Content uploaded by A. Chawla
Author content
All content in this area was uploaded by A. Chawla on Feb 05, 2015
Content may be subject to copyright.
International Journal of Crashworthiness
Vol. 14, No. 4, August 2009, 339–348
Response of lower limb in full-scale car–pedestrian low-speed lateral impact – influence of
muscle contraction
Anurag Sonia, Anoop Chawlaa∗, Sudipto Mukherjeeaand Rajesh Malhotrab
aDepartment of Mechanical Engineering, Indian Institute of Technology, New Delhi 110016, India; bDepartment of Orthopaedics, All
India Institute of Medical Sciences, New Delhi 110016, India
(Received 21 June 2008; final version received 24 January 2009)
This paper investigates the effect of muscle contraction on lower extremity injuries in car–pedestrian lateral impact. A
full-body pedestrian model with active muscles has been developed. Finite element simulations have then been performed
using the full-body model and front structure of a car. Two pre-impact conditions, that of a symmetrically standing pedestrian,
representing a cadaver and an unaware pedestrian, have been simulated. Stretch-based reflexive action was included in the
simulations for an unaware pedestrian. The results show that due to muscle contraction (1) peak strain in all the knee ligaments
reduces, (2) von Mises stresses in tibia and fibula increase and may fail and (3) knee joint effective stiffness increases by
58% in lateral bending.
Keywords: A-LEMS; lower extremity model; finite element model; dynamic simulation; muscle contraction; standing
posture; car–pedestrian impact; knee injury
Introduction
Pedestrians constitute 65% of the 1.17 million people killed
annually in road traffic accidents worldwide [24]. Epidemi-
ological studies on pedestrian victims have indicated that
together with the head, the lower extremity is the most
frequently injured body region [5, 15]. Pedestrian Crash
Data Study (PCDS) [5] reports that passenger cars have
the biggest share in vehicle–pedestrian accidents. Further,
the front bumper is the major source of injury to the lower
extremity when injuries are caused by a vehicle [15]. This
has posed an interesting challenge for vehicle designers to
design pedestrian-friendly car front structures. To devise
effective pedestrian protection systems, it is essential to
understand the injury mechanism.
So far the lower limb injury mechanism in car–
pedestrian crashes has been studied through tests on hu-
man cadaver specimens [2,6–9] and simulations using val-
idated passive finite element (FE) models [3,13,17, 18, 20,
21]. However, the major shortcoming in these existing ex-
perimental and computational studies is that they do not
account for muscle action. Therefore, effects of pre-crash
muscle contraction on the response of lower limbs in car–
pedestrian crashes remaines unclear.
Of late, Soni et al. [20] have investigated the probable
outcome of muscle contraction using a lower limb (single-
leg) FE model with active muscles (A-LEMS). More re-
cently, Chawla et al. [4] have performed a study using the
A-LEMS and reported that with muscle contraction, the
∗Corresponding author. Email: achawla@mech.iitd.ernet.in
risk of knee ligament failure is likely to be lower than that
predicted through the cadaver tests or simulations with the
passive FE models. However, in these studies a single-leg
model (A-LEMS) has been used and the geometry of the
upper body has not been considered. In addition, an im-
pactor of mass 20 kg is used which is much lighter than a
physical car. As a result, the conditions simulated in these
studies may not correspond to an actual crash.
The present study extends our earlier studies to inves-
tigate the effect of muscle contraction on the response
of lower limbs in a full-scale car–pedestrian lateral col-
lision using FE simulations. For this, a single-leg model
(A-LEMS) has been extended to a full-body model. The
real-world car–pedestrian lateral impact has been simulated
using the full-body model and front structures of a validated
car FE model. A pedestrian standing symmetrically with
legs positioned in a side-by-side stance has been simulated
for two pre-impact pedestrian conditions, i.e. with deac-
tivated muscles and with activated muscles mimicking an
unaware pedestrian. Stretch-based involuntary reflex action
has been included in the simulation for an unaware pedes-
trian. Strains in knee ligaments, von Mises stresses in bones
and tibia displacements for two levels of muscle activation
are then compared to assess the effect of muscle contraction.
Description of A-LEMS
In the present study, A-LEMS has been used. The A-LEMS
includes 42 muscles modelled as 1-D bar elements, in
ISSN: 1358-8265
Copyright C
2009 Taylor & Francis
DOI: 10.1080/13588260902775033
http://www.informaworld.com
340 A. Soni et al.
addition to the passive structures such as the cortical and
the spongy parts of the femur, tibia, fibula, and the patella.
The cortical part of the bones is modelled by shell elements,
while the spongy part is modelled by solid elements. Apart
from these, passive muscle response and skin are also mod-
elled using solid elements and membrane elements respec-
tively. Knee ligaments, anterior cruciate ligament (ACL),
posterior cruciate ligament (PCL) and lateral collateral lig-
ament (LCL) have been modelled using solid elements.
Because of a smaller thickness compared to the width, the
medial collateral ligament (MCL) has been modelled using
the shell elements. The articular capsule, i.e. ‘knee capsule’,
which encloses the knee joint and maintains joint integrity,
has also been included in this model.
Since all the available data is for cadaver tests, the
passive version of the A-LEMS was validated against
loading and boundary conditions reported in [8, 9, 10].
These validation results have been presented in detail in
[20]. The passive model correctly reproduces impactor
forces, knee kinematics and ligament failures reported
from the experiments. Forty-two active muscles, modelled
as 1-D bar elements with the Hill material model, are
subsequently activated to simulate the effect of muscle
contraction. Details regarding activation levels and Hill
model parameters (such as maximum muscle force (Fmax),
optimum muscle length (Lopt), maximum contraction
velocity (a Vmax)), fraction of fast fibres (Cfast) and initial
activation level (Na) are available in [19, 20].
Full-body model development
In the present study, a full-body pedestrian model with
active muscles in both legs has been developed. For this
the A-LEMS [20] was adopted as the base model. In
order to extend the A-LEMS to a full-body model, an
additional leg and the upper extremity were introduced.
FE mesh of the additional leg was obtained through mirror
transformation of the existing lower extremity mesh in the
A-LEMS about the sagittal plane using HYPER-MESHTM.
Material properties of each mirrored segment and the
contact definitions between them were kept identical as
in the A-LEMS. As a result, a two-legged FE model (legs
positioned symmetrically in a side-by-side stance) with 42
muscles in each leg was obtained.
Though the focus in the current study is to investigate
the response of lower extremity, the upper extremity is im-
portant to take into account the effects of mass and moment
of inertia of the upper body. Therefore, it has been decided
to model upper extremity segments using geometrically ac-
curate rigid body definitions. Hence, the upper extremity
segments in standard H-III dummy models available with
PAM-CRASHTM have been separated and integrated with
the two-legged FE model.
The assembled full-body model thus obtained was then
aligned in a symmetric standing posture of the pedestrian
Figure 1. Comparison of (a) alignment of body landmarks and
(b) partial COG positions of the PMALE segments with the cor-
responding positions reported in [23].
whose legs are in a side-by-side stance. It was ensured that
pre-impact loading conditions with respect to the forces and
moments at the knee joint corresponded to the conditions
for a symmetrically standing pedestrian. Two aspects of the
alignment have been considered here: (1) anterior-posterior
position of the knee joint, the hip joint and the shoulder
joint relative to the ankle joint and (2) anterior-posterior
position of partial centre of gravity (COG) above the knee
and the hip joints.
Relative positions of the body segments required to de-
fine the alignment of the symmetric standing posture were
taken from [23]. A series of FE simulations were performed
with the assembled full-body model to acquire a statically
competent alignment between its segments. We will refer to
the aligned full-body model as Pedestrian Model with Ac-
tive Lower Extremity (PMALE). Figure 1 shows the com-
parison of the alignment of body landmarks (Figure 1a)
and positions of partial COG (Figure 1b) of the PMALE
segments with the corresponding positions given in [23].
Simulations for standing posture
Simulation set-up
Figure 2 shows the simulation set-up used in the present
study. Here, the real-world car–pedestrian impact has been
reproduced using the PMALE and front structures of a
validated car FE model. PMALE is configured as standing
International Journal of Crashworthiness 341
Figure 2. Simulation set-up used in the present study.
freely with legs in a side-by-side stance on a rigid plate in a
gravity field. The coefficient of friction between the shoes
and the ground is set to 1.0 as suggested for grooved rubber
on road [12].
Pedestrian accident studies [5, 16] have shown signif-
icant incidence of bone fracture for high-speed impacts.
Since bone fracture unloads the knee joint, we decided to
simulate low-speed impact. On the basis of PCDS [5], which
reports a range of 20–30 kmph for low-speed impacts, an
impact speed of 25 kmph was selected. A car model with a
total mass of 1158 kg (mass of the front structure is 355 kg,
and 803 kg is modelled as the added mass to account for
the remaining car structures) was thus propelled towards the
PMALE to impact its left leg laterally. The PMALE was
placed in front of the car model to interact with the middle
of the bumper, with the bumper height above the ground
such that it corresponds to the car rolling on its tyres.
In the present study, each simulation was run for 100 ms.
For the initial 50 ms (stabilisation duration), the PMALE
was allowed to stabilise under gravity. The car model came
in contact with the PMALE only at the end of the stabilisa-
tion duration (i.e. after the first 50 ms).
Pedestrian pre-impact conditions
Two pre-impact pedestrian conditions, i.e. with deactivated
muscles and with activated muscles (including reflex ac-
tion), for an unaware pedestrian have been simulated in the
present study. We call these conditions cadaveric and reflex
conditions respectively. These conditions differ in terms
of initial activation levels in muscles and whether the re-
flex action is enabled. By enabling the reflex action for a
muscle, the activation level in that muscle rises with time
during the simulation; this increases the force produced by
that muscle.
Cadaveric condition: In this condition, a freely standing
cadaver was simulated. To model a cadaver in FE simula-
tion, all the muscles in PMALE were assigned the mini-
mum value of 0.005 as an initial activation level. The reflex
action was also disabled. As a result, in this condition, ac-
tivation levels in each muscle remained at the minimum
value (i.e. 0.005) for the entire duration of the simula-
tion. Therefore, all the muscles function at their minimum
capacity.
Reflex condition: In this condition, a symmetrically
standing pedestrian who is unaware of an impending crash
has been simulated. To model an unaware standing pedes-
trian, activation levels for maintaining the stability of the
standing posture of a pedestrian in gravity field are the in-
put. Kuo and Zajac [11] report activation levels in lower
limb muscles for near-erect standing posture with flex-
ion/extension about the ankle joint (i.e. ankle strategy)
keeping other joints motionless. It is important to highlight
here that standing posture is not a stable posture. While
standing erect, the human body sways back and forth in
sagittal and frontal plane about the ankle joint. In the present
study, muscle activation levels for the standing posture with
ankle in extension were selected. These activation values
are listed in Table 1. Thus, the chosen set of muscle activa-
tion levels would be able to correctly represent the normal
standing posture of a pedestrian.
A stretch-based involuntary reflex action has also been
enabled in this condition. For enabling the reflex, a thresh-
old value of elongation is to be defined in the Hill material
card of a muscle. When the elongation in muscle crosses
the threshold value, the stretch reflex in a muscle gets acti-
vated. However, the increase in muscle force starts only af-
ter a certain time known as reflex time. This delay between
the activation of stretch reflex and the onset of increase
in muscle force represents the time taken by the signal to
travel through the central nervous system circuitry (muscle-
spinal cord-muscle). A delay of 20 ms was assigned to all
the muscles in PMALE [1]. This mimics the ability of live
muscle to respond to a small stretch produced by an outside
agency. In medical terms this kind of reflex action is known
as ‘stretch reflex’ [22].
Data analysis
Two nodes each in the femur and tibia were selected
(Figure 3a) to compute the relative tibia displacements and
the knee joint angles. Figure 3b shows the sign conventions
used in the current study. Springs of very low stiffness
were modelled on each knee ligament to calculate ligament
strains. Element elimination approach has been enabled
to simulate the failure in the ligaments and the bones. A
342 A. Soni et al.
Table 1. Values of muscle parameters used to define Hill muscle
cards of 42 muscles.
Lower extremity Fmax Lopt
muscles (N) (m) Cfast aVmax Na
Vastus lateralis 1871 0.084 0.52 5.85 0.005
Vastus intermedius 1365 0.087 0.55.10 0.005
Vastus medialis 1294 0.089 0.53 5.36 0.005
Rectus femoris 779 0.084 0.619 5.55 0.005
Soleus 2839 0.03 0.25 2.67 1a
Gastrocnemius medialis 1113 0.045 0.518 5.74 1a
Gastrocnemius lateralis 488 0.064 0.518 5.69 1a
Flexor hallucis longus 322 0.043 0.55.17 0.005
Flexor digitorum longus 310 0.034 0.54.58 0.005
Tibialis posterior 1270 0.031 0.54.65 1a
Tibialis anterior 603 0.098 0.27 3.28 0.5a
Extensor digitorum longus 341 0.102 0.527 5.31 0.005
Extensor hallucis longus 108 0.111 0.54.32 0.005
Peroneus brevis 348 0.05 0.375 4.59 1a
Peroneus longus 754 0.049 0.375 4.35 0.005
Peroneus tertius 90 0.079 0.375 4.76 0.005
Biceps femoris (LH) 717 0.109 0.331 3.55 1a
Biceps femoris (SH) 402 0.173 0.331 3.91 1a
Semimembranosus 1030 0.08 0.55.61 1a
Semitendinosus 328 0.201 0.54.76 1a
Piriformis 296 0.026 0.55.71 0.005
Pectineus 177 0.133 0.54.62 0.005
Obturatorius internus 254 0.049 0.55.71 0.005
Obturatorius externus 109 0.030 0.55.71 0.005
Gracilis 108 0.352 0.55.13 0.005
Adductor brevis 1 286 0.133 0.55.17 0.005
Adductor brevis 2 286 0.133 0.55.22 0.005
Adductor longus 418 0.138 0.54.69 0.5a
Adductor mangus 1 346 0.087 0.416 5.07 0.005
Adductor mangus 2 444 0.121 0.416 5.07 0.005
Adductor mangus 3 155 0.131 0.416 5.07 0.005
Glutaeus maximus 1 382 0.142 0.476 5.53 0.005
Glutaeus maximus 2 546 0.147 0.476 5.53 0.005
Glutaeus maximus 3 368 0.144 0.476 5.53 0.005
Glutaeus medius 1 546 0.054 0.55.71 0.005
Glutaeus medius 2 382 0.084 0.55.71 0.005
Glutaeus medius 3 435 0.065 0.55.71 0.005
Glutaeus minimus 1 180 0.068 0.55.71 0.005
Glutaeus minimus 2 190 0.056 0.55.71 0.005
Glutaeus minimus 3 215 0.038 0.55.71 0.005
Sartorius 104 0.579 0.504 5.03 0.005
Tensor fasciae latae 155 0.095 0.55.71 1a
aThese values of muscle activation levels have been taken from [10].
transverse plane has been defined at the knee joint of the
impacted leg (i.e. left leg of PMALE) to calculate the knee-
bending moment in the simulations. Knee kinematics, strain
time history of each knee ligament, von Mises stress con-
tours in bones and lateral bending stiffness of knee joint of
the impacted leg, i.e. the left leg of the PMALE, have been
recorded from the simulations. Response in cadaveric and
reflex conditions was then compared to determine the role
of muscle contraction.
Figure 3. (a) Locations of nodes at which nodal time histories
have been obtained in the current study and (b) sign conventions
used to calculate relative tibia movements.
Results
In the present study, two simulations, each of 100 ms dura-
tion, have been performed. For the first 50 ms (stabilisation
duration), PMALE has been stabilised under gravity load
in both the simulations. At the end of the first 50 ms, the car
front impacts the left leg of the stabilised PMALE. Knee
joint contact force, knee kinematics, and von Mises stresses
in bones and ligament strains have been recorded from the
simulations to assess the effect of muscle contraction. Re-
sults presented here are for the impact duration, and the
initial time (i.e. 0 ms) corresponds to the time of impact.
Knee joint contact force
Figure 4 compares the contact force time history calculated
at the knee joint (i.e. between tibia plateau, meniscus and
femur condyle) of the impacted leg (left leg of PMALE)
Figure 4. Comparison of knee contact force–time history.
International Journal of Crashworthiness 343
for both cadaveric and reflex conditions. It is apparent that
due to the muscle action in reflex condition, contact force at
the knee joint remains significantly higher (ranged between
1.3 and 3.1 KN) as compared to the cadaveric condition
(ranged between 0.79 and 0.25 KN) in the entire duration
of the simulation. This indicates that in the reflex condi-
tion, active muscle forces pull the tibia towards the femur
and therefore the knee joint remains tightly enclosed. This
eventually results in higher compressive forces between the
tibia plateau, the meniscus and the femur condyles in the
reflex condition as compared to the cadaveric condition.
Knee joint kinematics
Linear and angular tibia displacement time histories rela-
tive to femur for both cadaveric and reflex conditions are
compared in Figures 5 and 6 respectively. It is evident that
active muscle forces have significantly affected the knee
joint kinematics in reflex condition as compared to the ca-
daveric condition.
In simulations for both cadaveric and reflex conditions,
the car front impacts the lateral side of the knee joint of the
left leg. As a result, both the tibia and the femur at the knee
level are forced to move in the medial direction. Due to this,
the relative medial tibia displacement (Figure 5b) and the
medial knee-bending angle (Figure 6b) increase for the ini-
tial 10–13 ms. After 13 ms, the momentum of the car front
further pushes the tibia and the femur at the left knee joint
to move together in the medial direction. As a result, the
medial knee-bending angle (Figure 6b) increases further.
However, the relative medial tibia displacement (Figure 5b)
remains nearly unaltered till 28 ms.
Later, after around 28–30 ms, the left foot loses contact
with the ground. Since the left leg is rigidly attached to
the massive upper body via the pelvis at the femur end, this
event frees the lower leg to move away from the femur in the
inferior direction and eventually sets the left leg in motion.
This results in a sudden increase in the inferior tibia dis-
placement (Figure 5c) and the medial bending (Figure 6b)
at around 28 ms.
By 30 ms, the moving left leg gradually establishes
contact with the medial side of the right leg, and the contact
force between them peaks after around 38–40 ms. During
this interaction (30–40 ms), the right leg along with the
friction at the right foot inhibits the motion of the left leg.
However, the inferior tibia displacement (Figure 5c) and the
medial bending (Figure 6b) keep on increasing and reach a
peak at around 40 ms. After 40 ms, the car front carries away
both the legs together in the lateral direction. Due to this,
the right leg also comes into motion and then rotates the
pelvis. As the pelvis is rigidly connected to the upper body
at the sacrum, this event eventually sets the medial rotation
in the upper body. As a result, during 40–50 ms, medial
tibia displacement (Figure 5b) suddenly shoots up, whereas
the inferior tibia displacement (Figure 5c) and medial knee
bending (Figure 6b) become constant.
Figure 5. Comparison of relative tibia displacements of the impacted leg (i.e. left leg) in (a) anterior-posterior, (b) medial-lateral and (c)
inferior-superior directions.
344 A. Soni et al.
Figure 6. Comparison of relative knee angles of the impacted leg (i.e. left leg): (a) extension-flexion, (b) medial-lateral bending and
(c) internal-external rotations.
In the reflex condition, active muscle forces pull the tibia
towards the femur. As a result, the inferior tibia displace-
ment (Figure 5c), the knee external rotation (Figure 6a) and
the medial bending angle (Figure 6b) are significantly re-
duced in the reflex condition as compared to the cadaveric
condition. The peak inferior tibia displacement reduces by
a factor of 2, the peak knee external rotation by a factor of
1.62 and the peak medial bending angle by a factor of 1.45.
However, the peak value of the anterior tibia displacement
(Figure 5a) is increased by 2.67 times in the reflex con-
dition as compared to the cadaveric condition. This could
be attributed to the higher active force in the gastrocne-
mius muscles in the reflex condition (peak value 1200 N)
as compared to the cadaveric condition (peak value 40 N),
which eventually forces the femur at the knee joint to move
in the posterior direction causing higher anterior tibia dis-
placement in the reflex condition.
Strain in knee ligaments
Figure 7 illustrates the calculated strain time history in knee
ligaments of the impacted leg (i.e. left leg of the PMALE)
for both cadaveric and reflex conditions. It is evident that
strains in knee ligaments reduce significantly in the reflex
condition as compared to the cadaveric condition.
ACL : Figure 7a compares the strain time history in
ACL for both the conditions. It is observed that till 16 ms,
ACL remained nearly unstrained in both the conditions.
Then, in the reflex condition, strain in ACL suddenly in-
creased to approximately 2% and remained higher than
in the cadaveric condition (1.5%) till 29 ms. This sud-
den rise in ACL strain can be attributed to the increased
anterior tibia displacement (Figure 5a) in the reflex con-
dition as compared to the cadaveric condition. After 29
ms, strain in ACL is observed to be higher in the cadav-
eric condition (peak value 6.51% at 40 ms) than in the
reflex condition (peak value 4.95% at 44 ms). This is at-
tributed to higher inferior tibia displacement (i.e. away
from femur) in the cadaveric condition (Figure 5c). It is
observed that peak strain value in ACL dropped by a factor
of 1.3 in the reflex condition as compared to the cadaveric
condition.
PCL: Strain time history in PCL is compared for both
the conditions in Figure 7b. It is observed that reduction in
strain due to muscle action is more prominent in PCL than in
the ACL. In the cadaveric condition, strain in PCL reached
up to 4.03%, whereas in the reflex condition it remained
nearly at 0% for the entire duration of the simulation. Rel-
ative displacement of tibia in the inferior direction is the
major source of strain in the PCL. Figure 5c illustrates that
in both cadaveric and reflex conditions, the tibia is moving
away from the femur. However, active muscle forces in the
reflex condition have pulled the tibia towards the femur and
therefore tibia displacement in the inferior direction (i.e.
away from the femur) has reduced (Figure 5c). This has
slackened the PCL in the reflex condition.
International Journal of Crashworthiness 345
Figure 7. Comparison of strain–time history in knee ligaments: (a) ACL, (b) PCL, (c) MCL and (d) LCL of the impacted leg (i.e. left
leg).
MCL: Strain in MCL for both the conditions is shown in
Figure 7c. It is observed that peak MCL strain has reached
the ligament failure limit of 15% in both the conditions.
However, in comparison to the cadaveric condition (30 ms),
failure is delayed by 5 ms in the reflex condition (35 ms).
The delay in MCL failure can be attributed to the com-
bined effect of reduced inferior tibia displacement (Figure
5c) and medial knee bending (Figure 6b) in the reflex con-
dition as compared to the cadaveric condition. Rupture in
MCL in both the simulations agrees with the results of
statistical analysis of real-world accidents in which MCL
is reported as the most frequently injured knee ligament
[14].
LCL: It is observed that strain in LCL (Figure 7d) re-
mains below 1% in both the conditions. This can be ascribed
to the lateral impact which forces the tibia to bend medially
and consequently keeps the LCL slackened.
Von Mises stresses in bones
Figures 8 and 9 show the von Mises stress distribution on
the bones of the impacted leg (i.e. left leg). It is apparent
that stresses in bones increased significantly in the reflex
condition as compared to the cadaveric condition.
von Mises stress distribution on femur and tibia at
the knee joint of the impacted leg at different stages of
the simulation for both cadaveric and reflex condition is
compared in Figure 8. It is observed that in the reflex
condition (Figure 8b), stresses in the bones (especially at
the femur lateral condyle region) are significantly higher
(between 104 and 138.1 MPa), and larger areas of the
bones are stressed as compared to the cadaveric condition
(Figure 8a). This can be attributed to the higher compres-
sive forces (shown in Figure 4) between the tibia plateau and
the femur condyle caused by the muscle pull in the reflex
condition.
Later, at 48 ms in the reflex-enabled situation (Figure
9a), the medial sides of the mid tibia and fibula were stressed
to their ultimate stress limit (138.1 MPa) and consequently
fractured in the simulation. However, in the cadaveric con-
dition (Figure 9b) stress in tibia and fibula remained below
the failure limits.
Knee lateral bending stiffness
Figure 10 compares the knee lateral bending moment and
the angle response of the impacted leg for both cadaveric
and reflex conditions. Peak bending moment is found to be
significantly higher in the reflex condition (135 Nm) than
in the cadaveric condition (96 Nm). Knee-bending angle
(Figure 6b) at the time of MCL failure decreased by approx-
imately 1◦in the reflex condition (10.2◦) as compared to
the cadaveric condition (11.3◦). Linear regression was per-
formed on the bending moment – angle response shown in
Figure 10 to obtain a representative lateral bending stiffness
of the knee joint. It is found that lateral bending stiffness
346 A. Soni et al.
Figure 8. Comparison of von Mises stress distribution on bones at the knee joint of the impacted leg (i.e. left leg) at different stages of
the simulations for (a) cadaveric and (b) reflex conditions (meniscus and other soft tissues are not shown here).
of the knee joint significantly increased by approximately
58% in the reflex condition (18.93 Nm/deg, R2=0.8849)
as compared to the cadaveric condition (11.99 Nm/deg,
R2=0.986).
Figure 9. Comparison of von Mises stress distribution on femur,
tibia and fibula of the impacted leg (i.e. left leg) at 48 ms for (a)
cadaveric and (b) reflex conditions.
Discussion
In the present study, effects of active muscle forces on the
response of lower extremity in a full-scale car–pedestrian
lateral impact have been investigated using FE simulations.
Two pre-impact conditions, representing a cadaver and an
unaware pedestrian, have been considered. It is observed
that active muscle forces have significant effects on the
lower extremity loading. Based on these FE simulations,
Figure 10. Comparison of knee-bending moment – angle re-
sponse of the impacted leg.
International Journal of Crashworthiness 347
Figure 11. Role of active muscle forces.
the role of active muscle forces has been summarised in
Figure 11.
Active muscle forces pull the tibia close to the femur
and hence hold the knee joint tightly. The tightly enclosed
knee joint has two effects:
rOn one hand, it restricts the relative movement between
the tibia and the femur and consequently alters the post-
impact knee kinematics. This eventually reduces the
strain in knee ligaments. The restricted relative tibia
movements at knee level along with friction at the foot
of the axially compressed leg restrain lower leg motion.
This increases the bending in the tibia and fibula and
results in fracture.
rOn the other hand, muscle pull causes higher compres-
sive forces between the tibia plateau and the femur
condyles, which results in higher stresses in these bones
in the knee region.
The present study indicates that active muscles have
potentially important effects on lower extremity injuries
during a car–pedestrian lateral impact and FE simulations
can be a viable tool to study these effects. However, to
correctly model active muscle forces in the simulations,
activation levels in the muscles representing the pre-crash
pedestrian conditions are required as an input. This signifies
the future direction of pedestrian safety research, since un-
like car occupants, pedestrian crashes occur in a variety of
postures (like stationary, walking, running or jogging). For
different human body postures, the central nervous system
either recruits entirely different sets of muscles or activates
the same set of muscles at different levels. Therefore, it
becomes essential to quantify patterns of pedestrian pre-
crash conditions in terms of activation levels, which can
then be used as an initial input to study muscle effects on
lower extremity loading for a pedestrian undergoing crash
in different postures.
Conclusions
In the present study, a full-body pedestrian model (named
as PMALE) with 42 active muscles in each leg has been
developed. PMALE is then used to investigate the effects
of active muscle forces on the response of lower extremity
in a car–pedestrian lateral impact. Differences in response
between a cadaver and an unaware pedestrian who is stand-
ing symmetrically with legs in a side-by-side stance (with
active muscles) have been studied. To assess the effect of
muscle activation, knee joint contact force, strains in knee
ligaments, VonMises stress in bones and knee lateral bend-
ing moment – angle response of the impacted leg (i.e. left
leg of the PMALE) have been compared. The following
conclusions can be drawn:
(1) Active muscle forces in the reflex condition pull the
tibia close to the femur and eventually alter the knee
kinematics.
(2) In the present study, peak strains in all knee ligaments
were found to be lower in the reflex condition (with
active muscles). This reinforces our previous findings
that the risk of ligament failure in real-life crashes is
likely to be lower than that predicted through cadaver
tests or simulations.
(3) MCL failed, whereas LCL remained nearly unstrained.
This implies that in lateral impacts, MCL could be
considered as the most vulnerable ligament and LCL
as the safest.
(4) Increased stresses in bones at the lateral side of the knee
joint and failure occurred at mid tibia and mid fibula
in the reflex condition. It leads to the conclusion that
chances of bone fracture increase with muscle contrac-
tion.
(5) Knee lateral bending stiffness increased by 58% in
the reflex condition. This suggests that due to mus-
cle contraction, the knee joint becomes stiffer in lateral
bending.
Acknowledgement
The authors would like to acknowledge the support from
the Transportation Research and Injury Prevention Program
(TRIPP) at the Indian Institute of Technology, Delhi, and
the Volvo Research Education Foundation.
References
[1] U. Ackerman, PDQ Physiology, 2002. Available at
http://www.fleshandbones.com/readingroom/pdf/226.pdf.
[2] K. Bhalla, Y. Takahashi, J. Shin, C. Kam, D. Murphy, C.
Drinkwater, and J. Crandall, Experimental investigation of
the response of the human lower limb to the pedestrian
impact loading environment, Proceedings of the Society of
Automotive Engineer World Congress, Paper No. 2005-01-
1877, 2005.
[3] A. Chawla, S. Mukherjee, D. Mohan, and A. Parihar, Vali-
dation of lower extremity model in THUMS, Proceedings of
the IRCOBI Conference, 2004, pp. 155–166.
[4] A. Chawla, S. Mukherjee, A. Soni, and R. Malhotra, Effect
of active muscle forces on knee injury risks for pedestrian
348 A. Soni et al.
standing posture at low speed impacts, Proceedings of the
IRCOBI Conference, 2007, pp. 95–112.
[5] A.B. Chidester and R.A. Isenberg, Final report – The
pedestrian crash data study, Proceedings of the 17th En-
hanced Safety of Vehicle Conference, Paper No. 248,
2001.
[6] J. Kajzer, S. Cavallero, J. Bonnoit, A. Morjane, and S.
Ghanouchi, Response of the knee joint in lateral impact:
Effect of bending moment, Proceedings of the IRCOBI Con-
ference, 1993, pp. 105–116.
[7] J. Kajzer, S. Cavallero, S. Ghanouchi, and J. Bonnoit, Re-
sponse of the knee joint in lateral impact: Effect of shearing
loads, Proceedings of the IRCOBI Conference, 1990, pp.
293–304.
[8] J. Kajzer, H. Ishikawa, Y. Matsui, and G. Schroeder, Shear-
ing and bending effects at the knee joint at low speed
lateral loading, Proceedings of the Society of Automo-
tive Engineer World Congress, Paper No. 1999-01-0712,
1999.
[9] J. Kajzer, G. Schroeder, H. Ishikawa, Y. Matsui, and U.
Bosch, Shearing and bending effects at the knee joint at
high speed lateral loading, Proceedings of the Society of
Automotive Engineer World Congress, Paper No. 973326,
1997.
[10] J. Kerrigon, K. Bhalla, N. Madeley, J. Funk, D. Bose, and
J. Crandall, Experiments for establishing pedestrian impact
lower injury criteria, Proceedings of the Society of Auto-
motive Engineers world Congress, Paper No. 2003-01-0895,
2003.
[11] A.D. Kuoand F.E. Zajac, A biomechanical analysisof muscle
strength as a limiting factor in standing posture,J.Biomech.
26 (1993), pp. 137–150.
[12] K.W. Lia, H.H. Wub, and Y.C. Linb, The effect of shoe sole
tread groove depth on the friction co-efficient with different
tread groove widths, floors and contaminants, App. Eng. 37
(2006), pp. 743–748.
[13] T. Maeno and J. Hasegawa, Development of a finite element
model of the total human model for safety (THUMS) and
application to car–pedestrian impacts, Proceedings of the
17th Enhanced Safety of Vehicle Conference, Paper No.
494, 2001.
[14] Y. Matsui, Biofidelity of TRL legform impactor and injury
tolerance of human leg in lateral impact,StappCarCrashJ.
45 (2001), pp. 495–510.
[15] Y. Mizuno, Summary of IHRA pedestrian safety WG activ-
ities – Proposed test methods to evaluate pedestrian pro-
tection afforded by passenger cars, Proceedings of the 18th
Enhanced Safety of Vehicle Conference, Paper No. 580,
2003.
[16] Y. Mizuno, Summary of IHRA pedestrian safety WG
activities-proposed test methods to evaluate pedestrian pro-
tection afforded by passenger cars, Proceedings of 19th In-
ternational Technical Conference on the Enhanced Safety of
Vehicles, Paper No. 138, 2005.
[17] K. Nagasaka, K. Mizuno, E. Tanaka, S. Yamamoto, M.
Iwamoto, K. Miki, and J. Kajzer, Finite element analysis
of knee injury in car-to-pedestrian impacts,Int.J.Traffic
Inj. Prev. 4 (2003), pp. 345–354.
[18] J.P. Schuster, C.C. Chou, P. Prasad, and G. Jayaraman, Devel-
opment and validation of a pedestrian lower limb non-linear
3-D finite element model, Stapp Car Crash J. 44 (2000), pp.
315–334.
[19] A. Soni, A. Chawla, and S. Mukherjee, Effect of muscle
active forces on the response of knee joint at low speed
lateral impacts, Proceedings of the Society of Automotive
Engineers world Congress, Paper No. 2006-01-0460, 2006.
[20] A. Soni, A. Chawla, and S. Mukherjee, Effect of muscle
contraction on knee loading for a standing pedestrian in
lateral impacts, Proceedings of the 20th Enhanced Safety of
Vehicle Conference, Paper No. 467, 2007.
[21] Y. Takahashi and Y. Kikuchi, Biofidelity of test devices and
validity of injury criteria for evaluating knee injuries to
pedestrians, Proceedings of the 17th Enhanced Safety of
Vehicle Conference, Paper No. 373, 2001.
[22] A.J. Vander, J.H. Sherman, and D.S. Luciano, Human Phys-
iology: The Mechanisms of Body Function, Tata McGraw-
Hill, New York, 1981.
[23] A.M. Woodhull, K. Maltrud, and B.L. Mello, Alignment
of the human body in standing, Euro. J. Appl. Physiol. 54
(1985), pp. 109–115.
[24] World Bank Road Safety, 2001. Available at http://www.
worldbank.org/transport/roads/safety.html.
