Content uploaded by Kristopher Seluga
Author content
All content in this area was uploaded by Kristopher Seluga on Aug 31, 2016
Content may be subject to copyright.
Seluga 1
ANALYSIS AND PREVENTION OF CHILD EJECTIONS FROM GOLF CARS AND PERSONAL
TRANSPORT VEHICLES
Kristopher Seluga
Technology Associates
Timothy Long
Accident Research & Biomechanics
USA
Paper Number 09-0186
ABSTRACT
United States Consumer Products Safety Commission
statistics indicate there are approximately 13,000 golf
car related emergency room visits in the United
States annually. Of these, approximately 40%
involve children (i.e. age < 16) and 50% of these
involve a fall from a moving car. Evidence also
indicates that many passenger ejections occur during
left turns. Children are especially susceptible to
ejection because of their small size and reliance upon
the hip restraint for stability. While adult ejections
have been studied, the present study analyzes
mechanisms of child ejection during left turns.
Dynamic tests are presented wherein an
anthropomorphic Hybrid III 6 year old dummy in the
front passenger seat is ejected during a moderate left
turn and ejection kinematics are analyzed. An
Articulated Total Body (ATB) occupant simulation is
also presented, which compares favorably with
experimental results. Additional simulations are
presented wherein a seatbelt is found to be effective
in preventing ejection with minimal belt force
requirements. While experimental and simulated
occupant dummies do not include muscular reactions,
the potentially rapid onset of vehicle acceleration
indicates that real occupants, particularly young
children, may not have time to react before the
ejection process has begun. Results indicate that
current hip restraints are not large enough to prevent
the ejection of small children during a moderate left
turn. Additionally, seatbelts or straps are effective in
preventing ejection during driver induced
accelerations. The small belt force requirements
indicate that seatbelts designed for use in automobiles
and meeting Federal Motor Vehicle Safety Standards
(FMVSS) may not be necessary. Based on these
results, it is recommended that children be prohibited
from riding in golf cars without a seatbelt type
restraint when driven on golf courses and that
seatbelt type restraints be provided for each occupant,
especially children, when driving outside the golf
course setting.
INTRODUCTION
Research and data compiled across the country
indicate that the use of golf cars1 and Personal
Transport Vehicles (PTVs) is rapidly expanding, as
are the numbers or injuries related to their use.
Recent research conducted by the University of
Alabama at Birmingham [1] has indicated that about
1,000 Americans are injured in golf car related
accidents each month. Another study completed by
the Center for Injury Research and Policy at
Nationwide Children’s hospital in Columbus, Ohio
[2] stated that annual injury rates for golf cars
increased 130 percent over 16 years ending in 2006.
This study suggested that rules should be in place
banning children under 6 years old from riding in
golf cars. These studies and their underlying data
also indicate that passenger ejection is a dominant
mode of injury in golf car and PTV accidents,
especially when children are involved. The testing
and simulations in the present study investigate the
effectiveness and load requirements for preventing
ejection of children seated in golf cars.
In addition to golf cars operated on golf courses,
resort and retirement communities in the United
States, as well as other local municipalities, now
allow golf cars and Personal Transport Vehicles
(PTVs) on streets as primary means of local
transportation [3, 4, 5, 6]. In fact, local transportation
of passengers is the express purpose of PTVs.
Advertising for many PTVs produced by the major
manufacturers (i.e. Club Car, E-Z-Go and Yamaha)
specifically indicates that these vehicles are intended
for “playing golf or cruising your neighborhood” [7]
and “hauling kids” [8] and feature photos of young
children riding in the vehicles. In response to the
trend of using golf cars and PTVs off the golf course,
1 While the term ‘‘golf cart’’ is used by general
public, the manufacturers of those vehicles use the
term ‘‘golf car.’’
Seluga 2
the U.S. National Highway Traffic Safety
Administration implemented requirements for safety
equipment on Low Speed Vehicles (LSVs) that
operate on public roads including mandatory
seatbelts for all passengers [9, 10, 11]. However,
these regulations define a Low Speed Vehicle as one
having a top speed between 32 and 40 kph (20 and 25
mph). As a result, vehicles with top speeds below 32
kph (20 mph), such as golf cars and PTVs remain
unregulated.
Golf cars and Personal Transport Vehicles are often
substantially similar, are manufactured by the same
group of companies, and are virtually
indistinguishable to the common observer. However,
the manufacturers differentiate these vehicles based
on maximum speed and intended usage, which can
lead to confusing or ambiguous distinctions. For the
purposes of this study, it is sufficient to understand
that according to American National Standards
Institute (ANSI) standards, the term “golf car”
applies to vehicles with a top speed of less than 15
mph that are “specifically intended for and used on
golf courses for transporting golfers and their
equipment” [12] while “Personal Transport Vehicle”
(PTV) applies to vehicles with a top speed of less
than 20 mph which are “operated on designated
roadways, or within a closed community where
permitted by law or by regulatory authority rules” not
including golf cars [13].
Previous research performed by Seluga et al [14] and
Long et al [15] has demonstrated the ineffectiveness
of most hip or handhold restraint systems typically
found on existing golf cars and PTVs. In fact, it has
been demonstrated that these types of restraints can
exacerbate the problem by acting as a tripping
mechanism, increasing the likelihood that an ejected
occupant will strike the ground head first.
Additionally, the documented increase in golf car and
PTV injuries is consistent with the data presented by
Long et al [15], which indicated an increase in the
number of injuries due to increased vehicle usage and
the lack of any seatbelt requirements. This study also
demonstrated the effectiveness of seatbelts in
preventing passenger ejections. Thus, if seatbelts
were provided and users exhibited comparable
compliance rates to those for automobiles (i.e.
approximately 80% [16,17]), then approximately
80% of ejection accidents could be prevented by
providing seatbelts.
The debate concerning restraint systems on golf cars
and PTVs has had opposing opinions both for and
against seatbelts. The opinion that golf cars and
PTVs should not have any type of seatbelt system has
been primarily put forth by the National Golf Car
Manufacturers Association (NGCMA), a non-profit
corporation consisting exclusively of golf car
manufacturers and organized “to promote the
common business interest of its members” [18].
During the 1997 NHTSA rulemaking process related
to the newly designated motor vehicle category of
“Low Speed Vehicle” (LSV) [9], the NGCMA
viewed the seatbelt requirement as “antithetical to the
personal safety of drivers and occupants of golf cars”
[10] and cited ANSI/NGCMA Z 130.1-1993 [19]
which required a Rollover Protective Structure
(ROPS) for any golf car containing seatbelts.
Additionally, the NGCMA suggested that existing
hip restraints do not prevent occupants from jumping
or leaping out of golf cars to avoid injury when the
car is about to rollover. Accordingly, the NGCMA
Golf Course Safety Guidelines [20] state that “use of
seatbelts without adequate overhead protection may
result in severe injury or death.” The investigation
by NHTSA regarding the establishment of the “Low
Speed Vehicle” (LSV) classification included
research of golf car safety; until it was determined
that NHTSA would only regulate Low Speed
Vehicles intended for on-road use and with a
minimum speed of 20 mph. Hence golf cars and
PTVs with a maximum speed of less than 20 mph are
not currently regulated by any federal agency and the
decision to require seatbelts in golf cars and PTVs is
left to state and local jurisdictions. It should also be
noted that NHTSA in its final ruling concluded that
“the conjecture by some commenters that it would be
valuable to be able to jump out of an LSV are
unsubstantiated speculation that is especially
unpersuasive given the volume of data showing that
ejection is extremely dangerous and that seatbelts are
remarkably effective at preventing ejection” [10].
Accident Statistics
It is estimated that there were, on average,
approximately 13,000 golf car related injuries
requiring emergency room treatment in the United
States per year from 2002 to 2007, not including
fatalities that did not involve emergency room
treatment. The estimated number of accidents
steadily increased from roughly 11,000 in 2002 to
over 17,000 in 2007 [21]. Of these, approximately
40% (i.e. over 5,000 per year) involved an ejection
from a moving car, representing by far the most
common type of accident. In cases where the
location of the injury was reported, approximately
70% occurred at sports or recreational facilities (e.g.
golf courses) while the remainder occurred at
Seluga 3
locations such as private homes or public streets,
indicating that these statistics do not make the
distinction between golf cars and PTVs that the
NGCMA does. When this data is filtered to include
only children (i.e. age < 16 years), it is found that this
age group is involved in approximately 40% (i.e.
over 5,000 per year) of all documented accidents.
Furthermore, children are substantially more
susceptible to ejection than adults based on the fact
that slightly over 50% of child accidents (i.e. over
2,600 per year) involve a fall from a moving car. Of
these child ejection accidents, approximately 50%
occurred at a sports or recreational facility, while the
remaining 50% occurred either at home, on a street,
or at some other public property. In light of these
statistics, ejections from a moving car represent a
significant number of serious golf car and PTV
accidents involving children, the reduction of which
would significantly improve occupant safety [1,2,22].
It should also be noted that according to the same
statistics, approximately 10% of golf car and PTV
accidents involve a rollover. Therefore, even if
seatbelts did present some increased danger for
passengers in rollover events as supposed by the
NGCMA, the relative number of ejections accidents
to rollover accidents (i.e. approximately 4 to 1)
indicates that the addition of seatbelts could still offer
an overall improvement of golf car and PTV safety.
Furthermore, there are design opportunities available
for reducing the number of rollover events [23].
In addition to the statistical injury data, CPSC case
narratives also include some details regarding each
accident. One common scenario for a passenger
ejection accident is when a golf car or PTV, traveling
near its maximum speed, is turned to the left. CPSC
data from 2002-2007 contains many accident
narratives that match this scenario closely, such as
“riding with dad in golf cart, dad made a sharp turn
and [patient] fell out” or “patient on golf cart at
home, brother turned and threw him off cart.” Many
more of the “fall from cart” type accidents may also
involve ejection during a left turn, but the accident
narratives are too vague to make this determination in
most cases. This theme is repeated in numerous
news articles that report many serious head injuries,
including some fatalities, that involve both child and
adult passengers falling out of a golf car or PTV
during a left turn [24, 25, 26, 27, 28, 29].
Current Designs and Standards
The major golf car and PTV manufacturers (i.e. Club
Car, E-Z-Go and Yamaha) do not provide seatbelts as
standard equipment with their golf cars and PTVs,
though personal communications with many
authorized dealers indicated that they will provide a
seatbelt if the customer requests one. While it may
be generally assumed that golf car users on a golf
course are not likely to make use of seatbelts due to
their need to frequently exit and re-enter the vehicle,
the same may not be true for a PTV or a golf car used
away from the golf course. In support of this
contention, many private communities and
municipalities where golf cars and PTVs are used as
the primary means of transportation do require
seatbelts. Obtaining some form of a seatbelt for a
golf car or PTV is not difficult, since most golf car
and PTV outfitters offer after market seatbelts
[30,31] to meet the market demands for seatbelts that
are not being met by the original equipment
manufacturers. The community of Palm Desert in
California was a pioneer in recognizing the use of
golf cars on their roadways and adopted a
transportation plan in 1993 requiring seatbelts in golf
cars. Some communities, such as Bald Head Island,
North Carolina, have recognized the safety benefits
of seatbelts but rather than requiring belts on all
vehicles, they only recommend that occupants utilize
them if present. It should be noted that, contrary to
the supposition of the NGCMA that “use of seatbelts
without adequate overhead protection may result in
severe injury or death,” the authors are not aware of
any incidences at these or other communities, where
the use of a seatbelt had a negative impact on the
injury outcomes of a rollover accident.
Unlike the Federal Motor Vehicle Safety Standard
(FMVSS) #500 for Low Speed Vehicles, which
requires a seatbelt be provided for each intended
occupant, neither ANSI standard Z130.1-2004 “Golf
Cars – Safety and Performance Specifications” [12]
nor ANSI Z135-2004 “Personal Transport Vehicles –
Safety and Performance Specifications” [13] require
that any seatbelts be provided. In lieu of seatbelts,
ANSI Z130.1 and Z135 require “a hand hold or
combination hand hold/hip restraint, anchored
securely to the [vehicle], creating a barrier to help
prevent an occupant from sliding outside of the
[vehicle]” [12, 13]. However, these ANSI standards
provided neither design requirements nor test
procedures to determine the effectiveness of the
provided restraints. It has previously been shown
experimentally and analytically that the existing
restrains, typically no more than 6” tall and 12” long
are ineffective for preventing passenger ejections [14,
15]. In addition to the fact that the top of the
handhold is often lower than the seated occupant’s
center of gravity, the location of the handhold (i.e. at
Seluga 4
the outboard edge of the seat) is the fulcrum about
which an ejected passenger will tend to rotate. As a
result, this type of handhold does not provide the
passenger sufficient leverage to prevent ejection.
Due to the ineffectiveness of these designs, occupant
ejections are by far the most common type of golf
car/PTV accident. Furthermore, the ANSI standards
do not require the manufacturer to provide a
recommended minimum occupant age. While both
standards require a warning label to be affixed to all
vehicles stating “remain fully seated and hold on
when in motion,” it is highly foreseeable that
occupants will not always hold on while the car is in
motion, which is especially true with regard to
children. Additionally, small children whose feet
cannot reach the floor may not have sufficient
strength to prevent ejection during a moderate turn.
METHODS
Dynamic Child Dummy Testing
A series of tests were conducted utilizing a 2004
Club Car Villager PTV and a Hybrid III 6 year old
dummy weighing 21.4 kg (47 lb). The test vehicle
had designated seating positions for four occupants,
two facing forward and two facing rearward. In each
test, the child dummy was placed unrestrained in the
right front seated position with a driver (see Figure
1).
Figure 1: Test vehicle with Hybrid III dummy
The vehicle in each test was brought up to full speed
(i.e. approximately 21 kph [13 mph]) by the driver
and the accelerator was then released and the car
steered into a moderate but easily controllable left
turn. In each test the occupant kinematics were
recorded with digital video and still images.
Two methods of collecting performance data for the
tests were employed. A tri-axial array of
accelerometers (IC Sensors 3031-050) was affixed
near the vehicle’s center of gravity. All
accelerometer data were collected following SAE
Recommended Practice: Instrumentation for Impact
Test – J211/1Mar95. The axis system was in
accordance with SAE J1733 Information Report with
positive X, Y and Z axes forward, rightward, and
downward, respectively. All accelerometer data were
collected at 1000 Hz and filtered using a SAE Class
60 filter. In addition to the accelerometer data,
vehicle performance data were measured using a
GPS-based system (VBOX, Racelogic LTD,
Buckingham, England). Three-dimensional speed
and positional data were collected at 100 Hz.
Biomechanical Simulations
Three-dimensional computer simulations of the test
vehicle and the child dummy were created using the
Articulated Total Body (ATB) simulation software
[32, 33] for comparison with the experimental results.
ATB is a simulation program that models the
dynamic response of systems of connected or free
bodies such as the human body during a dynamic
event. It can be used to model a dynamic
environment of surfaces and bodies that interact with
one another according to the physical laws of motion
and has been used previously to study ejections
during motor vehicle accidents [34]. In addition to
providing detailed numerical force and motion
results, the program also produces graphical
depictions of the simulation results.
To simulate the dynamic ejection experiments, a
model of the test vehicle was combined with a model
of a child dummy occupant based upon the geometry
of each. The child occupant model was created using
the Generator of Body Data (GEBOD), which is a
companion program to ATB that generates a model
of the human body for use in ATB simulations [35].
GEBOD utilizes regression equations to calculate the
geometric and inertial properties of body segments
based on the proportions associated with a 50th
percentile child of a given age, height and weight
[36]. The relevant geometry of the test vehicle
(primarily the seat and hip restraint geometry) was
measured directly from the test vehicle. Some of the
relevant measurements are shown in Figure 2.
Seluga
5
Figure 2: Test vehicle dimensions
The motion of the simulated vehicle was obtained
from the experimental vehicle acceleration data
described above. The coefficient of friction between
the simulated passenger and seat was 0.5 based on
averages obtained from testing typical clothing
materials on PTV seat surfaces.
The simulations were first conducted with no
attachments between the passenger and the vehicle
(i.e. occupant not tethered or holding on) for
comparison with the dummy experiment. For the
purposes of these and all subsequent simulations,
ejection was defined as a condition where the
passenger’s lower torso body segment moved over or
around the hip restraint and traveled outside of the
car. Next, the simulations were repeated with a
spring-damper connection added between the
occupant’s right hand and the hip restraint,
representing an occupant holding onto the hip
restraint. These simulations were used to determine
the effect of the occupant holding onto the hip
restraint. Subsequent simulations were completed
with a spring-damper connection between the
passenger’s left hand and the center of the seat,
simulating the occupant holding onto a handhold,
strap or the equivalent mounted near the center of the
bench seat, though such a handhold was not provided
on the test vehicle. Center seat handholds have been
previously investigated [14] and these simulations
were used to determine the grip and arm strength
necessary to prevent child ejection in conjunction
with such a device. The necessary grip strength was
then compared to typical child strength capabilities to
determine if it was feasible for a child occupant to
avoid ejection by making use of a central handhold.
Finally, additional ATB simulations were created
wherein a simulated lap seatbelt was added to
determine its effectiveness in preventing ejections
and to quantify the belt strength requirements
necessary to prevent ejection.
RESULTS AND DISCUSSION
Vehicle Dynamics/Occupant Kinematics
In Test 1 the PTV was brought up to a speed o
f
approximately 21 kph (13 mph) followed by
a
moderate left turn which produced peak latera
accelerations of approximately 0.6 g (see Figure 3)
This peak lateral acceleration was reache
d
approximately 0.5 seconds after the onset o
f
noticeable lateral accelerations. It should also b
e
noted that during the left turn maneuver pea
k
longitudinal decelerations of approximately 0.1
g
were developed. The radius of the resulting turn wa
s
approximately 20 ft.
Figure 3: Test 1 recorded vehicle acceleration
The occupant kinematics demonstrated in the tes
t
show the child dummy moving laterally initiall
y
followed by a combined movement of the dumm
y
moving laterally and forward (see Figure 4).
Figure 4: Test 1 observed occupant kinematics
5
f
a
l
.
d
f
e
k
g
s
t
y
y
Seluga
6
The hip restraint during the ejection phase acts as a
tripping mechanism placing the child dummy into a
head first dive onto the asphalt track. Due to the
lower inertial properties of the child dummy relative
to an adult dummy, the ejection process is of a
significantly shorter duration than that observed by
the authors in adult ejections. This rapid onset of
ejection and subsequent trip produced by the hip
restraint leaves the occupant little remedy in avoiding
ejection. It should also be noted that due to a child’s
small stature, an active child would have little
opportunity to jump from the vehicle by pushing off
the floorboards since the child’s feet cannot reach the
floorboards.
In Test 2 the PTV was again brought up to a speed of
approximately 21 kph (13 mph) followed by a
moderate left turn, but in this sequence, one potential
driver response to the child dummy ejection was
demonstrated. During the turn the driver remained
watching the child dummy and at the first observable
signs of a potential ejection, the driver commenced
maximum effort braking. This maneuver produced
peak lateral accelerations of approximately 0.5 g’s
along a turning radius of 27 ft followed by brake
induced longitudinal decelerations of approximately
0.5 g’s (see Figure 5).
Figure 5: Test 2 recorded vehicle acceleration
Once again, as demonstrated in the previous test, the
hip restraint acts as tripping mechanism putting the
child dummy into a head first dive onto the test track
(Figure 6). Additionally demonstrated in this test is
that once the ejection process has started, a braking
action by the driver will not prevent an ejection.
Figure 6: Test 2 observed occupant kinematics
These experiments demonstrate the effects of
a
moderate left turn and the resulting lateral an
d
longitudinal accelerations which act upon a righ
t
front passenger and lead to ejection. Passenge
r
ejection is most likely to occur during a left tur
n
since a right turn will tend to force the passenger t
o
his left, towards the center of the car. Chil
d
passengers are especially susceptible to ejectio
n
because of their small size and consequent relianc
e
upon the hip restraint to prevent ejection. While th
e
experimental child dummy occupant does not includ
e
muscular reactions, the potentially rapid onset o
f
vehicle acceleration (i.e. 0.5 seconds or less
)
indicates that real occupants, particularly youn
g
children, may not have time to react before th
e
ejection process has begun. The results of thes
e
experiments indicate that current hip restraints are no
t
large enough to prevent the ejection of small childre
n
during moderate left turns. It should be noted tha
t
driver ejections, while still possible, are generall
y
less likely due to the fact the driver will inherentl
y
anticipate all steering maneuvers and is also able t
o
use the steering wheel as a handhold.
Biomechanical Simulations
Unbelted Occupant
The simulated kinematics of the unbelted occupan
t
ejection show excellent correlation to th
e
experimental dummy results from both tests, as ca
n
be seen by comparing Figure 4 and Figure 6 wit
h
Figure 7 and Figure 8 (see Appendix A). Both th
e
direction and the timing of the simulated occupan
t
motions match the experiments. In the simulation o
f
Test 1, the unbelted occupant leans and slide
s
towards the passenger side hip restraint, due to th
e
6
a
d
t
r
,
o
d
n
e
e
e
f
)
g
e
e
t
n
t
y
y
o
t
e
n
h
e
t
f
s
e
Seluga 7
lateral acceleration generated by the turning vehicle.
Next, the occupant rotates over the top and around
the front of the hip restraint and is ejected out the
passenger side of the vehicle. Following ejection, the
occupant’s head strikes the ground. It should be
noted that although the experimental and simulated
dummy rest orientations are somewhat different (i.e.
the dummy’s feet are facing in different directions
after it comes to rest), these differences occur after
the impact with the ground. The occupant kinematics
during the ejection phase and at ground impact show
close correlation.
Figure 7: Test 1 simulated occupant kinematics
(unbelted and untethered)
In the simulation of Test 2, the unbelted occupant
again initially leans and slides towards the passenger
side of the vehicle due to the lateral acceleration.
Then, as a result of the braking induced longitudinal
deceleration, the occupant slides forward, beyond the
highest portion of the hip restraint, which is only 3
inches long. Finally, as in Test 1, the occupant
rotates over the hip restraint and is ejected out the
passenger side of the vehicle, striking his head on the
ground.
Figure 8: Test 2 simulated occupant kinematics
(unbelted and untethered)
The close correlation between the experimental
occupant kinematics and the simulated motions
indicate that the ATB model can be utilized to
accurately simulate golf car and PTV occupant
ejection motions. These simulations also reveal that
just before impact, the occupant’s head has a speed of
approximately 15-25 kph (9-15 mph), including a
vertical component of velocity of approximately 10-
12 kph (6-7 mph) which is equivalent to a fall height
of 0.4-.5 meters (1.4-1.5 ft). Research regarding fatal
falls from play equipment indicates that children who
fall from heights as low as 0.6 meters (2 ft) onto soil
or grass can receive fatal head injuries [37]. Thus,
the ejection of a child from a golf car or PTV poses
significant risk of serious, possibly fatal head injury,
especially if the child lands on a paved surface.
Occupant Holding Outboard Hip Restraint
The simulated kinematics of the unbelted occupant
holding the outboard hip restraint demonstrated a
high risk of ejection, consistent with the findings of a
previous study [14]. The ejection process of a child
holding onto the hip restraint is depicted in Figure 9.
Figure 9: Test 1 simulated occupant kinematics
(unbelted and tethered to hip restraint)
As can be seen from the simulated occupant motion,
holding onto the hip restraint handhold located at the
outboard edge of the seat is ineffective because that
point is also the fulcrum about which an ejected
passenger will tend to rotate. Therefore, this
arrangement requires the occupant to generate large
torques about the hand hole to counteract the lateral
acceleration forces, which will be difficult since the
point of force application (i.e. the outboard handhold)
offers very little leverage about the occupant’s center
of rotation over the top of the handhold. Generating
such torques will be difficult for adults and even
Seluga 8
more difficult for small children, since they are less
likely to be able to attain a “power grip” around the
handhold (i.e. fingers flexed around the handhold to
form a clamp against the palm) due to its size relative
to their hands. Therefore, this type of outboard
handhold may not provide the passenger sufficient
leverage to prevent ejection, regardless of the
occupant’s grip strength.
Occupant Holding Central Handhold
The simulated kinematics of the unbelted occupant
holding the proposed centrally located handhold
demonstrated that with sufficient grip strength, such a
handhold could effectively mitigate the risk of
ejection (see Figure 10).
Figure 10: Test 1 simulated occupant kinematics
(unbelted and tethered to central handhold)
In this case, the minimum peak hand force required
to prevent passenger ejection was simulated for each
test. For the simulated 21 kg (47 lb) 6 year old
occupant, a peak hand force of approximately 67-107
N (15-24 lb) was required to prevent ejection. It
should also be noted that during the simulated left
turn, the central handhold caused the occupants left
arm to be loaded in tension. Therefore, the only
action required of the occupant is to hold onto the
handhold, since active shoulder and elbow efforts are
not necessary to prevent ejection. Child strength data
indicates that children as young as 3-5 years old are
routinely capable of hanging from a bar with arms
straight for 45-90 seconds [38]. This data indicates
that children are capable of supporting roughly half
their body weight with each arm under tension when
a sufficient handhold is provided, on par with the
tensile arm force required to prevent ejection with a
central handhold. Since the recorded lateral vehicle
accelerations during a left turn last only 3 seconds, it
is reasonable to assume that many children would
have sufficient strength to hold themselves in a golf
car or PTV during a moderate left turn if a centrally
mounted handhold were provided. Therefore, a
center-mounted left handhold would be an effective
countermeasure for mitigating the risk of ejection and
seems to be a prudent and inexpensive safety feature
that also facilitates compliance with ANSI standard
Z130.1. The limitation of such a handhold is that it is
not a passive safety device as it does require that the
occupant utilize the handhold.
Belted Occupant
Finally, the occupant simulations with a safety belt
included demonstrated that a seatbelt is extremely
effective at preventing the ejection of even a passive
occupant (see Figure 11).
Figure 11: Test 1 simulated occupant kinematics
(belted)
This is also consistent with previous dynamic dummy
testing [15]. Furthermore, the simulations indicate
that the peak force at the inboard seatbelt anchor
point is approximately 220-490 N (50-110 lb) for the
simulated 21 kg (47 lb) occupant (i.e. approximately
1-2 times the occupant’s weight, see Figure 12).
Simulated belt forces at the outboard anchor point are
negligible. One explanation for the inboard lab belt
forces sometimes exceeding the occupant weight is
that the geometry of the seatbelt causes the tension to
act at non-horizontal angle, requiring larger forces to
generate the necessary horizontal loads to prevent
ejection. The initial slack in the belt and the resulting
magnitude of the interaction between the hip restraint
Seluga
9
and the occupant also play a role in how much force
must be provided by the seat belt to prevent ejection.
Figure 12: Test 1 Simulated inboard belt force
This belt load is significantly less than the loads
experienced by automobile belts during impact
events. Therefore, the strength of a golf car or PTV
lap belt need not be build to automotive standards to
be effective in preventing occupant ejection. Since,
if a lap belt is provided, it would be desirable to
provide one that could also prevent adult ejections,
additional simulations were conducted using a 95 kg
(209 lb) 95% simulated adult male occupant to
characterize the peak belt loads that would be
generated by a larger occupant. These simulations
resulted in peak belt loads of approximately 980 N
(220 lb), again indicating that an automotive strength
belt need not be provided if the goal of the design is
to prevent ejections during driver induced
accelerations and not to offer protection in collisions.
In fact, providing a safety strap that will break free
under high acceleration conditions may be more
appropriate since the proposed safety strap/belt’s
purpose is solely to prevent an occupant ejection
during a maneuver and not to offer crash protection.
CONCLUSIONS
Summary
The coordinated experimental dynamic dummy
testing and biomechanical computer simulation
program presented in this study indicate that current
golf car and PTV designs create a situation where
young passengers are especially susceptible to
ejection during moderate left turns. Furthermore,
when passengers use the provided outboard hip
restraint as a handhold, little protection is provided
because the ejected passenger can easily rotate abou
t
the hip restraint due to the small size of the hi
p
restraint and the insufficient leverage provided whe
n
holding onto the outboard handhold with the righ
t
hand. While a previously proposed center-mounte
d
left handhold does offer better ejection protectio
n
when used, this feature cannot protect a passiv
e
occupant. Therefore, a lap belt restraint, which i
s
extremely effective at preventing ejection, is the bes
t
method for preventing child ejections. Furthermore
the lap belt need only withstand minimal forces t
o
prevent ejection during a non-impact event and thu
s
automotive strength seatbelts meeting current Federa
Motor Vehicle Safety Standards are not necessary t
o
prevent occupant ejections.
Recommendations
In light of these results, it is recommended tha
t
children be prohibited from riding in golf car
s
without seatbelt type restraints when used on gol
f
courses. If children are allowed to ride on golf car
s
with no seatbelts then, at the very least, a centrall
y
mounted handhold should be provided to reduce th
e
likelihood of ejection. Furthermore, passive hi
p
restraint effectiveness should be improved on all gol
f
cars and PTVs by increasing the size of the restrain
t
in order to improve occupant retention when
a
seatbelt is either not provided or not used. When gol
f
cars or PTVs are driven outside a golf course setting
seatbelt type restraints should be provided for al
occupants, especially when those occupants ar
e
children. The community of Palm Desert i
n
California offers one example of the type of safet
y
rules that should be implemented in loca
communities.
9
t
p
n
t
d
n
e
s
t
,
o
s
l
o
t
s
f
s
y
e
p
f
t
a
f
,
l
e
n
y
l
Seluga 10
APPENDIX A: ENLARGED KINEMATICS FIGURES
Figure 4: Test 1 observed occupant kinematics
Figure 7: Test 1 simulated occupant kinematics (unbelted and untethered)
Seluga 11
Figure 6: Test 2 observed occupant kinematics
Figure 8: Test 2 simulated occupant kinematics (unbelted and untethered)
Seluga 12
REFERENCES
[1] McGwin, G., Zoghby, J., Griffin, R., and Rue,
L., Incidence of Golf Cart-Related Injury in
the United States. J Trauma; 64:1562-1566,
2008.
[2] Watson, D., Mehan, T., Smith, G., McKenzie,
L., Golf Cart-Related Injuries in the U.S. Am J
Prev Med 2008; 35 (1):55-59, 2008.
[3] Insurance Institute for Highway Safety (IIHS),
Souped-up golf carts hit the streets. Status
Report, www.iihs.org, Vol 37, No 4, April,
2002.
[4] Insurance Institute for Highway Safety (IIHS),
Low-Speed Vehicles. 2009,
http://www.iihs.org/laws/LowSpeedVehicles.a
spx.
[5] Suhr, J., Cheap ride: More towns let golf carts
putter about. Associated Press, August 6,
2008.
[6] Passaro KT; Cole TB; Morris PD; Matthews
DL; MacKenzie WR, “Golf cart related
injuries in a North Carolina island
community,” Injury Prevention, 1996 Jun;
2(2):124-5
[7] E-Z-Go Freedom RXV description,
http://www.ezgo.com/personal/leisure/freedom
_rxv.html, accessed February 16, 2009.
[8] Yamaha web site: Personal Vehicles,
http://www.yamahagolfcar.com/, accessed
February 16, 2009.
[9] National Highway Traffic Administration, 49
CFR Part 571, Docket No. 96-65; Notice 2,
1996.
[10] National Highway Traffic Administration, 49
CFR Part 571, Docket No. 98-3949, 1998.
[11] U.S. Department of Transportation (DOT),
Title 49 – Transportation, Chapter V –
National Highway Traffic Safety
Administration, Department of Transportation,
Part 571 - Federal Motor Vehicle Safety
Standards, 571.500 Standard No. 500; Low–
speed vehicles, 2003.
[12] American National Standards Institute (ANSI),
American National Standard for Golf Cars-
Safety and Performance Specifications
Z130.1-2004.
[13] American National Standards Institute (ANSI),
American National Standard for Transport
Vehicles - Safety and Performance
Specifications Z135-2004.
[14] Seluga, K., Ojalvo, I. & Obert, R., “Low
Speed Vehicle Passenger Ejection Restraint
Effectiveness,” J Accident Analysis &
Prevention 2005; 37:4:801-806.
[15] Long, T., Fugger, T., Randles, B., “Vehicle
Performance Characteristics and Seatbelt
Effectiveness in Low Speed Vehicle and Golf
Cars”, 19th International Technical
Conference on the Enhanced Safety of
Vehicles (ESV), Paper # 05-0431, June 2005.
[16] Glassbrenner, D., Safety Belt Use in 2004 –
Overall Results, National Center for Statistics
& Analysis, DOT HS 809 783, 2004.
[17] National Highway Traffic Safety
Administration (NHTSA), Seat Belt Use in
2006 - Use Rates in the States and Territories,
Traffic Safety Fact Sheet, DOT HS 810 690,
2006, www.nhtsa.dot.gov.
[18] National Golf Car Manufacturers Association
(NGCMA), Mission Statement,
http://www.ngcma.org/mission.htm, 2009.
[19] American National Standards Institute (ANSI),
American National Standard for Golf Carts-
Safety and Performance Requirements,
Z130.1-1993.
[20] National Golf Car Manufacturers Association
(NGCMA), Golf Course Safety Guidelines.
Atlanta, GA, 2006.
Seluga 13
[21] Consumer Products Safety Commission
(CPSC), National Electronic Injury
Surveillance System (NEISS) Database.
http://www.cpsc.gov/Neiss/oracle.html, 2007.
[22] O’Brien, C., Rutherford, G., and Marcy, N.,
Hazard Screening Report: Sports Activities
and Equipment (Excluding Major Team
Sports). Consumer Products Safety
Commission (CPSC), May 2005.
[23] Seluga, K., & Ojalvo, I., “Braking Hazards of
Golf Cars and Low Speed Vehicles,” J
Accident Analysis & Prevention 2006;
38:6:1151-1156.
[24] WYFF4 Greenville South Carolina News, 14-
Year-Old Dies After Fall From Cart. January
21, 2008,
http://www.wyff4.com/news/15102943/detail.
html.
[25] Cady, P., Golf Cart Accident Hospitalizes
Girl., EudoraNews.com, Eudora, KS, July 20,
2006.
[26] Meyers, D., Child Dies in Golf Cart Accident.,
The Neshoba Democrat, Philadelphia, MS,
September 7, 2005.
[27] McGrath, S., Alex Man Critical After Golf
Cart Fall., The Herald Bulletin, Anderson, IN,
April 5, 2008.
[28] Hollis, J., Golf Cart Accidents Skyrocketing.,
Atlanta Journal-Constitution, Atlanta, GA,
June 22, 2008.
[29] Saginaw News, Saginaw Township Police
Detective Michael L. Cohee Dies After Freak
Accident. Saginaw, MI, June 23, 208.
[30] Buggies Unlimited, Seatbelt Kit,
http://www.buggiesunlimited.com/golf-
cart/buggies-unlimited-seat-belt-
kit/SBT%20UNV%200413, accessed February
20, 2009.
[31] GolfCarCatalog.com, Golf Cart Seatbelt Kit,
https://secure.golfcarcatalog.com/catalog/inde
x.cfm?fuseaction=product&theParentId=1839
&id=1915, accessed February 20, 2009.
[32] CFR Part 571, Docket No. 98-3949 Cheng, H.,
Rizer, A, Articulated Total Body Model
Version V User’s Manual, Air Force Research
Laboratory, Wright-Patterson Air Force Base,
Ohio 45433-7901, 1998.
[33] Grimes, W., Using ATB in Collision
Reconstruction, Society of Automotive
Engineers (SAE) Technical Paper No. 950131,
1995.
[34] Cheng, H., Obergefell, L., Rizer, A., ATB
Model Simulation of a Rollover Accident with
Occupant Ejection, Society of Automotive
Engineers (SAE) Technical Paper No. 950134,
1995.
[35] Cheng, H., Obergefell, L., Rizer, A., Generator
of Body Data (GEBOD) Manual Air Force
Materiel Command, Wright-Patterson Air
Force Base, Ohio 45433-6573, 1994.
[36] Tilley, A., The Measure of Man and Woman:
Human Factors in Design, New York, NY,
Henry Dreyfuss Associates, 1993.
[37] Plunkett, J., “Fatal Pediatric Head Injuries
Caused by Short-Distance Falls”, The
American Journal of Forensic Medicine and
Pathology 22(1):1-12, 2001.
[38] Peebles, L., Norris, B. Childata: The
Handbook of Child Measurements and
Capabilities – Data for Design Safety, UK
Dept. of Trade Industry, 1995.