Document Sample
					Journal of the Chinese Institute of Engineers, Vol. 26, No. 6, pp. 835-843 (2003)                                         835


      Li-Tung Chang, Guan-Liang Chang, Ji-Zhen Huang, Shyh-Chour Huang, De-Shin Liu and
                                      Chih-Han Chang*

                      A motorcycle helmet is the best protective headgear for the prevention of head
                injuries caused by direct cranial impact. A finite element model based on realistic
                geometric features of a motorcycle helmet is established, and explicit finite element
                LS-DYNA code is employed to simulate dynamic responses at different impact
                velocities. Peak acceleration and Head Injury Criterion values derived from the
                headform are used to assess the protective performance of the helmet. We have con-
                cluded that the dynamic responses of the helmet dramatically vary with impact velocity,
                as well as the mechanical properties of the outer shell and energy-absorbing liner. At
                low velocities, e.g., less than 8.3 m/s, the shell stiffness and liner density should be
                relatively low to diminish head-contact force. At high velocity, e.g., 11.1 m/s, a stiffer
                shell and denser liner offer superior protection against head injuries.

                Key Words: head injury, impact, motorcycle helmet, finite element analysis.

                   I. INTRODUCTION                                      other countries differ from Taiwan, head injury of the
                                                                        rider remains a significant health problem (Hurt et
      Motorcycles and scooters are popular and pro-                     al., 1981; Matzsch and Karlsson, 1986; Sood, 1988;
vide an important means of transportation in both                       Kraus and Peek, 1995).
developing and developed countries. Approximately,                            Wearing a motorcycle helmet is the best way to
50% of the Taiwanese population owns a motorized                        prevent head injuries from traffic accidents (Kraus
two-wheeled vehicle. Therefore, it is not surprising                    and Peek, 1995; Gopalakrishna et al., 1998; Tsai and
that 45% of all traffic-accident fatalities in Taiwan                   David, 1999). To ensure protective performance,
involve motorcyclists, with most deaths resulting                       shock absorption test codes have been established in
from severe head injuries (Lee et al., 1990; Lin et                     many countries (CNS 2396, 1998; JIS T 8133, 2000;
al., 1998). Although patterns of motorcycle usage in                    BS 6658, 1985; E.C.E. R22, 1998; FMVSS 218,
                                                                        1998; Snell Memorial Foundation, 2000). In most of
                                                                        these standards, the peak acceleration of a magnesium
  *Corresponding author. (Tel: 886-6-2757575 ext. 63427; Fax:           alloy headform within the helmet should not exceed
886-6-2343270; Email:                      300 G (G=9.8 m/s 2), and the Head Injury Criterion
  L. T. Chang is with the Department of Childhood Education and         should be less than 2400 (E.C.E. only) with the im-
Nursery, Chia Nan University of Pharmacy and Science, Tainan,
Taiwan 717, R.O.C.
                                                                        pact energy ranging from 80-150 J (depending on
  G. L. Chang, J. Z. Huang, and C. H. Chang are with the Institute      the standards). Therefore, commercial helmet struc-
of Biomedical Engineering, National Cheng Kung University,              tures are principally designed to meet the specifica-
Tainan, Taiwan 701, R.O.C.                                              tions set out in these standards. In general, a motor-
  S. C. Huang is with the Department of Mechanical Engineering,         cycle helmet consists of a hard outer shell [isotropic
National Kaohsiung University of Applied Sciences, Kaohsiung,
Taiwan 807, R.O.C.
                                                                        molded plastic or fiber-reinforced plastic (FRP)], an
  D. S. Liu is with the Department of Mechanical Engineering,           energy-absorbing liner (expanded polystyrene foam
National Chung Cheng University, Chia-Yi, Taiwan 621, R.O.C.            or polypropylene foam) and inner comfort foam
836                             Journal of the Chinese Institute of Engineers, Vol. 26, No. 6 (2003)

           Outer shell
                                            Comfort foam




Fig. 1   A typical cross-sectional view of motorcycle helmet con-

(open-cell polyurethane or PVC; Fig. 1). The outer
shell serves mainly to distribute contact forces, while
the polystyrene liner absorbs the impact energy. The
comfort foam distributes the static contact force to
avoid headaches (Gilchrist et al., 1988).
      Many studies have been conducted to evaluate                                                         Anvil
the protective performance of helmets during direct
head impact, with constant-rate compression and
drop-impact tests which are typically used to investi-
gate the protective contribution of individual helmet
components (Kingsbury and Rohr, 1981; Gale and                         Fig. 2 Experimental setup for helmet shock absorption test
Mills, 1985; Mills and Gilchrist, 1991; Gilchrist and
Mills, 1994a). Gilchrist and Mills (1994b) used one-
dimensional mathematical approaches to examine the
dynamic response of helmets. Yettram et al. (1994)                           II. METHODS AND MATERIALS
established a finite element (FE) model to study the
influences of construction material on helmet perfor-               1. Shock Absorption Test
mance. Chang et al. (2000) used FE analysis to in-
vestigate the influence of helmet chin-bar design on                      To validate the FE model, experimental devices
head injuries. In these simulation studies, helmet                  were set up according to the procedures outlined in
geometry was simplified, with either spherical or                   the CNS 3902 code (1998). A tridirectional acceler-
regular shapes adopted. Thus, these mathematical                    ometer was installed at the center of a 5-kg magne-
models differ from real-world situations. In addition,              sium alloy headform, with the y-axis set parallel to
most of the studies utilized impact velocities ranging              the direction of gravity. The specimen helmet was
from 5.6-7.7 m/s (20-28 km/hr), as required by the                  then placed on the headform, and the composite
various helmet standards. However, the impact ve-                   dropped along two cables from a set height to strike
locity adopted for a particular standard probably re-               a flat steel anvil (Fig. 2). The acceleration-time trace
flects an underlying economic rationale rather than                 for the headform was recorded during impact. To con-
an attempt to define the conditions experienced in real             firm the impact energy, the velocity at the instant of
accidents. At higher velocities (over 30 km/hr), the                impact was measured using two optical sensors
dynamic response of the helmet and the influence of                 (Chang et al., 1999).
material on helmet performance are seemingly
unclear.                                                            2. FE Model
      The objectives of this study were, therefore, to
establish a FE model based on realistic geometric fea-                   The FE model was established according to the
tures of a motorcycle helmet and known material                     standard shock absorption test outlined above. The
properties, and to assess the protective performance                model consisted of three components, helmet,
of the helmet with respect to head injuries at distinct             headform, and flat anvil. In order to realistically rep-
impact velocities.                                                  resent the geometric features of the motorcycle
      L. T. Chang et al.: Finite Element Analysis of the Effect of Motorcycle Helmet Materials Against Impact Velocity    837

             Energy-absorbing                                    Table 1 The mechanical properties of the differ-
                                     Outer shell                         ent shell materials
                                                                   Shell             Elasticity    Hardening      Yield
                                                                   material          modulus        modulus       stress
                                                                                      (GPa)          (GPa)        (MPa)
                                                                   FRP                  20              6           150
                                                                   PC                    4.5            1.5          75
                                             Headform              ABS (baseline)        1.7            1.2          35
                                                                   HDPE                  1              0.5          30

                                                                 3. Material Properties

                                                                       The outer shell of the helmet used in this study
                                                                 was made from the isotropic molded plastic, ABS
      Fig. 3 The cross-sectional view of the FE model            (acrylonitrile, butadiene, and styrene copolymer). To
                                                                 determine the material properties of the shell, fast
                                                                 tensile tests were performed using the Material Test
                                                                 System (MTS). Analysis of the generated stress-strain
helmet, computer tomography was used to obtain a                 curves demonstrated that the material had a two-stage
series of transverse sectional images (6 mm interval)            mechanical behavior. Therefore, a bilinear kinematic
of a commercially available full-face helmet (KC-560;            hardening plasticity model, defined by elasticity
SUPA, Taiwan). The contours of the helmet were                   modulus (1.7 GPa), yielding stress (35 MPa), and
detected using an in-house image processing program.             hardening modulus (1.2 GPa), was used to represent
Nodes were generated from these contours and linked              the shell element.
into elements to form the FE model. The average                        In the simulation, ABS was selected as the
thickness of the shell was 4 mm, while that of the               baseline material for the shell component. FRP, PC,
energy-absorbing liner was 37 mm. The headform                   HDPE (high-density polyethylene) were also em-
mesh was based on the medium size headform of the                ployed as the shell material, for comparison with
Chinese standard (CNS 2396, 1998). The anvil mesh                ABS. The material property values were inferred
represented a square block 20-mm thick. A cross-                 from Waterman and Ashby (1997) as listed in Table
sectional view of the FE model is presented in Fig. 3.           1. Comparing the elasticity modulus values for these
The helmet and anvil model consisted of eight-node               materials with ABS, it can be seen that FRP is ap-
solid elements, while the headform consisted of four-            proximately ten times stiffer, while PC is over twice
node shell elements. The numbers of elements for                 as stiff and HDPE is nearly half as rigid.
each component were 4672 elements for the shell,                       The energy-absorbing liner in the FE model was
4986 for the liner, 300 for the headform and 16 for              defined as crushable foam material, with a curve of
the anvil. Contact elements were placed between                  volumetric strain against stress required to define the
the headform and the liner, as well as between the               mechanical properties of the material. An FE model
shell and the anvil. In addition, to reduce the com-             based on Yettram’s (1994) impact experiment, in
putational efforts required by the FE model, both                which a steel impactor was used to strike a liner
the headform and the anvil were defined as rigid                 (100×100×20 mm) placed on an anvil and impact ve-
bodies.                                                          locity of 6.9 m/s was employed, was established to
      To evaluate the dynamic responses of the model,            obtain this curve. After a series of trials and errors,
three impact velocities, low, medium and high, were              the mechanical behaviors of the liners with three dif-
employed [5.6 m/s (20 km/hr), 8.3 m/s (30 km/hr)                 ferent densities (24, 44 and 57 kg/m3) were established
and 11.1 m/s (40 km/hr), respectively]. To reduce                (Fig. 4).
the computational time required, the helmet was
placed very close to the anvil and given an initial              4. The Evaluation Indexes
impact velocity instead of being dropped. The ex-
plicit finite element code LS-DYNA (Rev. 960,                         The purpose of a motorcycle helmet is to pre-
Livermore Software Technology Corporation;                       vent or reduce the severity of head injuries in the event
Livermore, CA, U.S.A.) was used to simulate the                  of impact. To determine the protective performance
impact response of the helmet.                                   of a helmet, most test standards evaluate the peak
838                                            Journal of the Chinese Institute of Engineers, Vol. 26, No. 6 (2003)

               1.5               Expt.     24 kg/m3                                                   300                                                    Expt.
                                 FEA                                                                                                                         FEA

Stress (MPa)


                                                                                   Acceleration (G)

                  0.0      0.2           0.4          0.6    0.8      1.0
                                          Strain (%)
                                                                                                             0                         5                      10
                                                                                                                                        Time (ms)
               1.5               Expt.     44 kg/m3
                                                                                   Fig. 5                   Acceleration-time traces of the headform in the experi-
                                                                                                            ment and finite element model at medium-velocity impacts
                                                                                                            (8.3 m/s)

Stress (MPa)

                                                                                   Injury Criterion (HIC) (Versace, 1971), which is a
                                                                                   widely accepted index for assessing head injury, was
                                                                                   also derived. The calculation of HIC considers not
                                                                                   only the effects of the peak acceleration but also the
                                                                                   distribution and duration of the acceleration. The HIC
               0.0                                                                 is defined as the maximum value obtained from the
                  0.0      0.2           0.4          0.6    0.8      1.0          equation:
                                          Strain (%)                                                                        t2
                                                                                                       HIC = [ t 1 t             a(t)dt] 2.5 ⋅ (t 2 – t 1)
                                                (b)                                                             2– 1       t1

               1.5               Expt.     57 kg/m3                                where a(t) is the acceleration of the headform (in Gs),
                                 FEA                                               and t 1, t 2 are the bounds of all possible time intervals
                                                                                   defining the total duration of impact (in seconds).
                                                                                                                          III. RESULTS
Stress (MPa)

                                                                                         To validate the FE model, shock absorption ex-
                                                                                   periments were performed at impact velocities of 5.6
                                                                                   and 8.3 m/s. The acceleration-time traces for the
                                                                                   headform during the impact experiments were com-
                                                                                   pared with those for the corresponding simulations.
               0.0                                                                 In general, the pattern and peak value of the simu-
                  0.0      0.2           0.4          0.6    0.8      1.0          lated acceleration curves agreed well with the experi-
                                                                                   mental data. At the impact velocity of 8.3 m/s, the
                                          Strain (%)
                                                                                   peak acceleration was 288 G at 5.1 ms in the
                                                                                   simulation, which was consistent with the experimen-
Fig. 4           The simulated stress-strain curve of the polystyrene liner        tal outcome (255 G at 4.9 ms; Fig. 5). The maximum
                 compared with Yettram’s impact experiment. (a) 24; (b)            differences in peak acceleration and HIC values be-
                 44; (c) 57 kg/m3
                                                                                   tween the experimental and simulated impacts for
                                                                                   velocity of 5.6 m/s were 7% and 18%, respectively,
                                                                                   and for 8.3 m/s were 13% and 14%, respectively.
acceleration of the headform during impact. In addi-                                     The peak accelerations and HIC values for all
tion to the peak acceleration, the value of Head                                   investigated parameters in the simulations are listed
      L. T. Chang et al.: Finite Element Analysis of the Effect of Motorcycle Helmet Materials Against Impact Velocity        839

 Table 2 The peak acceleration (G) of the headform for a variety of shell stiffnesses and liner densities
  Shell Mat.               HDPE                       ABS                         PC                        FRP
  Liner Den.        24      44        57       24      44       57       24       44       57       24       44          57
  Low Vel.         114      157      186      126      176      185      132     196      208      141      219      248
  Medium           164      274      301      174      288      290      175     286      299      179      292      334
  High              −       466      537       −       456      514       −      412      491       −       371      437
Low Vel.: 5.6 m/s; Medium Vel.: 8.3 m/s; High Vel.: 11.1 m.s

         Table 3 HIC values of the headform for a variety of shell stiffnesses and liner densities
  Shell Mat.              HDPE                        ABS                         PC                        FRP
  Liner Den.        24     44         57      24       44       57       24       44       57       24       44          57
  Low Vel.         628      872     1074      637      977     1134      631     1004    1175      769     1414     1738
  Medium          1503     2939     3290     1508     2757     3334     1435     2884    3525     1638     3551     4533
  High              −      7500     8959       −      6660     8261       −      6406    7849       −      6819     8814

in Tables 2 and 3. At the high velocity, the lowest-             the simulations (Fig. 5). This discrepancy is possi-
density 24-kg/m 3 liner was over-compressed by the               bly a reflection of differences in the material proper-
headform, with a severe bottoming-out phenomenon                 ties of the real and modeled components, and small
causing a divergence in the FE simulations. As this              cracks discovered in the liner, which were induced
caused a divergence in the simulations, the data of              by headform compression but could not be simulated.
the lowest-density liner during high-velocity impacts            The second difference was that the simulated accel-
were excluded. At the low velocity, the values of                eration curve decayed more rapidly than the experi-
both evaluation indexes increased as helmet-shell                mental analog. This discrepancy was attributed
stiffness increased. The maximum variations in peak              mainly to the steepness of the unloading slope for the
acceleration and HIC values comparing different shell            simulated liner material (crushable foam) using LS-
constructions were 40% and 62%, respectively. At                 DYNA (Fig. 4). Different unloading paths will re-
the medium velocity, a consistent trend was not dem-             sult in inaccuracy of FE modeling, but it will not af-
onstrated for peak acceleration or HIC as shell stiff-           fect the contributions of the shell and liner in the FE
ness varied, with only small differences in both in-             model. In the study of Chang et al. (2001) in which a
dexes noted. At high velocity, however, the peak ac-             helmet model was used to analyse the effect of fit, it
celeration was lower for the stiffer shells, with a maxi-        was demonstrated that the above differences had little
mum variation of 26% demonstrated; the same trend                effect on the dynamic response of the helmet.
was noted for HIC index, with a maximum variation                      At the low velocity (5.6 m/s; 80 J), shell struc-
of 17%.                                                          tures that were less stiff provided better protection,
      Varying liner density significantly altered the            with reductions in peak acceleration and HIC values
peak acceleration and HIC, with both values increas-             of up to 40% and 62%, respectively (see Tables 2 and
ing as liner density increased, regardless of shell stiff-       3). This finding was consistent with those of other
ness and impact velocity varied. The maximum varia-              model and simulation studies (Mills and Gilchrist,
tions of the peak acceleration and HIC for the low               1991; Gilchrist and Mills, 1994a; Yettram et al., 1994;
velocity were 76% and 126%, respectively, for the                Chang et al., 2000) where reduced peak acceleration
medium velocity were 87% and 177%, respectively,                 and HIC values were demonstrated in flat-surface
and for the high velocity were 19% and 29%,                      impacts using helmets with shells that were less stiff.
respectively.                                                    One reason for this reduced acceleration was that the
                                                                 greater shell flexibility resulted in larger
                  IV. DISCUSSION                                 deformations, absorbing more impact energy. This
                                                                 effect was reflected in the internal energy history of
      Comparing the acceleration-time traces for the             the shell and energy-absorbing liner (Fig. 6). As
headform from the experimental and simulated                     larger elastic deformations of the shell delayed im-
results, two major differences were apparent. The                pact resilience which occurred after the time point of
first was the lower peak accelerations derived from              peak force of headform, reducing the interaction be-
analysis of the experimental data in comparison to               tween the energy-absorbing liner and the moving
840                                             Journal of the Chinese Institute of Engineers, Vol. 26, No. 6 (2003)

                        100                                       FRP                                 25                FRP
                                                                  PC                                                    PC
                                                                  HDPE                                                  HDPE
  Internal energy (J)

                                                                                         Force (kN)


                          0                                                                            0
                              0                     5                        10                            0       10          20       30        40        50
                                                Time (ms)                                                                 Displacement (mm)
                                                   (a)                                                                            (a)

                        100                                                                           25                FRP
  Internal energy (J)

                                                                                       Force (kN)

                                                                      PC                               5

                          0                                                                            0
                              0                     5                        10                            0      10           20      30        40        50
                                                Time (ms)                                                                 Displacement (mm)
                                                   (b)                                                                            (b)

Fig. 6                   Internal energy traces of the shell and energy-absorbing   Fig. 7             The relationship of headform force and displacement with
                         liner at low-velocity impacts (5.6 m/s)                                       different shell materials within peak force. (a) at low-ve-
                                                                                                       locity impacts, the design of the HDPE shell extended ef-
                                                                                                       fective headform displacement through reduction in con-
                                                                                                       tact area with energy-absorbing liner and resulted in more
                                                                                                       gradual curve before peak force; (b) at high-velocity
headform, headform peak acceleration was decreased.                                                    impacts, the stiffer shell provided an increase in the vol-
Moreover, with larger deformation of the shell, the                                                    ume of crushed liner and a decrease in the compression
contact area between the headform and the liner was                                                    strain of the liner, hence, stiffer shell had smaller peak
                                                                                                       force than the less-stiff shell
reduced, thereby decreasing the contact force on the
headform and, thus, the acceleration. This effect is
elaborated further in the diagrammatic relationship
of headform force and displacement (Fig. 7(a)). The
displacement rate for the headform in the HDPE shell                                provided an increase in the volume of crushed liner,
was greater than for the FRP variant, with the design                               due to a reduction in bending deformation of the shell
of the HDPE shell effectively extending headform                                    and liner (Fig. 8(a)). The combination of liner with
displacement through reduction in contact area, hence,                              stiffer shell thus absorbed a larger proportion of en-
decreasing the peak acceleration and HIC values for                                 ergy during impact than the less-stiff analog (Fig.
the headform at the low velocity.                                                   8(b)). This vital effect dictated the maximum force
      At the high velocity (11.1 m/s; 310 J), the trends                            on the headform during high-energy shock (Fig. 7(b)).
for both indexes contrasted with those at low velocity,                             As the compression strain on the liner increased over
with lower index values derived for helmets with a                                  about 60%, the contact force to headform increased
stiffer shell in comparison to less-stiff analogs. The                              dramatically (Fig. 4). The compression strain on the
reductions in peak-acceleration and HIC values were                                 liner in the stiffer shell was lower than that on the
26% and 17%, respectively, because the stiffer shell                                less-stiff analog, hence, both indexes were lower for
                          L. T. Chang et al.: Finite Element Analysis of the Effect of Motorcycle Helmet Materials Against Impact Velocity   841

                                    Shell                             FRP
                                                                                      reduction was that a lower-density liner offered
                                                                      PC              greater physical compliance, providing improved dis-
                                                                      HDPE            tribution of headform force during impact. This re-
                          300                                                         duced headform force and delayed the occurrence of
    Internal energy (J)

                                                                                      the peak value, thus reducing both indexes of
                                                                                      headform. For the high-velocity simulations in this
                                                                                      study, however, the 24-kg/m 3 liner was only 37 mm
                                                                                      thick, and the density was too low for complete dis-
                          100                                                         sipation of impact energy. Bottoming out in the con-
                                                                                      tact region was severe, with the functional equiva-
                                                                                      lent of metal-to-metal contact between headform and
                            0                                                         liner producing a high impact force, which could not
                                0                     5                        10     be simulated in this study. Therefore, the reductions
                                                  Time (ms)                           in peak-acceleration and HIC values for the less-dense
                                                     (a)                              liner were based on the assumption that the bottom-
                                                                                      ing-out effect during impact was not significant. To
                          400                                                         overcome this problem of shock transmission in high-
                                                                                      energy impacts, increasing the thickness of a less-
                                                                                      dense liner appears to offer an acceptable design
                          300                                                         improvement, however, utilization of a thicker liner
  Internal energy (J)

                                                                                      increases both the volume and mass of the helmet,
                                                                                      with obvious disadvantages with respect to loading
                                                                                      of the cervical spine (Huston and Sears, 1981; Huang,
                                                                                      1999). For the high-velocity trials in this study, the
                          100                                          FRP            best shock absorption was demonstrated for the 44-
                                                                       PC             kg/m3 liner (see Tables 2 and 3). Hence, increasing
                                                                       HDPE           the liner density should offer a superior approach for
                            0                                                         improved protection during high-energy impacts.
                                0                     5                       10      Analysis of the simulation data reveals that the low-
                                                  Time (ms)                           est indexes for low and medium velocities were for
                                                     (b)                              the helmet with the 24-kg/m 3 liner and the less-stiff
                                                                                      HDPE shell. Differences in the helmet design for low
Fig. 8                     Internal energy traces of the shell and energy-absorbing
                           liner at high-velocity impacts (11.1 m/s)                  and medium-velocity trials produced variations in
                                                                                      peak acceleration and HIC values of up to 118% and
                                                                                      177% and 104% and 202%, respectively (see Tables
                                                                                      2 and 3). By contrast, at high velocity both indexes
the former relative to the latter (Tables 2 and 3).                                   were lower for the 44-kg/m3 liner and the stiffer shells
      At the medium velocity (8.3 m/s), neither                                       (PC or FRP), with variations in peak acceleration and
index value was altered significantly as shell stiff-                                 HIC values of up to 45% and 40%, respectively. Based
ness varied (see Tables 2 and 3). In comparison to                                    on these findings, it appears reasonable to suggest that
the low and high-velocity variants, both indexes at                                   helmet design should reflect a projected range of im-
the medium velocity reflected an apparent turning                                     pact velocities, with shell and liner properties selected
point. This finding demonstrated that impact veloc-                                   to provide the best possible protection against head
ity was one of the most important parameters in                                       injury at the predicted speeds.
helmet design. Further, this variation in dynamic                                           In the real world, however, the velocity of the
characteristics was an indication that, in order to op-                               motorcycle in an accident and the terminal velocity
timize protective performance, projected use and,                                     of the rider’s head at impact are unknowns, with the
therefore, probable impact velocity, should be taken                                  latter depending on the former as well as on indeter-
into account in helmet design.                                                        minate conditions, while factors which might reason-
      The results of our parametric study of liner den-                               ably predict the former include geographical region,
sity for low and medium impact velocities were con-                                   road regulations, traffic conditions, motorcycle type,
sistent with the findings of previous helmet studies                                  age, personality of the rider and braking of the mo-
(Mills and Gilchrist, 1991; Yettram et al., 1994;                                     torcycle before the crash. These observations sug-
Chang et al., 2000), with lower-density (lower-                                       gest that, for a specific rider, the choice of helmet
stiffness) liners reducing headform peak acceleration                                 should reflect riding conditions and an ultimately id-
and HIC during impact. An explanation for this                                        iosyncratic assessment of risk.
842                        Journal of the Chinese Institute of Engineers, Vol. 26, No. 6 (2003)

      The specific aim of this study was to discuss the            Users,” British Standard Institution, London, UK.
issue of helmet design with respect to cranial impact          Chang, C. H., Chang, L. T., Chang, G. L., Huang,
at different impact velocities. The neurological and               S. C. and Wang, C. H., 2000, “Head Injury in Fa-
soft-tissue sequelae were considered beyond the scope              cial Impact: A Finite Element Analysis of Hel-
of the investigation, with only the engineering                    met Chin Bar Performance,” Transactions of the
indexes, peak acceleration and HIC employed to                     American Society of Mechanical Engineers: Jour-
evaluate the relative probability and comparative se-              nal of Biomechanical Engineering, Vol. 122, pp.
verity of any resultant head injuries.                             640-646.
                                                               Chang, L. T., Chang, C. H., and Chang, G. L.,
                 V. CONCLUSION                                     1999, “Experimental Evaluation of Chin Bar on
                                                                   Head Injury in Facial Impact,” Japan Society of
      In many Asian countries, motorized two-wheel-                Mechanical Engineers International Journal,
ers provide an important mode of transportation. As                Solid Mechanics and Material Engineering, Vol.
most traffic in the region is confined to narrow road-             42, No. 2, pp. 294-300.
ways crowded with vehicles and pedestrians, the                Chang, L. T., Chang, C. H., and Chang, G. L.,
speed of these two-wheeled vehicles is usually low,                2001, “Fit Effect of Motorcycle Helmet : A Fi-
with most riders needing a motorcycle helmet de-                   nite Element Modeling,” Japan Society of Me-
signed to provide protection against lower-velocity                chanical Engineers International Journal, Solid
impact. By contrast, motorcycles in the more-devel-                Mechanics and Material Engineering, Vol. 44,
oped countries are mainly used for leisure or                      No. 1, pp. 185-192.
entertainment. They tend to have larger-capacity en-           CNS 2396, 1998, “Protective Helmets for Motor
gines and freeway access, thus, their speed can reach              Cyclists,” Chinese Standard Institution, Taipei,
that of automobiles. These motorcyclists need a hel-               ROC.
met designed for higher-velocity impact. On the ba-            CNS 3902, 1998, “Method of Test for Protective Hel-
sis of our simulated results, therefore, the outer shells          mets of Cyclists,” Chinese Standard Institution,
of helmets designed for lower velocities (<8.3 m/s)                Taipei, ROC.
should be less stiff to reduce the contact area between        E.C.E. R 22, 1998, ”Uniform Provisions Concerning
the head and the energy-absorbing liner, and to sepa-              the Approval of Protective Helmets and of Their
rate the time interval between shell rebound and head              Visors for Drivers and Passengers of Motorcycles
movement, thus the force to the head will be reduced               and Mopeds,” Geneva, European Union.
during impact. Moreover, the energy-absorbing liner            FMVSS 218, 1998, “Motorcycle Helmets,” Federal
should be more compliant, reducing the contact force               Motor Vehicle Safety Standard, Washington,
to the head. By contrast, at higher velocities (above              USA.
8.3 m/s) the shell should be stiffer to increase the           Gale, A., and Mills, N. J., 1985, “Effect of Polysty-
energy absorption of the liner, and the liner should               rene Foam Liner Density on Motorcycle Helmet
be denser to increase capacity of energy absorption                Shock Absorption,” Plastics and Rubber Process-
and reduce bottoming-out effect.                                   ing and Applications, Vol. 5, No. 2, pp. 101-
             ACKNOWLEDGMENTS                                   Gilchrist, A., and Mills, N. J., 1994a, “Impact Defor-
                                                                   mation of ABS and GRP Motorcycle Helmet
       This study was supported by the National Sci-               Shells,” Plastics and Rubber Processing and Ap-
ence Council of the Republic of China (NSC-89-2218-                plications, Vol. 21, No. 3, pp. 141-150.
E-041-004-) We deeply appreciate the National Bu-              Gilchrist, A., and Mills, N. J., 1994b, “Modelling of
reau of Standard Inspection in Tainan for providing                the Impact Response of Motorcycle Helmets,” In-
the equipment for the shock absorption performance                 ternational Journal of Impact Engineering, Vol.
tests.                                                             15, No. 3, pp. 201-217.
                                                               Gilchrist, A., Mills, N. J., and Khan, T., 1988, “Sur-
                 NOMENCLATURE                                      vey of Head, Helmet and Headform Sizes Related
                                                                   to Motorcycle Helmet Design,” Ergonomics, Vol.
a(t) acceleration of the headform                                  31, No, 10, pp. 1395-1412.
t 1, t 2 bounds of all possible time intervals defining        Gopalakrishna, G., Peek-Asa, C., and Kraus, J. F.,
         the total duration of impact                              1998, “Epidemiologic Features of Facial Injuries
                                                                   among Motorcyclists,” Annals of Emergency
                   REFERENCES                                      Medicine, Vol. 32, No. 4, pp. 425-430.
                                                               Huang, S. C., 1999, “Numerical Simulation of Hu-
BS 6658, 1985, “Protective Helmets for Vehicle                     man Head-Neck Dynamics,” Bio-Medical
      L. T. Chang et al.: Finite Element Analysis of the Effect of Motorcycle Helmet Materials Against Impact Velocity   843

    Materials and Engineering, Vol. 9, pp. 61-71.                Matzsch, T., and Karlsson, B., 1986, “Moped and
Hurt, H. H. Jr., Ouellet, J. V., and Thom, D. R.,                   Motorcycle Accidents-Similarities and Dis-
    1981, “Motorcycle Accident Cause Factors and                    crepancies,” Journal Trauma, Vol. 26, No. 6, pp.
    Identification of Countermeasures,” Report DOT                  538-543.
    HS-805 862, U.S. Department Transportation.                  Mills, N. J., and Gilchrist, A., 1991, “The Effective-
Huston, R. L., and Sears, J., 1981, “Effect of Protec-              ness of Foams in Bicycle and Motorcycle Hel-
    tive Helmet Mass on Head/Neck Dynamics,”                        mets,” Accident Analysis and Prevention, Vol. 23,
    Transactions of the American Society of Mechani-                No. 2-3, pp.153-163.
    cal Engineers: Journal of Biomechanical Engi-                Snell Memorial Foundation, 2000, “Standard for Pro-
    neering, Vol. 103, pp. 18-23.                                   tective Headgear,” CA, USA.
JIS T 8133, 2000, “Protective Helmets for Drivers                Sood, S., 1988, “Survey of Factors Influencing In-
    and Passengers of Motorcycles and Mopeds,”                      jury Among Riders Involved in Motorized Two-
    Japanese Industrial Standard, Tokyo, Japan.                     Wheeler Accidents in India: A Prospective Study
Kingsbury, H. B., and Rohr, P. R., 1981, “Structure                 of 302 Cases,” Journal Trauma, Vol. 28, No. 4,
    Characteristics of Motorcycle Helmets,” Paper                   pp. 530-534.
    No. 810372, Society of Automotive Engineers,                 Tsai, M. C., and Hemenway, D., 1999, “The Effect
    Inc.                                                            of the Mandatory Helmet Law in Taiwan,” Injury
Kraus, J. F., and Peek, C., 1995, “The Impact of Two                Prevention, Vol. 5, pp. 290-291.
    Related Prevention Strategies on Head Injury                 Versace, J., 1971, “A Review of the Severity Index,”
    Reduction among Nonfatally Injuried Motorcycle                  Proceedings of 15th Stapp Car Crash Conference,
    Riders, California, 1991-1993,” Journal of                      New York, USA.
    Neurotrauma, Vol. 12, pp. 873-881.                           Waterman, N. A., and Ashby, M. F., 1997, The Ma-
Lee, S. T., Lui, T. N., Chang, C. N., Wang, D. J.,                  terials Selector, Chapman and Hall, London, UK.
    Heimburger, F., and Fai, H. D., 1990, “Features              Yettran, A. L., Godfrey, N. P. M., and Chinn, B. P.,
    of Head Injury in a Developing Country-Taiwan                   1994, “Materials for Motorcycle Crash Helmets :
    (1977-1987),” Journal Trauma, Vol. 30, No. 2,                   A Finite Element Parametric Study,” Plastics and
    pp. 194-199.                                                    Rubber Processing and Applications, Vol. 22, No.
Lin, F. F., Wang, M. H., Yeh, T. H., and Tien, Y. M.,               4, pp. 215-221.
    1998, “The Impact Estimation for Approving Mo-
    torcycle with Engine Larger than 150CC to be                                  Manuscript Received: Jan. 10, 2003
    Used on Streets,” Research Report S8717, Insti-                                 Revision Received: Jul. 01, 2003
    tute of transportation, Ministry of Transportation                                  and Accepted: Aug. 28, 2003
    and Communications, ROC.