The Severity of Bus Rollover Accidents
The Severity of Bus Rollover Accidents
The Severity of Bus Rollover Accidents
Mátyás Matolcsy
Scientific Society of Mechanical Engineers
Hungary
Paper Number: 07 0989
Matolcsy 1.
• Projection. Due to the uncontrolled movement d) Serious rollover. More than 2 rotations. The
of the occupants inside the bus, their body im- level difference between the road and the
pacts the structural parts of the passenger com- ground, where the bus finally stops is more
partment. than 10 m.
• Complete ejection. During the rollover proc- e) Combined rollover. The rollover is followed
ess the occupants could be ejected through the by a fire, or before the rollover a severe frontal
broken or fallen windows and crushed by the collision occurred, or after the rollover the bus
rolling bus. falls into a river or lake, etc.
• Partial ejection. During the rollover process Sometimes category “b” (turn into a ditch) is listed
parts of the passenger’s body come contact either in category “a”, or category “c”.
with outside surface and can be strongly Categories “a”, “b” and “c” may belong to the pro-
scratched or parts of the body (head, arms, tectable rollover accidents (PRA) and it is a realistic
chest) get under window column or waist rail public demand to assure high level survival prob-
and are pressed by it. ability for the bus occupants in these kinds of roll-
over. One of the most important requirements is
that in PRA-s the bus superstructure shall have cer-
tain strength to avoid its collapse or large scale de-
formation, to avoid the intrusion type casualties. It
has to be mentioned that the 2 rotations and the 10
m level difference in category “c” are not theoreti-
cal, but practical figures. There were more real ac-
cidents (as well as full scale rollover tests) validat-
ing these figures. It is important to emphasize that
the approval test specified in R.66 can assure an
appropriate strength for the superstructure to sur-
vive this type of rollover.
Table 1.
Summary of rollover statistics
Matolcsy 2.
Footnotes to Table I. Remarks to Table 3.
(1) countries may be involved as manufacturer, ap- • This statistics is projected by the Hungarian
proval authority, operator or the scene of the acci- media. It means that the Hungarian figures are
dent. almost complete (90-95%), so it may be said
(2) in the media reports this category is called: minibus, mi- that it is a representative sample from Hungary.
crobus, small bus, midi bus, club bus, ambulance bus,
• Assuming a proportional figure in Europe,
etc. without exact specification
(3) in many cases children, students were transported by based on the fleet sizes of buses (18.000 in
normal coaches, these accidents are counted as coach Hungary and 500-550 thousand in Europe) the
accidents. estimated number of the rollover accidents in
(4) serious deformation means the damage of the survival
Europe could be in the range of 380-480 roll-
space, (the collapse of the superstructure obviously be-
longs to this category). overs/year. If so, the European figures in this
(5) slight deformation means that the survival space very statistics cover only 2-4% of the total, which is
likely is not damaged in the rollover accident. not representative sample. It may be said that it
is a useful, usable signal from Europe.
The last presentation shown and analysed on the • The rollovers outside Europe may be used as
last EAEC Congress in Belgrade (2005) was based individual information, but they can be in-
on 222 rollover accident happened worldwide re- volved into the statistical evaluation of certain
ported by the Hungarian media [3]. Meantime this questions, special aspects.
statistics has been increased, the new version con- Table 4. summarizes the number and the rate of
tains already 338 accidents. Table 1. gives a sum- PRA-s in this statistics.
mary of this statistics analysing the 338 accidents
from different point of views. Table 4.
The rate of PRA-s in the regions.
Table 2.
Rollovers in three major regions. Regions All rollover PRA
Regions Before 2001- 2004- Total accidents number %
2001 2003 2006(3) Hungary 94 88 94%
Europe(1) 91 58 64%
Hungary 10 39 45 94 World(2)
Europe(1) 30 29 32 91 153 45 29%
World(2) 18 59 76 153 Total 338 191 57%
(1) without Hungary
Total 58 127 153 338 (2) without Europe
(1) without Hungary
(2) without Europe Remarks to Table 4:
(3) only the first 9 months in this year
• In Hungary the 94% of the rollover accidents
Table 2 shows the distribution of these accidents belong to PRA (No big mountains, precipices,
among three interesting regions. all rollover accidents are reported even if there
It is interesting to mention that the rates of the acci- was no fatality, no serious injury, etc.) As it
dent types (their severity) in this statistics strongly was said before, this statistics is representative,
depend on the region of the accident. An example: a related to Hungary
“turn on side” of a minibus without fatalities is re- • Related to Europe, this rate is 64% but it is ob-
ported by the Hungarian media only if it happened vious that the Hungarian media do not report
in Hungary, but it is not news if it happened in Bra- the less severe rollover accidents from Europe.
silia or China. This is proved by Table 3. The con- • Considering countries having more and bigger
clusion of this effect is that the more severe rollover mountains, too, the estimated rate of PRA-s is
accidents are over-represented in this accident sta- between these two values, probable closer to
tistics considering the whole world. the Hungarian one. It seems to be a reasonable
estimation that 80-85% - as an European aver-
Table 3. age – of the rollover accidents belong to PRA.
The rates of accident types in the regions • In other words, if we can provide high level
probability of survival and reduce the casualty
Regions Turn Rollover Serious Comb. risk in PRA-s, the passenger protection will be
ed on from the roll- roll- total
side road over over significantly increased in rollover accidents of
Hungary 45 43 0 6 94 buses.
48% 46% 0% 6% 100%
Europe(1) 18 40 13 20 91 THE ROLLOVER PROCESS
20% 44% 14% 22% 100%
World(2) 2 43 59 49 153
1% 28% 39% 32% 100% It is important to see clearly the rollover process,
Total 65 126 72 75 338 the factors influencing this process and to under-
19% 37% 21% 23% 100% stand the problem of severity in case of rollover.
(1) without Hungary (2) without Europe
Matolcsy 3.
friction coefficient bigger turning moment. The
Start of the rollover friction coefficient, more exactly the reaction force
(FR) could be increased by certain circumstances
The start of the rollover process mechanically is (see Figure 3.)
simply and more or less similar in all accidents. A
turning moment (M) starts the process (see Fig.2.)
which may be generated on two ways:
Matolcsy 4.
the motion of the bus (sliding or rotation) will When studying the further motion of the bus in a
continue. rollover accident, it has to be recognized that the
The further motion of the bus depends on the sur- surroundings and structural stiffness have common
roundings ( general geometry of the scene of the effects, too. [3]
accident, soil properties, locality of the ground,
etc.) and on the properties of the bus (shape, CG
position, stiffness of the superstructure, etc.) Let us
consider the two essential influences.
Matolcsy 5.
this statistics does not give direct information about • On the basis of these statistical data it may be
the efficiency of the approval of buses regarding said that the casualty risk of intrusions can be
ECE-Rg.66. Very few information are available, drastically reduced by the requirement of the
whether the bus having a rollover accident was ap- intact SS, by the required strength of the super-
proved on the basis of R.66 or not. But indirectly an structure.
interesting comparison may be done. As it was de- • It is interesting to mention - on the basis of Ta-
fined above, PRA-s cover those accidents in which ble 5. - that the slight injury rates are not
the passengers should be protected, the survival closely related to the kind of rollover groups. It
space (SS) shall be maintained. It has to be under- may be assumed that this type of injuries are
line that the required strength of the superstructure caused mainly by projection (the inside colli-
helps to avoid the intrusion type injuries, to reduce sion of the passengers) when they are leaving
drastically this type of fatalities, but it is less effec- their seats, seating position during the rollover
tive in the projection and ejection type injuries. process. The main tool to reduce this kind of
Among the 388 rollover accidents there are 191 injuries could be the use of seat belts. (It has to
PRA-s and among these accidents there are 142 in be emphasized that the seat belt can reduce the
which we have information about the behaviour of number of fatalities and serious injuries, too,
the superstructure: 82 accidents did not cause dam- and also the ejection of the passengers.)
age in the SS and in 60 accidents the SS was When starting to work with R.66 (in the mid of
harmed, including the total collapse, too. An inter- ‘70s) one of the most important and long discussed
esting comparison is shown in Table 5., in which question was to find on appropriate standard ap-
the casualty rates (casualty per accident, CR) are proval rollover test. At that time there was no clear
given for four kinds of rollover accident groups: idea about the PRA-s, but there was a demand for a
– All the 388 accidents giving a very general aver- “good” approval test which separates the strong su-
age perstructures from the weak ones. Figure 7. shows
– PRA-s in which the passengers should be pro- three kind of rollover tests used in Hungary.
tected
– PRA-s in which the SS remained intact (studying
the pictures, photos, videos available)
– PRA-s in which the SS damaged, the superstruc-
ture collapsed.
Table 5.
Casualty rates in rollover accidents
Matolcsy 6.
week superstructure
reinforced superstructure
Figure 10. Comparison with test “c”
Matolcsy 7.
a slope given in Figure 12. The level difference was shown on Figure 15. After 3 ¼ rotations – the level
around 9-10 m, the number of rotation 2 ¼ and after difference was 17-18 m – the survival space re-
this accident no significant deformation could be mained intact, the intrusions were avoided. Nine
observed on the superstructure. [5] (see Figure 13.) dummies were used in this test, 7 of them had 3pts
safety belt, 2 of them were without belt. The belted
dummies remained in their seats, (no projection
type injury) but the two unbelted dummies flew in
the passenger compartment and had untraceable
motion. According to our definition, this rollover
accident is out of the PRA group (more than 2 rota-
tions, more than 10 m level difference) it belongs to
the severe rollovers. But having the required
strength of superstructure and wearing seat belt, the
survival probability of the occupants is strongly in-
creased even in severe rollover accidents, too.
VO’
Matolcsy 8.
TWO EXAMPLES driver did not recognize the situation, it was gloom,
night.) After uncontrolled manoeuvre the bus
Thinking about the severity of rollover accidents, it turned on its side, slipped away on the double-way
could be interesting to study in details the following roundabout 20-25 m and hit the other side of a ditch
two accidents. next to the roundabout. (see Figure18.) The roof
structure completely collapsed as it may be seen on
Figure19. The result: 20 fatalities, 17 serious inju-
ries and 14 slight injuries. [7] The tip over (turned
on side) is the less severe rollover based on the 2nd
approach. But the first approach says, it is a very
severe accident. But if the superstructure should
have had the required strength, both approaches
could say that this is not a severe accident. The
public opinion says: it is unacceptable that in a
similar accident (tip over) the casualty rates are so
high. And that is the goal of the international regu-
latory work: to increase the safety, to avoid this
Figure 16. The path of the rollover kind of results in PRA-s.
Switzerland, Grand St.Bernard Pass, 17.04.2005
CONCLUSIONS
Matolcsy 9.
• The severity of the rollover should be defined
on the rollover process itself and not on the
measure (number) of the casualty figures.
• The survival space concept and the belonging
existing requirements are very effective. Statis-
tical data prove that the all casualty rate is 3 - 4
times lower, the fatality rate is lower with one
order (10 times) when the survival space re-
mains intact in a PRA.
• There are four important injury mechanisms
which should be considered enhancing the pas-
senger safety in rollover. The most dangerous
one is the intrusion, when due to the large scale
structural deformation structural parts intrude
into the passenger, or compress them (lack of
the strength of superstructure)
REFERENCES
Matolcsy 10.