Medical Engineering & Physics 29 (2007) 729–738

Injury risk of a 6-year-old wheelchair-seated occupant in a frontal

motor vehicle impact—‘ANSI/RESNA WC-19’ sled testing analysis
DongRan Ha a,c,∗ , Gina Bertocci b,c,d
a Department of Rehabilitation Science & Technology, University of Pittsburgh, Pittsburgh, PA, United States
b Injury Risk Assessment and Prevention (iRAP) Laboratory, Department of Mechanical Engineering,

University of Louisville, Louisville, KY, United States

c Injury Risk Assessment and Prevention (iRAP) Laboratory, Department of Pediatrics,

University of Louisville, Louisville, KY, United States

d RERC on Wheelchair Transportation Safety, United States

Received 13 December 2005; received in revised form 30 August 2006; accepted 31 August 2006

Abstract

Children with disabilities are transported on a daily basis to schools and developmental facilities. When they travel, they often remain seated
in their wheelchairs in vehicles. To study injury risk of pediatric wheelchair users in motor vehicle crashes, three of the same pediatric manual
wheelchairs were sled impact tested with a seated Hybrid III 6-year-old ATD using a 20 g/48 km/h frontal crash pulse. The sled test results
were compared to kinematic limitations and injury criteria specified in the ANSI/RESNA WC-19, FMVSS 213 and FMVSS 208. All sled test
results were below the limits specified in the ANSI/RESNA WC-19 standard and FMVSS 213. All tests exceeded the Nij limit of 1 specified
in FMVSS 208, and one test exceeded the limit of peak neck tension force. Chest deflection resulting from one of three tests was at the limit
specified in FMVSS 208. Our results suggest that children with disabilities who remain seated in their wheelchairs in vehicles may be at risk
of neck injury in a frontal impact motor vehicle crash. However, limitations in the biofidelity of the Hybrid III ATD neck raise concern as to
the translatability of these findings to the real world. Additional studies are needed to investigate the influence of neck properties and ATD
neck biofidelity on injury risk of children who travel seated in their wheelchairs.
© 2006 IPEM. Published by Elsevier Ltd. All rights reserved.

Keywords: Pediatric wheelchair; Six-year-old Hybrid III ATD; Injury risk; Wheelchair transportation safety; Children; Frontal impact; Wheelchair transportation

1. Background
Injuries related to motor vehicle crashes (MVCs) are the
leading cause of death for children over the age of one in the
United States [1]. Therefore, to protect children from injuries
and death in MVCs, extensive automotive safety research has
been conducted, and federal and state laws related to child
protection in MVCs have been established. Federal motor
vehicle safety standard (FMVSS) 213 regulates the child
restraint systems designed for children weighing 50 lbs or
less [2]. And recently, there has been an increase in concerns
∗ Corresponding author at: University of Pittsburgh, Department of Reha-
bilitation Science & Technology, Forbes Tower, Suite 5044, Pittsburgh, PA
15260, United States. Tel.: +1 412 656 1177.
E-mail address: dongranha@hotmail.com (D. Ha).
related to children who have outgrown booster seats but who
have not yet reached adult stature. The result has been a pro-
posal to extend FMVSS 213 regulation to children weighing
more than 50 lbs [3]. In June of 2003, National Highway Traf-
fic Safety Administration (NHTSA) extended the FMVSS
213 to cover children weighing 50–65 lbs [4]. FMVSS 208
also addresses children, outlining the air bag deployment
requirements in order to minimize the risk of injury resulting
from deployment of an air bag [5]. This regulation specifies
injury criteria for different sized dummies, including child
dummies (the 12-month-old CRABI dummy, the 3-year-old
child dummy, and the 6-year-old child dummy).
Children with disabilities often cannot be seated in stan-
dard booster seats or automobile seats because of physical
deformities or poor trunk and head control; they may differ
anatomically from children who do not have disabilities or

1350-4533/$ – see front matter © 2006 IPEM. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.medengphy.2006.08.011
730 D. Ha, G. Bertocci / Medical Engineering & Physics 29 (2007) 729–738
may lack sufficient balance while sitting due to trunk or head
instability. The results of a survey study on transportation of
children with disabilities conducted by Everly et al. indicated
that a large percentage of children (44%) transported daily
have poor head and trunk control and are therefore unable to
sit upright without support [6]. Therefore, children with dis-
abilities who must travel seated in their wheelchairs are often
excluded from the protections dictated by the FMVSS 213,
as well as by other laws relating to child protection in MVCs.
Federal laws such as the Americans with Disabilities
Act (ADA) and the Individuals with Disabilities Education
Act (IDEA), which prohibit discrimination against adults
and children with disabilities, have increased the number
of disabled travelers [7,8]. Children with disabilities are
transported daily to schools and developmental educational
facilities. The survey study conducted by Everly et al. shows
that a majority of children using transportation services are
school aged children, 6–17 years old [6]. When children are
transported, they often remain seated in their wheelchairs in
vehicles, such as school buses and family vans.
Not only children with disabilities, but also the overall
number of disabled travelers seated in wheelchairs in public
or private transportation has increased since the passage of
the ADA. In order to improve the safety of wheelchair-seated
travelers and other vehicle occupants, voluntary standards
have been established by national and international organi-
zations [9–12]. The Society of Automotive Engineers (SAE)
J2249 Wheelchair Tiedowns and Occupant Restraint Sys-
tems (WTORS) [9] and International Standards Organization
(ISO) 10542 Wheelchair Tiedown and Occupant Restraint
Systems [10] specify design requirements, test methods,
and performance requirements for WTORS. The American
National Standards Institute (ANSI)/Rehabilitation Engi-
neering and Assistive Technology Society of North America
(RESNA) WC-19 Wheelchairs for Use in Motor Vehicles
[11] and ISO 7176-19 Wheelchairs Used as Seats in Motor
Vehicles [12] contain design and performance requirements
as well as test procedures for wheelchairs used as forward-
facing seats in motor vehicles. The ANSI/RESNA WC-19
standard applies to transit wheelchairs designed for adults
and children with a body mass of at least 22 kg. Children are
not currently addressed in the ISO 7176-19 standard. Both
standards, ANSI/RESNA WC-19 and ISO 7176-19, require
that a wheelchair be provided with two front and two rear
securement points for attachment to a four-point, strap-type
tiedown system. They also require that the wheelchair, includ-
ing the wheelchair frame and seating system, be sled-impact
tested using a 20 g/48 km/h frontal crash pulse.
Research conducted to-date on wheelchair transportation
safety has focused largely on adult wheelchair users. There-
fore, studies on pediatric transit wheelchairs and the injury
risk associated with pediatric wheelchair users in a crash are
needed. In this study, the injury risk of a child seated in a
manual pediatric wheelchair in a frontal impact motor vehi-
cle crash was assessed by using frontal impact sled testing. A
Hybrid III 6-year-old (6HybridIII) ATD was used in the study
to investigate the injury risk of a 6-year-old wheelchair-seated
occupant in a frontal motor vehicle impact. A 6HybridIII
ATD has more advanced instrumentation capabilities than its
predecessor, the Hybrid II 6-year-old (6HybridII) ATD. A
6HybridII ATD can be equipped to measure head accelera-
tion, chest acceleration, pelvis acceleration, and femur forces.
With a 6HybridIII ATD, chest deflection and neck forces and
moments can also be measured in addition to the instrumenta-
tion available with the 6HybridII ATD. The 6HybridIII ATD
was developed by the SAE Hybrid III Dummy Family Task
Force, based on the information available on anthropometry
and mass distribution characteristics of 6-year-old children
in the United States [13–18]. Because “there [was] virtually
no literature dealing with biomechanical impact response of
children,” [19] biomechanical impact response requirements
for head, neck, chest, and knees of 6-year-old ATD [20] were
scaled from biomechanical response corridors of the mid-size
adult male, which were constructed from test data of human
cadavers and volunteers [18,21–25].
2. Methods
Among pediatric manual wheelchairs available on the
market, one with the transit option (transit wheelchair) was
chosen for this study. A transit wheelchair is defined as
a wheelchair that has been tested in accordance with the
ANSI/RESNA WC-19 and provides four tiedown attachment
points [11]. Using a transit wheelchair allows assessment
of injury risk for manual pediatric wheelchair occupants
in a frontal impact motor vehicle crash independent from
injury due to the failure of the wheelchair. That is, the transit
wheelchair has been shown to be crashworthy. The Sunrise
Medical Zippie (Longmont, CO), one of the most commonly
used transit pediatric manual wheelchairs, with transit-tested
standard conventional seating, consisting of a padded solid
seat and solid back, was used in the study.
Fig. 1. Sled test setup.
 D. Ha, G. Bertocci / Medical Engineering & Physics 29 (2007) 729–738
Fig. 2. Sled deceleration pulses with ANSI/RESNA WC-19 corridor.
Three Zippie wheelchairs having identical configuration
(frame width and depth, caster size, rear wheel size, and
seating systems) were sled tested with a seated Hybrid III
6-year-old anthropomorphic test device (ATD). The pediatric
wheelchair was placed on the sled platform, and the instru-
mented 6HybridIII was seated in the wheelchair (Fig. 1).
The wheelchair was secured to the sled platform using a
surrogate four-point, strap-type tiedown, and the 6HybridIII
was restrained with a surrogate, vehicle-anchored, three-
point belt in accordance with the WC-19 standard [11]. The
shoulder belt was taped to the dummy’s shirt at the upper
chest area to keep the shoulder belt positioned during the
initial sled acceleration.1 Sled tests were then conducted
under 48 km/h and 20 g average impact conditions [11]. The
deceleration pulses used for the three sled tests along with
the pulse requirements stated in ANSI/RESNA WC-19 are
shown in Fig. 2.
The following variables were collected from the instru-
mented ATD during each sled test: head acceleration, chest
acceleration, chest compression, and forces and moments at
the dummy’s upper neck. Data were collected at the rate of
10,000 s−1 . Moreover, the entire impact event was recorded
using high-speed (1000 frames/s) motion cameras positioned
at the side of the sled track to measure the wheelchair and
the ATD excursions after the test. All signals were filtered
following the requirements of SAE J211-2, instrumentation
for impact testing [26].
ANSI/RESNA WC-19 has a number of criteria that must
be met to pass the frontal impact test. One criterion requires
the ATD to be kept in an upright seated posture in the
wheelchair at the end of the test, as defined by “the ATD
torso being oriented at not more than 45◦ to the vertical when
1 A piece of tape was needed because the sled test method specified in
the ANSI/RESNA WC-19 standard requires 75 mm of slack in the surrogate
shoulder belt to simulate pay-out of a retractor. Due to the slack, the shoulder
belt was likely to slide off the ATD’s shoulder. Therefore, tape was used to
keep the shoulder belt in place during initial sled deceleration.
731
viewed from any direction” [11]. Posture of the ATD was
examined at the end of each test.
Injury criteria and kinematic limits specified in the
ANSI/RESNA WC-19 [11] and automotive regulations,
FMVSS 213 [2] and FMVSS 208 [5], were applied to col-
lected sled test data to determine the injury risk of a pediatric
occupant in a wheelchair. The injury criteria and kinematic
limits used for comparison with sled test data are shown in
Table 1. FMVSS 213 injury criteria include the head injury
criterion (HIC)2 and a maximum resultant acceleration of
the upper thorax sustained for three consecutive milliseconds
(3 ms clipped peak). Injury criteria for the 6HybridIII speci-
fied in FMVSS 208 include HIC15 , chest acceleration (same
as the FMVSS 213 criterion), chest compression deflection,
Nij neck injury criterion, peak neck tension force, and peak
neck compression force. HIC was calculated using Eq. (1).
Nij was calculated using Eq. (2).Eq. (1), calculation of HIC:
 2.5
 2
1
HIC = ar dt (t2 − t1 ) (1)
t2 − t 1 1
HICunlimited is the any two moments, t1 and t2 , during the
impact [2], HIC36 the two moments, t1 and t2 , separated by
not more than 36 ms [4], and HIC15 is the two moments, t1 and
t2 , separated by not more than 15 ms [5].Eq. (2), calculation
of Nij :
Fz Mocy
Nij = + (2)
Fzc Myc
Fz is the axial force, Mocy the occipital condyle bending
moment, Fzc the 2800 N when Fz is in tension, Fzc the 2800 N
when Fz is in compression, Myc the 93 Nm when a flexion
moment exists at the occipital condyle, Myc is the 37 Nm
when an extension moment exists at the occipital condyle
[5].
2 On June 2003, HICumlimited was replaced by HIC36 in FMVSS 213 [4].
732 D. Ha, G. Bertocci / Medical Engineering & Physics 29 (2007) 729–738

Table 1 head contacted its knee between 100 and 120 ms. Fig. 4
ANSI/RESNA WC-19, FMVSS 213, and FMVSS 208 injury criteria and shows the post-test pictures of the wheelchair and the ATD.
kinematic limits of the 6-year-old ATD and a wheelchair
In all three sled tests, the ATD was kept in a seated posture
Limit in the wheelchair as required by the ANSI/RESNA WC-19
ANSI/RESNA WC-19 [11] standard. The ATD torso was kept within 45◦ to the ver-
Xwc a (mm) 150 tical when viewed from any direction. Table 2 shows the
Xknee b (mm) 300
XheadF c (mm) 450
sled test results compared to ANSI/RESNA WC-19 hori-
XheadR d (mm) −350 zontal excursion limits and other criteria. The maximum
Xknee /Xwc e ≥1.1 horizontal excursions of the wheelchair and the ATD for
(Hpre − Hpost )/Hpre f <0.2 all three tests were within the WC-19 limits. All tests com-
FMVSS 213 [2,4] plied with the limit, Xknee /Xwc ≥ 1.1, which assures that the
HICunlimited 1000 wheelchair did not load the ATD. And none of the three tests
HIC36 1000 exceeded the (Hpre − Hpost )/Hpre limit of 0.2, which evaluates
Chest acc (g) 60
the integrity of seat surface and seat attachment hard-
FMVSS 208 [5] ware with the intent to prevent the occurrence of occupant
HIC15 70 submarining.
Chest defl. (mm) 40
Nij 1
Table 3 shows the sled test results compared to FMVSS
Neck tension (N) 1490 213 and FMVSS 208 injury criteria. Fig. 5 shows the time his-
Neck comp. (N) 1820 tories of the head resultant acceleration of the three sled tests
a The horizontal distance relative to the sled platform between the contrast which were used in the calculation of the HIC values. During
target placed at or near point P on the test wheelchair at time t0 , to the point P each sled test, the ATD’s head contacted its knee between 100
target at the time of peak wheelchair excursion (point P = a wheelchair seat and 120 ms (Fig. 5). High peaks shown in the time histories
reference point located on the wheelchair reference plane approximately of ATD head acceleration are the result of ATD head-knee
50 mm above and 50 mm forward of the projected sideview intersection of
the undepressed backrest and undepressed seat cushion).
contact. FMVSS 213 injury criteria assessed in this study
b The horizontal distance relative to the sled platform between the dummy included HICunlimited , HIC36 , and peak chest acceleration. In
knee-joint target at time t0 , to the knee joint target at the time of peak knee all tests, the HICunlimited and HIC36 values remained below
excursion [11]. the limit of 1000. And no tests exceeded the maximum chest
c The horizontal distance relative to the sled platform between the most
acceleration limit of 60 g. FMVSS 208 specifies limits for
forward point on the dummy’s head above the nose at time t0 , to the most for-
ward point on the dummy’s head at the time of peak forward head excursion
HIC15 , chest deflection, neck tension, neck compression, and
[11]. Nij , neck injury criteria. All the HIC15 values resulting from
d The horizontal distance relative to the sled platform between the most the sled tests remained below the 700 limit. Chest deflection
rearward point on the dummy’s head at time t0 , to the most rearward point resulting from sled test No. 3 was at the limit specified in the
on the dummy’s head at the time of peak rearward head excursion [11]. regulation, which was 40 mm. Sled test No. 2 chest deflection,
e The wheelchair shall not impose forward loads on the ATD, which is

considered to be achieved if the peak ATD knee excursion exceeds the peak
wheelchair point P excursion by 10% [11].
f The post-test height of the average of left and right ATD H-points relative
to the wheelchair ground plane shall not decrease by more than 20% from
the pretest height (H-point = a point located on the left and right sides of
the buttock/pelvis region of a weighted manikin or ATD that represents
the approximate human hip joint location relative to the back and bottom
surfaces of the pelvic flesh) [11].
3. Results
The time-lapse images of sled test No. 3 at 40 ms inter-
vals are shown in Fig. 3. During each sled test, the ATD’s
39.3 mm, was also close to the 40 mm limit. Fig. 6 shows the
Nij values for each sled test as compared to the limits specified
in FMVSS 208. All tests exceeded the Nij limit of 1 at the ten-
sion extension limit. The peak Nij was observed at 52.8 ms in
sled test No. 1, 48.3 ms in test No. 2, and 47.4 ms in test No. 3
(see Fig. 7). Sled test videos showed that until the time of the
peak Nij , the ATD’s upper torso was not fully restrained due
to 75 mm shoulder belt slack (see footnote 1). It appears that
at the time of the peak Nij , the shoulder belt fully engaged
and restrained the dummy’s torso. The peak neck tension
force of test No. 1, 1582 N, exceeded the limit of 1490 N at
68.1 ms. Sled test No. 2 neck tension force, 1435 N, was also

Fig. 3. Images of sled test No. 3 at 40 ms intervals.

 D. Ha, G. Bertocci / Medical Engineering & Physics 29 (2007) 729–738 733

Fig. 4. Post-sled test pictures.

Table 2
Comparison between sled test results and ANSI/RESNA WC-19 injury criteria and kinematic limits
ANSI/RESNA WC-19

Xwc (mm) Xknee (mm) XheadF (mm) XheadR (mm) Xknee /Xwc (Hpre − Hpost )/Hpre
Limit 150 300 450 −350 ≥1.1 <0.2
Test 1 17 56 283 −136 3.3 −0.05 (5% increase)
Test 2 13 57 230 −139 4.4 −0.03 (3% increase)
Test 3 13 65 252 −132 5.0 −0.03 (3% increase)

Table 3
Comparison between sled test results and FMVSS 213 and FMVSS 208 injury criteria
FMVSS 213 FMVSS 208

HICunlimited HIC36 Chest acc (g) HIC15 Chest defl. (mm) Nij Neck ten. (N) Neck comp. (N)
Limit 1000 1000 60 700 40 1 1490 1820
Test 1 520 344 52.0 208 39.3 1.4a 1582 1128
(41–109 ms) (55–91 ms) (62–77 ms)
Test 2 406 276 49.5 155 37.0 1.4a 1435 669
(35–109 ms) (40–76 ms) (62–77 ms)
Test 3 385 275 48.0 158 40.0 1.3a 1231 241
(34–95 ms) (46–82 ms) (64–79 ms)
Note: Value exceeded the FMVSS limit was given in italics.
a Tension extension.

Fig. 5. Head resultant acceleration.

734 D. Ha, G. Bertocci / Medical Engineering & Physics 29 (2007) 729–738

Fig. 6. Nij —sled test results.

close to the 1490 N limit. No tests exceeded the independent
compressive neck force limits.
4. Discussion
To assess the injury risk of manual pediatric wheelchair
occupants in a frontal impact motor vehicle crash, three pedi-
atric manual wheelchairs, Zippies, were sled tested with a
seated 6HybridIII ATD in accordance with the ANSI/RESNA
WC-19 standard. The sled test results were then com-
pared to kinematic limitations and criteria specified in the
ANSI/RESNA WC-19, FMVSS 213 and, FMVSS 208.
In May 2002, NHTSA proposed to replace the 6HybridII
ATD with 6HybridIII ATD in FMVSS 213 to improve the
evaluation of child restraint system performance [3]. After
the notice of proposed rulemaking (NPRM) was released [3],
NHTSA received comments that expressed concerns about
the biofidelity of the 6HybridIII ATD’s neck and upper chest
areas. Several commenters stated that the 6HybridIII ATD
had a more flexible neck and ribs than the 6HybridII ATD
[4]. However, NHTSA stated in the Final Rule that “the neck
of the HIII 6-year-old is currently performing within the spec-
ifications established by the Hybrid III Dummy Family Task
Force of the Society of Automotive Engineers (SAE),” and
the agency believed “the current neck on the HIII 6-year-
old dummy [provided] improved biofidelity over the current
dummy [Hybrid II 6-year-old dummy]” [4]. As a result, the
6HybridIII ATD was adopted into FMVSS 213 in the Final
Rule in June 2003.
During the FMVSS 213 revision in 2002, adopting the
scaled injury limits of FMVSS 208, such as chest deflection
and Nij , into FMVSS 213 was also proposed [3]. However,
because concerns about the biofidelity of the Hybrid III
dummies had been raised, NHTSA decided not to include
the scaled injury limits into FMVSS 213 at that time [4].
Although a 6HybridIII ATD provides improved biofidelity
over that of a 6HybridII ATD, the biofidelity of the 6HybridIII
ATD has not been confirmed by biomechanical impact
response of child data. Rhule and Hagedorn conducted an
accident reconstruction study to verify the injury criteria
performance limits of child dummies in FMVSS 208 [27].
Reconstruction of incidents was chosen as an alternative
approach to child cadaver testing, since “child cadaver avail-
ability [was] unknown, potentially very politically sensitive
and of undetermined timeliness” [28]. Three accident cases
were reconstructed in the study; one case involved a 7-year-
old female child and two cases involved 5-month-old infants.

Fig. 7. Sled test pictures at maximum Nij .

 D. Ha, G. Bertocci / Medical Engineering & Physics 29 (2007) 729–738
The study focused on head and neck injuries to a child occu-
pant. Three reconstruction tests were conducted for each case,
and a 6HybridIII ATD was used in the tests representing the
accident involving a 7-year-old child. The vehicle carrying
the 7-year-old child was in an intersection crash, and the child
was injured from an air bag deployment. Based on the test
results, the authors concluded that “the magnitude of [the
loads measured from the neck load cell of a 6HybridIII ATD]
is consistent with the neck injuries received by the 7-year-
old victim in this case who suffered transection of the spinal
cord in the neck” [27]. The Nij values in the tests, which
evaluate the potential for neck injury, were also well above
the FMVSS 208 performance limit of 1. HIC values, which
measure risk of head injury, were below the limit of 700 in all
three tests. Authors stated that “[the results were] consistent
with the minor head injuries of the case occupant, with the
exception of the fractured left mandible. However, HIC is not
expected to be a good predictor of mandibular injuries” [27].
Biofidelity of the 6HybridIII ATD was also evaluated in
the study conducted by Sherwood et al. [29]. About 49 km/h
sled tests were conducted using a 6HybridIII ATD, and sled
test results were compared to a 12-year-old cadaver test
conducted at the University of Heidelberg (Heidelberg, Ger-
many) [30]. The study focused on cervical spine injury risk
for a 6-year-old child in frontal crashes. The authors con-
cluded in the study that “the thoracic spine of the Hybrid III
6-year-old dummy is not biofidelic in restrained frontal crash
tests” [29]. The authors also stated that “the stiff thoracic
spine of the dummy results in high neck forces and moments
that are not representative of the true injury potential” [29].
A limitation of this study was that the size of the 12-year-old
cadaver was different from that of the 6HybridIII ATD. The
authors indicated in the study that “ideally comparisons to the
dummy tests would be done with a cadaver of the same age
and size, but this was not possible due to the small number
of child cadaver tests available for comparison” [29].
The 6HybridIII was originally developed using the
biomechanical impact response requirements derived from
adult data. Therefore, verification of the performance of
a 6HybridIII ATD with the actual biomechanical impact
response data of 6-year-old children is needed through stud-
ies such as those conducted by Hagedorn and Rhule [27] and
Sherwood et al. [29], and if needed, design changes should
be made to the 6HybridIII ATD to improve the biofidelity of
the ATD.
The results presented in this study showed that a 6-year-
old wheelchair seated occupant may be at risk of neck injury
during a frontal car crash. All tests in this study exceeded
the Nij injury criteria of 1, and sled test No. 1 exceeded the
peak neck tension force limit of 1490 N. Injury risk curves,
representing probability of risk of injury at various injury
measures, such as Nij and chest deflection, are presented in
Proposed Amendment to FMVSS No 213 Frontal Test Proce-
dure released by NHTSA [31]. The probability that a vehicle
occupant would receive a certain level of abbreviated injury
scale (AIS) injury [32] can be calculated using the injury risk
735
curves. For a 6-year-old ATD, the Nij limit of 1 is equivalent
to a 22% risk of AIS ≥ 3. “Blood in the synovial fluid of the
occipital condylar joint capsules was rated as an AIS = 3 neck
injury and was defined as the threshold of undesirable neck
trauma” [33]. It should be noted, however, that the results pre-
sented in this study on neck injury risk (Nij and neck tension)
need to be interpreted with caution since there are concerns
regarding the biofidelity of the 6HybridIII ATD’s neck area
as presented in the study conducted by Sherwood et al. [29].
In one of the comments that NHTSA received (in response
to the NPRM), it was stated that due to the flexible neck of
6HybridIII ATD, head-to-chest or head-to-knee contact was
observed during sled tests and resulted in “unrealistic and
unacceptably high HIC” [4]. NHTSA’s response to the com-
ment was that “none of the sled testing conducted with the
HIII 6-year-old dummy [initiated or conducted by the agency
had] indicated that head-to-chest or head-to-knee impacts
[were] an issue. Such impacts [were] not typical” [4]. The
agency indicated that during one test, the shoulder portion of
the occupant belt slipped off the ATD’s shoulder and resulted
in head-to-knee contact. Therefore, “NHTSA believes that if
head-to-knee contact occurs, there are likely design concerns
with respect to the particular child restraint that should be
addressed to eliminate such contact” [4]. NHTSA also stated
that if head-to-knee contact occurred during the sled test and
resulted in a spike in head acceleration, then the spike needs
to be included in the HIC computation.
Head-to-knee contact occurred in all three sled tests of a
pediatric wheelchair with the 6HybridIII ATD in this study
(see Fig. 3). As recommended by NHTSA [4], the spike in
head acceleration (Fig. 5) was included in the calculation of
the HIC value for each test in this study. All calculated HIC
values remained below the limits specified in the standards.
Similar to one of the sled tests conducted by NHTSA [4], the
shoulder belt slipped off the ATD’s shoulder during all three
sled tests (see Fig. 8). The occupant restraint system in all sled
tests was setup in accordance with WC-19 recommendations.
In the frontal impact test method section of the standard, it
is stated to “bolt the upper anchorage of the surrogate shoul-
der belt assembly to the rigid support structure at a location
that provides a good fit of the shoulder belt to the ATD’s
chest and shoulder as illustrated in FigureA.3” [11]. How-
ever, the “FigureA.3” provided in the WC-19 standard is for
the midsize-male ATD, which may not be the optimal position
for an ATD other than the midsize-male. A study conducted
by Bertocci et al. showed that shoulder belt upper anchorage
location does have a significant effect on the wheelchair occu-
pant kinematics and wheelchair occupant crash protection
[34]. In the study, the upper anchor point location of the shoul-
der belt was varied in three directions (x: forward–rearward
in sagittal plane, y: inboard–outboard in frontal plane, and z:
upward–downward), and the occupant kinematics and injury
criteria were determined using a computer model repre-
senting a 50th percentile male Hybrid III ATD seated in a
wheelchair. The study concluded that “variation in the anchor
point of the shoulder belt led to notable changes in occupant
736 D. Ha, G. Bertocci / Medical Engineering & Physics 29 (2007) 729–738
Fig. 8. Shoulder belt position—post-sled test.
restraint effectiveness, and hence, occupant dynamics” [34].
The study states that “although variations in physical size of
the occupant were not explored as a part of this study, a fixed
shoulder belt anchor point will lead to variations in occupant
belt fit with different-sized occupants. These variations in
belt fit will produce different, and in some cases undesirable,
levels of crash protection” [34]. The test location of shoulder
belt upper anchor point specified in WC-19 may not be the
optimal position for the 6HybridIII ATD, and the poor fit of
the shoulder belt during the sled tests in this study is likely
due to the improper position of the shoulder belt upper anchor
point. The preferred and optional zones for the shoulder belt
upper vehicle anchor point for different sizes of occupants
are provided in the Society of Automotive Engineers (SAE)
Recommended Practice (RP) J2249, Tiedowns and Occupant
Restraint Systems [9]. These zones are specified relative to
the top of a wheelchair occupant’s shoulder. Locations of the
zones are determined based on “Shoulder Height”, “Shoulder
Breath”, “Neck Breath”, and “Seat Height” of the wheelchair
occupant. Use of the preferred zones for 6-year-old occupant
size specified in the SAE RP J2249 in sled testing might have
prevented the shoulder belt from slipping off the 6HybridIII
ATD’s shoulder during sled testing. It is recommended that
WC-19 be modified to include shoulder belt upper anchorage
points for various sized occupants.
Shoulder belt slack required in WC-19 could also have
caused the shoulder belt slippage during the sled tests. WC-
19 requires 75 mm of shoulder belt slack (simulating the belt
pay out with a retractor) in the frontal impact sled test [11].
Compared to the adult ATDs, the 6HybridIII ATD has very
narrow shoulders. Although the requirement of shoulder belt
slack works well with the adult ATDs in sled tests, it might
not be an appropriate requirement for pediatric ATDs. Further
investigation of the occupant restraint system setup, including
shoulder belt anchor point location and shoulder belt slack,
for the pediatric population is needed in the future.
Chest deflection resulting from sled test No. 3 was at the
limit specified in FMVSS 208, 40 mm. Sled test No. 1 and
No. 2 chest deflections, 39.3 and 37 mm, also approached
the limit. As stated previously, the occupant shoulder belt
slipped off the ATD’s shoulder during all three sled tests.
The slippage of the shoulder belt displaced the belt away
from the chest deflection potentiometer as shown in Fig. 8.
Chest deflections would likely have been higher if the shoul-
der belt did not slip off the shoulder. The results showed that
a 6-year-old wheelchair seated occupant may have a risk of
chest injury in a frontal impact motor vehicle crash. “Tho-
racic injuries associated with peak sternal deflection are rib
and sternal fractures, which are rated as AIS ≥ 2, and tho-
racic organ damage produced by crushing forces, which is
rated as AIS ≥ 4” [33]. The chest deflection limit of 40 mm is
equivalent to a 72% risk of AIS ≥ 2 and 11% risk of AIS ≥ 4
[31]. Again, biofidelity of a 6HybridIII ATD has not been
confirmed by biomechanical impact response of child data,
and therefore, the results on a risk of chest injury need to be
interpreted with caution.
To study injury risk associated with manual pediatric
wheelchair occupants in a frontal impact motor vehicle crash,
a 6HybridIII ATD was used. Currently, a 6HybridIII ATD,
which represents the lower range of school aged children,
is the largest child dummy available for the FMVSS com-
pliance tests. In the automotive industry, because there has
been an increase in concerns related to children who have
outgrown booster seats but not yet reached adult stature, the
Society of Automotive Engineers (SAE) began development
of Hybrid III 10-year-old ATD in 2000 [35]. “The 10-year-
old was chosen because it is the transitional size at which
[standard vehicle] belt fit and seat design may be adequate
and a booster/safety seat may no longer be necessary” [36].
To assess injury risks for children who travel seated in their
wheelchairs in vehicles, testing of pediatric wheelchairs with
a Hybrid III 10-year-old ATD is also needed in future studies.
Several limitations are associated with this study. First,
a Hybrid III 6-year-old ATD (6HybridIII ATD) was used
in this study, and occupant injury risk was assessed from
the 6HybridIII ATD. The 6HybridIII ATD was originally
developed using data derived through the scaling procedures
from adult data, and biofidelity of the 6HybridIII ATD has
not been confirmed by biomechanical impact response of
child data. There are concerns regarding the biofidelity of the
6HybridIII ATD, and therefore, the results presented in this
study need to be interpreted in caution. Second, a 6HybridIII
ATD represents an average 6-year-old child (23.4 kg) with-
out disabilities. Children having anthropometric and inertial
characteristics differing from that of the 6HybridIII ATD
would likely lead to outcomes that vary from those reported
in this study. Third, the FMVSS injury criteria used in this
study were developed to assess the injury risk of non-disabled
children, who have normal muscle tone and balance, in motor
vehicle crashes. Because children with disabilities often have
 D. Ha, G. Bertocci / Medical Engineering & Physics 29 (2007) 729–738
less trunk or head stability than that of an average 6-year-
old child without disabilities, children with disabilities seated
in wheelchairs may be more susceptible to severe and fatal
injuries in circumstances that would not be injurious to chil-
dren without disabilities. Therefore, the injury risks presented
in this study may be underestimated when considering dis-
abled children. Fourth, sled tests conducted in this study
represent a 6-year-old child in a wheelchair restrained with a
three-point occupant restraint system (shoulder and lap belts)
and the wheelchair secured to the sled platform (representing
vehicle floor) using a four-point, strap-type tiedown system.
The occupant restraint system and wheelchair tiedown sys-
tem were setup in accordance with WC-19 recommendations.
In an actual transport situation, occupant restraint belts and
wheelchair tiedown belts are not always used in accordance
with WC-19. Often times, they are misused or not used in
vehicles. Therefore, the sled test setup used in this study may
not reflect all actual transport situations. Lastly, wheelchairs
used in this study represent one type of manual pediatric
wheelchair available on the market. Inertial and geomet-
ric characteristics can vary greatly across different types of
wheelchairs, especially power wheelchairs, and can have an
effect on wheelchair and occupant response to impact. There-
fore, the results presented in this study do not necessarily
represent all the pediatric transit wheelchairs available in the
market.
5. Conclusions
When children with disabilities are transported to schools
and developmental facilities, they often remain seated in their
wheelchairs in vehicles, such as school buses and family
vans. Children with disabilities who travel seated in their
wheelchairs are often excluded from the protections dic-
tated by the federal and state laws related to child protection
in MVCs. To study the injury risks of manual pediatric
wheelchair occupants in a frontal impact motor vehicle crash,
three pediatric manual wheelchairs were sled tested with a
seated 6HybridIII ATD in accordance with the ANSI/RESNA
WC-19 standard.
During three sled tests, the shoulder belt slipped off the
ATD’s shoulder, but the ATD was kept in a seated posture
in the wheelchair as required by the ANSI/RESNA WC-19
standard. Injury criteria and kinematic limits specified in the
ANSI/RESNA WC-19 standard, FMVSS 213 and FMVSS
208 were compared to collected sled test data. All three tests
complied with the criteria specified in the ANSI/RESNA
WC-19 and FMVSS 213. Compared to injury criteria speci-
fied in FMVSS 208, all tests exceeded the Nij injury criteria
limit of 1 at the tension extension limit, and the peak neck
tension force of sled test No. 1 exceeded the limit of 1490 N.
Chest deflection resulting from sled test No. 3 was at the
40 mm limit specified in the regulation.
Study results presented in this paper suggest that chil-
dren with disabilities who remain seated in their wheelchairs
737
in vehicles may be at risk of injury, especially to the neck
and chest areas, in a frontal impact motor vehicle crash. It
should be noted, however, that the results need to be inter-
preted with caution since biofidelity of a 6HybridIII ATD
has not been confirmed by biomechanical impact response of
child data and specifically there are concerns regarding the
biofidelity of the 6HybridIII ATD’s neck. During all three
sled tests, the Nij (neck injury criterion) limit exceeded the
tension extension limit during the early phase of an impact.
Sled test videos showed that up to the time of the peak
Nij , the ATD’s upper torso was not fully restrained due to
shoulder belt slack (required by WC-19 test setup). And, it
appears that the shoulder belt fully engaged and restrained
the dummy’s torso at the time of the peak Nij . The risk of
neck injury may be reduced if the shoulder belt engages and
restrains the ATD’s upper torso from time of initial impact
and if the ATD’s upper torso is better controlled in early
phase of an impact. Investigation of the occupant restraint
system setup, including shoulder belt slack and shoulder belt
anchor point location, for the pediatric population is needed in
the future.
Chest deflection values resulting from all three sled tests
approached the limit of 40 mm. For the pediatric population,
three-point occupant restraint systems, which were used in
this study, may not be the best occupant restraint system to
provide protection during frontal crashes. Occupant restraint
systems that allow the crash force to be distributed over
larger contact areas, such as a four-point harness, can possibly
reduce the risk of chest injury. Studies of occupant restraint
systems for children seated in wheelchairs are also needed in
the future.
To date, no study has been published that evaluates injury
risk of pediatric wheelchair users in motor vehicle crashes.
This preliminary study using frontal impact sled testing pre-
dicted that 6-year-old children seated in a manual wheelchair
may be at increased risk of neck and chest injuries in a frontal
impact motor vehicle crash. Additional studies are needed to
investigate the influence of neck properties and ATD neck
biofidelity on injury risk of children who travel seated in
their wheelchairs. Furthermore, studies on injury risk of pedi-
atric wheelchair users in motor vehicle crashes are needed
to promote the same level of safety and protection offered
to children seated in OEM vehicle seats to children with
disabilities seated in wheelchairs in vehicles.
Acknowledgements
This work was funded by the National Institute of Dis-
ability and Rehabilitation Research (NIDRR), Rehabilitation
Engineering Research Center on Wheelchair Transportation
Safety, grant # H133E010302. The authors thank Miriam
Manary, MS and Lawrence Schneider, PhD for their exper-
tise in conducting these sled tests. The opinions expressed are
those of the authors and do not necessarily represent those of
NIDRR.
738 D. Ha, G. Bertocci / Medical Engineering & Physics 29 (2007) 729–738
References
[1] National Center for Injury Prevention and Control. Leading causes
of death reports; February 2003. Available at: http://webapp.cdc.gov/
sasweb/ncipc/leadcaus.html.
[2] National Highway Traffic Safety Administration (NHTSA). FMVSS
213 child restraint systems (October, 2000 ed. Vol. 49 CFR 571.213);
1979.
[3] National Highway Traffic Safety Administration (NHTSA). Notice
of proposed rulemaking (NPRM) on federal motor vehicle safety
standards; child restraint systems (Docket No. NHTSA-02-11707):
NHTSA; 2002.
[4] National Highway Traffic Safety Administration (NHTSA). Final rule
on federal motor vehicle safety standards; child restraint systems
(Docket No. NHTSA-03-15351): NHTSA; 2003.
[5] National Highway Traffic Safety Administration (NHTSA). FMVSS
208 occupant crash protection (October, 2000 ed. Vol. 49 CFR
571.208); 1971.
[6] Everly JS, Bull MJ, Stroup KB, Goldsmith JJ, Doll JP, Russell R. A
survey of transportation services for children with disabilities. Am J
Occup Ther 1993;47(9):804–10.
[7] Equal Employment Opportunity Commission and the U.S. Department
of Justice. Americans with Disabilities Act Handbook. Washington:
U.S Government Printing Office; 1991.
[8] U.S. Department of Justice-Civil Rights Division. A Guide to Disability
Rights Laws. Washington, DC: U.S. Department of Justice-Civil Rights
Division; 2000.
[9] Society of Automotive Engineers (SAE). SAE J2249: Wheelchair
Tiedowns and Occupant Restraint Systems for Use in Motor Vehicles:
SAE; 1997.
[10] International Standards Organization (ISO). ISO 10542: Wheelchair
Tiedown and Occupant Restraint Systems: ISO; 2001.
[11] ANSI/RESNA Subcommittee on Wheelchairs and Transportation
(SOWHAT). ANSI/RESNA WC/Vol. 1: Section 19 Wheelchairs—
Wheelchairs Used as Seats in Motor Vehicles: ANSI/RESNA; April
2000.
[12] International Standards Organization (ISO). ISO 7176-19: Wheelchairs
Used as Seats in Motor Vehicles: ISO; 2000.
[13] Weber K, Lehman RJ, Schneider LW. Child anthropometry for restraint
system design (No. UMTRI-85-23). Ann Arbor, MI: University of
Michigan; 1985.
[14] Snyder RG, Schneider LW, Owings CL, Reynolds HM, Golumb DH,
Schork MA. Anthropometry of infants, children and youths to age 18
for product safety design (No. SP-450). Warrendale, PA: Society of
Automotive Engineers; 1977.
[15] Reynolds HM, Young JW, McConville JT, Snyder RG. Development
and evaluation of masterbody forms for three-year old and six-year old
child dummies (No. DOT HS-801-811). Ann Arbor, MI: University of
Michigan; 1976.
[16] Anthropomorphic Test Dummy (No. DOT-HS-299-3-569). Final
Report, Volume 1, NHTSA; December 1973.
[17] Anthropometry of Motor Vehicle Occupants (No. DOT-HS-806-715).
Final Report, Volume 1, NHTSA; April 1985.
[18] Hubbard RP, McLeod DG. Definition and development of a crash
dummy head. SAE Paper No. 741193; 1974.
[19] National Highway Traffic Safety Administration (NHTSA). Technical
Report: development and evaluation of the Hybrid III type six-year-old
child dummy (No. Docket 98-3972); June 1998.
[20] Irwin A, Mertz HJ. Biomechanical basis for the CRABI and Hybrid III
child dummies. SAE Paper No. 973317, 261-272; 1997.
[21] Culver CC, Neathery RF, Mertz HJ. Mechanical necks with humanlike
responses. SAE Paper No. 720959; 1972.
[22] Mertz HJ, Neathery RF, Culver CC. Performance requirements and
characteristics of mechanical necks. In: General motors research
laboratories symposium, human impact response-measurement and
simulation. 1972.
[23] Neathery RF. Analysis of chest impact response data and scaled per-
formance recommendations. SAE Paper No. 74118; 1974.
[24] Horsch JD, Patrick LM. Cadaver and dummy knee impact response.
SAE Paper No. 760799; 1976.
[25] Hodgson VR, Tomas LM. Comparison of head acceleration injury
indices in cadaver skull fracture. SAE Paper No. 710854; 1971.
[26] Society of Automotive Engineers (SAE). SAE J211—instrumentation
for impact tests, SAE Handbook; 1995.
[27] Rhule DA, Hagedorn AV. Child injury tolerance through case recon-
struction. In: Paper presented at the 17th international technical
conference on the enhanced safety of vehicles. 2001.
[28] NHTSA VRTC Pedestrian and Applied Biomechanics Division. Child
Injury Tolerance Through Accident Reconstruction; 2004. Available
at: http://www-nrd.nhtsa.dot.gov/vrtc/bio/child/childinjurytol.htm.
[29] Sherwood CP, Shaw CG, Van Rooij RWK, Crandall JR, Orzechowski
KM, Eichelberger MR, et al. Prediction of cervical spine injury
risk for the 6-year-old child in frontal crashes. Traffic Injury Prev
2003;2:206–13.
[30] Kallieris D, Schmidt G, Barz J, Mattern R, Schulz F. Response and
vulnerability of the human body at different impact velocities in sim-
ulated three-point belted cadaver tests. In: Paper presented at the third
IRCOBI conference. 1978.
[31] National Highway Traffic Safety Administration (NHTSA). Proposed
Amendment to FMVSS 213 Frontal Test Procedure: NHTSA; 2002.
[32] States JD. Abbreviated and the comprehensive research injury scales.
SAE, SAE Paper No. 690810; 1969.
[33] Mertz HJ. Accidental injury—biomechanics and prevention. 2nd ed.
New York: Springer-Verlag; 2002.
[34] Bertocci GE, Digges KH, Hobson DA. Shoulder belt anchor location
influences on wheelchair occupant crash protection. J Rehab Res Dev
1996;33(3):279–89.
[35] NHTSA VRTC Pedestrian and Applied Biomechanics Division.
Child Crash Test Dummies; 2004. Available at: http://www-nrd.nhtsa.
dot.gov/vrtc/bio/child/hybIII10ysum.htm.
[36] Denton ATD Inc. Hybrid III 10-year-old child; 2004. Available at:
http://www.dentonatd.com/dentonatd/pdf/HIII10F.PDF.