0.5
2

Geometry

Peak Bifurcation
Velocity (m/s)

Peak Normal or
Stenotic Velocity (m/s)

V (m/s)

Normal
Stenotic
Normal
Stenotic

0.5
0.55
2.0
2.2

0.55
1.25
2.2
4.8

0.05
0.7
0.2
2.6

Section 2 Group 8

94

c. Rebecca
Figure 33 through Figure 48 represent the velocity and pressure profile data for the stenosis simulations
from part 1 of the fluids module. For each of the stenosis simulations, the inlet, outlet, and initial
condition input data can be found in Table 12 through Table 15. For the 1 LPM steady stenosis, the
velocity profile can be seen in Figure 19 and the pressure profile in Figure 20. The 4 LPM steady stenosis
simulation results can be seen in Figure 21 (velocity profile) and Figure 22 (pressure profile). The
simulation using pulsatile 1 LPM results are seen in Figures 23-25 (velocity profiles) and Figures 26-28
(pressure profiles). Finally, the 4 LPM pulsatile simulation results can be seen in Figures 29-31 (velocity
profiles) and Figures 32-34 (pressure profiles). For each of the velocity results, regardless of the stenosis
simulation, an increase in velocity through the stenotic (narrow) region can be seen. Also, for the
pressure data, regardless of the stenotic simulation, a pressure drop can be seen through the stenotic
region with a slight recovery toward the exit of the stenosis. Seeing these results proved that the
simulations follow the laws of fluid dynamics which were mentioned earlier in this paper.

Figure 33: Rebecca's 1LPM Steady Velocity Profile

Section 2 Group 8

Figure 34: Rebecca's 1 LPM Steady Pressure Profile

Figure 35: Rebecca's 4 LPM Steady Velocity Profile

95

Section 2 Group 8

Figure 36: Rebecca's 4 LPM Steady Pressure Profile

Figure 37: Rebecca's 1 LPM Pulsatile Velocity Profile at 0 Seconds

96

Section 2 Group 8

Figure 38: Rebecca's 1 LPM Pulsatile Velocity Profile at 3 Seconds

Figure 39: Rebecca's 1 LPM Pulsatile Velocity Profile at 6 Seconds

97

Section 2 Group 8

Figure 40: Rebecca's 1 LPM Pulsatile Pressure Profile at 0 seconds

Figure 41: Rebecca's 1 LPM pulsatile pressure profile at 3 seconds

98

Section 2 Group 8

Figure 42: Rebecca's 1 LPM pulsatile pressure profile at 6 seconds

Figure 43: Rebecca's 4 LPM Pulsatile Model velocity profile at 0 seconds

99

Section 2 Group 8

Figure 44: Rebecca's 4 LPM Pulsatile Model velocity profile at 3 seconds

Figure 45: Rebecca's 4 LPM Pulsatile model velocity profile at 6 seconds

100

Section 2 Group 8

Figure 46: Rebecca's 4 LPM Pulsatile pressure profile at 0 seconds

Figure 47: Rebecca's 4 LPM pulsatile pressure profile at 3 seconds

101

Section 2 Group 8

102

Figure 48: Rebeccas 4 LPM pulsatile pressure profile at 6 seconds

The velocity profile and pressure profile results for the simulations using a normal carotid and a 43%
stenotic carotid geometries are seen in Figure 35 through Figure 68. The first condition tested was the
normal carotid at 0.5 LPM with 75 BPM pulsatility. The velocity profiles at 0, 3, and 6 seconds along with
a graphical representation of the velocity trend over time can be seen in Figures 35 through 38
respectively. It can be seen in Figures 35-37, the velocity increases in the left and right branches of the
carotid because of the decrease in diameter. The velocity graph over time (Figure 38) shows that the
entrance velocity and the entrance velocity of each of the branches is not much different from one
another. Table 22 shows the statistics for the velocity magnitude values between the flows of the
entrance of the carotid and the entrances of the carotid branches. The standard of deviation of the flow
velocity between each of those areas at time points 1 to 6 seconds is very low. Table 23 shows the
standard of deviation between the averages of each flow and again, this value is very low. The standard
of deviation between the values being low is normal for the normal carotid because the change in
diameter entering into the branches is not significant enough to cause extreme velocity changes.
Figures 39 through 41 shows the pressure profile data for the model at different time points. Because
the model is pulsatile, the pressure values change throughout the model (from 0 to 6 seconds) and the
velocity does somewhat decrease through the carotid branches.

Section 2 Group 8

Figure 49: Normal Carotid condition 2 velocity profile at 0 seconds

Figure 50: Normal carotid condition 2 velocity profile at 3 seconds

103

Section 2 Group 8

Figure 51: Normal carotid condition 2 velocity profile at 6 seconds

Figure 52: Normal carotid condition 2 velocities at the entrance, and entrance to the left and right carotid branches

104

Section 2 Group 8

105

Table 23: Velocity statistics between the entrance, left entrance and right entrance branches for the normal carotid condition 2

Velocity
(m/s)

Entrance
Left
Entrance
Right
Entrance
Average
STD

Time (s)
3
0.8

4
0.35

5
1.6

6
0.4

0
0

1
1.45

2
0.7

0.88

0.25

0.46

0.21

0.91

0.25

0
0
0

1.1
1.143333
0.28746

0.28
0.41
0.251595

0.6
0.62
0.17088

0.29
0.283333
0.070238

1.15
1.22
0.350286

0.3
0.316667
0.076376

Table 24: Velocity statistics for the average of each of the flows between the entrance, left and right branches of the carotid

Average Velocity (m/s)

Entrance
Left Entrance
Right
Entrance
STD

0.757142857
0.422857143
0.531428571
0.170529388

Figure 53: Normal Carotid condition 2 pressure profile at 0 seconds

Section 2 Group 8

106

Figure 54: Normal carotid condition 2 pressure profile at 3 seconds

Figure 55: Normal carotid condition 2 pressure profile at 6 seconds

Figures 42 through 48 are the velocity and pressure profile results for the normal carotid with a 2 LPM
flow rate and a 75 BPM pulsatility (condition 3). Figures 42 through 44 show the velocity increase
through the left and right branches of the carotid artery. Figure 45 is a graphical representation of the
velocities of each location at time points 1 through 6 seconds. Tables 24 and 25 show the statistics for

Section 2 Group 8

107

the velocity flow through each of the entrances. Table 24 shows the standard of deviation increasing
over time for the flow between each entrance at specific time points. Table 25 shows the standard of
deviation of the velocity flow of the average velocity flow values at each of the entrances. This standard
of deviation is larger than the standard of deviation in Table 23. This is because the flow rate for the
normal carotid condition 3 was increased to 2 LPM while keeping the 74 BPM pulsatility. Figures 46
through 48 show the pressure profile data for the normal carotid condition 3. This pressure profile data
shows a decrease in pressure in each of the branches of the carotid.

Figure 56: Normal Carotid condition 3 velocity profile at 0 seconds

Section 2 Group 8

Figure 57: Normal carotid condition 3 velocity profile at 3 seconds

Figure 58: Normal carotid condition 3 velocity profile at 6 seconds

108

Section 2 Group 8

109

Figure 59: Normal carotid condition 3 velocities at the entrance, and entrance to the left and right carotid branches
Table 25: Velocity statistics for the normal carotid condition 3 between each of the carotid entrances.

Velocity
(m/s)

Entrance
Left
Entrance
Right
Entrance
Average
STD

0
0

1
10

2
10.5

3
3

6.5

0
0
0

8
8
2

8.5
8.5
2

2.5
2.5
0.5

Time (s)
4
6.5

5
9

6
2

5.5

5.2
5.233333
1.250333

7.5
7.333333
1.755942

2.5
1.833333
0.763763

Table 26: Velocity statistics of the average velocities between each of the carotid entrances

Average Velocity (m/s)

Entrance
Left Entrance
Right
Entrance
STD

5.857142857
3.571428571
4.885714286
1.147134851

Section 2 Group 8

Figure 60: Normal carotid condition 3 pressure profile at 0 seconds

Figure 61: Normal carotid condition 3 pressure profile at 3 seconds

110

Section 2 Group 8

111

Figure 62: Normal carotid condition 3 pressure profile at 6 seconds

Figures 49 through 51 show the velocity and pressure profile data results for the normal carotid at 2
LPM flow rate at steady state. Figure 49 specifically shows the velocity profile where the velocity is
increasing slightly through the branches of the carotid artery. Figure 50 shows a graphical
representation of the velocity data and reflects that there is no pulsatility in the model. Finally, Figure
51 shows the pressure profile data and that the pressure decreases as the flow enters the smaller
diameter branches of the carotid artery.

Section 2 Group 8

Figure 63: Normal Carotid condition 4 velocity profile

Figure 64: Normal carotid condition 4 velocity magnitude at the entrance and the left and right branch entrances

112

Section 2 Group 8

113

Figure 65: Normal Carotid condition 4 pressure profile

Figures 52 through 58 show the velocity and pressure profile data for the stenotic carotid artery at 0.5
LPM flow rate and 75 BPM pulsatility (condition 2). Figures 52 through 54 show that through the
stenotic region of the carotid, the velocity significantly increases. Figure 55 is a graphical representation
of the velocity data. Tables 26 and 27 represent the statistical analysis of the velocity data between the
carotid entrance and the left and right carotid branch entrances. Table 26 shows that the standard of
deviation between the flows in each of the entrances is large. Table 27 also shows that the standard of
deviation between the average flow velocities of each entrance is high. This can be contributed to the
stenotic region in the carotid along with the pulsatility of 75 BPM. Tables 56 through 58 show the
pressure profile data of the flow through the stenosis. As seen in those figures, the pressure drops
faster in the stenotic region than in the carotid branch without a stenosis.

Section 2 Group 8

Figure 66: Stenotic Carotid condition 2 velocity profile at 0 seconds

Figure 67: Stenotic carotid condition 2 velocity profile at 3 seconds

114

Section 2 Group 8

Figure 68: Stenotic Carotid condition 2 velocity profile at 6 seconds

Figure 69: Stenotic carotid condition 2 velocities at the entrance, and entrance to the left and right carotid branches

115

Section 2 Group 8

116

Table 27: Velocity statistics for the flow through the stenotic carotid at condition 2 between each of the entrances.

Velocity
(m/s)

Entrance
Left
Entrance
Right
Entrance
Average
STD

0
0

1
2
1.25 0.5

Time (s)
3
0.6

4.9 1.5

2.5

1.5

5 1.25

0
0
0

2.25
1
2.8
1
1.886133611 0.5

1.25
1.45
0.96566

0.75
0.833333
0.629153

2.25 0.75
2.8333 0.75
1.9419 0.5

4
0.25

5
6
1.25 0.25

Table 28: Statistics of the average velocity data between each entrance for stenotic carotid condition 2

Average Velocity (m/s)

Entrance
Left Entrance
Right
Entrance
STD

0.585714286
2.378571429
1.178571429
0.91340173

Figure 70: Stenotic Carotid condition 2 pressure profile at 0 seconds

Section 2 Group 8

117

Figure 71: Stenotic carotid condition 2 pressure profile at 3 seconds

Figure 72: Stenotic carotid condition 2 pressure profile at 6 seconds

Figures 59 through 65 are the results for the velocity and pressure profile data for the stenotic carotid at
a flow rate of 2 LPM and a 75 BPM pulsatility (condition 3). Specifically, Figures 59 through 62 show the

Section 2 Group 8

118

velocity profile data. It can be seen that the velocity increases through the stenotic region of the
carotid. Figure 62 shows a graphical representation of the velocity data through the carotid entrances
over time. Tables 28 and 29 represent the statistical analysis of the velocity data. Table 28 shows a high
standard of deviation in the beginning and it decreases over time. Table 29 also shows a high standard
of deviation between the average velocity values for each carotid entrance. The high standard of
deviation can be contributed to the 75 BPM pulsatility and the 2 LPM flow rate along with the stenotic
region disturbing the flow. Figures 63 through 66 show the pressure profile data for the model and it
can be seen that the pressure drop in the stenotic region occurs faster than the pressure drop in the
carotid branch without stenosis.

Figure 73: Stenotic carotid condition 3 velocity profile at 0 seconds

Section 2 Group 8

Figure 74: Stenotic carotid condition 3 velocity profile at 3 seconds

Figure 75: Stenotic carotid condition 3 velocity profile at 6 seconds

119

Section 2 Group 8

120

Figure 76: Stenotic carotid condition 3 velocities at the entrance, and entrance to the left and right carotid branches
Table 29: Velocity statistics for the stenotic carotid condition 3 between each of the carotid entrances

Velocity
(m/s)

Entrance
Left
Entrance
Right
Entrance
Average
STD

0
0

1
3.2

2
1.5

12.5

0
0
0

6
7.233333333
4.771093515

2.5
3
1.802776

Time (s)
3
1
4 3,.2
1.8
2.266667
1.553491

Table 30: Statistics for the average velocity data for the stenotic carotid condition 3

Average Velocity (m/s)

Entrance
Left Entrance
Right
Entrance
STD

1.742857143
6.666666667
3.014285714
2.556048952

4
1.5

2
1.75
0.353553

5
3.5

6
1.5

13

5.5

6.3
7.6
4.881598

2.5
3.166667
2.081666

Section 2 Group 8

Figure 77: Stenotic carotid condition 3 pressure profile at 0 seconds

Figure 78: Stenotic carotid condition 3 pressure profile at 3 seconds

121

Section 2 Group 8

122

Figure 79: Stenotic carotid condition 3 pressure profile at 6 seconds

Finally, Figures 66 through 68 represent the velocity and pressure data for the stenotic carotid with a 2
LPM flow rate at steady state (no pulsatility). It is seen in Figure 66 that the velocity increases in the
stenotic region of the carotid. Figure 67 is a graphical representation of the data in Figure 66 and it
shows the steady state of the system by reaching a plateau around 0.015 seconds. Figure 68 is the
pressure profile data. Again, like the previous stenotic carotid models, the pressure drops faster in the
stenotic region of the carotid faster than the branch without stenosis.

Section 2 Group 8

Figure 80: Stenotic carotid condition 4 velocity profile

Figure 81: Stenotic carotid condition 4 velocity magnitude at the entrance and the left and right carotid branch entrances

123

Section 2 Group 8

124

Figure 82: Stenotic carotid condition 4 pressure profile

Discussion
d. Alex
The plastic stenosis flow loop provided a good opportunity to become familiar with the concepts of
fluid mechanics before tackling more complicated phenomenon. The 1 L/min and 4 L/min steadystate models were an excellent way to see the basic ideas behind the mechanics. Table 19
(computational results) shows two such ideas: conservation of mass and Bernoullis equation.
In a closed tube, all the fluid flowing goes in the inlet and out the outlet. As a consequence,
whenever the cross-sectional area decreases, the average velocity increases to main the same
volumetric flow rate (Q = vA). Table 19 clearly shows an increase in velocity when moving from the
first tap at the larger diameter to the second tap at the smaller, stenotic diameter. When the fluid
emerges from the stenosis, the cross-sectional area increases, decreasing the velocity once again.
The change in velocity from middle to downstream shows this phenomenon as well. Overall, these
velocity measurements confirm this fundamental idea.
Bernoullis equation then dictates what happens to the pressure exerted by the fluid when the
velocity changes. Through conservation of energy, this equation shows that for an idealized laminar
flow (a simplification of reality in many cases), the pressure decreases when the velocity increases.
In other words, the kinetic energy that is gained by the fluid in a smaller-cross sectional area takes
away from the pressure energy that it can exert. Once again, Table 19 shows the basic idea:
pressure drops from upstream to the stenosis, and then recovers somewhat after exiting the

Section 2 Group 8

125

stenosis. This pressure recovery is usually not complete due to recirculation of the fluid after the
stenosis.
Looking at Table 17, the experimental data generally shows larger pressure fluctuations than predicted
by the COMSOL models and Bernoullis equation. My calculations with Bernoullis equation for the
change from upstream to the middle expected a drop of 0.0122 mmHg for the 1 L/min model and about
0.196 mmHg for the 4 L/min model. The percent errors of these calculations are displayed in Table 31.

Table 31: Percent errors for the stenotic pressure measurements

Flow condition
1 L/min
4 L/min

P (experimental)
(mmHg)
-0.144
0.0245

P (theoretical) (mmHg)

% error (%)

-0.0122
-0.196

1080
113

These percent errors could have been large for several reasons. First and foremost, Bernoullis equation
only applies to an idealized laminar flow. The 4 L/min flow rate actually turned out to be turbulent. But
even smaller pulsatile flow rates have the potential to deviate from ideal. These experiments and
calculations show that Bernoullis equation is not easily applicable to a lot of biological systems,
especially those with pulsatile flow. This necessitates the need for more advanced methods in fluid
mechanics (e.g. Navier-Stokes equation).
Air could have also potentially ruined the pressure transducer measurements. During the experiments,
we noticed many tiny air bubbles moving through the system. When we got negative pressure readings
after converting over from current, this indicates that insufficient clearing of air from the system might
also be a significant source of error.
The hemoflow loop and PIV data provided a very different chance to apply the concepts of fluid
mechanics to a more practical and realistic model than the plastic stenosis. Just as in the stenosis, the
steady state models provided a good baseline sense of the change in the velocity and pressure between
normal and occluded arteries. Table 18 shows that in a stenotic artery, the increase in velocity is a full
order of magnitude greater (V = 1.4 m/s for stenosis compared to 0.1 m/s for a normal artery) than in a
normal one. As a consequence of the increase in velocity, there is a corresponding decrease in pressure
of about 20 mmHg. Since human blood pressure can differ by about 40 mmHg between systole and
diastole,10 it is a somewhat believable pressure difference. A 43 % stenosis is clearly a serious occlusion
that could potentially have fatal consequences down the road.
Using Bernoullis equation to calculate the velocity change that would result from a 20 mmHg drop in
pressure yields 1.3 m/s. Therefore, at the very laminar flow rate of 2 L/min (refer back to Table 1, R
720), Bernoullis is a good approximation of the average flow field. Since this equation only applies to
idealized laminar flow, it provides additional confirmation of the relative simplicity of the velocity field.
The velocities before the bifurcation were measured both experimentally and computationally. The
information in Figures 9, 13, and 17 was used to approximate the peak velocity during systole (9 and 13)
or the average velocity (17). Table 32 compares the values acquired through the experiment and the
computations.

Section 2 Group 8

126
Table 32: Percent errors for the PIV velocities

Condition

Experimental velocity
(mm/s)
4
10
80

0.5 L/min (pulsatile)

2 L/min (pulsatile)
2 L/min (steady-state)

Computational
velocity (mm/s)
0500
2000
1100

% difference (%)
197
189
173

This wildly huge error could be due to several factors. First, the simulations used the properties of blood
and not that of 45 % glycerin. It is possible that slightly different mechanical properties would alter the
outcome of the experiment. PIV is also a fickle technique. It was difficult to acquire and analyze the
data. Beyond the mechanics of the loop system, fluid, and seed, there is also a lot of uncertainty with
the post-processing techniques. Because of this ambiguity with what is considered good data, these
sources are also likely causes of error.
Regardless of how well the simulations or theoretical predictions matched reality, it is clear that stenosis
plays large role in cardiovascular mechanics, and therefore, health. With the widespread prevalence of
heart disease and stroke, its useful to learn more about the theory and techniques behind fluid
mechanics in order to potentially devise new solutions and treatments in the future.

e. Rebecca
The ten simulations were built using SolidWorks software. The four stenotic models
were designed using the CAD software dimensions supplied to us on Angel. All of the
dimensions were in inches so conversions were needed to calculate the velocity and
other input and initial value data. The remaining six simulations were developed using
the given SolidWorks designs for the normal and 43% stenotic carotid artery. The input,
output, and initial condition data was obtained and calculated using the flow rate and
pulsatile conditions provided in the fluids module protocols that were posted on Angel.
When designing the models using the COMSOL software, turbulent flow was
determined by the Reynolds number and whether or not the model was time
dependent was determined by the pulsatility. If the condition being used to develop the
model had a BPM value, the model was set to be time dependent. If the model did not
have BPM value, the model was determined to be stationary. Also, various assumptions
were made for the models such as incompressible Newtonian fluid, the fluid was at the
same height, and the alpha value was the same so that the Bernoulli equation ended up
being

After speaking with Dr. Leung and Bryan Good, I learned that the input of each model
should be a velocity and the output of each model should be set to atmospheric
pressure. However, some of the models would not run if the input values were a
velocity so the values were converted to mass flow and entered into the input values of
the simulations (Table 12 and Table 13). Also, the initial conditions should be set to

Section 2 Group 8

127

atmospheric pressure. All of the conditions for the models were set to have an entrance
and exit length of 8 inches longer than what was initially given in the CAD software
drawing dimension details. The 1 LPM steady stenosis data was not extended because
there was a condition in the inlet data that allowed for compensation of entrance
effects shown in Table 12. This input option was also available and used for all of the
carotid artery models. Extending the entrance and exit lengths in the remaining four
stenosis models (Tables 13 through 15) will eliminate the entrance and exit effects that
would be present from the flow field because the flow would be fully developed by the
time it reached the stenotic area. Also, the material (water) was added for the four
stenotic simulations because that is what we were instructed to do. The experimental
data in the flow loop experiment used water which would make it easier for comparison
to the computer simulated results. For the 1 LPM stenotic simulations, the Reynolds
numbers that were calculated were low enough that the simulations could be run using
a laminar flow physics. The 4 LPM flow rates resulted in simulations that used a
turbulent flow physics. For all of the files using turbulent flow, the COMSOL multiphysics
turbulent flow chosen was the Algebraic y plus (spf) file because the description stated
that it was the simplest turbulent flow to run. Our simulations were not very complex
and therefore they did not need a complex turbulent flow physics for the simulation
calculations. The simulations that were time dependent and not steady state were
chosen to run from 0 to 6 seconds with a time step of 0.1 seconds. This is because
approximately 10 cycles of data for our experimental simulations were run at 6 seconds.
Also, using a small time step of 0.1 seconds would make the calculated data from the
simulations more accurate. The pressure results obtained from the four stenotic
COMSOL simulations followed the results obtained from the flow loop pressure
calibration. This is because the pressure decreased from the upstream to the
downstream pressure tap locations on the physical and the simulated models.
The remaining six simulations were of the carotid artery. Three of the six simulations
were of the normal carotid artery and the three remaining simulations were of the 43%
stenotic carotid artery. Each of these simulations also used an initial condition and
outlet value of atmospheric pressure. The inlet values were all set to be a laminar
inflow velocity. The inlet values were determined partially by the data obtained by
experimental results and partially by the hemoflow protocol. The portion obtained by
experimental data was the amplitude of the sine wave which was taken from the
velocity magnitude graphs shown in Figure 9, Figure 13, and Figure 17. The magnitude
was taken as the first systolic velocity magnitude data point (cm/s) and converted to
m/s. The remaining input values determined by the hemoflow protocol were the
pulsatility (conditions 2 and 3) and the flow rate converted to velocity which was set as
the sine wave offset value. For each of the normal carotid and stenotic carotid
simulations, blood was used as the material type to run the simulation and calculate the
Reynolds number. For all of the carotid files, normal and stenotic, the Reynolds
numbers was calculated to be low enough to run all of the simulations at a laminar
physics. All of the Reynolds numbers calculated were done by using an online Reynolds
number calculator.11

Section 2 Group 8

128

All of the velocity and pressure profile results for the stenosis and the normal and
stenotic carotid simulations showed an increase in velocity through the stenotic region
and a pressure drop through the stenotic region as explained above in the results
section. Because the simulations showed a proper output, it was determined by
speaking with Dr. Leung and Bryan Good that my simulations were good. All of us
evaluated my input, output and initial condition data and determined my values to be
appropriate.
The results sections for my stenotic carotid data results (Figures 52 through 68) show
that the velocity increased in the stenosis area and the pressure dropped in the stenotic
area faster than other areas of the carotid artery. The problem with those models is
that the areas outside of the stenotic regions remained blue for the velocity profiles.
After discussing my results with Bryan, it was determined that the change in velocity in
the stenotic region was so drastic that the remaining sections of the carotid appeared to
be blue (which represented a 0 velocity) and the program was not able to properly scale
the velocity change. COMSOL did not allow me to change the scale of my velocity
results and therefore I was not able to prove this assumption.
An important observation of the carotid models is that from the normal carotid to the
stenotic carotid, the left stenotic branch has a large increase in velocity for the models
with the stenosis. This is because the smaller the diameter of the artery, the more flow
is restricted and the higher the velocity value. These results can be observed by
comparing Tables 22, 23, 24, and 25, to Tables 26, 27, 28, and 29 respectively. It can be
seen that a larger standard of deviation exists between the flows at different points in
the models over time as the flow rate increases and a stenotic region exists in the
models.
Errors in the first experiment of the fluids module could have happened because of
bubbles in the system. All of the large air bubbles were removed from the stenosis but
some may not have been visible close to the pressure transducer. This would have
distorted the pressure data results. Streamlines were added to each of the models to
show the recirculation zones of the fluid which should be present after the stenosis.
Because none of the streamline images showed recirculation zones, the images were
not included in this report. Errors which could have caused recirculation zones not to
occur could have resulted from my experimental results. This is because the input
values for the computer simulations were taken from the experimental results. For
example, the amplitude of the sine was taken to be a peak systole value from each of
the velocity magnitude (cm/s) vs. frame number charts (Figures 9 and 13). The graphs
shown in Figures 9 and 13 were dependent on when the button on the computer was
clicked to record the data because the conditions were pulsatile. During experimental
testing, the pump was moved to allow for more tracer particles to enter the system and
obtain clearer flow images. Also, our experimental results were collected with the
assistance of Dr. Leung and Erik because the PIV systems were not working properly.

Section 2 Group 8

129

Initially, our results were taken using a hemoflow system with a hair stuck in the carotid.
The hair being in that area could have somewhat disturbed the flow and our results.
This was another error that occurred and resulted in Dr. Leung having us re-do our data
at a different hemoflow station.
In conclusion, the goals of the experiment were met because we were able to observe
and obtain reasonable data approved by Dr. Leung, Erik, and Bryan. My experimental
and computer simulated results followed the laws of fluid dynamics by increasing
velocity and decreasing pressure through a stenotic region. This was shown through the
tables and figures that were referenced various time throughout the text in the results
and discussion section. Understanding the stress that is put on the arteries by the
increased velocity and decreased pressure helps me understand the various
cardiovascular diseases which are caused by stenotic vessels. Knowing where turbulent
flow occurs and what can cause it to occur can lead to new technology to prevent
plaque buildup which causes cardiovascular disease. It can also lead to treatments and
new surgical procedures to help people who suffer from these conditions.
IV.

Summary
The goals of the experiment were to compare our experimental pressure data results to our
simulation velocity and pressure results and to be able to develop models which accurately
represent stenotic vessels that can cause cardiovascular diseases and disorders. We also
were able to learn how to us the hemoflow PIV data technology. Each of the results were
compared and statistically analyzed. The obtained results were also able to be analyzed by
using our knowledge from the fluid mechanics course. The results also helped us
understand the mechanics behind atherosclerosis and other cardiovascular diseases. By
examining all of the tables and figures present in this report, it is clear that we have an in
depth understanding of the software being used and the biology background needed to
examine and analyze the results.

V.

References

1.

What Is Atherosclerosis? - NHLBI, NIH. http://www.nhlbi.nih.gov/health/healthtopics/topics/atherosclerosis. Accessed November 11, 2015.

2.

Bapat A. Predicting Stenosis in Blood Vessels.

3.

Javadzadegan A, Yong ASC, Chang M, et al. Flow recirculation zone length and shear rate are
differentially affected by stenosis severity in human coronary arteries. Am J Physiol Heart Circ
Physiol. 2013;304:H559-H566. doi:10.1152/ajpheart.00428.2012.

4.

Zhao K, Dalton P, Yang GC, Scherer PW. Laminar & turbulent flows. Openchannel flow.

Section 2 Group 8

130

2002;47(2):1-30. http://www.ncbi.nlm.nih.gov/pubmed/11948732.
5.

Stather PW, Sylvius N, Wild JB, Choke E, Sayers RD, Bown MJ. Arterial Disease. 2013:490-497.
doi:10.1161/CIRCGENETICS.111.000053.

6.

Manning K. Hemoflow Protocol. 2015:1-18.

7.

Manning, Keefe. "Measuring Pressure in a Symmetric Stenosis Under Pulsatile Flow." State
College, PA: The Pennsylvania State University, 2015. Print.

8.

"Whole Blood." Viscopedia: http://www.viscopedia.com/viscosity-tables/substances/wholeblood/

9.

"Density of Blood." The Physics Factbook:

http://hypertextbook.com/facts/2004/MichaelShmukler.shtml

10.

High blood pressure (hypertension). Mayo Clinic, 2015. http://www.mayoclinic.org/diseasesconditions/high-blood-pressure/expert-answers/pulse-pressure/faq-20058189

11.

Reynolds Number Calculator. http://www.efunda.com/formulae/fluids/calc_reynolds.cfm.

Accessed November 13, 2015.
VI.

Acknowledgements
a. Alex
I owe thanks to several people. I worked closely with Rebecca Nagurney on the
experiments, group data analysis, and writing the report. I discussed modeling and
interpretation of results with Rebecca, Max Greenberg, Michael King, Alex Miller, Jessica
Barrett, and Samuel Cramer.
b. Rebecca
Throughout this fluids module, I worked with many people to collaborate on results and
simulations. I worked with my partner, Alex on all of the graphs, data, and experimental
discussions. I worked with Jessica Barrett, Erin Hartmann, Kaylene Killeen, and Joe
Roberto to construct the COMSOL simulations and compare results.