View accepted manuscript - NRC Publications Archive

Transcription

View accepted manuscript - NRC Publications Archive
NRC Publications Archive
Archives des publications du CNRC
Tornado Damage in the Barrie/Orangeville Area, Ontario, May 1985
Allen, D. E.
This publication could be one of several versions: author’s original, accepted manuscript or the publisher’s version. /
La version de cette publication peut être l’une des suivantes : la version prépublication de l’auteur, la version
acceptée du manuscrit ou la version de l’éditeur.
Publisher’s version / Version de l'éditeur:
Building Research Note, 240, p. 23, 1986-01
NRC Publications Record / Notice d'Archives des publications de CNRC:
http://nparc.cisti-icist.nrc-cnrc.gc.ca/npsi/ctrl?lang=en
http://nparc.cisti-icist.nrc-cnrc.gc.ca/npsi/ctrl?lang=fr
Access and use of this website and the material on it are subject to the Terms and Conditions set forth at
http://nparc.cisti-icist.nrc-cnrc.gc.ca/npsi/jsp/nparc_cp.jsp?lang=en
READ THESE TERMS AND CONDITIONS CAREFULLY BEFORE USING THIS WEBSITE.
L’accès à ce site Web et l’utilisation de son contenu sont assujettis aux conditions présentées dans le site
http://nparc.cisti-icist.nrc-cnrc.gc.ca/npsi/jsp/nparc_cp.jsp?lang=fr
LISEZ CES CONDITIONS ATTENTIVEMENT AVANT D’UTILISER CE SITE WEB.
Contact us / Contactez nous: nparc.cisti@nrc-cnrc.gc.ca.
Ser
m
B82
no. 240
c. 2
B
m
G
National Research
Council Canada
Consdl nafional
de rechemhes Canada
Institute for
Research in
Construction
lnstitut de
recherche en
construction
Building Research Note
Tornado Damage in the BarrieIOrangeville Area,
Ontario, May 1985
by D.E, Allen
BRN 240
ANALYZED
TORNADO DAMAGE IN THE BARRIE/ORANGEVILLE AREA,
ONTARIO, MAY 1985
by D.E. Allen
Building Structures Section
Institute for Research in Construction
BRN 240
ISSN 0701-5232
Ottawa, January 1986
@National Research Council Canada 19&6
ServCng the comtr~1ction
industty
IRC
Au service de la construction
TABLE- OF CONTENTS
DAMAGE IN GRAND VALLEY
DAMAGE TO MONO SEOPPING PLAZA
DAMAGE IN BARRXE
Kesidential Buildings
Non-Residential Buildings
ESTIMATE OF HAXIHUM WIND SPEEDS
H
W
SAFETY IKFLICATIOHS OF THE TQIWADO DAMAGE
Residential Buildings
Non-Residential Buildtngs
NATIONAL BUILDING
REFUGE AREAS
APPENDIX A
CODE
OF CANADA
TORNADO DAMAGE
IN THE BARRIE/OBAPIGEVILLE AREA, ONTARIO, MAY 1985
by
D,E, Allen
ABSTRACT
The b u i l d i n g damage caused by a serles of"iornadoes in the
~arrielorangevillearea on 31 May 1985 is described and related to the human
casualties that occurred. The damage suggests certain changes are needed in
Canadian building practice, and the National Buildfng Code of Canada. Safe
refuge areas during a tornado are discussed,
Ce texte dgcrit l e s dommages causgs aux bstiments par la sgrie de
tornades qui s'est abattue sur la rsgion Barrie-Orangeville le 31 mat 1985
et trace le rapport entre ces domuages et l e s v&ccimes. L'elsamen des
domages occasionn€is f a l t ressortir la 116cessit6 d'apparter certains
changernents aux pratiques de construction canadiennes et au Code national du
bBtlment du Canada. k texte comprend aussi une Stude des endroits qui sont
sgcuritaixes pendant l e s tornades.
INTRODUCTION
A series sf tornadoes followed parallel paths approximately 100 m wide
for about 80 km in an Em direction (see Fig. 1) between 4:15 and 5 p.m. on
31 Hay 1985, caus3ng severe damage in the mun~cipalitiesof Grand Valley,
Tottenham and Barrie and fn the farm areas nearby. Thelve people d i e d ;
rhree suffered very serious i n j u r i e s (3.e. they w e r e hospitalized, but d i d
n o t completely recover); approximately f i f t y suffered serious injuries (5.e.
they spent at least one day in hospital, but recovered) and approximtely
one hundred fifty suffered minor i n j u r i e s (it@. they spent less rhan one day
i n hospital).
More than $100 million damage occurred to more than 1000
residential and noa-residential buildings. Most of the casualties and
damage occurred in Barrie, a city w i t h a population of 45,000. A major
power l i n e was knocked out during the early stages of the tornadoes
resulting I n a blackout and loss of comunicatlon fn Sarrie about 15 minutes
before the arrival of the tornado. From the pattern of damage, rdaxfmum w i n d
speeds were estimated to be of the order of 200 Icm/h.
Figure 1.
Damage tracks of the Hay 31, 1985 tornadoes (courtesy Climatic
Perspectives, Environment Canada)
This nare briefly d e s c r i b e s t h e damage that -occurred, the associated
human casualties, and indicates what this damage .implies f a r existing
structural requirements of buildings. A more detailed study of human s a f e t y
i n s - i d e the buildings damaged by the tornadoes is being carried our by t h e
Ontario Ministry of Health (HOH) with cooperation from DBK/NRCC.
DAMAGE IN GRAND VALLEY
The tornado travelled through
Grand Valley along Amaranth Street (see
Fig. 2), severely damaging about 30 buildings on the north side of the
street and about ten on the south side. Two people died, one inside a house
and the ather in a truck, and apptoldmarely six suffered serious injuries.
& s t of the b u i l d i n g s on Amaranth Street are relatively o l d , some dating
back t b the beginnfng of the century. The following describes t y p i c a l
damage in more detail.
One house, recently built, was completaly lifted off its basement wall,
k i l l i n g one person inside and injuring another. Figure 3 shows a front
opening in the basement which allowed wind t o enter the house, helping t o
l i f t off the floor. Figure 4 shows lack of anchorage of t h e flqor in t h i s
house t o the basement foundation w a l l .
F i p r e 2.
Grand Valley.
Star)
House damage on Amaranth Street (courtesy Toronto
Figure 3.
Grand Valley.
Ebuse lifted off basement walls.
Figure 4.
Grand Valley. Same house
as F i g . 3.
Note lack of
nailing of floor into
sill plate
Figure 5 shows an older wood-fram house which racked as a result of
openings in t h e front and back walls and lack of crass partitions. Must of
t h e structural damage to houses c o n s i s t e d of lift-off a£ a roof (Fig. 6 ) or
No s e r i o u s injuries occurred in houses with t h i s kind of
a wall (Fig. 7).
damage.
Figure 5.
Figure 6 .
Grand Valley.
Grand Valley.
Racking collapse of f i r s t floor of house
Roofs (and part of walls) f i f t e d off brick houses
Figure 7.
(Xand Valley.
Toronto Star)
Wall b l o m off 85-year-old brick house (courtesy
FSgures 8 to i O show old brick
library, t w o church buildLngs and a
escaped from these buildings wZth a
buildings were heavy (three courses
probably failed as a result of roof
unsupported brick walls.
Figure 8.
Figure 9.
b u i l d i n g s that were demolished: a
house. Approximately seven p e o p l e
few minor i n j u r i e s at most. Ttte
of brickwork In the l i b r a r y ) and
lift-off followed by collapse of
Grand Valley.
Grand Valley.
kmolished l i b r a r y
Demolished church and church hall
Figure 11 shows damage t o a large church not directly in the path of
the tornado. Apparently the large front windows were initially broken fti by
missiles, which allowed the wind to enter and help blow o f f the roofing.
The damage at Grand Valley and the resulting casualties indicate the
need for anchoring light frame houses t o the foundation and for providing
Figure 20.
Figure 11.
Grand Valley.
Severely damaged brick house
Grand Valley. Church windows on right were probably broken by
missiles which allowed d u d pressure inside to help lift off
cladding
sufficient racking resfstance. Old brick buildtngs appear to be fafrly safe
provided p e o p l e go to places of refuge, such as a basement or a small room
~ I the
I
ground floor.
DAMAGE IN TOTTENRAM
The tornado followed a concession l i n e just south of Tottenham,
severely damaging approximately 20 houses, many of them recently b u i l t . Two
people died, one inside a house and one i n s i d e a shed, and approximately six
suffered serious fnjurfes.
Figure 12 show the anchorage detail of a house w h i c h was lifted off
the basement foundation wall, killing one person inside. Although anchorage
bolts were provided i n the foundation walls, they were not very effective
because the washers required for oversized bolt holes were not installed.
Five other ltght houses were completely l i f t e d off the foundations,
resulting in one very serious injury.
A number of houses were severely damaged with everything lifted o f f
except the floor and part of the walls.
Most of the structural house damage
consisted of l i f t a ff of a roof or wall. A silo with no foundation,
approximately one-third f u l l , overturned.
Figure 12.
Tattenham. &use lifted
off basement wall. Mote
lack of washers below
nuts of anchor b o l t s
DAMAGE M HONO SHOPPING P M A
A large shopping plaza of block-all
and steel construction, situated
north of Orangeville, was severely damaged as the tornado travelled from
Grand Val1 ey to Tot t enham,
The south s i d e of the shopping centre, which contained a large building
supplies store, collapsed completely {Fig. 13) whereas the north stde
suffered nonstructural damage or collapse of the front and back block walls
only (Figs. 1 4 , 15). There are two reasons for t h e difference Ln damage:
(i) the centre of the tornado, characteristically about 100 m wide, passed
through the south side of the shopping plaza;
(ii) the north s i d e of the plaza contained smaller stores w i t h Interior
steel columns anchored into the foundation p l u s infill block walls
which provided racklng resistance. The building s u p p l i e s store, on the
other hand, contained larger open areas with f m r steel columns
anchored into the ground and fewer i n f i l l block walls,
None of the 40 or so people inside the plaza during the t o m a d o w a s
killed, and only one was seriously injured. The number of fatalities waslow, partly because of the presence of stored goods ( F i g . 16) which
prevented the collapsing structure from reachlng the floor. Some exterior
block walls, which d i d collapse inwards (Figs. 13 and 15), would have been a
serious hazard t o anyone standing nearby.
A hause near the shopping plaza lost everything except the f l o o r and a
portion of the w a l l s , but the two blind people inside escaped sertous
injury. Nearby houses in the path of the tornado were undamaged.
F5gure 13.
Mono Shopping Plaza. South side (bottom h a l f o f photo)
collapsed. Note inward collapse of block wall (courtesy
Identification U n i t , Ontario Provincial Police, Barrie)
Figure 14.
Figure
15,
&no Shopping Plaza.
East wall {leeward)* -age
to block wall
near roof l i n e at north end. C o l l a p s e of south end
mno Shdpping Plaza.
b r t h side. Windward block wall blown in
after roofing lifted oEf. Interior steel structure plus lafill
block w a l l s prevented further collapse
Figure 16.
Mona Shopping Plaza. South sfde. Stored g w d s protected
inhabitants from collapsing structure
DAMAGE IN BARRIE
The Barrie tornado followed a northeasterly path 100 to 200 m wide
through the south end of the c i t y as indicated in Fig. 17, destroying many
buildings. In all, 8 people died, approximately 33 had serious injuries and
approxlraaSely 120 had minor injuries. M the 857 buildfngs that were
damaged, 237 nust be t o t a l l y reconstructeds An aerial v i e w of typical house
damage is shown in Fig. 18. The following descrgbes the damage in more
detail.
Figure 17.
Barrie.
Tornado .path
Figure 18.
Barrie.
and Mail )
&use
damage on Murray Street (courtesy Toronto Globe
R e s i d e n t i a l Buildings
Fifteen llght f r a w houses were pieked off t h e i r f oundatiods and
deposited downwind killing five persow, very seriously injuring one, and
sertously injuring six. As Figs, 19 and 20 show these houses were not
effectively anchored t o t h e i r foundations. A l l but one of these houses w e r e
o u t s i d e the old city liaits and not subject to the building code uhen they
were b u i l t -
Approximately 50 houses or apartment units suffered heavy damage, the
tornado removing the roof and m a t of the walls; eleven serious injuries
were sustained, none very s e r l w s . Typical cases are shown fn F.ig. 18,
Figure 19,
Barrie. Eouse lifted o f f
foundation. Note lack of
anchorage
Figure 20.
hrrie. House l i f t e d
off faundatlw. Note
weak anchorage d e t d 1
Seven of the injuries occurred inside row houses (shown in Fig. 21).
Approximately 100 houses or apartment units lost a roof, or a roof and part
of a wall; at most, a few of the occupants suffered serious ffijuries.
Non-Residential Buildings
+proximately 10 n o n ~ e s i d e n t t a lbuildings conststing of unreinforced
block walls supporting timber or steel-truss roofs collapsed, resulting in
one death and a few serious i n j u r i e s . Figure 22 shows a typical example*
Approximately five nrrn-residential large-rea buildings consisting of
unreinforced exterior block walls, plus an intertar steel frame supportfng a
steel-joist roof, collapsed; a few serious fnjurles at most were sustained.
In most of these buildings the s t e e l structure collapsed w t o objects such
as storage r a c k located i n s i d e the buildings (Fig. 23). A few allateta1
buildings, although suffering considerable damage t o the cladding, did not
collapse sufficiently to injure anyone inside* A l l the collapsed
non-resfdential buildings were industrial. Fortunately, most of the workers
had left before the tornado struck at 5 p .m
Figure 21.
Barrie. Severely damaged row houses on Adelaide Street
(courtesy Canapress Photo Service)
Figure 22.
Barrie. a l l a p s e d industrial building made of unreinforced
block walls supporting a truss roof
Figure 23.
8arrie. Collapse of storage building consisting of exterior
unrainEorced block walls and interior steel structure. Note
lateral failure of steel columns
ESTIMATE OF HAXIMUM WIND SPEEDS
What wind speeds are necessary to cause failure of t y p i c a l small houses
and large block-wall buildings? Calculations in Appendix I indicate t h a t ,
assuming, as often happens, that internal pressure is created by failure of
windows and doors facing the tornado, roof lift-off for both types of
buildings occurs at roughly 130 h / h . This speed is usually less than that
required t o overturn a small house (150 km/h, and greater if the house is
anchored).
Once a house roof f l i e s o f f , however, the unsupported walls will
start to collapse or pull off, making the partly demolished house lighter
and less resistant t o overturning or floor l f f t w f f , As the hause
disintegrates in t h i s way, the anchorage of the floor t o the foundation
becomes increasingly important to anyone on the floor. fhce the roof of a
large black-all building l i f t s off (see incipient failure In Figure ( 1 4 ) ) ,
the unsupported external walls will collapse. If these block walls also
provided raclring resistance t o laterdl loads, the unsupported internal steel
structure will f a i l laterally, and w i l l c o m e down where it was previously
supported by the block w a l l s .
The BarrielWangeville tornado damage indicates that:
(i) unreinforced block-wall butldiags f a i l e d ,
(ti) house roofs l i f t e d ,
(iii) house roofs and w a l l s flew away, and
(iv) wood-frame houses {including floors) that were not w e l l anchored t o
the foundations w e r e swept away.
-
This damage severity is typical for strong tornadoes In Eastern Canada (1).
A review of the failure w5nd speeds in Appendix I indicates that the maximum
tornado w i n d speeds were of the order of 160 km/h, posszbly up t o 200 h l h .
Also the fact that cars were overturned and became airborne, which occurs at
wind speeds of approximately 180-220 h / h , indicates that maxinnrm wind
speeds were of the order of 200 kdh.
HUHAN SAFETY IMPLICATIONS OF TIZE TORNADO DAMBGE
A survey of the damage and the human casualties indicates that htman
of this intensity, wRich is the
s a f e t y Inside buildings exposed t o tornadoes
strongest entensity typical for Rastern Canada, depends very much on
construction d e t a i l s , e s p e c i a l l y with regard to anchorage and ties, and on
the location of people i n s i d e buildings.
Table 1 glves an estimte of the human casualties and the number of
building u n i t s associated wlth different types o f b u i l d i n g damage. More
d e t a i l e d information relatlng injuries to the location of people and type of
damage will be obtained from the MOB survey, The following observations
are, however, apparent from the information available.
Residential Buildings
Unanchored light houses without basements are a very high hazard to t h e
occupants during a tornado because t h e b u i l d i n g s , including the floor,
become airborne, causing casualtiee upon impact with the ground. It appears
TABLE 1.
Earrie/Orangeville Tornadoes:
Estimate of Casualties vs.
B u i l d i n g Damage
-
- - ..
.
- -
.
- - - -- - - .
Deaths and
Very S e r i o u s
Ikgree of B u i l d i n g Damage
'Injuries
Serious
Injuries
Number of
Building
U n i t s (1)
Residential buildings
Everything lifted o f f , including
floor
Roof and w a l l s l i f t e d off
Only roof or w a l l l i f t e d off
Industrial and comnercial buildings
Total collapse (floor remaining)
Only roof or wall collapsed
Barns, sheds, garages, etc.,
destroyed
Cars, trucks, cycles overturned
Total
Note:
4
10
-
15
50
550
(1) A u n i t corresponds, for example, to a house, an apartment or a
store.
to be safer to s i t inside a car Gfith seatbelts on, and considerably safer t o
lie down outside, flat on the ground. Racking collapse is a l s o hazardous,
but occurs much less frequently.
If only the f l o o r remains, t h i s considerably increases t h e s a f e t y of
the occupants. I£ the floor and some walls rernain, the safety is increased
even further. Table 1 indicates that in such cases t h e r i s k to human life
is 1/10 or less, of the risk for unanchored light houses.
If only the roof or a wall i s lost, the s a f e t y for p e o p l e inside is
very high, as it is if p e o p l e go to their basements. Table I indicates that
in such cases the risk to human life is 1/100 or less, of the risk for
unanchored houses.
Non-Residential Buildings
Unreinforced block walls supporting truss roofs (either steel or wood)
The hazard is less severe than for
a l s o present a danger to the occupants.
unanchored light houses because the floor remains, and the structure often
collapses onto objects before injuring people. Open-assembly areas,
however, do not provide such protection.
Large-span metal buildings in which the steel roof structure is tied to
the foundations through the columns are much safer than unreinforced block
wall b u i l d i n g s . Those with roofs supported Tnternally by steel columns with
infill block walls, and externally by unreinforced block walls, are also
safer, but there is a danger to p e o p l e l o c a t e d near the windward walls.
Open-assembly areas would provide a s p e c i a l danger i n this regard, as noted
previously.
NATIONAL BUILDING CODE OF CANADA
The p a t t e r n of damage and human casualties caused by the
~arrie/Orangevilletornadoes indicates that for residential buildings the
anchorage requirement of Clause 9.23.6 of the National Building .Code of
Canada 1985 (HBCC) is necessary and, i f properly a p p l i e d , effective for
human safety; a l l except possibly one o f the deaths and very serious
injuries inside residential buildings occurred fn eases where Clause 9.23.6
was apparently n o t s a t f s f i e d . The anchorage requirements of Part 9 of the
NBCC should, nevertheless, be reviewed for small l i g h t houses without
basements. Anchorage of the floor and w a l l s to the foundation is especially
Lnportant for structures inhabited by handicapped or e l d e r l y people. Extra
tie capacity through basement walls is also desirable around large doors
(see Fig. 3).
Racking collapse is a potential hazard for houses that lack cross
partitions and have large windows or doors in exterior walls. Most houses
have s u f f i c i e n t inherent racking resistance, but the requirements of Part 9
of the ElBCC shauld be reviewed to ensure sufficient racking resistance for
weak configurations.
The damage indicates that: the adequacy far human s a f e t y of buildings
constructed of unreinforced black w a l l s supporting the roof requtres more
attention. Buildings of t h i s type which do not contain Irrternal goods or
equipment t o protect the people i n s i d e , open-assembly buildings in
particular, provide a high r i s k t o the occupants during a tomado.
Preliminary calculations for tarnado u p l i f t indicate that such buildings
require minimm vertical reinforcing in the block walls extending into the
foundation, similar to that required for NBCC Seismic Zone 2.
Tornado wind speeds of 160-200 km/h exert loads that correspond
approximately to the ultimate resistance of buildtng structures designed
according to Part 4 of the NBCC. This indicates that existing NBCC design
requLrementg for wind loads are satisfactory provided (1) internal pressure
is d e t e d n e d asswlng failure of windows and doors facing the wind (for
u p l i f t ) , (2) the safety factors are applied correctly using the limitstatesdesign method (dead-load factor of 0.85 for u p l i f t and overturning).
The only major deficiency with Part 4 is the application oE the empirical
design rules contained in CAN3-S305+484(2) t o block-all buildings used for
assembly occupancfes ,
REFUGE AREAS
I
Basements provide a safe refuge f o r people during a tornado. When a
tornado approaches, p e o p l e upstairs should go downstairs, Znto the basement,
if possible. S m a l l rooms, preferably without: exterior windows, are a l s o
good refuge areas provided the floor remains on the foundation; t h i s is
especially important: for handicapped and elderly people. For people inside
houses or cottages that are not anchored into the foundation and that do not
have basements, lying outside on t h e ground appears to be b e t t e r than
staying inside, but advic.e on where ro go once outside is needed.
ACKNOWLEDGEMENT
The author would l i k e to thank Messrs. Chris Spanis, Ernie frord and
B i l l k f t h e a d of the City of Barria, Mr. Peter Oomea of the County of
Dufferin and Mr, Peter Robertson of the T m s h i p of Tecumeth f o r providing
information on the extent of building damage caused by the tornadoes.
REFERENCES
(1)
Allen, D.E.
Tarnado Damage at Blue Sea Iake and Nicabong, Qaebec, July
1984, Building Research Note Elo. 222, National Research Council of
Canada, O t t a w a , ORtario,'November 1984.
( 2)
CSA Standard CAM3-S304-M34.
Masonry Design for Buildings, Canadian
Standards Association, Rexdale, Oatario, 1984.
APPENDIX A
Wind Speeds Required to Cause U p l i f t or Overrurnfng
The fallowing equations were developed in Appendix A of Ref. (1).
where:
V is the wind speed (kmJh) and q is the velocity pressure (kPa).
Roof Uplift (assumes an internal pressure coefficient of +O.5)
where:
r
e
is the weight of the roof (kPa), P
and W is the span of roof j o i s t
u
capacity f@jm)
is the anchor
Overturning
%lag
Qo =
where: %ldg
wall (m).
1
+
+
1.2
'anchor
/w
(512
W
is the weight of the building (kPa) and B is the height of the
Estimated Minimum Weights
Roof
house
block wall industrial building
House
whole
partly demolished
Estimated Anchorage Resistance
Toenailing
(assumes two toenails resist 0.8 kN u p l i f t )
Anchorage t o uureinforced block walls
(assumes tensile bond strength of 0.1 MPa)
Wgnd Speeds Required to Cause Roof Uplift
House
Block building
0.5 kPa
0.8 W a
Wind Speeds Required t o Cause Overturning
-11
house
no f oundarion anchor
foundation anchor
toenails weak l i n k
block wall weak link
Partly demolished bouse
no foundation anchor
foundation anchor
toenails weak l i n k
b,lock w a l l weak l i n k
5
3
0.8