Characterization of Material Properties of Aorta from Oscillatory Pressure Tests

V.V. Romanov, K. Darvish, and S. Assari

Biomechanics Laboratory, Temple University, Philadelphia, USA
Abstract Traumatic Aorta Rupture (TAR), also known as
Traumatic Rupture of Thoracic Aorta (TRA), is one of the
major causes of fatalities in motor vehicle accidents. This issue
has been studied in the past; however, the results of the suggested mechanisms are speculative and inconclusive. One
cause for these speculations is an incomplete understanding of
the material properties of aorta. Therefore, the goal of this
experiment was to characterize the dynamic structural response of the aorta to a biaxial loading. Seven samples were
subjected to a pressure oscillation input ranging from 7kPa to
76kPA at a frequency ranging from .5Hz to 5Hz. The results
were presented in the form of pressure versus volumetric
strain which showed an increase in both phase and magnitude
of the linear modulus. Initial modeling was completed utilizing
hyperelastic constitutive equations. The results of this study
will be used to determine a viscoelastic constitutive model and
a finite element model for the aorta material.

ascending and descending aortas, as well as the aortic arch,

see injury as well. The general aortic anatomy is diagrammed in Figure 1.

Ascending Aorta

Aortic Arch

Peri-Isthmus

Descending Aorta

Keywords traumatic aorta rupture, dynamic aorta test,

material properties of aorta, viscoelasticity, hyperelasticity.

I. INTRODUCTION
Traumatic Aorta Rupture (TAR), also known as Traumatic Rupture of Thoracic Aorta (TRA), is a major cause of
deaths in motor vehicle collisions. According recent studies,
TAR was diagnosed in 12% to 29% of autopsied fatally
injured occupants. [1] Furthermore, when people experience
such trauma, only 9% (7500-8000 victims in US and Canada) survive from the scene of the accident and the overall
mortality rate is 98%. [2] In 94% of the cases, the shearing
forces of high speed impacts have been associated with
transverse tears at the peri-isthmic region which is subjected
to the greatest strain. [3]
Thoracic aorta consists of three major segments: ascending aorta, aortic arch, and descending aorta. Ascending aorta
originates from the heart at the aortic valve. It then becomes
the aortic arch which is suspended by brachiosephalic, left
common carotid, and left subclavian arteries, which supply
the blood to upper extremities. The last section, the descending aorta, supplies the blood to the lower limb and is
fixed to the spine with the intercostals arteries. The periisthmic region is located between the aortic arch and the
descending aorta, at the point where the vessel becomes
unattached from the spine. It should be noted that even
though most of the rupture occurs at the peri-isthmic region,

Fig. 1 Anatomy of the aorta [4]

Several rupture mechanisms have been recommended in
the literature. Presumably, the earliest proposal came from
Rindfleisch, as cited by Richens, who suggested that the
injury was caused by the sudden stretching of the aorta. [5]
[2] Therefore, as the body experiences a rapid deceleration,
the heart moves forward creating stress between fixed descending aorta and aortic arch at the peri-isthmus. The
second mechanism is the rupture due to pressure increase in
the vessel when the thorax and the abdomen compress. This
theory, however, was dismissed by a number of investigators who argued that if the aorta were an isotropic axial
cylinder vessel under pressure, it would rupture axially
rather than transversely. Another mechanism proposed was
the osseous pinch which suggests that the rupture of the
aorta is caused by high local stress created by the pinch
between highly compressed thorax and the spine. [2] However, though these mechanisms provide plausible explanations of rupture, a lot of the studies conducted are conflicting and insubstantial.

K.E. Herold, W.E. Bentley, and J. Vossoughi (Eds.): SBEC 2010, IFMBE Proceedings 32, pp. 380384, 2010.
www.springerlink.com

Characterization of Material Properties of Aorta frrom Oscillatory Pressure Tests

381

This study is one of the series with a com

mmon final goal
to determine the most feasible cause of aorttic rupture. The
specific aim of this research is to understan
nd the dynamic
material properties of the aorta through the pressure
p
oscillation tests. A noncontact experiment was deveeloped to better
understand the structural properties of the ao
orta while keeping it intact.

II. MATERIALS AND METHODDS

A. Experimental Procedure
m 6 month-old
For the experiments, seven samples from
pigs were acquired from the local slaughter house. Porcine
aorta was chosen because it has been widely used in the past
in the cardiovascular related research as a substitute for a
human aorta due to their similarities. [6] Fu
urthermore, human samples would have been difficult to
o obtain and it
would have been highly unlikely that the samples would
have been healthy.
To have all of the samples tested under the
t same conditions, aortas were ordered attached to the heart.
h
Once received from the slaughter house, the samplees were kept in
saline solution at 5 C until the beginning of the experiment.
Though out the test, the aorta was kept hydraated in the same
solution but at the room temperature (25 C).
Preparation work for each sample went th
hrough the same
following procedure. The descending aorta waas cut at 152mm
from the aortic valve. The excessive fatty tissu
ue was removed
from the aorta. Careful attention had to be paid
p
during this
procedure in order to avoid cutting off the inttercostal arteries
at the base. Once the tissue was removed, inttercostal arteries
were tied off at the base to prevent pressure loss.
To keep track of the geometrical chan
nges, ultrasonic
measurement sensors (Sonometrics Corp., Caanada, 2R-34C40-NB) were utilized. Each of the sensors sends
s
and receives a pulse every 31ns to 2ms. Based on the speed of sound
in the medium and the time of travel, Sonom
metrics software
determined the distance between the two sensors
s
with an
accuracy of 30m. Six sensors were suture to the adventitia, the outside layer of the aorta.
The sensors were attached in pairs, on either side of the
aorta. First pair of sensors was placed at 31m
mm from the cut
edge. Second and third pair of sensors was then sutured at
22mm and 30mm, respectively, from the preceding
p
pair.
This position of the sensors allowed trackin
ng the axial and
transverse changes of the vessel. This po
osition allowed
calculating volumetric strain, one of the paraameters used in
the study. It is defined as the ratio of the chaange of volume
to the initial volume of the aorta.
It should be noted that volumetric strain
n is a structural
property rather than material property of the aorta but is still
suitable for the purpose of this study.

Fig. 2 Sensor Position Diagram

To determine the pressure insidde the vessel fiber optic
pressure sensors (FISO Technologgies Inc., Canada, FOPMIV) were used. With a pressure aaccuracy of 0.1kPa, these
sensors are designed around Fabry--Perot interferometer and
use fiber optic cables to send the signal. One sensor was
placed at both ends of the vessel. Figure 3 below is the
photograph of the sensor position.

Displacement Sensoors

Descending Aorta

Pressure Sensors

Fig. 3 Sensor Poosition

Aorta was then attached to two tubes, one of which was
sealed, while the other had a connnection available for the
supply tank. Next, aorta was placedd inside the display case

IFMBE Proceedings Vol. 32

382

V.V. Romanovv, K. Darvish, and S. Assari

and coupled to the supply tank. Saline was purged

p
through
the system, and then poured into the display
y case submerging the aorta. Figure 4 gives an example of th
he experimental
set up.
The pressure was supplied through the airr regulator from
(Control Air Inc, USA, 550X). The regulatorr had a range of
0 psig to 30 psig (0kPA to 206.8kPa) which linearly corresponded to a current input 0mA to 20mA
A. Repeatability
accuracy of the regulator was 2.07kPa. Th
he pressure was
applied in the form of a sinusoidal wave with
w a range of
7kPa to 76kPa and a frequency range of .5H
Hz to 5Hz. This
specific pressure range was chosen to fin
nd the material
properties that account for the maximum systolic and the
failure pressures.
LabView program was used to control th
he current input
and to record pressure output from the prressure sensors.
Sonometrics software was used to record th
he deformation
change of the vessel. To insure that the two
o programs recorded at the same time, a trigger was used to
o active both of
the programs. The entire experiment was also recorded
using a high speed camera capable record
ding at 2200fps
(Vision Research, USA). The movie clips weere used to keep
track of the experiment.

paths. This hysteresis represents thhe energy loss per cycle

and it is almost independent of the strain rate within several
decades of the strain rate variation. [7]
This independency of the tissuee to the strain rate, can
serve as the basis for the simplificaation of the material behavior for the initial modeling proccess. That is if the stressstrain relationship is truly indepenndent of the strain rate,
then the material can be treated aas nonlinear elastic, also
known as hyperelastic or pseudoelasstic. [7]
One of the more accepted ways to describe the stressstrain relationship of a hyperelastic material is to utilize the
exponential form of the strain energgy function as the following. [8]
2

exp

(1)

The constants C (with units of stresss) and , , (dimensionless) are the material constantss.
,
are the strains
corresponding to a pair of stresses
,
chosen to the be
the offset of the dynamic oscillationns.
The strains are calculated as definned by Green:
1
2

1 ;

1
2

(2)

where
, are the stretch ratios oof the blood vessel in the
circumferential and axial directions..
For this research, the radial stress can be considered negligible comparing to the stresses inn the axial and circumferential directions. [8, 9] Thereforee, the arterial wall was
considered as a two-dimensional body subjected only to
, , which are defined as follow
ws
(3)

(4)
Stresses
, can be found using tthe deformation data from
the displacement sensors and utilizingg Lames theory. [10]

Fig. 4 Experimental Set Up

B. Hyperelasticity
Biological tissues have complex structurees and are capable of undergoing large deformations. Fu
urthermore, the
stress-strain relationship of the tissue is in general nonlinear. If the slope, or Youngs Modulus, off a stress-strain
curve is to be calculated one would see that with low loading the modulus is lower than at high loading
g. [7]
Another characteristic of the biological tisssue is the hysteresis that would occur between the loading
g and unloading

III. RESULTS AND DDISCUSSION

A. Initial Result Analysis
As shown in Figure 5, where thhe results for two sample
frequencies are shown, a hysteresis behavior was seen in the
pressure-volumetric strain graphs. T
The change of area inside
the graph is proportional to the energgy loss and is representative of the internal friction (dampinng effect) of the material.

IFMBE Proceedings Vol. 32

Characterization of Material Properties of Aorta from Oscillatory Pressure Tests

80

40
35
Linear Modulus

The material also exhibits frequency dependency, as can be

seen by the change of slope in the graph from low to high
frequency. These are the characteristics of the viscoelastic
material. [11]

383

30
25
20
15

70

10

Pressure (kPa)

60

0.1

1
Linear Modulus

50
40

10

Frequency (Hz)

30

Fig. 6 Linear Modulus versus Frequency

20
10
35

0
1.7

1.9
0.5 Hz

2.1

2.3
4 Hz

2.5

30

2.7
Phase Shift (deg)

1.5

Volumetric Strain

Fig. 5 Volumertic Strain versus Pressure

The change in average strain that can be seen from the .5
Hz frequency to the 5 Hz can be attributed to the relaxation
of the material. [11] Preconditioning procedure for the test
only included preconditioning the material once before the
entire experiment. However, it is possible that the time
spent between changing frequencies was sufficient for the
material to re-set back to its original state and required preconditioning once again. For future experiments, preconditioning protocol will be changed.
To study the frequency dependency of the material properties, Fast Fourier Transform (FFT) was used in LabView
to extract the fundamental amplitude and phase for both
pressure and volumetric strain. A linear modulus has been
defined as the ratio of the pressure amplitude to the amplitude of volumetric strain at each frequency. The results
showed an increase in linear modulus and phase shift as the
frequency increased (Figures 6 and 7). This stiffening behavior confirmed the viscoelastic assumption.
The results of this study will be used to optimize and validate a finite element model that is being developed in
parallel to this research.

25
20
15
10
5
0
0.00

1.00

2.00

3.00
4.00
Phase Shift

5.00

6.00

Frequency (Hz)

Fig. 7 Phase Shift versus Frequency

B. Hyperelastic Modeling
The hyperelastic constitutive model was successful in
modeling stress in both axial and circumferential directions
at low frequencies, where the phase shifts between the stress
and the strain were small. (Figure 7 and 8) The constants
used for a sample hyperelastic model shown below are
listed in Table 1.
Table 1 Constants of sample hyperelastic model
Frequency
.5

C
1.87E+01

5.7284

2.14E-05

4.8635

As mentioned above, however, with higher strain rate

phase shift increases, a condition that could not be predicted
with the hyperelastic model alone. For future work, a quasilinear viscoelastic constitutive model will be utilized to
model aorta to account for the phase shift in addition to the
nonlinear elastic response.

IFMBE Proceedings Vol. 32

384

V.V. Romanov, K. Darvish, and S. Assari

250

Stress (kPa)

200
R2=0.910619
150
100
50
0
0

4
Sth 1

Sth 1 (model)

Time (sec)

Stress (kPa)

Fig. 8 Circumferential Stress

160
140
120
100
80
60
40
20
0

R2=0.915655

4
6
Sz 1
Sz 1 (model)
Time (sec)

2. Richens D, Field M, Neale M, Oakley N (2002), The Mechanism of

Injury in Blunt Traumatic Rupture of the Aorta. European Journal of
Cardio-Thoracic Surgery, pp. 288-293.
3. Katyal, D., McLellan, B., Brenneman, F., Boulanger, Bernard R.
Sharkey, P., Waddell, J., (1997) Lateral Impact Motor Vehicle
Collision: Significant Cause of Blunt Traumatic Rupture of the
Thoracic Aorta. The Journal of Trauma; Injury, Infection, and Critical
Care, Vol. 42, pp. 769-772.
4. The Descending Aorta, Yahoo! Education,Yahoo! Inc, n.d., Web,
23 April, 2010
5. Rindfleisch, E. (1893), Zur Entstehung und Heilung des Aneurysma
Dissecans Aortae. Virschows Arch Pathol Anat, Vol. 131, pp. 374378.
6. Crick S., Sheppard M., Ho S., Genstein L., Anderson R., (1998),
Anatomy of the Pig Heart: Comparison with Normal Human Cardiac
Structure. Journal of Anatomy, Vol. 193, pp. 105-119.
7. Fung, Yuan-Cheng., (1975) On Mathematical Models of Stress-Strain
Relationships for Living Soft Tissues. Riga : Plenum Publishing
Corporation, UDC 611.08:539.4.
8. Fung, Y.C., Fronek, K., Patitucci P., (1979) Pseudoelasticity of
Arteries and the Choice of its Mathematical Expression. American
Journal of Physiology, 1979, Vol. 237.
9. Bass, C.R., Darvish K, Bush B, Crandall JR, Srinivasan SC, Tribble
C, Fiser S, Tourret L, Evans JC, Patrie J, Wang C. (2001) Material
Properties for Modeling Traumatic Aortic Rupture.: Stapp Car Crash
Journal. Vol. 45.
10. Singh D.K., (2008) Strength of Materials, Boca Raton: CRC Press,
ISBN-10: 1 42006 916 0
11. Fung, Yuan-Cheng., (1993) Biomechanics, Mechanical Properties of
Living Tissue. New York: Sringer-Verlag New York, ISBN 0-38797947-6

Author:
Institute:
Street:
City:
Country:
Email:

Fig. 9 Axial Stress

ACKNOWLEDGEMENT
The support for this study was provided by the NHLBI
under Grant Numbers K25HL08651201A2 and Temple
University College of Engineering.

REFERENCES
1. Bertrand S, Cuny S, Petit P, Trosseille X, Page Y, Guillemot H, Drazetic P. (2008), Traumatic Rupture of Thoracic Aorta in Real-World
Motor Vehicle Crashes. Traffic Injury, Vol. 9, pp. 153-161.

IFMBE Proceedings Vol. 32

Kurosh Darvish
Temple University
1947 North 12th Street
Philadelphia
USA
kdarvish@temple.edu