Th. Hummel, K. Langwieder, F. Finkbeiner, \X/. Hell
Institute for Vehicle Safety, GDV, Germany
Second Child Occupant Protection Symposium
November 12, 1997
Lake Buena Vista, Florida
The compulsory use of child restraint systems (CRS) in
cars which came into force on 1st April 1993 led to a con-
siderable increase in the belting rate of children in
Germany, but between 30% and nearly 60% of the
children aged 0<12 years are only restrained by an adult
belt (lap or three-point belt).
On the basis of a new accident material of the German
Motor Insurers (593 restrained children 0 to 12 years
involved m 448 car accidents) the injury risk of children
being belted with an adult belt only is compared to those
injury risks of children being restrained in different types
of CRS (4/5-point belt, 3-point belt, impact shield, booster
cushion). The form of restraint ,,child with an adult belt
only involves disproportionately high risks.
In retrospective accident studies it is relatively difficult
to get detailed information about the frequency and the
exact kind of CRS misuse. Therefore 250 users of CRS
were observed and interviewed. Only in one third of these
observations the CRS were correctly mounted; depending
on different types of CRS the misuse rates lay between 20
and 90%.
To check up the effect of the most frequent and most
severe kinds of misuse 20 crash tests with wrong fitted
CRS were carried out; for this reason tests with an ECE
sled and with a cut-off car body have been performed.
Depending on the kind of CRS a partly considerable
reduction of the safety effect of CRS could be observed in
case of misuse. Protection against misuse by a proper
design has to be given highest priority for the safety
improvement of future CRS.
The number of children killed in car accidents in West
Germany [18] has been reduced significantly during the
past 25 years (Fig. 1). In view of the total number of
Th. Hummel, K. Langwieder, F. Finkbeiner, W. Hell
Institute for Vehicle Safety GDV, Germany
400 | | |
350 | A -- o
300 |
200 |... SN
100 |
50 .. ...
F i ]
| ;
1970 1975 1980 1985 1990 1995
Source: Federal Statistical Office of Germany
Figure 1: Children (from 0 to 14 years of age) killed in car accidents in
West Germany
children killed in traffic accidents, however, fatalities
involving children in automobile accidents continue to
rank first compared with children in pedestrian or bicycle
A comparison of individual countries indicates that
children in Middle Europe and Canada (Table 1) are rela-
tively frequently restrained when travelling in automo-
Table 1: International comparison of securing quotas of children
in cars
Total securing quota
Country Year Age Quota
Germany [1] 94 up to 12 yrs. 87%
Sweden [2] 94 up to 15 yrs. 87%
France [3] 91-92 |upto 12 yrs. 61%
Canada [4] 92 up to 9 yrs. 73%
* inner City
Annual surveys conducted by the Federal German
Highway Research Institute (BASt) [1], however,
confirmed that children are often restrained on an irregular
basis with adult seat belts only. This led German
legislators to introduce in April 1993 a statutory obligation
for the mandatory use of child restraint systems (CRS) in
automobiles. It 1s stated in this law that children who have
not yet completed the age of 12 and are shorter than
150 cm may only ride in automobiles if secured in child
restraint systems that are suitable for children and have
official approval. This measure resulted in a significant
increase In the percentage of children being restrained
while travelling in cars (1992 = 72%; 1996 = 88%) and
also led to a drop in the number of children killed in car
accidents (28% decrease from 1992 to 1996) .
A more detailed analysis of the percentage of children
restrained in cars (Fig. 2) shows that children up to the
age of five years are usually placed in child protection
systems and that only a few are restrained using an adult
seat belt and "only" 8% travel without any kind of protec-
tion. In the case of children between the ages of six and
eleven years, on the other hand, only 27% use a special
child protection system, 50% use an adult seat belt and
almost one out of every four children in this age group
does not use any kind of safety protection.
Cup to Syears @6 - 11 years
CRS Adult belt only unrestrained
Source: Federai German Highway Research Institute
Figure 2: Percentage of children restrained in cars according to age
groups (Germany, city traffic, 1996)
It has been demonstrated on the basis of real life
accidents [5] that the risk of suffering fatal and serious
injuries 15 approximately seven times higher for unres-
rained children than for restrained children.
The Institute for Vehicle Safety (formerly the Depart-
ment for Automotive Engineering and Accident Research)
has determined in a series of sled tests and accident studies
the loads to which children are subjected when they wear
adult seat belts at too early age [6]. The use of adult seat
belts on children, both lap belts and three-point belts,
causes considerable risk of injury, particularly to the
abdominal region.
In order to obtain additional information about the risk
of injury as a function of the type of restraint system in
real accident situations, new statistical accident data based
exclusively on restrained children was set up and analysed
within the scope of a research project entitled "Improving
the Protection of Children in Automobiles" [7]. In addi-
tion, parents were observed and interviewed as to how
they secure their children in cars and whether or not they
make mistakes (misuse in the sense of "incorrect installa-
tion of the child protection system" and "improper restraint
of children in restraint devices"). The most frequent or
most serious types of misuse were reproduced in sled tests.
A field study was devoted to determine whether or not
mistakes made during installation of a child restraint sys-
tem can be avoided by using ISOFIX, a standardised rigid
snap-type connection for mounting child protection
systems [8].
All automobile accidents involving at least one
restrained child less than 12 years of age were retrieved
from the "Vehicle Safety 90" accident data [9]. This
created a new accident database entitled "Child Safety 90"
and included 448 car accidents with details about 593
restrained children between 0 to 12 years of age. This new
accident material was based on approximately 18,000 car
accidents involving personal injury in the period 1990 and
The relatively well-balanced age distribution that exists
(Fig. 3) indicates the great importance of child restraint for
all age groups, since even the numerically smallest age
group (6 years) had an incidence of N = 35 in the statisti-
cal data.
59 57 60
DI 53 50 53
46 — 46 44 144
35 3
O 1 2 3 4 5 6 7 8 9 10 N
Age in years
Figure 3: Age distribution (N = 593 restrained children)
A comparison of different accident databases at the
Institute for Vehicle Safety reveals that the statistical data
contained in the "Child Safety 90" database relates to
accidents with a relatively high extent of vehicle damage
and relatively serious accidents.
Head-on collisions occurred in 340 cases (57%) and
were thus around three times more frequent than side
collisions (120 cases, 20%) or rear-end collisions (126
cases, 21%). Rollover accidents were observed in only 7
However, if the maximum abbreviated injury scale
(MAIS) [10] is included in these considerations, a totally
different picture emerges (Fig. 4). In the case of front-end
collisions, only 16% of the restrained children suffered
MAIS 2+ injuries compared with 26% of those involved in
side collisions. The risk to children on the struck side of
the vehicle is particularly high.
In order to reduce the high risk to such children, child
restraint systems should in the future be subjected to the
most realistic possible side collision tests and be further
improved by incorporating appropriate design measures.
International work is currently in progress on a suitable
inspection and test protocol [11,12].
The consequences of injury sustained during rear-end
collisions and rollover accidents were comparatively
noncritical in this case material (refer to Fig. 4), since
MAIS 2 injuries were the most serious degree of injury
observed in these accidents.
Front Side Rear Rollover
(N = 340) (N = 120) (N = 126) (№= 7)
Figure 4: Frequency of MAIS 2+ injuries as a function of the impact
area (N = 92)
An analysis of the severity of injury as a function of the
type of restraint (Table 2) shows that children who were
restrained in child protection systems were injured much
less frequently (42%) than children restrained with a lap
belt or three-point belt (13%). The range of moderate to
fatal injuries shows clearly (see Table 2) that a frequency
of MAIS 2+ injuries to children in a child protection
system (11%) 1s less than to children restrained by three-
point/lap belts (26%). It is clearly obvious that there is a
trend to a greater severity of injury in children who were
restrained with an adult seat belt only. These trends are
also confirmed by an analysis of the frequency of injury to
different body parts and the frequency of incidence of
multiple injuries.
Table 2: MAIS distribution as a function of the type of restraint
CRS 3-point belt/
lap beit
No. % No. %
MAIS 0 83 41.5 25 13.3
MAIS 1 95 47.5 114 60.6
MAIS 2 17 8.5 35 18.6
MAIS 3 - - 5 2.7
MAIS 4/5 2 1.0 8 4.3
MAIS 6 3 1.5 1 0.5
total 200 100 188 100
Yet another phase of assessment dealt with questions
about any conspicuous features observed in the use of
different types of child protection systems as far as the risk
of injury and safety were concerned. The exact type of
child protection system involved in the accident was
known in a total of 120 cases (ECE-Group 0-11, defined in
APPENDIX C). In order to obtain statistically significant
results, this comparison investigated only head-on
collisions that involved relatively severe injury (without
disastrous intrusion; degree of damage 3+4, defined in
APPENDIX A). Distinct differences were found between
the individual types of child protection systems. Table 3
shows a great degree of safety in the ECE-Group I,
particularly when the child 1s restrained with an impact
shield system (no MAIS 2+ injuries). Children sustain
fewest injuries (50%) with this type of child seat. There is
a significantly higher tendency to severe injuries when
4/5-point systems (harness belts) are used. There were
MAIS 2 injuries in four out of 20 cases: head injuries
(concussions, lacerations, gashes) in two cases and leg
injuries (tibia fractures) in another two cases. There are
different reasons for these injuries: Pushing front seats too
far back, rearwardly inclining flat backrests or placing a
child seat in the sleeping position all serve to increase the
Table 3: MAIS distribution as a function of the type of child restraint
system; only head-on collisions in serious accidents (degree
of damage 3+4)
rearward | 4/5-point impact 3-point
facing system shield system
No. No. No. No.
MAIS O - 4 1
MAIS 1 2 12 7
MAIS 2 2 4 - 1
MAIS 3-6 - - - -
total 4 20 14 9
risk that a child will suffer both head and leg injuries as it
impacts against the headrest or backrest during the acci-
dent. Another negative factor arises if the seat is very loose
and if there is considerable slack in the seat belt.
APPENDIX B contains two real accidents comparing
an impact shield system and a 4-point system with the
different injury severity already described above.
In the case of the rearward facing systems in ECE-
Group 0, there are two cases involving MAIS 2 injuries
that are particularly prominent. These involved a concus-
sion and cranial fracture, both of which are quite likely the
result of head impact against the lateral edge of the seat
shell. Incorrect installation of the seats might well have
contributed to the injuries as well.
In order to obtain additional information about the
isolated child restraint system in combination with a three-
point seat belt, the severity of injury for the three-point
belt both with and without booster cushion is compared in
Table 4.
Table 4: MAIS distribution in head-on collisions as a function of age
groups; a comparison of restraint using a "three-point seat
belt plus booster cushion" and "three-point seat belt alone”
(without degree of damage 5)
3-point-belt + oniy 3-point-belt
booster cushion
2 - 5 years 6 - 11 years 2 - 5 years 6 - 11 years
No. % No. % No. % No. %
MAIS 0 2 18.2 3 30.0 - - 5 9.8
MAIS 1 8 27 6 60.0 9 69.2 32 62.7
MAIS 2 - - 1 10.0 3 23.1 10 19.6
MAIS 3 1 9.1 - - - - 2 39
MAIS 4/5 - - - - 1 7.7 2 3.9
MAIS 6 - - - - -
total 11 100 10 100 13 100 51 100
The MAIS distribution shows that children in both age
groups who were restrained solely by a three-point seat
belt were harmed much more frequently and suffered criti-
cal injuries more often. MAIS 2+ injuries were about three
times more frequent when the child was restrained using
only a three-point belt than when restrained with a three-
point belt combined with booster cushion. The clear
differences in the age groups ranging from 2 to 5 years are
confirmation that one should avoid restraining a child with
a three-point belt with or without booster cushions at too
early an age. In view of these results, it can be recom-
mended that even older children (from 6 to 11 years of
age) should continue to use booster cushions.
As already mentioned, the introduction of a statutory obli-
gation for the mandatory use of child restraint systems in
Germany had an obvious impact on child safety in
automobiles. As a result, the numbers of child protection
systems actually in use has plainly increased. However,
other deficits have emerged instead, as the statutory
obligation did not have the same positive influence on the
quality of child restraint use as it did on their quantity.
Incorrect use of child seats leads to a marked reduction in
the safety protection that is theoretically possible. Studies
carried out in Europe on the quality of child restraint
use [13] concluded that clearly more than half of all
children are incorrectly restrained when travelling in cars.
These realisations caused an observation and interview
study to be set up within the scope of a research project
entitled "Improvement in the Protection of Children in
Automobiles" [7]. In order to be able to ascertain all
aspects of misuse, technical and psychological aspects
were combined in a holistic methodology. An observation
protocol was drawn up based on the "Misuse Mode and
Effects Analysis - MMEA" study [14] as well as on the
basis of the observation forms developed by the ISO [15].
These were then extended and upgraded to a considerable
extent for the purpose of this study. Information was
collected that pertained to the vehicle and passengers,
additional traits of the interviewed persons as well as
detailed information about the children such as their age,
sex, seating position, type of restraint and misuse. The
interviews were conducted on the basis of a totally
redesigned concept containing not only the particulars of
the persons involved, the type of restraint, mounting and
purchase of the child protection system, but also subjective
assessments, judgements and expertise of the persons
using the child protection systems.
A random sampling containing a total of 250 observa-
tions/interviews provided information on 354 children
under 12 years of age. All restraint systems that were
representative of the German market (APPENDIX C) were
taken into consideration within the scope of the study.
Child protection systems belonging to ECE-Groups 0-111
amounted to 83% the other children were restrained with
the adult belt only.
Mistakes in installing the child seats or in securing the
children in the seat were made in 63% of all children who
were restrained using a child restraint system. As indicated
by Figure 5, serious mistakes that could seriously impair
the safety of the child were found in one-third of all cases.
slight misuse
Figure 5: Misuse in the use of child restraint systems (N = 354)
Depending on the kind of protection system, misuse
rates were observed that ranged between 20% and in
excess of 90% (Table 5). A comparison of 4- and 5-point
systems illustrates that misuse can indeed be reduced by
technical measures and further developments and in this
case amounted to almost 25% at the very least. In addition
to the misuse rate, the absolute number of mistakes made
when installing the seat or restraining the child was of
absolutely essential - in particular with a view to avoiding
misuse - (see Table 5). It must be established that systems
in which installing the seat and restraining the child are
two separate operations (e.g. 4/5-point systems, rearward
facing systems) exhibit a higher frequency of misuse than
the other systems. The hitherto good results achieved
using ECE-Group | impact shield systems were found to
have a high misuse rate of 92% and were thus called into
question. However, 1f one takes the frequency of mistakes
into consideration, it will be found that this system is
subject to only 1.3 mistakes per installation/restraint and
thus has the most positive results except for ECE-Group II
impact shield systems.
Table 5: Misuse and frequency of mistakes for the different types of
child restraint systems
system-type total misuse misuse frequency
No No. %
ECE-group 0
rearward facing 18 10 55.6 2.0
ECE - group |
4-point system 63 57 90.5 2.5
5-point system 62 41 66.1 2.0
impact shield system 13 12 92.3 1.3
ECE - group 11
3-point system 24 6 25.0 1.5
ECE - group ll
impact shield system 5 1 - 1.0
ECE - group [l/l]
booster cushion 107 56 52.3 1.7
total 292 183 62.7
The following serious forms of misuse were found for
the individual types of systems:
ECE-Group 0; rearward facing systems
=“ seat belt guide not used
= belt path incorrect (Picture 1)
=" seat installed in wrong direction
= seat installed in wrong position
= safety system not attached
= weight limit exceeded
ECE-Group I; 4/5-point systems
Picture 2
=» seat fastening loose (Picture 2)
Pictu re 3
= 4/5-point belt too loose (Picture 3)
Picture 4
= belt path incorrect (Picture 4)
Picture 5
belt buckle positioned too high (Picture 5)
seat attached incorrectly
weight above or below weight limit
seat defective/damaged
seat installed in the wrong position
parents used their own mounting construction
4-point shoulder belts passing beneath the arm
S-point shoulder belt hanging over the shoulder
belt in the child seat unbuckled
vehicle seat belt unbuckled
ECE-Group I; impact shield systems
used without impact shield
seat damaged
impact shield used alone
weight limit exceeded
ECE-Group I; 3-point systems
used with lap seat belt
ECE-Group IVIII; booster cushions
Picture 6
shoulder belt passing beneath the arm (Picture 6)
used with lap seat belt
used too early, child still too small
shoulder belt passing behind the back
The extent to which the person securing the child is
convinced that the restraint system will enhance the child's
safety is certainly important for the consistent and proper
use of child seats. In response to the question whether the
child seat currently being used provides adequate safety,
79% of the interviewees responded affirmatively, 18% of
those interviewed were only partly convinced and in 5% of
the cases, the person interviewed stated that (s)he was not
generally convinced of the protection afforded by child
seats. The reasons given for their mistrust were the inade-
quate mounting of the restraint system in the vehicle, the
size of the child as well as the conviction that a child seat
cannot prevent injury entirely.
The problem is that parents are often unaware that their
child has not been properly restrained. Parents who had
improperly restrained their child were asked: "Do you
think that your child is properly restrained and that every-
thing is OK ?" In 60% of all cases, the response was "yes",
1.e. they thought they had made no mistakes; only 21% of
the respondents replied "no", indicating they were aware
that a mistake had been made (Fig. 6).
Question: „Do you think your child is properly
restrained and that everything is OK?“
don’t know
yes | think so no
Figure 6: Interviews ol parents who had improperly restrained their
Asked to assess the possible consequences of incorrect
or improper restraint measures, 20% of the respondents
replied that they were in no position to judge the mistakes
they had made as far as the safety of their child was
concerned. 26% of the persons interviewed thought that
incorrectly restraining the child would have no
consequences whatsoever on the safety of the child. 32%
of the respondents assumed that it might have a minor
effect on or that it might slightly endanger the safety of the
child, 20% thought there would be a considerable
influence, and only 0.6% of those interviewed thought that
the mistakes they had made in restraining their child
would have a grave effect on or would gravely endanger
their child's welfare.
To determine what stress forces occur in situations in
which children are improperly restrained or in which the
restraint system is incorrectly installed, two series of sled
tests were performed to simulate the most frequently
observed and the gravest forms of misuse.
In the first series of tests (Picture 7), the child protec-
tion system was mounted on an ECE sled equipped with a
rear seat bench from a VW Golf III.
In the second series of tests (Picture 8), the child protec-
tion systems were mounted in a cut-off car body mounted
on a sled. Since it is impossible to detect all important risk
factors that might cause considerable stress forces in a real
accident, the sequence of events were also filmed using a
high-speed camera. A total of 20 tests were carried out
along the lines set forth in ECE-R 44 [16].
Picture 7: Test series /1; ECE test sled
Picture 8: Test series /2; car body
The measurement data for all 20 test runs is set forth in
Table 6.
ECE-Group 0 - (infant carrier) tests
Since all tests in ECE-Group 0 were performed using a
"new-born" dummy which was not equipped with measur-
ing sensors, the results can only be interpreted in this case
with respect to forward displacement of head and motion
The results for tests 1 to 3 with different forms of
"incorrect belt guide" reveal that in both test series the
permissible corridor of 550 mm based on the point C, as
defined in ECE-R 44 was exceeded but for one exception.
A marked difference was shown in test 3/1 in which the
seat disengaged completely from the seat attachment
(Picture 9).
Picture 9: Test 3/1; incorrect belt path
When the restraint system 1s installed so that it faces
forward (test 4), not only is there a danger that the forward
displacement of the head will be exceeded, but there is
also an additional risk that stress forces, which are likely to
be very high, will be applied to the cervical vertebrae
(Picture 10).
Picture 10: Test 4/2; incorrect positioning of a rearward facing
baby shell
The film evaluation of the tests using the car body (sec-
ond series of tests) clearly shows that in all ECE-Group 0
tests there is a danger of serious to fatal head injury when
the head collides with the instrument panel or windshield.
ECE-Group I - (toddler seat) tests
Both series of tests illustrate differences in the
measured forward displacement of the head. This is due to
differences In the stiffness of the seats and in the geometry
of the seat belt.
In test 5, a 5-point belt that was 25 mm too loose
produced different results in both test series. In the first
series {with the ECE sled), the acceleration to head and
chest was higher, whereas in the second series (using the
car body) a considerably higher forward displacement of
the head was measured. It exceeded the limit of 550 mm
by 88 mm and caused the legs of the test dummy to slam
against the back of the front passenger seat. Both series of
tests exhibited great overextension of the dummy's neck
which is the cause of the high stress forces applied to the
cervical vertebrae (Picture 11).
Picture 11: Test 5/2; loose belt in the seat
In test 6 with a loosened seat attachment (75 mm slack
in the belt), differences arose due to the use of a P3/4
dummy for the first test series and a three-year-old dummy
for the second test series. In spite of the lighter dummy in
the first test, the head and chest were subjected to critical
stress forces. The forward displacement of the head that
was measured, on the other hand, was clearly below the
limit. In the second test series, an absolutely critical accel-
eration value of 120 g was measured at the head and the
permissible forward displacement of the head was clearly
exceeded by 124 mm. The seat mounting was too loose
and caused great hyperflexion of the dummy's neck in
both test series, which is indicative of high stress forces
applied to the cervical vertebrae (Picture 12).
Table 6: Results of the misuse sled tests
Seat Sled-Measurements Dummy Dummy- Measurements Abdomen Head
Position Type Imprint Excursion
Test No. kon] | [men] | fo) arar” [mm]
1/1 51.0 530 - TNO-PO - - - -
1/2 front 48.2 660 25.9 TNO-PO - - - -
2/1 51.0 530 - TNO-PO - - - - 480
2/2 front 48.8 660 24.4 TNO-PO - - - -
3/1 51.0 550 - TNO-PO - - -
3/2 front 48.6 670 25.2 TNO-PO - - - -
4/1 - 51.0 570 - TNO-PO - - - -
4/2 front 48.6 630 28.4 TNO-PO - - - -
5/1 - 50.1 520 23.6 TNO-P3 891 54.3 very small
5/2 rear 48.8 660 24.4 | US-3yrs. - 74.7 -
6/1 - 50.3 520 22.7 | TNO-P3/4 very small
6/2 rear 48.6 670 25.2 | US-3 yrs. 11
7/1 - 50.3 550 22.5 TNO-P3 606 59.4 25.7 very small
7/2 rear 49.2 690 26.4 US-3 yrs. -
8/1 - 50.4 550 21.0 TNO-P3 no
8/2 rear 49.3 680 24.5 US-3 yrs,
9/1 50.1 540 22.5 TNO-P6
9/2 rear 49.2 690 26.4 | US-6 vrs. - 37.0 40.2 - 475
10/1 - 50.1 520 23.2 TNO-P6 40.4
10/2 rear 49.3 680 24.5 | US-Byrs. 43.6
(shaded contours define values over the ECE-R 44 limits)
Vg = Sled Speed Ares Head = Resultant Head Deceleration
5 = Breaking Path Bes Thorax = Resultant Thorax Deceleration
Amaxs = Maximum Sled Deceleration HIC = Head Injury Criterion
US-Dummy = Humanoid-Systems
Test Description
Test No. /1 = 1st test series ECE sled
Test No. /2 = 2nd test series car body
Test No. 1-3: ECE-Group 0;
ECE-Group 0;
Test No.
Test No.
Test No.
Test No.
Test No.
Test No.
Test No.
NN un
: ECE-Group I;
: ECE-Group 1;
: ECE-Group I;
: ECE-Group I;
ECE-Group I;
. ECE-Group 11/111;
incorrect belt path
seat installed in the wrong direction
5-point seat belt with 25 mm slack
loose belt attachment (3-point belt with 75 mm slack)
incorrect belt path: lap seat belt placed around the seat at the front;
shoulder belt not attached to the back of the seat shell
4-point belt buckle located too high
impact shield not used
booster cushions; shoulder belt passing beneath the arm
Picture 12: Test 6/2; loose seat attachment
Test 7 shows the serious consequences that can arise
from improper or incorrect belt positioning. Whereas the
measurement data of the first test series only illustrate this
risk based on excessive values for the forward displace-
ment of the head, the car body test (second test series)
depicts the dreaded risk very strikingly. The entire system
tilted forward virtually without decelerating, thus
slamming the dummy's body against the back of the front
passenger seat (Picture 13). The head acceleration was
measured at 261 g which indicates that critical to fatal
injuries would have been sustained if this had occurred in
a real accident. The chest acceleration also far exceeded
permissible limits, although in this case it did not result
from the forces exerted by the belt, but rather from the
impact against the seat back.
Picture 13: Test 7/2; incorrect belt path
Test 8 was designed to investigate the consequences
that would arise when harness belt systems are used if the
seat buckle 1s positioned too high (at the level of the costo-
sternal junction). Furthermore, the seat was moved to the
sleeping position as this was expected to produce even
higher forces during an accident. Both tests produced con-
siderable submarming effects (Picture 14). As the dummy
Picture 14: Test 8/2; helt buckle located too high; submarining
slid forward through the seat, the buckle of the 4-point belt
slid up into the region of the dummy's frontal neck leading
to great cervical overextension and causing the head to hit
against the buckle. The high acceleration forces applied to
the head as well as the HIC value measured in the first test
series indicate very serious head injuries would have been
sustained mn the event of a real accident. The forward
displacement of the restraint system and the submarining
which was experienced in the car body test (second series
of tests) resulted in a deep indentation in the back of the
passenger seat caused by the dummy's legs. This in turn
means that leg injuries must be expected in the event of a
real accident.
When the impact shield of an impact shield system is
omitted (test 9) and a six-year-old dummy 1s used (Which
exceeds the age group for which the restraint device is
actually intended), the values for head and chest that were
measured in both test series as well as the measured for-
ward displacement of the head were found to be within
defined limits. The moderate abdominal imprint that was
detected in the first test series, however, indicates that
injury might be possible in the abdominal region, some-
thing that is caused by the high position of the lap seat
ECE-Group IVIII - (booster cushion) tests
The dangers that are caused by a shoulder belt passing
beneath the arm when a child is restrained using booster
cushions are illustrated in test 10. In both test series, the
forward motion of the dummy caused the shoulder belt to
slip downwards into the abdominal region. The upper part
of the body jackknifed virtually without being decelerated
and this in turn caused the head to hit against the legs
(Picture 15). The permissible forward displacement of the
head was greatly exceeded and, in addition, the head was
subjected to critical acceleration forces caused by the head
hitting against the legs. The extreme constriction that the
seat belt caused left a deep abdominal imprint which
seems to predict severe injury to the abdominal region.
Picture 15: Test 10/2; shoulder belt beneath the arm;
jackknifing effect
A "standardised rigid plug-in type connection" repre-
sents 4 completely novel type of mount. The system
known as the ISOFIX system was developed by the Inter-
national Organisation of Standardisation (ISO) [8] and is
intended to simplify the installation of child seats and
secondly to reduce misuse. In crash tests already
conducted [17] the ISOFIX system was found to have
significantly lower HIC values compared to a convention-
ally attached system and also lower chest deceleration
values were measured. A major advantage of the ISOFIX
system is that forward displacement of the head has been
reduced in several tests by an average of approx. 140 mm,
thus resulting in smaller forces being applied to the thorax
(Fig. 7).
conventional attached
Figure 7: Comparison of head movement between existing child
restraints and ISOFIX seats in a crash (upright position)
In the same study [17], the user-friendliness of the
ISOFIX system was studied. Significantly fewer mistakes
were made when installing the ISOFIX system than when
installing conventional seats. Whereas approximately half
of all current child seats were mounted incorrectly, misuse
occurred in only about 10% of the cases when ISOFIX
was used.
150 adult test subjects were asked to install and remove
two child seats from an automobile using schematic
operating instructions: one seat was a conventional ECE-
Group 1 seat, the other was an ISOFIX system seat
(prototype, ECE-Group I) (Picture 16). The test vehicle
was equipped with four ISOFIX attachment points
provided on the passenger seat. Using an observation
questionnaire, a protocol was kept as the test subjects
attempted to mount the seat including any special points or
conspicuous incidents. At the conclusion of the test, the
test subjects were requested to answer a few questions
about the installation procedure they had just completed.
Picture 16: ISOFIX prototype
For the most part, parents with children (91%)
succeeded in installing the seat, some 80% of them already
having had experience with child protection systems. In
order to gather enough instances of misuse in the study,
different child seats were used as the comparison child
protection system. In 135 out of a total of 150 cases, the
test subjects succeeded in installing the comparison child
protection system 1 (5-point system); the comparison child
protection system 2 (5-point system), which in the opinion
of the experts was more difficult to install correctly, was
used in 15 cases.
Significantly fewer mistakes were made in installing the
ISOFIX system (Fig. 8) than when the conventional seats
were installed. Whereas the misuse rates for the conven-
tional child seats ranged between 60 and 80%, misuse was
registered in only 4% of the cases when the ISOFIX seat
was installed, 1.e. only 6 out of a total of 150 test subjects
failed to correctly instal! the ISOFIX system.
N = 134 N = 150
testseat 1 testseat 2 ISOFIX
Figure 8: Frequency of mistakes made while installing conventional
child seats and an ISOFIX seat
The test subjects were also questioned about the
advantages and disadvantages of ISOFIX. The majority
said that ISOFIX was easier to install than a comparable
conventional system. They were also convinced that
ISOFIX 1s more stable and that the anchoring 1s better and
that it would provide more protection for the child in the
event of collision. The higher weight was seen as bemg a
significant drawback as well as the fact that ISOFIX is
only available for new cars. The ISOFIX system gamed
high acceptance in the study and most of the test subjects
preferred it over the comparison child protection system
that was used.
The goal of the accident studies was to examine exclu-
sively for restrained children what impact the type of
restraint would have on injuries and the severity of injury.
A new accident database entitled "Child Safety 90" was
created and contains information pertaining to 593
restrained children in cars (between 0 to 12 years of age).
[rrespective of the type of restraint, there is a far greater
danger that children will be injured in side collisions than
in head-on collisions. For this reason, greater attention
must necessarily be paid to this type of collision both
when designing and developing child seats (side-wings,
increased stiffness, head protection) and when drafting test
regulations. Within the scope of the ISO Working Group
WG 1! Child Restraint Systems in Road Vehicles, an ad-hoc
group was convened several years ago and 1s currently
drawing up testing criteria and testing regulations for child
protection systems in side-impacts on the basis of a world-
wide accident database [11, 12].
Children who have been restrained only by a three-
point seat belt or lap belt more frequently suffer injury
than those restrained in a child restraint system. Not only
the frequency of injury, but also the severity of injury is
significantly higher when children are restrained solely by
means of an adult seat belt. Children younger than six
years of age have been found to suffer a higher risk of
injury - when restrained only by an adult seat belt - than
children between 6 and 11 years of age.
There are indications that point to differences in the
degree of protection afforded by different types of
systems. Compared with forward facing impact shields
and 3-point belt child seat systems, forward facing harness
belt systems show a trend on the whole to greater severity
of injury in children. The effectiveness of rearward facing
ECE-Group I systems has not yet been investigated as
their market share in Germany is still very small.
The protection afforded by child protection systems can
be reduced considerably in some cases due to the improper
or incorrect use of child seats (misuse). On the whole,
misuse was observed in 63% of the random samples stud-
ied where child protection systems were being installed.
Serious misuse, that may well cause fatal injury, was in
fact observed in one-third of all cases. Design modifica-
tions and user education on how to install child restraint
systems must be improved so that the protection afforded
by such devices 1s not impaired. In the case of conven-
tional child protection systems, particularly those systems,
in which restraining the child and installing the child seat
are two separate operations, are more likely to induce
misuse than other systems.
Problems involving the proper fit of the child protection
system in the vehicle (seat belt geometry, vehicle seats)
led to moderate to serious misuse in the majority of cases.
Here, cooperation between child seat manufacturers and
motor vehicle manufacturers must emphasise combina-
tions that fit or do not fit (e.g. positive or negative vehicle
list). One of the crucial results of the interviews was that
most parents are unaware of the existence of misuse and
the dangers inherent in misuse. Well-targeted educational
programs and better structions for use ought to address
this problem.
In order to determine what impact Incorrect restraint
would have on the forces applied to the dummies,
frequently observed forms of misuse were reproduced 1n
sled tests. Both the forward displacement of the head as
well as the forces of acceleration to which the head and
chest of the dummies were subjected and even the forces
acting in the abdominal region were detected and regis-
tered. The maximum load values laid down in the test
regulations (e.g. ECE-R 44) were exceeded in almost all
tests, in some cases considerably. This confirms that incor-
rect or improper use of a child protection system can
substantially reduce the protection afforded by this system.
Serious to fatal injuries could well occur even in accidents
that are less serious than the accident simulated in ECE-
R 44 as a result of misuse.
The ISOFIX system has been tried and tested and repre-
sents a decisive improvement in the number of mistakes
that are made during installation (only 4% mistakes)
compared with conventionally mounted child seats (more
than 50%). This in turn can drastically reduce by a third
the high percentage of severe misuse (33%) observed with
conventionally mounted child seats. Previous tests also
verify that ISOFIX subjects a child dummy to lower
biomechanical forces than conventional systems both in
the case of head-on [17] and even side collisions [12].
The respondents themselves evaluated the ISOFIX
system as being better than the comparable conventional
child seat. The majority stated that ISOFIX is easier to
install than a comparable conventional system. They were
also convinced that ISOFIX is more stable and provides
more protection to a child in the event of a collision; the
only point of criticism that was frequently made was the
aspect of greater weight.
ISOFIX therefore appears to be a central element in the
improvement of future child protection systems.
1. Child restraint systems should be designed in such a
way that mistakes made when installing the seat and
restraining the child are eliminated as far as possible.
In so doing it should be borne in mind that children as
well as adults must be regarded as the persons causing
possible misuse.
2. The instructions for use should make possible dangers
and risks due to misuse perfectly clear.
A misuse assessment should incorporated as a test
criterion into future test standards (ECE, FMVSS).
4. Child restraint systems should be designed and tested
in such a way that in none of the possible positions
(sleeping position) the defined test criteria should be
5. Children should not advance from a child seat to
booster cushion at too early an age. À body length of
at least 110 cm (approx. 5 years of age) appears
6. There should be more publicity than in the past
stressing that adult seat belts, when used as the sole
restraint for children up to 150 cm (12 years of age),
can provide only limited protection.
7. Parents should be given more information about the
advantages and disadvantages of the different types of
restraint systems, reference also being made to the fact
that they may well be purchasing enhanced comfort at
the expense of safety.
8. The lateral protection of child seats must be improved.
A generally recognised and accepted side-impact test
for child restraint systems must therefore be defined
as soon as possible and incorporated into future regu-
9. The child dummies required for a side collision test
should be developed as soon as possible for the
relevant age groups.
10. A "smart" passenger airbag - which is not activated, if
a rearward facing child restraint system has to be
mounted on the passenger seat - should be imple-
mented in all new vehicles as soon as possible.
11. Belt force limiters must be studied in particular for the
effect by installing conventional CRS (greater forward
displacement of the child restraint system).
12. ISOFIX should be marketed as a standard as soon as
possible. The results of both crash tests and field
observations indicate a significant increase in the
safety protection of child restraint systems. There
have already been presentations of preliminary
systems at the International Automobile Exhibition
(IAA, 1997) in Germany.
INSTITUTE: Anlegequoten von Sicherheitsgurten
und Benutzung von Kinder-Riickhaltesystemen in
Pkw. Bergisch Gladbach, Germany
(ETSC) (1996): Seat Belts and Child Restraints -
increasing use and optimising performance.
Brussels, Belgium
(1993): How Child Protection Devices are Used in
France: Followup Surveys in 1991-1992. Child
Occupant Protection, Special Session, SAE Interna-
tional SP-986, November 6-8, page 91. INRETS -
LESCO, France
(1993): Potential Improvements to the Canadian
Child Restraint Regulations. Child Occupant Protec-
tion, Special Session, SAE International SP-986,
November 6-8, page 71. RONA Kinetics and
Accociates Ltd., Transport Canada
[5] LANGWIEDER, K., HUMMEL, TH. (1989): Chil-
dren in Cars - Their Injury Risks and the Influence of
Child Protection Systems. 12th International Techni-
cal Conference on Experimental Safety Vehicles
(ESV). Goteborg, Sweden
[6] LANGWIEDER, K., HUMMEL, TH. (1994):
Biomechanical Risk Factors for Children in Cars and
Aggravation by Misuse of Restraint Systems. XIVth
International Technical Conference on Enhanced
Safety of Vehicles (ESV). Munich, Germany
Verbesserung des Schutzes von Kindern in Pkw.
Federal German Highway Research Institute, Publi-
cation M 73. Bergisch Gladbach, Germany
ISO/WD 13216 (1995): Standardised universal
attachment to vehicle (ISOFIX). ISO/TC22/ SC
12/WG 1 Child restraint systems (in road vehicles),
Secretariat: SIS/MMS, Sweden
(1994): Vehicle Safety 90 - Analysis of Car Acci-
dents Foundations for future research work. Institute
for Vehicle Safety, GDV. Munich, Germany
Association for the Advancement of Automotive
Medicine. Des Plaines, IL
(1996): Performance of Child Restraint Systems in
Real-Life Lateral Collisions. Stapp Conference, San
HULSKENS, C. (1996): Side impact to children in
cars. Experience from international accident analysis
and safety tests. 15th ESV Conference. Melbourne,
A. H. (1992): Misuse of restraint systems for children
in the Netherlands. IRCOBI Conference. Verona,
Misuse mode and effects analysis - an approach to
predict and quantify misuse of child restraint
systems. Accident Analysis and Prevention, 25,
HUIJSKENS, C. G. (1994): Road vehicles - Child
restraint systems - Reduction of misuse risk - Form
for field studies. ISO/WD 13215-1, Secretariat:
SIS/SMS - Sweden, Paper: ISO/22/12/WG 1 N 322
ECE-R 44: Einheitliche Bedingungen für die
Genehmigung der Rückhalteeinrichtungen für Kinder
in Kraftfahrzeugen. Economic Commission for
Europe (ECE)
(1994): ISOFIX: The potential of a universal
vehicle/child restraint interface for misuse reduction
and performance enhancement. Britax International
Ltd., 38th Stapp Conference. Fort Lauderdale
MANY: Verkehr, Fachserie 8, Reihe 7, Verkehrs-
unfille. Wiesbaden, Germany
Examples of degree of damage classification (I'rontal collision)
Degree of Damage 2
e 6079
Degree of Damage 4
Degree of Damage 3
Degree of Damage 5
Severity of injury to children restrained in different restraint systems
Child, rear left side
2 years, 1 month old, male
restrained in an impact shield system
No injury
Child, rear right side
2 years, 2 months old, female,
restrained in a 4-point-belt system
Tibial fracture (due to leg impact against back
of front seat)
Group 0: up to 10 kg Group 0/0+: up to 10/13 roup 0/1: up to approx. 15 kg
о e i
Seon se
a Basan.
Tu i EN
E о О
carrycot system
infant carrier system infant carrier
rearward facing rearward facing system
Group I: 9-18 kg Group I: 9-18 kg Group 1/11: 9-25 kg
‚ .
/5-point-belt system
impact shield system
Group II: 15-25kg Group II/IIT: 15-36 kg
e За
vy i) dr AE
ao AO O
impact shield system
ooster cushion
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF