ARTICLE IN PRESS

Journal of Biomechanics 41 (2008) 22532262

www.elsevier.com/locate/jbiomech
www.JBiomech.com

Inuence of angular accelerationdeceleration pulse shapes on regional

brain strains
Narayan Yoganandan, Jianrong Li, Jiangyue Zhang, Frank A. Pintar, Thomas A. Gennarelli
Department of Neurosurgery, Medical College of Wisconsin and VA Medical Center, 9200 West Wisconsin Avenue, Milwaukee, WI 53226, USA
Accepted 14 April 2008

Abstract

Recognizing the association of angular loading with brain injuries and inconsistency in previous studies in the application of the
biphasic loads to animal, physical, and experimental models, the present study examined the role of the accelerationdeceleration pulse
shapes on region-specic strains. An experimentally validated two-dimensional nite element model representing the adult male human
head was used. The model simulated the skull and falx as a linear elastic material, cerebrospinal uid as a hydrodynamic material, and
cerebrum as a linear viscoelastic material. The angular loading matrix consisted coronal plane rotation about a center of rotation that
was acceleration-only (4.5 ms duration, 7.8 krad/s/s peak), deceleration-only (20 ms, 1.4 krad/s/s peak), accelerationdeceleration, and
decelerationacceleration pulses. Both biphasic pulses had peaks separated by intervals ranging from 0 to 25 ms. Principal strains were
determined at the corpus callosum, base of the postcentral sulcus, and cerebral cortex of the parietal lobe. The cerebrum was divided into
17 regions and peak values of average maximum principal strains were determined. In all simulations, the corpus callosum responded
with the highest strains. Strains were the least under all simulations in the lower parietal lobes. In all regions peak strains were the same
for both monophase pulses suggesting that the angular velocity may be a better metric than peak acceleration or deceleration. In
contrast, for the biphasic pulse, peak strains were region- and pulse-shape specic. Peak values were lower in both biphasic pulses when
there was no time separation between the pulses than the corresponding monophase pulse. Increasing separation time intervals increased
strains, albeit non-uniformly. Acceleration followed by deceleration pulse produced greater strains in all regions than the other form of
biphasic pulse. Thus, pulse shape appears to have an effect on regional strains in the brain.
Published by Elsevier Ltd.

Keywords: Rotational acceleration; Brain trauma; Finite element model; Strain; Pulse shape

1. Introduction other modes, specically, effects of inertial angular accele-

Angular head motions have been causally linked with
diffuse brain injuries for over six decades. Based on
fundamental mechanics, angular acceleration has been
attributed to be more responsible than translation motion
for the ensuing pathology (Holbourn, 1943). Early experi-
mental animal studies were focused on whiplash-type
injuries, due to (hyper) extension of the headneck complex
from motor vehicle rear impacts (Ommaya et al., 1966,
1964). Authors of these studies analyzed the biomechanics
of cerebral concussion and laid a framework to investigate
Corresponding author. Tel.: +1 414 384 3453; fax: +1 414 384 3493.
E-mail address: yoga@mcw.edu (N. Yoganandan).
rations on brain motions and injuries.
Examinations of animal, physical, and numerical mod-
eling studies indicate that the externally applied angular
loading generally follows a biphasic pulse consisting of
varying accelerationtime followed by decelerationtime
histories with different separation time intervals (Fig. 1).
The rationale for the selection of the biphasic loading pulse
was driven by the technology available for the design of the
experimental device rather than any specic application,
motor vehicle-related or other events. The rst two loading
apparatus (head accelerating devices, HAD-I and HAD-
II), designed for inducing constrained motions to the head
of small-size animals consisted of a cylindrical cam cut on
the face (HAD-I) or periphery (HAD-II) of a y wheel
(Higgins and Schmall, 1967). The biphasic pulse was

0021-9290/$ - see front matter Published by Elsevier Ltd.

doi:10.1016/j.jbiomech.2008.04.019
 ARTICLE IN PRESS
2254 N. Yoganandan et al. / Journal of Biomechanics 41 (2008) 22532262
Fig. 1. Accelerationtime followed by deceleration-time histories showing the separating time interval along the abscissa between the two pulses.
approximately triangular in shape, and peak deceleration
magnitudes were equal to approximately one-half of peak
acceleration, with 3 ms separation time interval. Typical
durations of acceleration and deceleration pulses were 1.8
and 3.3 ms. This device was used to apply sagittal angular
accelerationdecelerations to squirrel monkeys, producing
diffuse brain injuries (Gennarelli et al., 1971, 1972).
To test larger size animals (macaque and Rhesus
monkeys, baboons), the later series of studies used a
HYGE pneumatic actuator coupled externally to a cam-
linkage mechanism to deliver a biphasic pulse (Abel et al.,
1978). Although typical pulses from this device had similar
separation time intervals between acceleration and decel-
eration pulses (4 ms) as the HAD device, peak acceleration
magnitudes were lower than peak deceleration magnitudes.
Acceleration and deceleration pulses had durations of 6
and 3.5 ms (Abel et al., 1978).
To allow independent control of the applied external
biphasic pulse, instead of the cam mechanism, metering
pins were used to improve the HYGE device (Thibault and
Gennarelli, 1990). Applied angular acceleration pulses had
0 ms separation time interval, peak acceleration magni-
tudes were less than one-half of peak deceleration
magnitudes, and acceleration and deceleration pulses had
pulse durations of 9 and 7 ms (Thibault and Gennarelli,
1985a, b). In another study that used the same device, the
applied one-half sinusoidal waveform was such that the
peak deceleration was approximately three times peak
acceleration (Gennarelli et al., 1987). This device has been
used to study diffuse axonal injuries in miniature pigs by
applying the pulse over 2030 ms (Smith et al., 1997). In
another miniature pig study, the applied pulse was less than
20 ms (Miller et al., 1998). A custom device used to deliver
loading to experimental rabbits used a biphasic pulse with
no separation time, and the acceleration peak was almost
10 times the deceleration peak with durations from 0.9 to
1.0 ms (Gutierrez et al., 2001).
To better understand deformations sustained by the
brain from controlled angular loading to experimental
animals, gel-based physical models were subjected to
biphasic pulses using the HYGE device and motions were
tracked optically (Margulies et al., 1985, 1990; Meaney et
al., 1994, 1990). The 22 ms duration acceleration pulse was
followed by 6 ms duration deceleration pulse, and the peak
magnitude of deceleration was approximately six times
peak acceleration, with 0 ms separating time interval
(Meaney et al., 1990). More recently, in a series of studies
conducted to investigate brain mechanics, physical model
tests used a biphasic pulse with peak accelerations
approximately six times peak decelerations with 90 ms
separating time interval, and pulse durations of 4.5 and
20 ms for the acceleration and deceleration phases (Brad-
shaw et al., 2001; Ivarsson et al., 2002, 2000). Finite
element modeling studies have used continuous (separation
time 0 ms) biphasic haversine pulses with identical peak
amplitudes and 30 ms combined pulse duration (Zhang et
al., 2001). Another numerical modeling study used squared
angular sinusoidal acceleration pulse with 5 ms duration
(Kleiven, 2006). All these data were used to dene human
tolerance by some scaling method (usually geometric/mass)
that neglected differences in pulse shape.
Peak angular acceleration and deceleration have been
used in the cited studies to describe injuries such as diffuse
axonal trauma and acute subdural hematoma sustained by
nonhuman primates and miniature pigs (Gennarelli et al.,
1979, 1981, 1987; Kotapka et al., 1991; Lee et al., 1987;
Miller et al., 1998). Animal studies have suggested that
single metric such as peak angular acceleration may not
fully characterize the diffuse brain injury spectrum, from
concussion to acute subdural hematoma to more severe
diffuse axonal injuries, implying region-specic responses
to external mechanical stimulus. Increasing peak angular
accelerations produced mild or more severe form of
concussion in non-human primates, while acute subdural
hematoma occurred with increasing acceleration levels, and
increasing pulse duration with constant acceleration
eliminated the hematoma (Gennarelli and Thibault,
1982). However, increasing angular accelerations at a
greater rate with longer pulse durations resulted in acute
subdural hematoma. Physical modeling studies have also
identied different candidates to describe differing types of
brain injuries, implying region-specic response of the
 ARTICLE IN PRESS
N. Yoganandan et al. / Journal of Biomechanics 41 (2008) 22532262
human brain (Bradshaw et al., 2001). Because injuries from
animal experiments are coupled with angular loading via
parameters such as strain, it is important to know how the
strain eld changes with alterations in the loading prole.
It is also important to know how changes in the
acceleration prole affect predicted human tolerances.
The present study was designed to seek the fundamental
answers to these questions, and also to determine the
magnitude of error that might be present in extrapolating
animal studies to human tolerances. Specically, the
objective was to use a validated computational model to
determine regional brain deformations under different
proles of accelerationdeceleration pulses with varying
separation time intervals.
2. Methods
The study was conducted using a nite element model of the human
head. The anatomical measurements of the model representing an adult
male was extracted from an experimental study (Bradshaw et al., 2001).
The two-dimensional model consisted of a cylindrical aluminum skull and
falx, cerebrospinal uid, and cerebrum (Fig. 2). The skull was 192 mm in
internal diameter, the falx was 8 mm wide and 60 mm deep, and the
cerebrospinal uid layer was a three-quarter circle with thickness of
25 mm. The skull and falx were assumed to be linear elastic with Youngs
modulus of 70.3 GPa, Poissons ratio of 0.345, and density of 2700 kg/m3
(Meyers and Chawla, 1984). A hydrodynamic material model, governed
by the MieGruneisen equation of state in linear Hugoniot form, was used
for the cerebrospinal uid. A linear viscoelastic material model was
adopted for the cerebrum. Material properties of these constituents were
obtained from literature (Bradshaw et al., 2001). The nite element model
(Fig. 2), meshed using a preprocessor (MSC Inc., Santa Ana, CA),
consisted of 4650 elements and 4595 nodes. A nonlinear explicit nite
element software (ABAQUS version 6.5, HKS Inc., Providence, RI) was
used to extract the output from the numerical stress analysis. The right-
handed rectangular cartesian axis of reference was used with the origin at
the rotation center of the model. The outer nodes of the skull were rotated
about the center of geometry. The following simulation matrix (Fig. 3) was
Fig. 2. Left: Structures and locations used to extract the stress analysis output. Right: Finite elements mesh.
2255
used to determine the role of the accelerationdeceleration pulse and
separation time interval.
A. Acceleration versus decelerationeffect of mono-phase pulse: the
acceleration pulse was 4.5 ms in duration with 7.8 krad/s/s peak
magnitude. The deceleration pulse was 20 ms in duration with
1.4 krad/s/s peak magnitude. The change in angular velocity was the
same for both mono-phase pulses. The pulse characteristics were the
same as used in the experimental study (Bradshaw et al., 2001).
B. Acceleration followed by decelerationeffect of biphasic profile with
025 ms separation time intervals: the acceleration and deceleration
pulses were identical to case A in terms of duration and peak except
that the prole was biphasic, and the separation time intervals were
incremented in steps of 5 ms for each simulation.
C. Deceleration followed by accelerationeffect of biphasic profile with 0
to 25 ms separation time intervals: the biphasic acceleration and
deceleration pulses were identical to case B except that the pulse
sequence was reversed.
Temporal maximum principal strain distributions were obtained at the
three components: the corpus callosum, base of the postcentral sulcus, and
cerebral cortex of the parietal lobe (Fig. 2). In addition, the cerebrum was
divided into 17 regions, and peak values of average maximum principal
strains were determined for each region.
3. Results
3.1. Acceleration versus decelerationeffect of mono-phase
pulse
Temporal maximum principal strains at the three
components for the deceleration-only pulse occurred later
than those from the acceleration-only pulse (Fig. 4). The
strain at the corpus callosum was lower in the deceleration-
only pulse than in the acceleration-only pulse. However,
strains at the base of the postcentral sulcus and cerebral
cortex of the parietal lobe were similar under both pulses.
In all 17 regions, peak strains were slightly greater in the
 ARTICLE IN PRESS
2256 N. Yoganandan et al. / Journal of Biomechanics 41 (2008) 22532262
Angular acceleration (krad/s/s)

Angular acceleration (krad/s/s)

8 8

6 6

4 4

2 2

0 0
0 10 20 0 5 10 15 20
Time (ms) Time (ms)

Angular acceleration (krad/s/s)

Angular acceleration (krad/s/s)

2
8
0
6 0 5 10 15 20 25 30
-2 Time (ms)
4
-4
2
-6
0
0 5 10 15 20 25 30
-2 -8
Time (ms)

Fig. 3. Simulated pulses. Left top: acceleration-only pulse, right top: deceleration-only pulse, left bottom: acceleration followed by deceleration biphasic
pulse, with varying separate time intervals, and right bottom: deceleration followed by acceleration pulse with varying separate time intervals.

acceleration-only pulse than the decelerationonly pulse, was followed by regions 911 and 1 and 4. Regions 57
and the highest difference occurred in region 7 (Fig. 5). sustained the highest strains in both pulses (Fig. 7). This
region was least affected by the separation time interval.

3.2. Acceleration followed by decelerationeffect of

3.3. Deceleration followed by accelerationeffect of
biphasic profile with varying separation time intervals
biphasic profile with varying separation time intervals
Comparison of peak principal strains between the
In general, responses were similar between this series of
acceleration-only pulse and acceleration-deceleration pulse
simulations and the previous biphasic prole series with
with no separation time interval indicated no differences at
leading acceleration prole, i.e., no differences at the
the corpus callosum (Fig. 6). However, strains at the base
corpus callosum for no separation time interval and lower
of the postcentral sulcus and cerebral cortex of the parietal
strains in the biphasic than in the mono-phase pulse at the
lobe were lower in the biphasic simulation than in the
base of the postcentral sulcus and cerebral cortex of the
acceleration-only pulse case. In addition, the biphasic pulse
parietal lobe (Fig. 8); existence of twin peaks in the
responded with two peaks (Fig. 6). This phenomenon was
biphasic prole (Fig. 8); and lower peak magnitudes in all
true for all separation time intervals at the corpus
17 regions at all separation time intervals than the mono-
callosum, base of the postcentral sulcus, and cerebral
phase pulse together with regional variations in the
cortex of the parietal lobe. Further, the rst peak strains
attainment of greater peaks with increasing time intervals
increased with increasing separation times, reaching values
(Fig. 9).
close to the acceleration-only pulse case. In contrast, the
rst peak strains in the corpus callosum region did not
change with separation time intervals. Strain amplitudes of 4. Discussion
the second peaks were smaller than the rst, except at 20 ms
time interval, for the peak at the cerebral cortex of the The impetus for this study arose because impulsive
parietal lobe. Peak strains in all 17 regions were pulse loading to the head is related to brain injury; in addition, a
shape- and region specic (Fig. 7). Regions 57 responded majority of experimental animal and physical modeling
with almost the same magnitudes of strains with separation studies delivered different pulses resulting in varying
times greater than 10 ms, as the monophase pulse. severities of trauma (see Introduction). Furthermore,
However, regions 1 and 4, 8 and 12, and 911 were more previous studies have been inconsistent regarding the
gradual in attaining the same level of strain as the specic biomechanical metric (angular acceleration peak,
monophase pulse (Fig. 7ad). Furthermore, regions 8 and maximum deceleration, rotational velocity change, etc.), or
12 responded with the least magnitude of strain, and this even the phase of loading, acceleration or deceleration,
 ARTICLE IN PRESS
N. Yoganandan et al. / Journal of Biomechanics 41 (2008) 22532262 2257

0.5 0.44

0.4
0.33

0.3
Strain

Strain
0.22
0.2

0.11
0.1

0 0
0 20 40 60 80 100 0 20 40 60 80 100
Time (ms) Time (ms)

0.3

0.2
Strain

0.1

0
0 20 40 60 80 100
Time (ms)

Fig. 4. Time histories of principal strains at the corpus callosum (left), base of the postcentral sulcus (right top), and cerebral cortex of the parietal lobe
(right bottom) for acceleration-only (solid curves) and decelerationonly pulse (dashed curves).

0.5

0.4
Peak strain

0.3

0.2

0.1

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Region

Fig. 5. Region-specic average maximum principal strains in the brain for acceleration-only (left side bars) and deceleration only pulses. See Fig. 2 for
region details.

responsible for the ensuing pathology (Gennarelli et al.,
1979, 1981, 1987; Gennarelli and Thibault, 1982; Kotapka
et al., 1991; Lee et al., 1987; Miller et al., 1998). Although it
has been recognized that head motions are related to
external loading, angular accelerations in the present
context, a systematic study of the effects of differing
angular acceleration-deceleration loading proles on re-
gional and temporal brain kinematics has not been
conducted (Yoganandan et al., 1995, 1998). The need for
such an analysis was noted in an earlier report (Thibault
and Gennarelli, 1985a). Through the use of parametric
analyses and computational modeling, this study
 ARTICLE IN PRESS
2258 N. Yoganandan et al. / Journal of Biomechanics 41 (2008) 22532262

0.5 0.44
Acceleration-only
0 ms Acceleration-only
10 ms 0 ms
10 ms
0.4 20 ms
20 ms
0.33

0.3
Strain

Strain
0.22
0.2

0.11
0.1

0 0
0 20 40 60 80 100 0 20 40 60 80 100
Time (ms) Time (ms)

0.3 Acceleration-only 0 ms
10 ms 30 ms

Strain 0.2

0.1

0
0 20 40 60 80 100
Time (ms)

Fig. 6. Time histories of maximum principal strains at the corpus callosum (left), base of the postcentral sulcus (right top), and cerebral cortex of the
parietal lobe (right bottom) for acceleration-only pulse (thicker curves) and biphasic, i.e., acceleration followed by deceleration pulse with varying
separating time intervals (thin curves).

5
0.35 0.56
6
0.52 7
0.3
Peak principal strain

Peak principal strain

0.48
0.25
0.44
0.2
1 0.4
4
0.15 0.36

0.1 0.32
Accel only 0 ms 5 ms 10 ms 15 ms 20 ms 25 ms Accel only 0 ms 5 ms 10 ms 15 ms 20 ms 25 ms
Acceleration-deceleration separation time interval Acceleration-deceleration separation time interval

0.3 0.5
8 9
0.25 12 0.45 10
Peak principal strain

Peak principal strain

11
0.2 0.4

0.15 0.35

0.1 0.3

0.05 0.25

0 0.2
Accel only 0 ms 5 ms 10 ms 15 ms 20 ms 25 ms Accel only 0 ms 5 ms 10 ms 15 ms 20 ms 25 ms
Acceleration-deceleration separation time interval Acceleration-deceleration separation time interval

Fig. 7. (ad) Region-specic average maximum principal strains in the brain for acceleration-only and acceleration-deceleration pulses with varying
separating time intervals. See Fig. 2 for region details.
 ARTICLE IN PRESS
N. Yoganandan et al. / Journal of Biomechanics 41 (2008) 22532262 2259

0.44 0.4
Deceleration-only Deceleration-only
0 ms 0 ms
10 ms 10 ms
20 ms 20 ms
0.33 0.3
Strain

Strain
0.22 0.2

0.11 0.1

0 0
0 20 40 60 80 100 0 20 40 60 80 100
Time (ms) Time (ms)

0.3 Deceleration-only
0 ms
10 ms
20 ms

0.2
Strain

0.1

0
0 20 40 60 80 100
Time (ms)

Fig. 8. Time-histories of maximum principal strains at the corpus callosum (left), base of the postcentral sulcus (right top), and cerebral cortex of the
parietal lobe (right bottom) for decelerationonly pulse (thick curves) and biphasic, i.e., deceleration followed by acceleration pulse with varying
separating time intervals (thin curves).

attempted to delineate the role of the input angular
acceleration-deceleration proles on strain distributions
in the idealized human brain.
A simplied nite element model was used in the analysis
because it was validated using experimental data, wherein
temporal motions at the corpus callosum, base of the
postcentral sulcus, and cerebral cortex of the parietal lobe
of the physical model were determined using optical
techniques (Bradshaw et al., 2001; Li et al., 2007). This
was the reason for choosing the three specic components
in the current analysis. Because of the condence achieved
in matching the biomechanical motion data of these
components, it was deemed appropriate to extend the
model to determine regional strains and as a function of
different loading proles. Evaluations of strains have a
biomechanical basis as this metric has been related to brain
trauma (Gennarelli, 1985). In the current idealized model, a
general oneoneone correspondence existed between the
chosen regions and the human brain; regions 1 and 4
represented the superior parietal, and 2 and 3 its
parasagittal areas; 5 and 7 represented the bilateral areas
of the areas of the corpus callosum centrally located in
region 6; 8 and 12 represented the lower parietal/temporal
lobes; 911 the basal ganglia, and 1416 its inferior aspects;
and 13 and 17 represented the inferior aspects of the
temporal lobes. Therefore, strains are representative of
these anatomical regions with angular loads applied in the
coronal plane, most susceptible to brain trauma from
inertia-induced impulsive loading (Gennarelli et al., 1987).
Simulations in all cases, i.e., from acceleration-only and
deceleration-only to both types of biphasic pulses indicated
that the corpus callosum is the highest strained component.
This nding parallels previous research in humans and
experimental animals (Adams et al., 1982; Gennarelli et al.,
1982). It should be noted that the exact prole of the
applied pulse used in experimental devices was not based
on any specic crash scenarios. For example, during the
development of HAD devices, acknowledging that the
study used military sponsorship, no mention of the basis
was made for choosing the biphasic pulse as a design
objective. Airbag loading in motor vehicle frontal impacts
generally have a long duration pulse in contrast to impact
of an unrestrained, or an out-of-position occupant with
vehicle interiors such as pillars and roof headers (NHTSA,
 ARTICLE IN PRESS
2260 N. Yoganandan et al. / Journal of Biomechanics 41 (2008) 22532262

0.35 0.56
1 5
4 0.52
0.3 6
Peak principal strain

Peak principal strain

0.48 7
0.25
0.44
0.2
0.4
0.15 0.36

0.1 0.32
Decel only 0 ms 5 ms 10 ms 15 ms 20 ms 25 ms Decel only 0 ms 5 ms 10 ms 15 ms 20 ms 25 ms
Deceleration-acceleration separation time interval Deceleration-acceleration separation time interval

0.3 0.5
8 9
0.25 12 0.45 10
Peak principal strain

Peak principal strain

11
0.2 0.4

0.15 0.35

0.1 0.3

0.05 0.25

0 0.2
Decel only 0 ms 5 ms 10 ms 15 ms 20 ms 25 ms Decel only 0 ms 5 ms 10 ms 15 ms 20 ms 25 ms
Deceleration-acceleration separation time interval Deceleration-acceleration separation time interval

Fig. 9. (ad) Region-specic average maximum principal strains in the brain for deceleration-only and decelerationacceleration pulses with varying
separating time intervals. See Fig. 2 for region details.

2003). Impact mode also tends to change the pulse prole, peak angular acceleration. Although this observation may
for example, side impacts have less ride-down than be applicable on a regional basis, as shown in Figs. 6 and 8,
frontal impacts. Occupant interaction with pretensioners longer duration pulses may induce secondary strains with
and load limiting belts may also affect the input pulse. peaks larger than the primary peak in more localized areas.
Active systems currently under research and development This topic needs further study.
as a safety technology for modern vehicles can also Peak accelerations or decelerations were different in the
inuence pulse shape. The current parametric study acceleration-only and deceleration-only pulses. However,
analyzed the effects of different angular loading proles under both pulses, strains in all regions were very similar
on internal brain mechanics using a simplied approach. (less than 10% change), suggesting that the angular
Translation effects were not introduced because experi- velocity may be a better metric instead of peak accelera-
mental and analytical modeling results have shown that tion, or deceleration amplitude for the mono-phase loading
strains are primarily controlled by angular loading even in condition. From this perspective, if real world events
the presence of translational accelerations (Holbourn, impart mono-phase pulses, as a rst step, angular velocities
1943; Huang et al., 1999; Lee et al., 1987; Ommaya and may be used to characterize brain injury biomechanics and
Gennarelli, 1974; Unterharnscheidt and Higgins, 1969; pathology. Under this scenario, because strains were not
Zhang et al., 2006). region dependent, the single velocity metric may be an
In general, peak principal strains increased with increas- appropriate hallmark for the likely different neurological
ing separation time interval for both types of biphasic outcomes.
pulses. At the widest separation time of 25 ms used in the In contrast, under both types of biphasic loading
current study, peak strains attenuated towards the case of a scenarios, strains were region- and pulse shape specic.
mono-phase pulse, i.e., acceleration-only pulse for the Secondary peaks, described in the earlier section, were also
acceleration followed by the deceleration pulse, and found in previous nite element modeling studies wherein
decelerationonly for the deceleration followed by the scaled and unscaled experimental animal model pulses were
acceleration pulse. This attenuating process tends to applied to the human brain models with homogeneous and
support the thesis that with a long interval, the two pulses inhomogeneous material properties (Abel et al., 1978;
can be considered as mutually independent with secondary Zhou et al., 1996, 1995). Although the corpus callosum
interactive effects. Thus, strains responsible for localized consistently responded with the highest strains under all
brain trauma in experiments applying long duration proles, magnitudes of strains in all regions were greater
biphasic pulse loading may be dictated by the primary in the acceleration-deceleration prole than in the decel-
pulse prole. This nding may offer a rationale for using eration-acceleration prole primarily because of the
the rst peak in such instances for correlating brain injury geometry, material properties of the brain material, and
with the primary angular biomechanical metric, example, a type of loading. Strain peaks occurring subsequent to the
 ARTICLE IN PRESS
N. Yoganandan et al. / Journal of Biomechanics 41 (2008) 22532262
completion of the rst pulse indicate that the brain
response lags the applied input. Similar delayed strain
responses have been reported in literature (Ivarsson et al.,
2002). The amount of lag not only depends on the
constitutive properties of the brain tissue but also on the
separation time interval. If the second pulse follows the
rst pulse quickly, peak responses decrease considerably,
again because of the effect of the onset of the negative
loading on the already deformed state of the brain.
Increasing the separation time interval tends to allow
initial deformations induced by the rst loading sequence
to grow and attain peak values. When the separation time
interval is large enough, the negative pulse has a secondary
effect. While this phenomenon was true for both biphasic
proles, the actual magnitudes were region specic,
suggesting that even in the current idealized model, strain
distributions are complex. Although experimental animal
studies have identied brain pathology, the role of the peak
acceleration or peak deceleration as a metric responsible
for causing the strain to induce the abnormality is only
inferential; researchers have suggested both peaks to be
likely candidates (Gennarelli and Thibault, 1982; Lee and
Haut, 1989; Zhou et al., 1996, 1995). The present study,
being computational in nature with absolute repeatability,
has delineated the roles of the two proles under specic
conditions and demonstrated that brain strains are both
region- and pulse shape specic.
Because the present analysis representing the adult male
human head was based on a two-dimensional nite element
model extracted from a simplied geometry and anatomy,
it is a rst step in the investigation of region-specic strain
distributions of the in vivo human brain. Further studies
are necessary to account for the effects of complex three-
dimensional anatomy and more realistic brain tissue
properties on strain distributions to better understand
internal kinematics concomitant with externally applied
angular loads.
Conict of interest statement
None of the authors have any nancial and personal
relationships as the study was supported by the Depart-
ment of Neurosurgery Funds and conducted at the VA
Medical Center.
Acknowledgment
This study was supported by VA Medical Research.
References
Abel, J., Gennarelli, T., Segawa, H., 1978. Incidence and severity of
cerebral concussion in the rhesus monkey following sagittal plane
angular acceleration. In: Proceedings of the 22nd Stapp Car Crash
Conference, pp. 3353.
Adams, J.H., Graham, D.I., Murray, L.S., Scott, G., 1982. Diffuse axonal
injury due to nonmissile head injury in humans: an analysis of 45 cases.
Annals of Neurology. 12 (6), 557563.
2261
Bradshaw, D.R., Ivarsson, J., Morfey, C.L., Viano, D.C., 2001.
Simulation of acute subdural hematoma and diffuse axonal injury in
coronal head impact. Journal of Biomechanics. 34 (1), 8594.
Gennarelli, T.A., 1985. The state of the art of head injury biomechanics.
In: Proceedings of the 30th American Association for the Advance-
ment of Automotive Medicine, pp. 447463.
Gennarelli, T.A., Thibault, L.E., 1982. Biomechanics of acute subdural
hematoma. Journal of Trauma 22 (8), 680686.
Gennarelli, T.A., Thibault, L.E., Ommaya, A., 1971. Comparison of
translational and rotational acceleration in experimental cerebral
concussion. In: Proceedings of the 15th Stapp Car Crash Conference,
pp. 797803.
Gennarelli, T.A., Thibault, L.E., Ommaya, A., 1972. Pathophysiologi-
cal responses to rotational and translational accelerations of the
head. In: Proceedings of the 16th Stapp Car Crash Conference,
pp. 296308.
Gennarelli, T.A., Abel, J.M., Adams, H., Graham, D., 1979. Differential
tolerance of temporal and frontal lobes to contusion induced by
angular acceleration In: Proceedings of the 23rd Stapp Car Crash
Conference, pp. 563586.
Gennarelli, T.A., Adams, J.H., Graham, D.I., 1981. Acceleration induced
head injury in the monkey. I. The model, its mechanical and
physiological correlates. Acta Neuropathologica Suppl 7, 2325.
Gennarelli, T.A., Thibault, L.E., Adams, J.H., Graham, D.I., Thompson,
C.J., Marcincin, R.P., 1982. Diffuse axonal injury and traumatic coma
in the primate. Annals of Neurology 12 (6), 564574.
Gennarelli, T.A., Thibault, L.E., Tomel, G., Wiser, R., Graham, D.,
Adams, J., 1987. Directional dependence of axonal brain injury due to
centroidal and non-centroidal acceleration In: Proceedings of the
Stapp Car Crash Conference, New Orleans, LA, pp. 4953.
Gutierrez, E., Huang, Y., Haglid, K., Bao, F., Hansson, H.A.,
Hamberger, A., Viano, D., 2001. A new model for diffuse brain
injury by rotational acceleration: I model, gross appearance, and
astrocytosis. Journal of Neurotrauma 18 (3), 247257.
Higgins, L.S., Schmall, R.A., 1967. A device for the investigation of head
injury effected by non-deforming head accelerations. In: Proceedings
of the 11th Stapp Car Crash Conference, pp. 3546.
Holbourn, A.H., 1943. Mechanics of head injuries. Lancet 2, 438441.
Huang, H.M., Lee, M.C., Chiu, W.T., Chen, C.T., Lee, S.Y., 1999. Three-
dimensional nite element analysis of subdural hematoma. Journal of
Trauma 47 (3), 538544.
Ivarsson, J., Viano, D.C., Lovsund, P., Aldman, B., 2000. Strain relief
from the cerebral ventricles during head impact: experimental studies
on natural protection of the brain. Journal of Biomechanics 33 (2),
181189.
Ivarsson, J., Viano, D.C., Lovsund, P., 2002. Inuence of the lateral
ventricles and irregular skull base on brain kinematics due to sagittal
plane head rotation. Journal of Biomechanical Engineering 124 (4),
422431.
Kleiven, S., 2006. Evaluation of head injury criteria using a nite element
model validated against experiments on localized brain motion,
intracerebral acceleration, and intracranial pressure. International
Journal of Crashworthiness 11 (1), 6579.
Kotapka, M.J., Gennarelli, T.A., Graham, D.I., Adams, J.H., Thibault,
L.E., Ross, D.T., Ford, I., 1991. Selective vulnerability of hippocampal
neurons in acceleration-induced experimental head injury. Journal of
Neurotrauma 8 (4), 247258.
Lee, M.C., Haut, R.C., 1989. Insensitivity of tensile failure proper-
ties of human bridging veins to strain rate: implications in
biomechanics of subdural hematoma. Journal of Biomechanics 22
(67), 537542.
Lee, M., Melvin, J., Ueno, K., 1987. Finite element analysis of subdural
hematoma. In: Proceedings of the 31st Stapp Car Crash Conference,
pp. 6777.
Li, J., Zhang, J., Yoganandan, N., Pintar, F., Gennarelli, T., 2007.
Regional brain strains and role of falx in lateral impact-induced head
rotational acceleration. Biomedical Sciences and Instrumentation. 43,
2429.
 ARTICLE IN PRESS
2262 N. Yoganandan et al. / Journal of Biomechanics 41 (2008) 22532262
Margulies, S., Thibault, L., Gennarelli, T., 1985. A study of scaling and
head injury criteria using physical model experiments. In: Proceedings
of the IRCOBI, pp. 223234.
Margulies, S.S., Thibault, L.E., Gennarelli, T.A., 1990. Physical model
simulations of brain injury in the primate. Journal of Biomechanics 23
(8), 823836.
Meaney, D., Thibault, L., Gennarelli, T., 1990. Physical model of cortical
brain deformation in response to high strain rate loading. In:
Proceedings of the IRCOBI, pp. 215224.
Meaney, D., Smith, D., Shreiber, D., Rahdert, D., Bogen, D., Ross, D.,
McIntosh, T., Gennarelli, T.A., 1994. Prediction of experimental
diffuse axonal injury severity using macroscopic head motion para-
meters. In: Proceedings of the AAAM, Lyon, France, pp. 345357.
Meyers, M., Chawla, K., 1984. Mechanical Metallurgy. Prentice-Hal Inc.,
Englewood Cliffs, NJ.
Miller, R., Margulies, S., Leoni, M., Nonaka, M., Chen, Z., Smith, D.,
Meaney, D., 1998. Finite element modeling approaches for predicting
injury in an experimental model of severe diffuse axonal injury.
In: Proceedings of the 42nd Stapp Car Crash Conference, Tempe, AZ,
pp. 113.
NHTSA, 2003. /www.nhtsa.dot.govS. United States Department of
Transportation.
Ommaya, A.K., Gennarelli, T.A., 1974. Cerebral concussion and
traumatic unconsciousness. Correlation of experimental and clinical
observations of blunt head injuries. Brain 97 (4), 633654.
Ommaya, A.K., Rockoff, S.D., Baldwin, M., 1964. Experimental
concussion; a rst report. Journal of Neurosurgery 21, 249265.
Ommaya, A., Hirsch, A., Martinez, J., 1966. The role of whiplash in
cerebral concussion. In: Proceedings of the 10th Stapp Car Crash
Conference, pp. 314324.
Smith, D.H., Chen, X.H., Xu, B.N., McIntosh, T.K., Gennarelli, T.A.,
Meaney, D.F., 1997. Characterization of diffuse axonal pathology and
selective hippocampal damage following inertial brain trauma in the
pig. Journal of Neuropathology and Experimental. Neurology 56 (7),
822834.
Thibault, L., Gennarelli, T.A., 1985a. Biomechanics of craniocerberal
trauma. In: Becker, D., Povlishock, J. (Eds.), Central Nervous System
Trauma NINCDS, vol. 1. Bethesda, MB, pp. 371389.
Thibault, L.E., Gennarelli, T.A., 1985b. Biomechanics of diffuse brain
injuries. In: Proceedings of the 11th Experimental Safety Vehicles
Conference, Oxford, England, pp. 7985.
Thibault, L.E., Gennarelli, T.A., 1990. Brain injury: an analysis of neural
and neurovascular trauma in the non-human primate. In: Proceedings of
the 34th American Association of Automotive Medicine, pp. 337351.
Unterharnscheidt, F., Higgins, L.S., 1969. Pathomorphology of experi-
mental head injury due to rotational acceleration. Acta Neuropatha-
logica 1969 (12), 200204.
Yoganandan, N., Pintar, F.A., Sances Jr., A., Walsh, P.R., Ewing, C.L.,
Thomas, D.J., Snyder, R.G., 1995. Biomechanics of skull fracture.
Journal of Neurotrauma 12, 659668.
Yoganandan, N., Pintar, F.A., Larson, S.J., Sances Jr., A., (Eds), 1998.
Frontiers in Head and Neck Trauma: Clinical and Biomechanical. IOS
Press, The Netherlands, 743pp.
Zhang, L., Yang, K.H., Dwarampudi, R., Omori, K., Li, T., Chang, K.,
Hardy, W.N., Khalil, T.B., King, A.I., 2001. Recent advances in brain
injury research: a new human head model development and validation.
Stapp Car Crash Journal 45, 369394.
Zhang, J., Yoganandan, N., Pintar, F.A., Gennarelli, T.A., 2006. Brain
strains in vehicle impact tests. Annual Proceedings Association for the
Advancement of Automotive Medicine 50, 112.
Zhou, C., Khalil, T.B., King, A.I., 1995. A new model for comparing
responses of the homogeneous and inhomogeneous human brain. In:
Proceedings of the 39th Stapp Car Crash Conference, pp. 121136.
Zhou, C., Khalil, T., King, A., 1996. Viscoelastic response of the human
brain to sagittal and lateral rotational acceleration by nite element
analysis. In: Proceedings of the IRCOBI, Dublin, Ireland, pp. 3548.