Liu XJ 2002 - A Study of Influences of Vehicle Speed and Front Structure On Pedestrian Impact With Mathemathical Models

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 13

This article was downloaded by: [Umeå University Library]

On: 17 November 2014, At: 01:07


Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

Traffic Injury Prevention


Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/gcpi20

A Study of Influences of Vehicle Speed and Front


Structure on Pedestrian Impact Responses Using
Mathematical Models
a a a
X. J. Liu , J. K. Yang & P. Lövsund
a
Crash Safety Division, Department of Machine and Vehicle Design , Chalmers University of
Technology , Göteborg, Sweden
Published online: 15 Sep 2010.

To cite this article: X. J. Liu , J. K. Yang & P. Lövsund (2002) A Study of Influences of Vehicle Speed and Front Structure on
Pedestrian Impact Responses Using Mathematical Models, Traffic Injury Prevention, 3:1, 31-42, DOI: 10.1080/15389580210517

To link to this article: http://dx.doi.org/10.1080/15389580210517

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
Traffic Injury Prevention, 3:31–42, 2002
Copyright ° C 2002 Taylor & Francis
1538-9588/02 $12.00 + .00

A Study of Influences
of Vehicle Speed and Front
Structure on Pedestrian
Impact Responses Using
Mathematical Models
Downloaded by [Umeå University Library] at 01:07 17 November 2014

X. J. LIU, J. K. YANG, and P. LÖVSUND


Crash Safety Division, Department of Machine and Vehicle Design, Chalmers University
of Technology, Göteborg, Sweden

A validated pedestrian multibody model was used to investigate the influences of impact speed and
vehicle front structure on the pedestrian dynamic responses in vehicle collisions. To predict the
injury risks of pedestrians at different impact speeds, the injury-related parameters concerning
head, chest and lower extremity areas were calculated from mathematical simulations. Four vehicle
types including large and compact passenger cars, minivans and light trucks were simulated
according to their frequency of involvement in real world accidents. The influences of various
vehicle front shape and compliance parameters were analyzed.
Based on the results from the parametric study, the possible benefits from speed control in urban
area were assessed, and a feasible speed limit was proposed to reduce the risks of pedestrian
injuries. Moreover, the possible countermeasures on basis of vehicle front design to mitigate the
injury severity of the pedestrians were discussed.

Keywords Mathematical Modeling; Pedestrian; Impact Speed; Vehicle Front Design

In the EU countries during 1996, there were 6,769 sign are two key aspects controlling pedestrian injury
pedestrian fatalities in vehicle accidents (ETSC, 1999). severity.
The rate of pedestrian deaths per total traffic fatalities Accident studies (Ashton, 1982; ETSC, 1999) indi-
varied from 12% in the USA (NHTSA, 1999), 16.2% in cated that the majority of the pedestrians are struck by
the EU countries (ETSC, 1999), 26% in China (TAPSM, the front of vehicles in urban and residential area at
1997) to 27% in Japan (Matsui and Ishikawa, 1998). crash speeds up to 40 km/h. Children under 15 years
Among all the factors related to pedestrian accidents, old and elderly people aged above 70 are overrepre-
impact speed and the safety performance of vehicle de- sented (NHTSA, 1999; Berg et al., 2000). Head and
lower extremities are the most frequently injured body
Received 18 May 2000; accepted 12 April 2001. regions in road accidents. Injuries to the head are the ma-
This study is sponsored by the Swedish National Road Administration jor causes of pedestrian fatalities (Mizuno and Kajzer,
(Vägverket).
2000), whereas injuries to the lower extremities are usu-
Address correspondence to X. J. Liu, Crash Safety Divisions, Dept. of
Machine and Vehicle Design, Chalmers University of Technology, Göteborg, ally non-fatal but often have long-term consequences
Sweden. E-mail: liuxj@mvd.chalmers.se with possible permanent disability (Schuster et al., 2000;
31
32 X. J. LIU ET AL.

Takahashi et al., 2000). Impact speed is the predomi- Many studies emphasized on the influence of vehi-
nant factor determining the pedestrian injury severity. cle front shapes on pedestrian kinematics (Niederer &
At a collision speed of less than 20 km/h, pedestrians Schlumpf, 1984; Janssen & Wismans, 1985; Ishikawa
usually sustain only minor injuries, whereas at a speed et al., 1991). Hood edge height appears to be an impor-
above 45 km/h pedestrians are most likely to be killed tant factor on the pedestrian kinematics after the initial
(Otte, 1999). An essential increase in injury severity was impact between bumper and lower extremity of pedes-
observed within the impact speed between 20 and 45 trian, while the deformation properties of vehicle front
km/h. It thus is important to investigate the injury mech- structure exhibits only secondary effect on the motion
anisms and to develop protection counter-measures in pattern. Lowering the level of the initial contact point
this speed range. and increasing the compliance of the bumper was con-
Besides the accident statistics, various evaluation sidered as the effective measures to reduce the injury
methods have been implemented in pedestrian protec- severity of legs and knee joint areas (Aldman et al.,
tion research, such as biological specimens (Bunketorp, 1985b; Kajzer & Schroeder, 1992; Cesari et al., 1994).
1983; Cavallero et al., 1983), mechanical dummies However, most of the previous studies only focused on
Downloaded by [Umeå University Library] at 01:07 17 November 2014

(Niederer & Schlumpf, 1984; Higuchi & Akiyama, the bonnet-type vehicle, i.e. passenger cars at a given
1991) and mathematical models (Van Wijk et al., 1983; impact speed.
Harris & Grew, 1985; Ishikawa et al., 1991). The most Due to the rapid growth of various vehicle types with
human-like responses and injury mechanisms of body different front shapes nowadays, the involvement of ve-
segments have been obtained with cadavers. However, hicle types is considerably different from that of the last
such tests are unsuitable for the development and eval- two decades. Accident statistics (NHTSA, 1999) indi-
uation of safety counter-measures due to the limita- cated that passenger car accounted for about 56% of
tion of specimen resources and prohibitive experiment vehicles striking pedestrian followed by van/Utilities
costs. Alternatively several types of mechanical dum- for 17%, pickup/trucks for 17%, and others for about
mies have been used as pedestrian substitute to assess 10%. Mizuno and Kajzer (2000) found that bonnet-type
the aggressiveness of different front structures (Aldman vehicles tend to incur serious injuries to pedestrian legs,
et al., 1985a; Kajzer & Schroeder, 1992; Huang et al., whereas mini vans are more aggressive to the head
1999; Björklund & Zheng, 2001). Nevertheless, none and thorax. However, the effects of impact speed were
of them had shown enough biofidelity to properly pre- excluded, and the aggressiveness of light trucks to
dict the dynamic responses of pedestrian bodies. Fur- pedestrians was not assessed.
thermore, whole body dummies were found difficult to In this study, the influences of vehicle impact speed,
handle in the assessment test, since the pedestrian im- front shape and compliance properties on pedestrian
pact responses are sensitive to initial standing posture dynamic responses were evaluated by using a fully val-
and position. Therefore, a sub-system test procedure idated human-body mathematical model with improved
was proposed to avoid such complexity and uncertain- lateral biofidility, especially in lower extremities with a
ties (EEVC, 1998). It consists of four separate impactors human-like knee joint and a breakable lower leg model
representing adult head, child head, pelvis of adult and (Yang et al., 2000). The effects of different vehicle
thorax of child, as well as lower extremity. However, types, including large and compact passenger cars, mini
these sub-system tests cannot evaluate the integrated vans and light trucks, were assessed. It is aimed to
safety performance of a vehicle design in terms of the demonstrate the feasibility and effectiveness of travel
overall responses of pedestrian. speed control and vehicle front design to mitigate pedes-
As a complement to the experimental studies, two trian injury severity.
types of mathematical models were developed. One was
based on the properties of mechanical dummies
(Van Wijk et al., 1983; Harris, 1989; Yoshida & METHOD AND MATERIAL
Matsuhashi, 1998), and another was based on human-
body characteristics (Gibson et al., 1986; Ishikawa et al., A parametric study involving a set of variables re-
1993). Human-body models had shown improved biofi- garding impact speed, vehicle front shape and compli-
delity and better correlation with the results from impact ance properties is conducted with the validated pedes-
tests with cadaver specimens than dummy-based mod- trian mathematical model (Yang et al., 2000). The effect
els (Ishikawa et al., 1993). However, it was pointed out of impact speed is examined from 20 to 50 km/h with
that a leg fracture model and realistic knee joint charac- four vehicle types according to their frequency of in-
teristics are necessary to improve the biofidelity of the volvement in real world accidents in terms of the calcu-
lower extremities (Yang, 1997). lated injury parameters. To concisely clarify the specific
INFLUENCES OF VEHICLE SPEED AND FRONT STRUCTURE 33

Figure 1 The baseline model setup for car–pedestrian simulations, BCH-Bumper Central Height, BL-Bumper Lead Length,
Downloaded by [Umeå University Library] at 01:07 17 November 2014

HEH = Hood Edge Height, HL = Hood Length, HA = Hood Slope Angle, WA = Windscreen Angle.

effects of geometric and stiffness variables, these two (Figure 1). To avoid the unrealistic protection effects re-
set of variables are discussed separately at a given im- sulting from the shoulder and elbow striking, the pedes-
pact speed of 40 km/h. trian body was rotated 30 degrees inward from a posi-
tion perpendicular to the centerline of vehicle. The right
and left upper arms are inclined rearwards at 30 degree
Model Setup
to the vertical plane, while the lower arms hang natu-
rally down.
The mathematical model used in this parameter study
Car front model consists of bumper, hood edge, hood
is a 50th percentile adult male pedestrian model formu-
top and windscreen ellipsoids to approximate the
lated in MADYMO (Yang et al., 2000), comprising the
exterior profile of a vehicle. Four wheels are repre-
human-like knee joint model (Yang & Kajzer, 1992) and
sented by four identical ellipsoids to produce the brak-
the breakable leg model (Yang et al., 1993). The human-
ing deceleration of the vehicle by the friction force
like knee joint allows more accurate simulation of knee
between wheels and ground. The vehicle model is
joint responses, whereas the breakable lower leg model
pitched to provide a geometric attitude equivalent to
can predict the bone fracture phenomena when the im-
0.6 g braking. Force-deformation characteristics of car
pact force exceeds the tolerance level. The spherical
front components are obtained from legform and head-
joint model is used to describe the neck, shoulder, tho-
form impactor tests. The friction coefficients are 0.6 for
racic, lumbar, hip and ankle joints, while elbow joint is
foot/ground and wheels/ground, and 0.5 for the contacts
defined as hinge joint model. The motion range of each
between body segments and car front structures.
joint is restricted to the physiology movements in the
sagittal, coronal and horizontal planes. The stiffness of
the joints is defined as moment-angle functions based Design of Parametric Study
on the biomechanical studies on human joints. Mean-
while the mechanical properties of body segments are The parametric study is divided into three parts. The
described as force-deformation function, taking into ac- first part concerns the influence of impact speed, taking
count the effects of soft tissues. into account the involvement of different vehicle mod-
The validity of the pedestrian model was well ver- els including large and compact passenger cars, mini
ified with cadaver tests in terms of overall kinemat- vans and light trucks. The main purpose is to predict
ics, dynamic responses of body segments, as well as the effect of impact speed on the injury risk of pedes-
the impact forces. The injury severity of head, pelvis trian exposed to the real world traffic accidents. The
and lower extremities were well predicted (Yang et al., injury patterns with regard to different vehicle types
2000). Therefore this pedestrian model is considered will also be compared. Secondly, the effects of vari-
adequate for the evaluation of the influences of vehicle ous vehicle front shape parameters on the injuries of
parameters on the pedestrian responses. pedestrians will be discussed with passenger car mod-
The pedestrian model is struck on the right side at els at impact speed of 40 km/h. It is aimed to reveal the
the central line of vehicle, with the right leg backward possible improvement of vehicle shape to mitigate the
and the left leg forward to represent a walking stance injury severity of pedestrians. Finally the influence of
34 X. J. LIU ET AL.

Table I Selected factors and levels of validation simulations (Yang et al., 2000). Some of
Levels the data, such as bumper, hood top and wind-screen
stiffness are obtained from subsystem tests according
Geometric and stiffness factors −1 0 +1 to the test procedures proposed by EEVC (1994), in-
BCH = bumper central height (mm) 300 400 500 cluding lower-leg impactor to bumper, and headform
BS = bumper stiffness (N/mm) 125 250 500 impactors to hood top tests. Hood edge stiffness is es-
BL = bumper lead (mm) 50 100 200 timated within the corridor reported by Ishikawa et al.
HEH = hood edge height (mm) 600 700 800 (1991). All these stiffness variables vary from 50% to
HES = hood edge stiffness (N/mm) 200 400 800
HL = hood length (mm) 500 700 1200
200% to simulate the different impact locations on ve-
HTS = hood top stiffness (N/mm) 75 150 300 hicle front structure.
HA = hood slope angle (deg) 10 25 45 Since the hood edge height and bumper center height
WA = windshield angle (deg) — 30 45 were recognized as two key factors affecting the pedes-
WS = windshield stiffness (N/mm) 300 600 800 trian overall kinematics and the injury severity to knee
joints, these two variables are varied in full factorial of
Downloaded by [Umeå University Library] at 01:07 17 November 2014

force-deformation properties of vehicle structure will three levels, which lead to 9 variations for both large
be discussed at impact speed of 40 km/h. and compact passenger cars. Other variables vary in re-
The effect of impact speed is examined at four levels, verse levels for the large and compact passenger cars, so
i.e. 20, 30, 40 and 50 km/h. The selected geometric and that the corresponding effects can be obtained without
stiffness variables and corresponding levels are listed in any confounding effects between these variables. The
Table I. geometric and stiffness variables for van and light truck
Bumper central height is chosen between knee joint only vary in two levels due to the obvious higher and
and the center of gravity of the lower leg for a 50th per- stiffer front profile than that of passenger cars.
centile adult male. The hood edge height varies between The influences of geometric variables of vehicle
hip and knee joints. The hood length has three levels in fronts on pedestrian responses are evaluated with the
order to simulate the different vehicle types, of which basic configuration of large passenger cars at the im-
1200 mm for large passenger car, 700 mm for compact pact speed of 40 km/h. The stiffness variables remain
passenger car and 500 mm for van and light truck. Hood constant at their middle levels as below,
slope angle also depends on the specific vehicle models.
• Bumper stiffness 250 N/mm
For instance, hood slope angle at 10 degree is assigned
• Hood lead edge stiffness 400 N/mm
for passenger cars, whereas 30 degree for mini van and
• Hood top stiffness 150 N/mm
45 degree for light truck. Likewise, windshield angle
• Windscreen stiffness 600 N/mm
is varied to fit the different vehicle models. In the case
of mini van and small passenger car, windshield angle
Geometric variables, including the bumper center
has two levels at 30 and 45 degree. For large passen-
height, bumper lead length, and hood edge height, vary
ger car and light truck, the windshield angle is 30 and
at three levels, which makes 27 runs in a full factorial
45 degree respectively. By varying the bumper height,
analysis.
bumper lead, hood edge height, hood length, hood slope
Similarly, the effects of different stiffness variables
angle, and windshield angle, different vehicle models
on pedestrian injuries are studied at the impact speed of
and front shapes can be simulated (Figure 2).
40 km/h with a fixed vehicle front shape as follow,
The medium level of force-deformation properties
for different structures is consistent with the input data • Bumper center height 400 mm
• Bumper lead length 100 mm
• Hood lead edge height 700 mm
• Hood length 1200 mm
• Hood slope angle 10 degree
• Windscreen slope angle 30 degree

To avoid the interference of the windscreen to the


pedestrian kinematics, the hood length is chosen at the
upper level of 1200 mm. While the windscreen stiffness
remains constant, other three stiffness variables includ-
ing bumper stiffness, hood edge stiffness and hood top
Figure 2 Different vehicle types used in parametric study. stiffness are varied according to the three-level full
INFLUENCES OF VEHICLE SPEED AND FRONT STRUCTURE 35

Table II Tolerance levels of selected injury parameters bending angle and 6 mm of knee lateral dislocation
serve as the knee joint tolerance levels.
Injury parameter Tolerance level

HIC 1000
Chest acc. (3 ms) 85 g RESULTS
Pelvis impact force 10 kN
Thigh impact force 5 kN Influence of Impact Speed
Tibia acc. (3 ms) 150 g
Knee lateral dislocation 6 mm
Knee lateral bending angle 15◦ Calculated injury parameters are presented in the cu-
mulative frequency of 25%, 50% and 75% to describe
the injury risks of pedestrians at different impact speeds
factorial design, leading to 27 runs totally. The effects (Figure 3). The effects of different vehicle types with
of specific variables and possible interactions on pedes- varying front shapes and compliance properties are
trian injury can be analyzed without any confounding included.
effects between these variables. The head injury (HIC value) increases steadily with
Downloaded by [Umeå University Library] at 01:07 17 November 2014

the impact speed. The probability of the HIC value ex-


Selected Injury Parameters ceeding 1,000 is more than 50% at the impact speed of
40 km/h, while less than 25% at 30 km/h (Figure 3a).
The injury risks of pedestrian are evaluated in terms The injury risks of thigh and pelvis area exhibit a
of the selected injury parameters and tolerance levels strong dependency on the impact speed. For instance,
(Table II). HIC (Head Injury Criterion) of 1,000 is as- at the impact speed of 30 km/h, there are only less
signed to evaluate the resultant head injury severity, than 25% of all cases exceeding the tolerance level of
while the TTI (Thoracic Trauma Index) of 85 g for 5 kN, while around 75% are serious injured at the im-
chest area. The impact force of 10 kN is chosen as the pact speed of 40 km/h (Figure 3b). There is almost no
injury tolerance level for pelvis side impact (Cesari & tibia bone fracture at the impact speed of 30 km/h, but
Rameet, 1982). The injury criteria concerning lower more than 75% at 40 km/h (Figure 3c). The knee joint
extremities are primarily based on the recent report by appears to be the most vulnerable area under the lateral
EEVC (1998). The thigh impact force of 5 kN repre- impact loading. Even at the impact speed of 20 km/h,
sents the 20% risk of an AIS2+ femur fracture, whereas the injury risk of the knee joint is about 25%. As the
tibia acceleration of 150 g indicates 40% risk of an AIS impact speed increases up to 40 km/h, the knee injury
2 + lower leg fracture. A proposed 15 degree of lateral risk reaches almost 50% (Figure 3d).

Figure 3 Injury risks at different impact speeds (a) HIC value, (b) thigh impact force, (c) tibia acceleration, (d) knee lateral dislocation.
36 X. J. LIU ET AL.
Downloaded by [Umeå University Library] at 01:07 17 November 2014

Figure 4 Effects of vehicle impact speed on pedestrian (a) head impact speed with large car, (b) head impact speed with compact car,
(c) head impact angle with large car, (d) head impact angle with compact car.

Figure 4 illustrates the dependency of head impact analysis involving both geometric and stiffness vari-
speed and angle on vehicle impact speed. The head im- ables are conducted to reveal the corresponded effects
pact speed means the value of head resultant velocity on the pedestrian injury severity.
prior to the impact event between the head and vehicle
structure. The head impact angle refers to the angle be-
tween the direction of the head impact speed and the Influence of Vehicle Types
ground reference level in downward and rearward di-
rection. These two items are the essential test conditions The resultant head velocities are significantly af-
for the head impactor to bonnet top test (EEVC, 1998). fected by the vehicle types (Figure 5). In case of the
It’s clearly that the head impact speed increases propor- light truck, the resultant head velocity approximately
tional with the vehicle impact speed. The ratio between remains constant during the first 50 ms after the initial
the head impact speed and vehicle travel speed varies impact, and then decreases sharply after the head impact
from 0.72 to 1.04 for the large passenger car, and 0.65 to with windshield structure at 10.0 m/s (55 ms), slightly
1.28 for the compact passenger car, which are approx-
imately consistent with the previous studies, 0.7–0.9
for the large passenger car (Pritz, 1983) and 1.1–1.4 for
the compact passenger car (Cavallero et al., 1983). The
head impact angle varies from 42 to 73 degree for the
large passenger car, and 24 to 74 degree for the compact
passenger car. However, EEVC (1998) only specified
the head impact speed of 40 km/h with 65◦ ± 2◦ to the
ground reference level as the test conditions for the head
impactor test.
The considerable variations in calculated injury pa-
rameters at a given impact speed imply that factors other
than impact speed are also important in determining the Figure 5 Resultant head velocity against different vehicle types
injury severity of pedestrians. Therefore, an intensive at 40 km/h.
INFLUENCES OF VEHICLE SPEED AND FRONT STRUCTURE 37

lower than the travel speed of vehicle (11.11 m/s). For are compared in terms of the thigh impact force and
the mini van and passenger cars, the resultant head ve- knee joint injury risk. Generally, pedestrian is exposed
locities increase progressively and reach the maximum to higher injury risk to the upper body area against the
at around 13.1 and 13.9 m/s, respectively. The head im- mini van and light truck, while the passenger cars are
pact speed against vehicle structure varies from 10.8 m/s more aggressive to the lower extremity. Similar results
(95 ms) for mini van, 13.1 m/s (105 ms) for compact have been found in pedestrian accidents (Mizuno &
passenger car to 10.2 m/s (116 ms) for large passen- Kajzer, 2000).
ger car.
The differences in head velocity changes can be
Influence of Vehicle Front Shape
mainly attributed to the different hood slope angles of
various vehicle types (Figure 2), which lead to different
To avoid the effects of the hood slope angle and hood
kinematics of pedestrian after the initial contact with
length on pedestrian kinematics, only large passenger
bumper. Due to the large hood slope angle of the light
cars are considered to assess the influences of bumper
truck, the pelvis and chest contact with hood earlier
center height, bumper lead length and hood edge on
Downloaded by [Umeå University Library] at 01:07 17 November 2014

than the mini van and passenger cars. Consequently the


the head and lower extremity injuries. The hood edge
pedestrian body has not been rotated but pushed for-
height has significant effect on the pedestrian kinemat-
ward along the direction of vehicle travel speed. When
ics and resultant head velocity. When the hood edge
impacted by the mini van or passenger cars, the pedes-
height varies from 800 to 600 mm, the Wrap Around
trian body is rotated downwards to the hood top, leading
Distance (WAD) increases from 1.78 m to 2.05 m, due
to a considerably rotational movement after the impact.
to the considerable “slipping” effect of pedestrian body
The resultant head velocities thus increase after the ini-
(Figure 6a). The head impact speed increases from 9.4
tial impact for both mini van and passenger cars. The
to 11.1 m/s (Figure 6b).
extent of this rotational movement differs with the slope
Compared to the effect of the hood edge height, the
angle and length of hood. Compared to the large pas-
influences of the bumper lead length and bumper height
senger car, the moment of head impact is earlier for the
on the resultant head velocity are slight. The apparent
compact passenger car, which results in a higher head
effect of the bumper lead is to delay the time of the
impact speed accordingly.
head contact with hood top, and consequently increase
Table III summarized the mean values of various in-
the WAD (Figure 6c). The effect of the bumper lead on
jury parameters at the impact speed of 40 km/h with
head impact velocity is slight (Figure 6d).
respect to different vehicle types. In the simulation with
With the increasing bumper height close to the knee
the van and light truck models, the contact locations of
joint, the head impact speed increases (Figure 6e). This
the pedestrian head are closed to the lower windshield
is mainly due to the large knee lateral deflection, which
frame, where is stiffer than the center of windshield
restricts the “slipping” movement but intensifies the ro-
and hood top. Therefore higher injury risk of head has
tation movement of upper body downwards to the hood
been found in terms of the HIC value. Moreover, a large
top.
hood slope angle results in the direct blow on pedestrian
The relative importance of these three factors is com-
chest and pelvis area, the injury risk to these body parts
pared within Figure 6 (e, f). It is clearly that the hood
are higher than that of passenger cars with apparent
edge height has greater effect than other two parame-
“slipping” movement along the hood top. The injury
ters. No significant interaction exists between the hood
risks to the lower extremity of different vehicle types
edge height and bumper center height.
In general, the head impact speed tends to increase
Table III Comparison of injury parameters with different with a decreasing hood edge height, and an increasing
vehicle types at 40 km/h bumper center height, but has minor relationship to the
bumper lead length.
Vehicle types
The influence of vehicle front shape on the dynamic
Injury parameters LPAa CPAa Van Trucks responses of knee joint is illustrated in Figure 7. When
HIC 890 514 1261 1317
the bumper lead length varies from 50 to 100 mm, the
Chest acc. (3 ms-g) 24 27 45 59 knee lateral dislocation ascends rapidly (Figure 7a),
Pelvis impact force (kN) 5 2.8 6.7 7.3 while the knee lateral bending angle tends to decrease
Thigh impact force (kN) 6.7 3.6 5.8 6.2 as the bumper lead increases (Figure 7b). When the
Knee lat. angle (deg) 25 12 20 20 bumper lead at 50 mm, the mean responses of the knee
Knee lat. dis. (mm) 5.1 8.6 5.0 5.0
lateral dislocation is always below the injury threshold
a
LPA: large passenger car, CPA: compact passenger car. of 6 mm.
38 X. J. LIU ET AL.
Downloaded by [Umeå University Library] at 01:07 17 November 2014

Figure 6 Influences of vehicle front shape parameters on pedestrian kinematics and head responeses: hood-edge height on (a) kinematics
(b) head impact speed, bumper lead length on (c) kinematics (d) head impact speed. (e) interactions between bumper center height and bumper
lead length on head impact speed, (f ) interactions between hood edge height and bumper center height on head impact speed.

Figure 7 Influences of vehicle front shape on knee injuries (a) Interaction between bumper center height and bumper lead length on knee
lateral bending angle. (b) Interaction between bumper center height and bumper lead length on knee lateral bending angle.
INFLUENCES OF VEHICLE SPEED AND FRONT STRUCTURE 39

The only detectable effect of the hood edge height vehicle front shape and impact speed. A soft hood top
on the knee injury parameters is observed in case of can provide a significant protection to the head area
the bumper height at 300 mm. When the hood edge (Figure 8a), compared with other two levels of hood top
height was raised from 600 to 800 mm, the knee lat- stiffness. Similar tendency is also found in the thigh im-
eral dislocation increased from 4.6 to 5.4 mm, whereas pact force (Figure 8b), tibia acceleration (Figure 8c) and
the knee lateral bending angle decreased from 6.2 to knee joint injuries (Figure 8d). An exception is found
3.6 degree. Meanwhile, no significant interaction exists in the 3 ms clip of tibia acceleration. It’s mainly due to
between the bumper center height and hood edge height the fact that an extremely high peak value of the tibia
in terms of the knee injury parameters. Therefore, it can acceleration causing tibia fracture has been filtered out.
be concluded that the hood edge height has little effect Compared to the influences of vehicle shape vari-
on both knee lateral bending angle and lateral disloca- ables, stiffness proprieties have great effect on the re-
tion. This result was consistent with the previous studies sulted injury severity, but little influence on the pedes-
conducted with mechanical substitutes (Ishikawa et al., trian motion pattern.
1994; Sakurai et al., 1994) and mathematical simulation
Downloaded by [Umeå University Library] at 01:07 17 November 2014

(Nagatomi et al., 1996).


DISCUSSION

Influence of Vehicle Stiffness Properties Several studies investigated the likely effects of the
reduced vehicle travel speed limit on the incidence of
In a similar way described above, the influence of pedestrian fatalities, based on the observations from
vehicle stiffness properties is analyzed with the pedes- real-world pedestrian accidents and statistical analyses.
trian kinematics, resultant head velocity and injuries to If the speed limit reduced from 60 km/h to 50 km/h,
head and lower extremity regions. a reduction of pedestrian fatalities and casualties by
Figure 8 illustrates the influence of the local stiffness 25% could be achieved (Walz et al., 1983; Fieldwick
of contact area on the injury severity of different body & Brown, 1987; Anderson et al., 1997). A literature
segments. The stiffness variables have significant ef- review on vehicle travel speed and pedestrian injuries
fects on head and lower extremity injuries, with a given summarized that only 5% of pedestrian would be killed

Figure 8 Effects and interactions of vehicle front stiffness variables on pedestrian injuries: (a) Hood top stiffness—HIC; (b) Hood-edge
stiffness—Thigh impact force; (c) Bumper stiffness—Tibia acceleration; (d) Bumper stiffness—Knee lateral bending angle.
40 X. J. LIU ET AL.

at the traveling speed of 32 km/h, while the death rate icantly affected by the impact speed, and can be miti-
would be up to 40%, 80% and 100% at the traveling gated by altering vehicle front shape and the stiffness
speed of 48, 64 and 90 km/h, respectively (Leaf & properties of structures.
Preusser, 1999). These results have proven the necessity Impact speed has the critical effect on the pedestrian
of reducing speed limit in urban area. However, vehi- injury severity. As the impact speed decreases from
cle travel speed is usually higher than impact speed, 40 to 30 km/h, the probability of severe head injury
which is the major concern in crash safety research. (HIC > 1000) decreases from 50% to lower than 25%.
In this study, a reduction of 25% of severe injuries However the injury risks to knee joint remains above
(HIC > 1000) is found if the vehicle impact speed 50% at 30 km/h. Considering the possible improve-
changes from 40 to 30 km/h, which shows a reason- ment of vehicle front structure to mitigate the injury
able agreement with the previous studies. to the knee joint, a speed limit of 30 km/h in urban area
The influences of vehicle front structures on pedes- is feasible way to reduce the pedestrian injury risk.
trian kinematics and injuries have been discussed in As to the kinematics and resultant head velocity of
many studies (Niederer & Schlumpf, 1984; Harris & pedestrian struck by passenger cars, the hood edge height
Downloaded by [Umeå University Library] at 01:07 17 November 2014

Grew, 1985; Ishikawa et al., 1991). However, these stud- has been identified as the dominant factor. The effect
ies were limited on the bonnet-type vehicle at a given of the bumper center height and bumper lead length
impact speed. The biofidelity of pedestrian substitutes is slight. In general, the head impact speed decreases
were limited, especially in lower extremities. The knee with a raising hood edge height and a lowering bumper
joint injuries were often omitted. In this study, the in- center height. The injury severity to the knee joint is
fluences of different vehicle are evaluated at the impact mainly influenced by the bumper central height and
speed from 20 to 50 km/h with the validated mathemat- stiffness properties. The bumper lead and hood edge
ical model (Yang et al., 2000). This validated model height only has slight effect on the lateral bending an-
can properly predict the overall kinematics, dynamic gle of knee.
responses of body segments, as well as the insight into The local stiffness of head contact area strongly con-
knee joint injuries. trols the resulted injury severity of head. The influ-
There are several limitations of the present study. ence of the force-deformation properties of bumper and
First, the initial impact location is restricted at the cen- hood edge on the pedestrian kinematics and head im-
tral line of vehicle model. Other accident situations are pact speed is slight, but significant on the resulted injury
not included. Second, the involvement of different vehi- severity to the involved body segments.
cle types is constant at different impact speeds, which
needs more detailed accident data to refine the simu-
lation matrix. Third, the adult model (Yang, 1997) was REFERENCES
well validated at the impact speeds of 25, 32 and 39 km/h
with two given vehicle models. However, the speed Aldman, B., Kajzer, J., Cesari, D., Bouquet, R., and Zac, R. (1985a).
in this study ranges from 20 to 50 km/h with wide A New Dummy for Pedestrian Test, Proc. 10th Int. Technical
variations of vehicle models. This may result in Conf. on Experimental Safety Vehicle, pp. 176–185.
Aldman, B., Kajzer, J., Anderlind, T., Malmqvist, M., Mellander, H.,
some inaccuracy of simulation results. Fourth, only the
and Turbell, T. (1985b). Load Transfer from the Striking Vehicle
50th percentile adult male is considered. Other body in Side and Pedestrian Impacts, Proc. 10th Int. Technical Conf.
sizes, like small female and children at different ages are on Experimental Safety Vehicle, pp. 620–637.
excluded. Anderson, R. W. G., Mclean, A. J., Farmer, M. J. B., Lee, B. H., and
In the future study, the most interest will focus on the Brooks, C. G. (1997). Vehicle Travel Speeds and the Incidence
development of child pedestrian models at various ages of Fatal Pedestrian Crashes, Accident Analysis & Prevention,
Vol. 29(5), pp. 667–674.
and the influences of car-front structures on the injury Ashton, S. J. (1982). Vehicle Design and Pedestrian Injuries, Ch. 6,
risks of child pedestrians. Pedestrian Accidents, Chapman et al., (eds.) John Wileys & Sons
Ltd.
Berg, F. A., Schmitt, B., Weis, S., Mattern, R., and Kallieris, D.
CONCLUSIONS (2000). Pedestrian Head Impact on the Windscreen of Compact
Cars—A New Test Rig and First Results. Proc. Int. IRCOBI Conf.
The validated pedestrian model based on human Biomechanics Impacts, pp. 293–305.
Björklund, M., and Zheng, Q. (2001). Development and Evalua-
body characteristics is an effective means to study the
tion of a Pedestrian Anthropomorphic Test Device—A Prelimi-
complex collision event between vehicle structure and nary Study on Responses of Pedestrian Models in Car Collisions.
human body. The results from this parametric study Master Thesis, Crash Safety Division, Chalmers University of
indicate that the injury severity of pedestrian is signif- Technology, Göteborg, Sweden.
INFLUENCES OF VEHICLE SPEED AND FRONT STRUCTURE 41

Bunketorp, O. (1983). Pedestrian Leg Protection in Car Accidents— Kajzer, J., and Schroeder, G. (1992). Examination of Different
An Experimental and Clinical Study. Ph.D Thesis, Dept. Bumper System using Hybrid II, RSPD Subsystem and Cadav-
Orthopedic Surgery II, Göteborg University, and Dept of Traffic ers. Proc. 36th Stapp Car Crash Conf, SAE Paper 922519, pp.
Safety, Chalmers University of Technology Göteborg, Sweden. 119–127.
Cavallero, C., Cesari, D., Ramet, M., Billault, P., Farisse, J., Seriat- Leaf, W. A., and Preusser, D. F. (1999). Literature Review of Ve-
Gautier, B., and Bonnoit, J. (1983). Improvement of Pedestrian hicle Travel Speeds and Pedestrian Injuries, Reports. DOT HS
Safety: Influence of Shape of Passenger Car-Front Structures 809 012, National Highway Traffic Safety Administration, U.S.
Upon Pedestrian Kinematics and Injuries: Evaluation Based on 50 Department of Transportation.
Cadaver Tests. Pedestrian Impact Injury & Assessment, SAE Int. Matsui, Y., and Ishikawa, H. (1998). Validation of Pedestrian Upper
Congress Exposition, P-121, SAE Paper 830624, pp. 225–237. Legform Impact Test—Reconstruction of Pedestrian Accidents.
Cesari, D., and Rameet, M. (1982). Pelvic Tolerance and Protection Proc. 16 Int. Technical Conf. on Enhanced Safety of Vehicles, pp.
Criteria in Side Impact. Proc. 26th Stapp Car Crash Conf., SAE 2152–2167.
Paper 821159, pp. 145–154. Mizuno, K., and Kajzer, J. (2000). Head Injuries in Vehicle-
Cesari, D., Bouquet, R., Caire, Y., and Bermond, F. (1994). Protec- Pedestrian Impact. Biomechanics research, SAE 2000 World
tion of Pedestrians Against Leg Injuries. Proc 14th Int. Technical Congress, SP-1494, SAE Paper 2000-01-0157, pp. 29–40.
Conf. on Enhanced Safety of Vehicles, pp. 1131–1138. Nagatomi, K., Akiyama, A., and Kobayashi, T. (1996). Bumper
EEVC (1994). Proposals for Methods to Evaluate Pedestrian Protec- Structure for Pedestrian Protection. Proc. 15th Int. Technical
Downloaded by [Umeå University Library] at 01:07 17 November 2014

tion for Passenger Cars, Report, European Experimental Vehicle Conf. on Enhanced Safety of Vehicles, pp. 593–601.
Committee, Working Group 10. NHTSA (1999). Traffic Safety Facts—Pedestrian, Report, National
EEVC (1998). Improved Test Methods to Evaluate Pedestrian Pro- Highway Traffic Safety Administration, Department of Trans-
tection Afforded by Passenger Cars, Report, European Experi- portation, Washington DC, U.S.A.
mental Vehicle Committee, Working Group 17. Niederer, P. F., and Schlumpf, M. R. (1984). Influence of Vehicle
ETSC (1999). Safety of Pedestrian and Cyclists in Urban Areas, Front Geometry on Impacted Pedestrian Kinematics. Proc. 28th
Report, European Transport Safety Council, Brussels. Stapp Car Crash Conf., SAE Paper 841663, pp. 207–219.
Fieldwick, R., and Brown, R. J. (1987). The Effect of Speed Limits Otte, D. (1999). Severity and Mechanism of Head Impacts in Car
on Road Casualties. Traffic Engineering and Control, Vol. 28, pp. to Pedestrian Accidents. Proc. Int. IRCOBI Conf. Biomechanics
635–640. Impacts, pp. 329–341.
Gibson, T. J., Hinrichs, R. W., and Mclean, A. J. (1986). Pedestrian Pritz, H. B. (1983). Experimental Investigation of Pedestrian Head
Head Impacts: Development and Validation of a Mathematical Impacts on Hoods and Fenders of Production Vehicles. Pedestrian
Model, Proc. Int. IRCOBI Conf. Biomechanics Impacts, pp. 165– Impact Injury & Assessment, SAE Int. Congress and Exposition,
176. P-121, SAE Paper 830055, pp. 67–76.
Harris, J., and Grew, N. D. (1985). The Influence of Car Design on Sakurai, M., Kobayashi, K., Ono, K., and Sasaki, A. (1994). Evalua-
Pedestrian Protection. Proc 10th Int. Technical Conf. on Experi- tion of Pedestrian Protection Test Procedure in Japan ∼ Influence
mental Safety Vehicle, pp. 1009–1022. of Upper Body Mass on Leg Impact Test. Proc. 14th Int. Technical
Harris, J. (1989). A Study of Test Methods to Evaluate Pedestrian Conf. on Enhanced Safety of Vehicles, pp. 1114–1130.
Protection for Cars. Proc. 12th Int. Technical Conf. on Experi- Schuster, P. J., Chou, C. C., Prasad, P., and Jayaraman, G. (2000).
mental Safety Vehicle, pp. 1217–1225. Development and Validation of a Pedestrian Lower Limb Non-
Higuchi, K., and Akiyama, A. (1991). The Effect of the Vehicle Linear 3-D Finite Element Model. Proc. 44th Stapp Car Crash
Structure’s Characteristics on Pedestrian Behavior. Proc. 13th Int. Conf., SAE Paper 2000–01-SC21, pp. 315–334.
Technical Conf. on Experimental Safety Vehicle, pp. 323–329. Takahashi, Y., Kikuchi, Y., Konosu, A., and Ishikawa, H. (2000).
Huang, T. J., McDonald, J., Artis, M., Rangarajan, N., Shams, Development and Validation of the Finite Element Model for the
T., White, Jr. R. P., Beach, D., Campbell, Jr. R., Akiyama, A., Human Lower Limb of Pedestrians. Proc. 44th Stapp Car Crash
Yoshida, S., Ishikawa, H., and Konosu, A. (1999). Development Conf., SAE Paper 2000-01-SC22, pp. 335–355.
of a Biofidelic Dummy for Car-pedestrian Accident Studies. Proc. TAPSM (1997). Road Traffic Accident Data in 1996 from P.R.
Int. IRCOBI Conf. Biomechanics Impacts, pp. 315–327. China. Traffic Administration of Public Security Ministry,
Ishikawa, H., Yamazaki, K., Ono, K., and Sasaki, A. (1991). Current Report, August 1997.
Situation of Pedestrian Accidents and Research into Pedestrian Van Wijk, J., Wismans, J., Maltha, J., and Wittebrood, L. (1983).
Protection in Japan. Proc. 13th Int. Technical Conf. on Experi- MADYMO Pedestrian Simulations. Pedestrian Impact Injury &
mental Safety Vehicle, pp. 281–293. Assessment. SAE Int. Congress & Exposition, P-121, SAE Paper
Ishikawa, H., Kajzer, J., and Schroeder, G. (1993). Computer Simu- 830060, pp. 109–117.
lation of Impact Response of the Human Body in Car-Pedestrian Walz, F. H., Hoefliger, M., and Fehlmann, W. (1983). Speed Limit
Accidents. Proc. 37th Stapp Car Crash Conf., SAE Paper 933129, Reduction from 60 to 50 km/h and Pedestrian Injuries. Proc.
pp. 235–248. 27th Stapp Car Crash Conf., SAE Paper 831625, pp. 311–
Ishikawa, H., Kajzer, J., Ono, K., and Sakurai, M. (1994). Simu- 318.
lation of Car Impact to Pedestrian Lower Extremity: Influence Yang, J. K., and Kajzer, J. (1992). Computer Simulation of Impact
of Different Car-Front Shapes and Dummy Parameters on Test Response of the Human Knee Joint in Car-pedestrian Accidents.
Results. Accident Analysis and Prevention, Vol. 26(2), pp. 231– Proc. 36th Stapp Car Crash Conf., SAE Paper 922525, pp. 203–
242. 217.
Janssen, E. G., and Wismans, J. (1985). Experimental and Math- Yang, J. K., Rzymkowski, C., and Kajzer, J. (1993). Develop-
ematical Simulation of Pedestrian-Vehicle and Cyclist-Vehicle ment and Validation of a Mathematical Breakable Leg Model.
Accidents. Proc. 10th Int. Technical Conf. on Experimental Safety Proc. Int. IRCOBI Conf. Biomechanics Impacts, pp. 175–
Vehicle, pp. 977–988. 186.
42 X. J. LIU ET AL.

Yang, J. K. (1997). Injury Biomechanics in Car-Pedestrian Colli- Pedestrian Impacts. Journal of Crash Prevention and Injury Con-
sions: Development, Validation and Application of Human-Body trol, Vol. 2(2), pp. 131–149.
Mathematical Models. Ph.D Dissertation, Chalmers University Yoshida, S., and Matsuhashi, T. (1998). Simulation of Car-
of Technology, Göteborg, Sweden. Pedestrian Accident for Evaluate Car Structure. Proc. 16th Int.
Yang, J. K., Lövsund, P., Cavallero, C., and Bonnoit, J. (2000). Technical Conf. on Enhanced Safety of Vehicles, pp. 2344–
A Human-Body 3D Mathematical Model for Simulation of Car- 2348.
Downloaded by [Umeå University Library] at 01:07 17 November 2014

You might also like