Академический Документы
Профессиональный Документы
Культура Документы
BQII 0
British CementAssociation
0
z
11
m
m
z
C,
m
Concrete car parks -
Design and maintenance
issues
m
'C,,
Conference held at
The Cavendish Centre
London
29 September 1997
This document
contains 104 pages
OFC
CONCRETE CAR PARKS
Design and maintenance issues
Conference held at the Cavendish Centre on 29 September 1997
This BCA conference was held to bring together the expertise of those who
have been investigating the collapse earlier this year of the Pipers Row car
park, Wolverhampton, and the apparent problems with similar structures, for
examplethe Castle Streetcar park in Banbury.
There is a significantnumber of flat slab car parks in the UK. The conference
was of interestto those involved with investigating, assessing and maintaining
these structures, including the owners, particularly public authorities. The
speakers were selected for their specialist knowledge of the variety of issues
discussed.
For further information and up-to-date literature on this and related subjects, contact the
BCA'sCentrefor Concrete Information on (01344)762676.
M21
''
PubEisheu 4OO7
ISBN0 7210 15239
Published by the British Cement Association
CenturyHouse,TelfordAvenue
Crowthome
BerkshireRG456YS
Tel (01344)762676
Price60.00 Fax (01344) 761214
E-mail: cement @bca.org.uk
C British Cement Association
The opinions expressed and the representations made by the speakers at BCA events are their own
and are not those of the BCA itself. The BCA accepts no liability(includingthat for negligence) for any
loss resultingfrom such opinions,representationsor otherstatements.
IFC
Papers
Keynote address - the challenge for Dr John Menzies
the future SecretaryofSCOSS - theStanding
Committee on Structural Safety
* Precast
multi-storey car parks Paul Bracey
Composite StructuresLtd
INTRODUCTION
There are over 4000 multi-storey car parks in the UnitedKingdom. Almost all have been built
since 1940 and the vast majority are of concreteconstruction. Over the years the standardsof
design and construction have improved following adverse experiences ofperformance ofsome
early structures.
The Standing Committee on Structural Safety (SCOSS) has had concerns about these
structures for more than five years, and has made recommendations for action in its two most
recent reports'2.
A boom in multi-storey car park construction occurred in the I960s. The structures were
usuallyfree standing, the internal space being devoted entirely to accommodating parked cars
and to their access and egress. Recently, multi-storey car parks have more often been
incorporated into large multi-purpose office and shopping complexes, their structure being an
integral part ofthe whole development.
A large number of multi-storey car parks have been erected with floors of in-situ or precast
concrete suspended on structural members of reinforced or prestressed concrete or of
structural steelwork. They are generally owned by local authorities, owners of shopping
developments or offices, private car park operators, transport authorities, or hotels. Because
of the choices available in layout and structural system many have been built on the basis of
competitive 'design and construct' proposals. Many were designed to the lowest first cost per
parking bay with the optimistic assumption that operators would provideappropriatecare and
maintenance.
Concretemulti-storey car parks are usually open to the weather, the large shaded deck areas
are often damp, and water contaminated with road salts and oil drips onto the horizontal
surfaces. Although road salt is only likely to be applied to ramps and top deck surfaces,
experience has shown that significant chlorides are carried in and deposited on multi-storey car
park decks by vehicles when roads have been salted. Drainage of contaminated surface water
is often poor, especially in the older structures. Car parks are thereforegenerally subjected to
an aggressive environment.
1
Concrete car parks have generally been safe structurally but the performance record of the
early low cost bare utilitarian structures has too often been marred by poor durability and
some incidents of loss of safety have occurred. They have often received a bad press both
popularand technical. The reasons are not hard to find:
The general public who perforce have to use them, do not like them because they often
provide a dark, dirty and forbidding environment. They are subject to vandalism and are
perceived, not withoutcause, to be rather dangerous places.
The boom in multi-storey car park construction in the 1960s led to a spate of remedial
actions in the 1970s and 1980s as structural distress was manifest by deflections, cracks,
vibrations and spalling of reinforced concrete. These and other defects were attributed at
the time to inadequate design and detailing, lack of adequate concrete cover to steel
reinforcement due to poor design and construction, little provision of waterproofing to
exposed surfaces, and inadequate maintenance often amounting to neglect.
Car park owners and operators are believed to be obligated under law to provideand maintain
car parks that are safe for their employees and users. However, SCOSS cannot be certain that
there is no gap in the regulatoryframework relating to the safety of existing multi-storey car
park structures and has drawnthe government's attention to this point.
Compared with most forms of commercial building multi-storey car parks have a number of
distinct characteristics.
The design module for clear span construction is 16 metres. This is a long span and the dead
load/live load ratio is higher than for most forms of normal concrete building structure. In
general a larger proportion ofthe live load is experienced for longerperiodsfrom the start.
Internal floors are subject to wetting and drying and the movement ofthe structuralframe due
to thermal and moisture changes requires consideration in design. This is especially important
for top decks, which are waterproofed with black surfacing materials, as they will experience
the largest thermal movement. Large areas of structure are required and movement reliefjoints
are unwelcome and if provided can fail to operate. Many have leaked or were wrongly
positioned in early concretecar parks.
2
An alternative span module was acceptable in early concrete car parks leading to shorter spans
between columns (approximately 9 metres) with cantilevers of approximately 3 metres on
either side. This was economical in depth of the deck slab and the quantity of concrete used
since the positive and negative moments in the parking deck were balanced. Many architects
also had a preference for avoiding the appearance of columns on the facade. Cantilever
performance is thereforecritical in suchdesigns.
Multi-storey car parks sometimes, as already mentioned, form part of a building with shops
below and possibly offices or flats above. This requires load transfer floors which would
change the colunm grid on non-parking floors, producing a heavy slab structure
uncharacteristic of the other parking floors which might resemble the low cost 'design and
build' multi-storey car parks. Parts of such structures may therefore be subject to powerful
movement differentials or stressesdue to the different stifThesses provided by the parkingand
non car park parts ofthe building.
The design of many concrete car parks is also complicated by the complex geometry of
ramp/circulation arrangements required to maximise access throughthe structures.
In more recent times, forms ofstructure not included in the above list, have been built. There
are a small number of concretecar parks built with bonded post-tensioned concretedecks. In
several cases insitu fl.ill span beams were post-tensioned over 20 years ago using Macalloy bar
through the precast external columns.
Partly due to relaxation of structural fire resistance requirements, more structural steel frame
multi-storey car parks with concrete floors acting compositely have been constructed.
Following developments ofthe technology and its general wider use, unbonded post-tensioned
flat slabs have also been used in more recent construction.
Deck structures
Premature corrosion of reinforced concrete decks has generally been due to lack of or poor
quality of top cover concrete in the deck, surface cracking and exposure to aggressive
environmental conditions including road salts. Generally there is adequate warning of
corrosionthrough the expansion ofcorrosionproductsand spalling of concrete. Suchwarning
may not howeverbe generated in some cases, especially wherepreviousrepairs of a cosmetic
nature have been undertaken. From the structural safety point of view, the critical feature is
likely to be the combination ofcantilever action and high shear at internal columns supporting
a flat slab approximately 3m from the perimeter ofthe structure.
The tip of the cantilever sometimes supports heavy parapet walls which may also serve as an
edge beam acting as a load distributor along the full length of the cantilever. The redundancy
inherent in such a system may ensure that substantial local strength losses produce no serious
immediate results. On the other hand, any consistent shortcomings in cantilever design or
construction would be made more serious. It has been known for incorrect or badly placed
reinforcement in the negative moment zone of the cantilever to have been detected after
construction and for additional reinforcement to be inserted and bonded in epoxy mortar to
compensate. In another case the deflection of the cantilever became excessive progressively
over 25 years and the corrosion of the top steel became so advanced that the structure was
demolished.
3
For early structures incorporating post-tensioned concrete components, the risks of tendon
corrosionand structuralfailure withoutwarning should be recognised similarlyas for pre-1992
post-tensioned bridge decks although the risk is probably less and a longer time may elapse
before safety isjeopardised. The continued integrity ofthe prestressing, whetherby bonded or
unbonded tendons, is dependent upon the ability of the waterproofing, the concrete and
grouting to resist chloride migration. Chlorides are known to migrate even through good
structural concreteand there is an eventual risk ofcorrosionif road salt is carried on to floors.
This risk may be significant only in the tong-term but ingress at anchorages and joints is a
moreimmediate concern.
The multi-layer protection approach, as now recommended .in the Concrete Society Technical
Report 47 for bonded post-tensioned concretebridges, would be appropriatefor new concrete
car parks using this form of construction. Where unbonded tendons are used it is equally
important that they are protected against an aggressive environment which may include road
salt.
Columns
The column head/beam zone is perhaps the most critically stressed area and most difficult to
detail effectively to avoid cracking. In flat slab construction, code rules changed with regard to
the shear provision around column heads in 1972 and 1984. Also in 1972 provisionwas made
to encourageprovision ofmorenegative moment reinforcement at internal supports.
Failure of a column and redistribution of load to adjoining columns following vehicle impact
would perhaps be more likely to cause disproportionate collapse of a multi-storey car park
structure than in a normal concrete building structure. Clearly consideration of accidental
loading is necessary in design ofa concretecar park giventhe likelihood ofimpact from errant
vehicles. The structure should be resistant to credible impacts of this type or protected from
them.
Totally precast forms of multi-storey car park construction may include wall-typecolumns in
combination with deck units of double tee or I-beamssimilar to those used in bridges. These
systems involve stress concentrations at the deck bearings, the condition of which should be
monitored by inspections during service.
The collapse of part of the top deck of the Pipers Row car park in March this year should
serve to emphasise many ofthe technical shortcomings in the current stock of older concrete
car parks which are mentioned above. Whilst concretecar parks have proved generally to be
safe structures, unacceptable risks to safety may arise in some cases and the causes are known.
Available knowledge is now sufficient to enable existing and new concrete car parks to be
made safer (and more pleasant places) for the public to use.
4
SCOSS made recommendations in its Tenth ReportWin 1994 on inspection and appraisal of
multi-storey car parks. However, it became apparent in the succeeding period that these
recommendations were not being given appropriate priority. Accordingly, SCOSS repeated
the recommendations in its Eleventh Report2, published in January 1997, as follows:
1) Owners and operatorsofexisting multi-storey car parks should commission inspections and
structural appraisals of their structures by engineers with appropriate experience before
carrying out repairs. Appraisal should extend beyond the areas of conspicuous
deterioration, particularly where water with road salt may have penetrated.
2) Appraisals should be made periodically following the principles adopted by bridge owners.
Structuralappraisal should include a reviewofresistance to progressive collapse.
Barriers
Barriersaroundthe edges ofthe concretedecks ofmulti-storey car parks are required to serve
two functions. They should protect pedestrians, especially small children, from accidentally
falling over the edge. Theyshould also prevent errant vehicles from being driven, inadvertently
or otherwise, over the edge of the deck (then to fall to the ground) or into exterior cladding
causing it to break free and also fall to the ground below. Incidents of these types have been
reported.
Accidents involving children falling from concrete car park barriers have not come to notice
but potential accidents have been reported. Some car park barriers provide an irresistible
temptation to 4 year oldswho can easily climb them and so be placed at risk. Whilst parental
watchfulness is the main safeguard, it is clearly desirable that barriers should be designed to
prevent children climbingon them as far as is reasonably possible.
Two accidents are known in which a car was driven through a car park barrier and plungedto
the ground below. The most recent occurred in Canterbury in January 1996 seriously injuring
the driver. The existing edge barrier was inadequate to restrain the car.
Existing edge barriers should be evaluated. Criteria and methods need to be established and
agreed on the basis of appropriate standards, and options for remedial action defined. Some
forms ofmetal edge barrier are susceptible to brittlefailure on impact and need to be replaced
or strengthened to ensure a ductile mode offailure at a higherload. Fixings can be particularly
vulnerable. Since barriers are likely to receive vehicle impact damage, they should also be
designed so that damaged sections can be easily replaced,
In 1994, SCOSS recommended action on edge barriers, and the recommendations were
extended in 1997 as follows:
5
2) The Institutions of Civil Engineers and of Structural Engineers
should urgently prepare
guidance on assessment and strengthening of existing edge barriers in multi-storey car
parks.
Technical guidance
During the boom in concrete car park construction during the 1960s, 'design and build'
contractswere frequently used based on specifications less demanding than was appropriate
for achievement of durable structures. The structural codes CP114, CP115, CP1 10 and
BS 449 were generally applied to these structures but the proportions of car parks were
unconventional compared with other building structures and their exposure conditions
ambiguous. It wasnot until 1976 that the Institutions of Structural Engineers and ofHighways
and Transportation published the first authoritative guide to multi-storey car park design but it
did not give in depthadvice on structural design.
Some of the early sub-standard structures have been demolished. In several other cases they
have been strengthened either by the use ofexternal post-tensioning, by provision ofadditional
top reinforcement, by rebuilding columns or, by the application ofbonded plate reinforcement.
Such experiences resultedin changes to the design codes and to the IStructE/IHTguide which
was updated in 1984. The guide drew attention to the risk of chloride-induced corrosion in
concrete decks. Recent concrete car park construction has been based on more up-to-date
structural codes. These car parks are likely to give better long-term structural performance.
The Institutionsare shortly to commence a revision ofthe 1984 guide. This action will provide
an excellent opportunity for improving guidance based on the extensive experience, both good
and bad, now available. It is highly desirable not only for the design and construction of
concrete car parks and barriers to be covered, but also their inspection, structural appraisal
and maintenance.
Past experience has revealed technical and other shortcomings in concretecar parks built over
the past 50 years. Generally concrete car parks have not provided attractive surroundings for
car park users or long-termlow maintenance structures for owners and operators. The lessons
learned providethe basis for achievingbetter performance and quality ofconcrete car parks in
the future.
For existing concrete car parks, some initiatives to improve the environment for users in
concrete car parks are already being taken through improvements to security and
refurbishment to provide brighter and more attractive decks and facades. At the same time
structural repairs, replacement or rebuilding are offsetting the worst effects of deterioration
and inadequacies to structures and barriers. In particular, the early structures built at minimum
cost and quality - both structuraland environmental - need to be closely watched in future.
More effort in these directions is needed backed up by standards and guidance for improved
quality. Provision ofsuch standards and guidance should be given morepriority.
For new concrete car parks, current practice and standards should be critically examined.
Compared to earlierpractice, higher quality should be specified for the users' environment and
for the structures and the barriersthemselves. Particular attention should be given to achieving
6
durable and maintainable construction using standards and codes akin to those used in
highway bridgeconstruction.
The maintenance of concrete car parks should not be an afterthought. Inspections and
structural appraisals should be scheduled and maintenance (and repair or strengthening as
necessary) carried out as needed in a manner akin to that used for highway bridges.
Repair and maintenance are crucial to the future of concrete car parks. Repairs must be
undertaken with care. Some thin concrete decks were poorly built, Removing and replacing
significant thicknesses of concrete may weaken them to an unacceptable extent. Subsequent
reinforcement corrosion may not give visible signs if cosmetic and surface treatments are
applied. Newer forms of repair, such as cathodic protection, may have a role to play.
SCOSS interest and concern about the safety of multi-storey car parks continues. The
regrettable incidents in recent years have, it is to be hoped, brought home to owners and
operators the necessity of giving continuing attention to the safety and durability of these
structures. There remains an urgent need for more pooling ofexperience to generate improved
information and guidance to assist theirtask.
Overall the challenge for the future is to make concrete car parks pleasanter and safer places
for the people using them and more durable and maintainable for owners and operators. The
knowledge and experience needed to meet this challenge now exists. This Conference provides
a timely opportunityfor the industry to pool experience and to determine the wayforward.
REFERENCES
(1) Standing Committee on Structural Safety. Tenth Report : July 1992 to June 1994,
SETO, London, October 1994.
(2) Standing Committee on Structural Safety. Structural Safety 1994-96 : Review and
Recommendations. Eleventh Report, SETO, London, January 1997.
(3) Institution of Structural Engineers and Institution of Highways and Transportation
Design recommendations for multi-storey and underground car parks. London, 2nd
edition 1984.
7
Pipers Row car park, Wolverhampton
Results of the investigation
John F Kellermann
Harris & Sutherland
INTRODUCTION
In the early hours of the morning of the 20 March 1997 part of a roof slab that formed the
seventh parking level of the Pipers Row multi-storey car park in Wolverhampton collapsed
onto the fifth parking level immediately belowit.
Immediately following the collapse, National Car Parks Ltd., who owned and operated the car
park, instructed Harris& Sutherland to visit the site and establish cause ofthe collapse.
In the time available I intend to briefly describe the Pipers Row car park, to illustrate the
nature ofthe collapse and the subsequent investigation.
It should be noted that the collapse is currentlythe subject of a Health and Safety Executive
investigation and I do notwant to prejudice this investigation in anything I say heretoday.
The Pipers Row car park was designed and built, in 1964-65, using Lift Slab System
techniques.
The Lift Slab Systeminvolves the initial erection of the columnsbefore constructing the floor
and roof slabs. The slabs are then cast one on top of theother at ground level, with structural
steel lifting collars around each column.
Following the completion of the concrete pours to floors and roof slabs, they are individually
lifted up the columns using a system of hydraulic jacks with threaded lifting bars. They are
then fixed to the columns with steel shear keys or 'wedges'. The voids between the columns
and theslabsare filled with fine aggregate concrete.
THE INVESTIGATION
When I arrived at the site on the 20 March the part of the rooflevel parking slab was resting
on the fifth level slab and mostof the columnsin thecollapsedarea werestill in place.
The slab below the roof did not fail and continued to supported its own weight and the roof
slab.
There was no sign of failure of the steelwork lifting collars that provided the connection
between the lifted slabs and the precast concretecolumns. The areas of concrete slab that
surrounded thesecollars had failed on a perimeterat/or beyond thesesteelworkcollars
Following the initial inspection of the collapse, Harris & Sutherland and the Health and Safety
Executive (HSE) carried out a detailed inspection of the collapsed area. Due to the concernas
to the stabilityof the structure all of the investigation work was carried out from a hydraulic
access platform. During the investigation areas of the intact concreteslab were identified as
being suitable for sampling for testing. These areas were marked out on the slab with paint
and were later cut outwith pneumatic picksand taken down from the structure with a crane.
Al
The concretein the collapsed area of the roof parking level slab and in particularthat around
the columns had fractured extensively and had, in places, completely disintegrated leaving
only loose aggregate.
The concrete consisted of rounded gravel coarse aggregateand pinkish colour sand. It was
noticeable that lumps of concretecould be easilybroken up by hand or with light blowsfrom a
chipping hammer. The extraction of concretesamples proved to be difficult, as the concrete
tendedto disintegrate, even at some distancefrom wherethesampleswere being extracted.
The reinforcement used in the roof parking level slab was generally square Y2" and " twisted
cold workedsteel bars.
The reinforcement in the collapsed area was generally uncorroded with a small amount of
surface rusting, although there were localised areas in which the reinforcing bars had
significant amountsof surface corrosion.
It was noted that there was very little concretebond to the reinforcement in thecollapsedarea
and that manybars had completelycome awayfrom the concrete. The reinforcing bars had a
white surface deposit on them, as did the concrete fracture planes and exposed aggregate
surfaces.
Concrete repairs had previously been carried out on the top surface of the slab around and
adjacentto columnsH2 and J2As noted earlier the ColumnHeads were all in place and whilst
some corrosion was observed on the collar steelwork, there was no indicationof any failureof
the steelworkor of its connection to the column.
At column J2 a drainagegulley was located adjacent to this column and a 6" diametercast
iron drainpipewas located in a slot in the edge of the slab. The drain gulley and drainpipehad
been installed in box-outs in the slab that had then been filled with concrete. This concrete
infill was still attached to the drainage gulley following the collapse. There were no indications
of failure in the steelworkor in its connection to the column. Therewere some signs of voids
around the collar steelworkand poor compaction of the concrete.
MATERIALSTESTING
Messers Sandberg were appointed to test and report on materials recovered from the Pipers
Row Car Park.
Due to the potentially unstable natureof the partiallycollapsed structure itwas decided that all
concrete coring would be carried out at ground level on samples cut from the structure.
These samples were cut using small pneumatic picks with the operators secured to the
access hoist by safety harnesses and lines. The samples were lowered to the ground by
crane.
A broad range of physical and chemical tests on the concrete from both the collapsed area
and the remaining areas ofthe car park. These tests included:-
Visual Examinations
Depth ofCarbonation
Compressive strength
Tensile splitting of core samples
Density of core samples
Chemical analysisof core samples
Aggregategrading
Low power microscopic examinations
Tensiletesting of reinforcing steel
Internal fracturetests
A2
Visual Examination
Visualexamination ofthecores and lumpsamplestaken from the collapsed area showed had
a distinctly pink colour imparted to it by the fine aggregate, whereas the concrete samples
from the remaining structurehad a greyish colour. It was concluded that the concrete in the
collapsed area had been mixed with differentfine aggregates.
The concrete from which it had been possible to obtain core samples was judged to have
been fairly well compacted, but there were indications that this was at least partially
attributable to a very fluid mix. It was noted that one core sample, which incorporated a 1/2"
square twisted bar, had a continuous void variously above and below that bar. Some of the
lumpsampleshad fine voids around aggregate particles that were sometimesfilled with white
material. It was further noted that two lump samples had suffered partial disintegration and
featured multiple horizontal laminations in what had been the top 80mm of the roof slab.
These laminations were old and contained whitedepositsof calcium carbonate. From a visual
examination of the samples, it was not possible to decidewhat had caused these laminations,
but it was concluded that they would have significantly reduced the strength and durabilityof
theconcreteand that theywould also have reduced the effectivedepth oftheslab.
Lump samples from the failed areas of the collapsed slab were observed to have fractured
almost exclusively around the aggregate particles indicating a lack of bond between the
cement paste and the aggregate. These samples had extensive white deposits on internal
surfaceswithin the concrete and on some of the contactsurfaceswith the reinforcing bars. It
was also noted that one core samplecontained a reinforcing bar which could be freely rotated
and moved along its axis, indicating a grievous lack of bond between the bar and the
surrounding concrete.
Depths of Carbonation
The depths of carbonation of the concrete in both the collapsed and remaining parts of the
structure varied from less than 1mm to a maximum of 30mm in both the top and bottom of the
slabs.
Compressive Strength
The estimated insitu cube strength of the concretein the intact parts of the collapsedarea of
slab varied from 18.5 to 35N/rnm2. It was not possible to take core samplesof the concrete in
the collapsed areas close to the columns, because it was tooweak and disintegrated during
coring.
The estimated insitu cube strength of the concretein the remaining, uncollapsed, areas of the
carpark varied from 16.5 to 48.5 N/mm2.
Internal fracture tests carried out on lumpsamplesfrom the collapsedarea indicated concrete
compressive strengthsas low as 8 N/mm2. Of note was that the two lowest strengthswere
from sampleswhichwere highly fragmented in what had been thetop 80mm ofthe roofslab.
The cement content for the concrete samplestaken from thecollapsedarea varied from 9.9%
to 13.1%, (223kg/rn3to 295kg/rn3). The cementcontentin samplestaken from the remaining
structure varied from 11.7% to 13.9%, (227kg/rn3to 270kg/rn3).
Chloride contents of sampleswere low and sulphate contents normal and were considered
unlikely to have contributed to the collapse.
A3
Examination of fluorescentdye impregnated samples in thin section using ultraviolet light
indicated that the sampleshad a very high unevenly distributed microporosity. This suggests
a high initial free water/cementratio, estimated to be in the region of 0.7 by comparisonto
known referencespecimens.
MESSRSSANDBERG'SCONCLUSIONS
Bleed waterfrom the high water/cement ratio concretemix accumulatedaround the aggregate
particlesand reinforcing steel during the compaction of the concrete and, afterthe concrete
had set and hardened, slowly dispersed throughout the concrete, leaving voids around the
aggregateparticlesand reinforcing bars.
It was considered that the high porosity of the concrete, with voids around the aggregate
particles and reinforcing bars would have with time allowed water to permeate into the slab
absorbing Portlandite from the concrete and, when temperatures dropped to near freezing,
depositing the Portlandite in the voids. Also when temperatures dropped below freezing, any
water in the voids would have frozen and expanded, causing damage and in particularthe
partial disintegration and laminations that wereobserved in the top ofthe slab.
DESIGN CHECK
General
A design checkon the collapsed area of the seventh parking level slab was carried out using
information taken from the original drawings, checkedwhere possible by site measurements
and observations. The slab was analyzed both as a series of simplified two-dimensional
frames,following the recommendations given in CPI 14 and using a finite elementprogramto
investigate the influence ofthe various boxouts adjacentto the column.
It can also be argued that the spacings between the groupsof tensile reinforcing bars on both
sides of the columns, in both directions, did not comply with the recommended maximum
spacing given in CPI 14, but this is a criticism that could be levelled against very many Lift
Slab System structures and it would undoubtedly be rebutted on the basis that the CPI14
recommendation does not apply to the spacingbetween separate groups of bars, only to bars
within a group. Experience has, in any case, shown that Lift Slab System structures, which do
not complywith this recommendation, behave satisfactorily.
Temperature Effects
In both the simple frame and the plate analyses, it was assumed that the temperature
differentialof ICoC was applied to an uncrackedslab and that the slab remaineduncracked.
This was done because it is very difficult, if not impossible, to take properaccountof the effect
of the cracking, which is always presentin reinforced concrete. It is however possible to say
that the cracking which existed in the slab without taking into account any temperature
differential and that which may have been caused by a temperature differential will have
A4
reduced the forces and therefore the stressesresulting from the temperature differential. The
reason for this is that the forces caused by temperature differentials are a consequence of the
restraints imposed on the slab and these will have been dissipated by anycracking.
For the above reasons, while an attemptwas made to ascertain what might be consideredas
an upper bound for the stresses that could have been caused by a temperature differential
between the top and the bottom of the roof level slab, the stresses calculated have been
ignored in checking compliance with CPI14. Nevertheless the role that cyclic stresses
caused by temperature differentials may have had in propagating cracks and/or inducing a
fatigue related typeoffailure hasto be recognized.
DISCUSSION
A part of the roof level slab collapsedon to the fifth parking level slab, which sustained the
load of the upperslab. The fifth parking level slab proved to be capableof sustaining a load of
more than twice the standard live load. The similarly designed roof level slab collapsed under
selfweight alone, indicating that the principal problem with the collapsed slab was likely to be
related to materials or workmanship rather than design. Potential differences between the
collapsed partand the remainderofthe car park that havebeen identified are:-
Laboratory tests have shown that the cementcontents in the concretefrom the collapsed zone
were as low as 223 kg/rn3. A typical targetvalue for a 3000psi concretemix would have been
about 320 kg/m3.
Correspondence relating to the construction of the car park shows that, during the
construction, the contractor changedfrom ready-mix concreteto site-mixconcrete. There are
no details available of the typeof concrete mixing equipmentthat the contractor set-up on site
orofthemix proportions.
Tests carried out by Messrs Sandberg show that the water cement ratios of the concrete in
the collapsed area varied between 0.69 and 0.84. Current engineering practice for the
production of durable concretelimits the water cement ratio to about 0.5, but higher values
were common in the 1960's. High water/cementratios produce concrete mixes which flow
more easily, but have lower strengths and lower durabilitythan mixes with low water/cement
ratios. Cement content is a significant consideration in terms of durability, thus a high
water/cement ratio arising as a function of a low cementcontentwill be significant in relation to
durability.
A5
Messrs Sandberg have also shown that the surface of some of the coarse aggregate was
coveredwith clay or silt which will have reducedthe bond between the cement paste and the
aggregate.
The estimated insitu cube strength of concretemeasured using cores cut from the intact part
of the collapsed area of slab varied from 18.5 to 35 N/mm2. The specified concrete strength
was 3000psi or 21 N/mm2. The strengthsobtained from tests are therefore comparable with
the specified cube strength and cannot of themselves account for the failure of the slab.
However, because of the total disintegration of the concretein the collapsed area of slab in
the vicinity of the columns, it was not possible to take intact concrete cores from these
locations. The core samplesthat were takenwere from the intact part ofthe collapsedarea of
slab are not thereforerepresentative of the concretethroughoutthe collapsedarea of slab.
Workmanship
Correspondence with the contractor during the construction period indicates that the
contractor was experiencing problems with the supply and quality control of the concrete.
Test resultsgiven in a letter of recorded a fine silt/clay contentof 4.25% of thetotal concrete
weight. This would be equivalent to approximately 15% of the sand content by weight. This
would have significantly reduced the strength and durability of the concrete. The
correspondence also recorded in the same letter, that all of this "doubtful" concrete had
removed.
Repairs
The repairs had been carried outusing polymermodified fine aggregateconcrete.
The repairs to the roof slab in the vicinity of the columns were carried out because the top
surface of the concretewas found to be disintegrating. The disintegration was the resultof the
poor durability of the concreteand was undoubtedly made worse by the layersand laminations
in the concrete that were observed in samples from the collapsed area. The repairs to the
slab involved the removal of the disintegrating concrete to expose a sound base and the
replacement of the disintegrated concretewith high quality repairconcrete. Examination of the
repair material following the collapse showed, in someareas, signsof poor compaction and a
lack of bond to the baseconcrete, in other areas the base concretefailedjust belowthe bond
surface.
CONCLUSIONS
We concluded that whilst there were several factorswhich would have strongly influenced the
structural capacity of the slab and the natureofthe collapse:
A6
The overall cause of the collapse was the poor quality and therefore the durability of the
concrete in the area of the roof that collapsed. The durability was lower than it should have
been and was also lower than that of the concrete samples taken from elsewhere in the
structure, as a consequence, it suffered progressive damageby frost which, with time, caused
the lamination and disintegration of the upperpart oftheroof slab.
The weightof evidence and, in particular, the analysis of the structure carried out following the
partial collapse of the roof slab, indicates that the partial collapse of the roof slab was most
probablyinstigated by a punching shear failure at one of the internal columnson grid line 2.
This was a consequence of the roof slab having been progressively weakened by frost
damage to the pointat which it could no longersupport its own weight. Finally collapsing when
subjected to the additional stressesthat are likely to have been induced by an overnightdrop
of temperature.
SUMMARY
The collapse was caused by a localized aberration and was not symptomaticof a general
deficiency of the structure.
The car park was subsequently demolished as itwas considered that there would be a lack of
confidence in a repaired structure.
AND FINALLY
I would like to thank NCP for assisting in the preparation of this paper and Wolverhampton
Borough Council fortheir cooperation during the investigation and demolition.
A7
Pipers Row - the DETR view
Dr Satish Desai
BuildingRegulationsDivision, DETR
ThETRODUCTION
Partial collapse ofthe roofslab ofPipers Row Car Park in Wolverhampton occurred on 27
March 1997. This incident couldhave been caused by the combined influence ofa number
offactors. It is not possible to determine the precise contribution ofeachfactor to the
failure ofthe roofslab since this may be vieweddifferently by different parties. I propose to
presentmy viewsbased on consideration ofthe following factors and on the information
extractedfrom the reportsand the records:
Typeofconstruction
Design and detailing
Quality ofconcrete
Effectiveness ofrepairs
Serviceabilityissues
Severe exposure
Temperature shock.
Type ofconstruction
Although the roofslab in question was a slabwithout beams, it was not designed and
detailed to the standards offlat slabs as we know today. Ifa flat slab is designed and
detailed according to the currentcodes ofpractice, I believe that it could safely sustain an
imposed load much higher than its estimated load-canyingcapacity. The roof slab in
question did not have suchattributesofa normal in-situ flat slab construction, for example,
a designed provision for an effective two-waydistribution ofthe load, steel in two
directions and a minimumamount ofsteel at top and at bottom.
Shear heads
The size ofshear-heads was reviewed andincreased by the specialist contractorin 1972.
The post-1972lift slabs have better column-slab connection and improved punching shear
capacity compared with the pre-1972lift slabs.
Absence ofrobustness
The roofslab did not have continuous horizontal tying in two directions. Also,there was
no monolithic connection or continuity betweenthe column and the slab or betweenthe
vertical and horizontal members ofthe frame. This was subsequently addressedin 1982 by
BI
adding post-tensioning to the lift slabconstruction, to improve its robustness.
Slab thickness
The slab had 230 mm overall depth and this meant that, for 7.3 m span, the span-to-
effective depth ratio was greater than 36. The lift slabsdesigned after 1972 were 280 nmi
thick overall. Whenpost-tensioning wasadded to the construction technique in 1982, the
slab depthof250 mm was adopted.
Withoutthe usual beneficial attributes ofa flat slabor even withoutthe post-1972
improvements, the pre-1972lift slab at lower level (level 5) carried the debrisload of5.5
kN/m2 or twice its design imposed load of2.5 kN/m2. This would suggest that the pre-1972
lift slab construction couldnot be isolatedas the solecause offailure. However, the roof
slab did not have the robustness and strength-in-hand available to a normal flat slab and it
failed in the end, on accountofa combination ofa numberofcircumstances as described
below.
Reinforcement
Fine aggregate concretewasused to fill in the box-outs,assuming that it could restore the
length ofshear perimeter around the column and that this couldmitigate the effect ofthe
opening on punching shear capacity ofthe slab. However, the filling was not effective in
this respect. The slab, therefore, had an openingwith its size larger than that permitted by
CP1 14 andthe nominal shear stress couldhave exceeded its permissible value.
Quality of concrete
Fromthe Consultants' report it appearsthat, some time duringthe construction ofthe roof
slab, the use ofready-mixed concretewas suspended and the concretewas mixed on the
site. However, it appearsthat the workmanship was questionable and that concretewith a
high water-cement ratio was used for facilitating placing ofconcretein the formwork. Also,
the contractoris reported to have experienced quality control problems during concreting,
as mentioned in a letter to NCP in September 1966. High claycontent in sandwasnoted at
one stage, aboutfive timesthe permissible 3% by weight. Although it is reportedthat the
B2
concretein question wassubsequently removed, this record does suggest that the
production ofconcretemay have suffereddue to the lack ofquality control.
Multiple horizontal laminations were observed in the top 80 mm depth ofthe slab samples,
containingwhite depositsofcalcium carbonate. Some samples ofthe roof slab concrete
showed poor adhesion between coarse aggregate andthe sand-cement paste. Compaction
ofconcretearoundbars was also poor, showing voids and white material in the voids. In
some cases, the bond between concreteand steelwas so inadequate that some bars could
be rotatedjust by hand.
Effectivenessofrepairs
Laminations were foundin the portion ofthe slab which had not been repaired. It would
seem that repairwas limited to the identifiableproblem area and not extendedto the
adjacent part susceptible to deterioration. This situation could have been worsenedby the
masking effect ofthe weather-proofing layer.
Serviceabilityissues
Exposureconditions
The roofs ofcar parks are known to sufferworse effects ofexposurecomparedwith the
other floors; for example, change in temperatures, rain, freeze-thaw and exposureto
chemicals or de-icing salts.
Most car parks are generally "un-occupied" buildings. In occupied buildings, the users
B3
would be expected to highlight the need for any repair, whichwould signifr some distress
to the structure,for example, leaking ofwater through the roof, etc. An open type car park
would be subject to the elements any way and sucha feedback from the users may not be
available.
Cracksnear columns
Before the partial collapse ofthe rooftookplace, some crackswere observed in the slab
near column J2. It has not been established whetherany ofthesecracks caused significant
reduction in the shear perimeter, additional to that causedby the openings. However, such
crackscould have significantlyreduced the load-carrying capacity ofthe slab.
Temperature shock
CONCLUSIONS
The investigations suggest that the partial collapse ofthe roofofPipers Row Car Park was
a unique andunfortunatecase. Thequality ofconcretewasfound to be generally poorand
the repairs were not effective. The steeldetailing was not satisfactory accordingto the
present-day standards. The box-outs created weakness and the filling in ofopeningswas
not good enough for compensating the reductionin shear perimeter. This situation was
worsenedby the cracks at the vulnerable locations or in the vicinityofcolumns and the
box-outs.
It can be arguedthat, perhaps, some ofthese defects could have been covered and forgiven
by the inherent qualities ofin-situpost-1972fiat slab construction, which are "hidden
extras" and not accounted for in the design. Also, the provisions against disproportionate
collapse in a post-1972construction would have giventhe structurethe requisite
robustness. However, it is not fair to conclude that the pre-1972lift slabconstructioncan
be isolated as a causeoffailure on its own. AlsO; there is no evidence to suggest that all the
other pre-1972 lift slab car parks may have the problems ofpoorquality concreteand bad
box-out details. The causefor concernwould seemto be the combination ofsome probable
factors, which are identified as follows:
B4
Hence, there is a need for inspecting such structures and for attending to the points made in
the HSE press release. This press release asks owners and operators ofsuch car parks to
inspect andmonitorthem, giving attentionto the distress and cracksin the slabs, box-out
details, etc. and it drawsattentionto the need for ensuring the suitability oftype and
method ofrepair.
Beyond this, it would not seem desirable to formulate anygeneral and detailed guidancefor
the authorities and ownersofall existing car parks, based on the Pipers Row incident,
which involves an old car park roofwith out-of-datetype ofconstruction and with a
numberofinadequacies and construction faults. However,it is not intended that this
experience should be ignoredcompletely and it is considered necessary to alert the
construction industry through warnings similar to those contained in the HSE pressrelease.
It is also necessary that the authorities and owners ofexisting car parks should continue to
take deliberate stepsfor monitoring and maintaining the car parks. However,it is equally
important to avoid any suggestion ofa general problem and to ensure that there is no
blight cast on the present reinforced concretein-situflat slabconstruction.
B5
Concrete car parks -
investigations and recommendations
Mike Courtney, John Figg, MalcolmJordan and Robin Whittle
OveArup & Partners
INTRODUCFION
Thevery publicnature ofthese two potentially, though thankfully not actually, catastrophic events
prompted manyconcrete car parks owners torequestengineering advice on the structural qualityoftheir
carparks,especially iftheywere known tobe lift slab construction technique carparks.
Thestructural stabilityand durability ofa concrete car park depends onthe stressesto whichits elements
areand havebeensubjected, the qualityand form ofthe structureandmaterialswhenconstructed and the
deterioration ofthat quality since construction, caused bythe environment towhichthe car park structure
has been exposedamelioratedby any maintenance or repair.
Concrete car parkshave to have adequate resistance to imposed loads,lateral as well as vertical, and
progressive collapse. They are exposed to strain and dilapidation ofthe concrete from thermal and
moisture movements, creep, freeze-thaw action, andcorrosion ofreinforcement followingthe actionof
chlorides in dc-icing salts and carbonation, exacerbated byinadequate cover.
Theresistance ofthe carpark structure to theseeffectsis improved bygood quality concrete and canbe
enhanced byroutine maintenance based on inspection.
It is apparent from the surveys that havebeen carried outthat car park owners, and/ormanagers, desire
that a concrete car park willprovideavery economic, notto say cheap, structure to supportvehicles and
theirsafe passage for an indefinite period oftime with effectivelynomaintenance ofthe concrete
structure, perhaps not surprisingly as the beliefhas been fostered thatconcrete, unlike steel, is a
maintenance free material.
Cl
Thedesigners and builders ofconcrete carparksknowhowever, at least theydo nowevenifperhapsthey
did not inthe 1960s and 1970s, thateconomy inconcrete structures that are exposed to the elements, as
are carparks,provides astructurethat has aminimal lifewithoutinspection and maintenance ofthe
concrete.
Assessments carriedout byOve Arup suggest however that concrete car parks atpresent are exhibiting a
reasonable service life withoutplanned maintenance perhapsbecause they are not subjected to the vehicle
load forwhichtheywereostensibly designed, theyhavenot beensubjected to the environmental
conditions thatthey mightand that the concrete qualitymaybe better than mightbe expected. The service
life ofsuchconcrete structures depends on the quality oftheconcrete that was achieved andit appears
that some atleast ofthe car parksofthe 60s and 70s havereached the end oftheirrepairfree service life.
Inthe 1970s a Lift Slab carpark was foundto havecracks in its slabs. Investigation and analysis showed
that the quality and durability ofthe concrete ofthe lift slabswas suchthattheir strengthhadnotbeen
significantly affected by the cracking. Deterioration ofthe concrete in the insituconcrete parts ofthe
structure was however foundduringthe survey.
The structural qualityofa 9 storeycarpark was investigated. Thecar park hada histoiyofspalling and
delamination ofthe concrete wearing surfaceat all levels, for which partial repairs hadbeen carriedout
over some years.
Owners ofthree car parkswith concrete structures formedby the lift slab technique were concerned with
theirstructural stabilityand durability following thepublicityofthe Pipers Rowcollapse. Information on
thedesignofthe structures wascollected and studiedand an investigation ofthe currentstructural form
and material quality planned and undertaken.
C2
Thestructural surveys involved visual examination, cover metersurveys, concrete strengthtests witha
lighthammer, measurementofcracklocations and widths, bar spacing andpositionat columns, tests on
samples ofthe concrete and P.U.N.D.LT tests for delamination ofconcrete. Tests were carried out on the
concrete samples for strength, carbonation depth, hydration, aggregate quality, HAC,alkali carbonate
reaction and alkali silica reaction.
Nodefects were identified in respectoftheconcrete, savethat the lowconcrete gradeused in the design
wasconfirmed. Thespacingofthereinforcement inthe column strip exeded therecommendations of
CP110. Therewasnodelamination orcracking within the concrete.
Cracks in and spalling ofthe uppersurfaceofthe concrete was identified as a defect. Radialcracks in the
top surfaceofthe slab ofwidth from 0.1mmto 0.9mmextending 2m to 3mfrom columns were foundbut
they did notpenetrateto the slabsoflit. These are believed tobe serviceability cracks dueto high stresses
in reinforcement. Localspalling ofthe concrete wasobserved over reinforcementbars wherethe cover
wasless than 15mm.
Recommendationswere madethatthe car parkswould have a reasonable service life ifthe local concrete
spalling was repaired; the radialcracks were sealed; the existing roofmembranes were removed so that
the roofslabs could be inspected and repaired ifnecessaiy; the roofmembranes were replaced with athin
bonded membrane; and that the concrete structures should beregularly inspected and properly
maintainecL
Concrete car park structures are particularly affected by deterioration ofthetop surfaces ofconcrete slabs
dueto freeze - thaw actionbecause the concrete is generallyunprotected from attack and the deterioration
effectivelyremoves the concrete over reinforcementexposing ittocorrosion but also reducing thebond
anchorage ofthe reinforcement. Concrete oflesserqualityismore susceptible to damage.
All thesemanifestations ofdamage are the inevitable response ofa relatively lowtensilestrength, brittle
material to a swelling pressure(volume increase) within the hardened concrete.
Fromthemost elementaryconsideration the bursting stressis generated by thephase change from liquid
waterto solid icewith a volume increase of 1.09%. Although thismaynotseemmuch, it is quiteenough
to cause damage (as anyone willknow whohas had a cracked car cylinder blockwhenthe engine did not
haveanti-freeze inthe radiator!)
Themost common incidence offreeze-thaw damage is to young (green) concrete cast inwinterand left
over nightwith inadequate protection. Type(i) damage usually resultsespecially delamination,
sometimes inmultiple layers, more or less parallel to the exposed surface.
Where the concrete contains flaky orporous aggregate, pop-outs(ii) are common.
C3
Withflakyaggregates alayerof(bleed) watercanaccumulate on theunderside and the expansion stress
as iceformsis sufficient to expelamoreorless conical spat! from theyoungconcrete. The offending
flaky aggregate particle is then seen at the apex ofthecone.
Withwater- saturated,porous aggregate the effects are similar, but as the aggregate particle itselfis
disintegrated inthe freezing process, partofthe aggregate particle remains intheconcrete andpartcan be
seen atthe apexofthe cone.
Scaling (general surface deterioration) is the mainproblemwith concrete slabs (iii),especially carparks.
It is a problemofmatureconcrete.
Scaling resultsfrom a succession offreeze-thaw cycles i.e. manynights offrost. The UK is more atrisk
than (say)Canada or Siberia, as in Britainwinter temperatures frequently alternate above andbelow
freezing,whereas in much colderclimates only a few cycles offreezing - thawingoccur and all winterthe
concrete is well belowzero.
Salting (and to a lesserextentuse ofother deicers) creates extra cycles offreezing andthawing. A single
applications ofsalt can cause seismic waveofexpansion andcontraction as the salinesolutionpenetrates
theconcrete, thawingice as it goesand refreezing as thesolutionbecomes diluted.
Chloride salts (mainly common salt, calcium chloride, or mixtures thereof) not onlycan causedamageto
theconcrete itself, but arethe principal factorcausing premature corrosion ofembedded reinforcement.
Reinforcementcorrosion is the principal cause ofinadequate concrete durability.
Adequate falls should be provided to flat slabswith a well-designed drainage system. Surface textures
must be deepenough and correctly oriented to encourage goodwaterflow.
Multi-storey car parksare particularly susceptible to frostscaling due to increased coolingfrom wind-
chill action. Saline drippingfrom vehicles is an exacerbating factor.
Lift slab car park structureshave been designedin accordancewith the general recommendations of
British StandardCodes ofPractice forinsitu reinforcedconcrete flat slabs. Considerationofthe
detailsofthe columnslabconnection oflift slabcar parks suggests the shear distributionat these
connections may differ significantly from that atthe connectionofflat slabs and columnleadingto a
C4
less than conservative structureifthe sameempiricalshear stress calculationsfor capacityand
effectiveapplied shear are carried out for a lift slab structureas would be carried out for a flat slab
structure.
The structuralmodelandanalysis was an elastic,uncracked, typical internal panel ofa Lift Slab.
This analysiswas madeto test a hypothesisthat the flexibilityofthe support systemwould leadto
an unevendistributionofshear stressesaround the critical shear perimeterwith some being
significantly greaterthan the meanvalue as conventionallycalculated.
It was foundthat the hypothesiswas true. Forthe EC2 perimeter(with curvedcorners),the worst
shear stress, averaged over a lengthofperimeterofthree times the effective depth,exceeded the
averagevalue calculatedon the BS8110 rectangularperimeterby 39%. This comparedwith the
correspondingvalue of 18% obtained by a similar analysisfor a conventionalin-situ flat slab. Or, if
it is preferredto relatethe "peak" stresses(averagedover 3d) to the EC2 curvedperimeter,the
corresponding overstresses are 24% and 2%.
Figure 1 showsatypical layout ofthesteelworkfor a Lift Slabshearhead. Figure 2 shows how the
steel elements and supportswere modelled.
The analyseswas for a 7.5m squarepanel andused 8-noded plate bendingelementsas shown in
Figure 3. One quarteronly ofthe panel, with symmetry boundary conditionson each edge,needed
to be analysed. Figures4a and4b show the arrangements close to the supportfor dead load andlive
loads respectively. The additional grid ofbeams shown in figure 4b representsthnfill portion of
the slab that wouldbe in positionwhen live load is applied. The ends ofthese beams are taken as
pinnedbut capableoftransferringtorque.
The factoreddead load was taken as 1.4 x 6kPa and the factoredlive loadas 1.6 x 2.5kPa.
E for concretewas takenas I8GPa for dead load and 24GPa for live load. The effectivedepth,
"d", was, for convenience,taken as the same in each directionat 204mm.
The same structural modeland loading was used to analysea conventionalin- situ flat slab, with
modifications close to the supportas shown in figure 4c
Shearstressesfrom the Lift Slab analysisare plotted, for ultimatefactored loading, in figure 5a.
Calculations were carried out for the conventionalrectangularperimeterofBS8110 (at 1.5d from
the support structure). Howeverin figure 5a the shear stressesare plottedalong the more realistic
perimeterwith curved corners. Distancesalongthe perimeterare expressedas multiplesofthe
effectivedepth "d". This shows peaks in line with the edges ofthe supportbehind. (The peaks are
evenmore marked using the BS8110 perimeter.) At 1.1 5MPathe greater peak exceedsthe average
valueof0.76 MPa by 50%. However,it would be unrealistic to use thispeak value as a criterionof
failure. With its very local characterit would be smoothedout over some
length ofperimeter. If
this length is taken as 3d, the worstaverage value as shown, 0.94MPa, gives a stress-concentration
C5
factorof 1.24. Howeversince in practicethe engineerwill have calculatedhis averageshear stress
over the rectangularBS8110 perimeter,the "peak" value should be related to that average.This
gives a stress-concentration factorof 1.39. The "peak" value occurs over the part ofthe perimeter
corresponding to the shorter edge ofthe supportsystem.
The conclusionis that the stress-concentration factor forthe Lift Slab may be significant. For this
examplethe ratio ofLift Slab stress- concentration factor to that ofa normal in-situ flat slab is
1.39/1.18 = 1.18 (say 1.2). This should, perhaps, be used in future calculationsfor assessing shear
resistance.
Since the conferencethe authorshave carried out further analyses ofthe slab with an iitu
boundaryequivalentto the perimeterofthe lift slab shear head. It gives very similar answersto that
ofthe lift slab configuration. This confirmsthat it is the aspect ratio which is the importantcriteria
notthe type ofconnection to the column as was firstthoughtpossible.
C6
Figure 1
V
I
Co
'U.
Co
.c N-
C-)
Cl)
0
Co
4ji
This area is a hole during lifting.
It is filled with concrete once the
slab is in its final position
Shearhead Elements
Pinned Joint
. S S S S
S S
S S
Pinned ,I
Pinned
I
S
Supports-- zSuPorts
I S"c" V'S S
S I S S S S S S
\ Pinned Joint
Pinned Joint ii
Structural Model C
Midspan S -C .-0- -
0---- -5 0- -5-
4 -4 -0- -)
0--C
0---$- -0-I.
5-- -5 -5- 5- I
I -5-. -5--4-
1 -0-- 4.-
4- .-_-.-
$ -0 -0-o- Centre-line
I -4
I 0 -S-S.
Support
S
-n
Centre-line Lines of CO
Support Symmetry
CD
8 Noded Finite Element Mesh c)
C9
Concrete
___ _____ Steel Beams Beams
l."Support t"Support
t_ Structure 4L Structure
a) Self weight Only b) Final Structure
Encastre
Supports
-n
- - '
Support c6
Structure
-5
CD
1 of Support 4_ of Support
J.
.L. ..?yerg 0.8
-. .. -. yQrqgejQr
T!TK!'T5:;::: BS8IIO perimeter
I of Support 4 of Support
88110 perimeter 0.6
1Lcurved part
of perimeter curved portion:
0.4
3d
of Support Support of
0.2 0.2 corner of support
behind
11
0 1 2 3 4 5 6 0 1 2 3 4 5
sid s/d
1
a) Lift Slab CD
b) Normal flat slab
Shear stress on EC2 Perimeter 01
CII
Banbury car park - a synopsis
Jim Kellett
Mouchel ConsultingLtd
Synopsis
This presentation summarisesthe results of a fill structural assessment completed resently of a multi-
storey car park in Banbury. Our commissioncame as adirect result ofthe partial collapseofthe rooflevel
ofthe Piper's Row car park in Wolverhampton,which was an early exampleofthe Lift Slab techniqueof
construction which the Banbuiy Car park is a later example. Both car parks relied on conventional
reinforcementwhich predated the current method ofconstruction.
Observationsand investigative results ofthe car park have led to an apparent contradiction in what is
widelybelievedto be the lower threshold levelsfor chloride ion initiatedcorrosion and the levels found in
areas of obvious active corrosion in these slabs. We have then researched the basis upon which these
threshold limits were based and it became obvious that as a profession we have been misinterpreting
informationfrom recognised reports and hence assumptions have then been made that may well be
severelyinadequatewhenconsideringchloridesin concrete.
CorrosionMechanismsin Concrete
The cement paste usedin the production of concrete is a mix ofextremely complex compounds which,
on the addition ofwater, react transformingthese compounds into a rigid matrix which is used to bond
generally inert aggregates together to form what is referred to as concrete. This mix is highlyalkaline due
mainly to the calcium hydroxide formed during the cement hydration. This, along with smaller quantities
of sodium and potassiumhydrates, provide an environmentthat will maintain the reinforcing steel in a
thermodynamically stable conditionin which a tightly adheringlayer ofgamma iron oxide (a form ofrust)
forms on the steel surface providing a non-porous seal which then shields the reinforcement from its
environment. It has also been suggested that, in addition to this layer, there exists a portlandite layer of
calcium hydroxide in contact with the steel which also contributes to the protective mechanism. When
bound in a concrete matrix, normal reinforcingsteel is entirelydependant on these two layers to prevent
corrosion.
Corrosion of reinforcement is the oxidation of the metallic iron which results in an iron oxide layer
formingon the surfaceofthe iron. This layer has only a very weak mechanicalbondwith the parent metal
and being an expansive reaction will "flake" off, thereby exposing fi.irther steel to the corrosive
mechanism. This expansive reactive will, in severe cases result in a build up ofinternal concrete stresses
which willcause the exterior layer ofcoverconcreteto spallaway from the structuralelement.
The two most significant factors that lead to a transformationofthis passive resistivityof the steel and
initiation ofthe corrosion process are carbon dioxide and chlorideions (i.e. chloridesin solution). Carbon
dioxide from the atmospherediffuses into concrete and in the presence ofwater forms a weak acid which
neutralisesthe concrete's alkalinity by reacting with the cement matrix. Carbonation in itselfdoes not
corrode the reinforcementbut by destroyingthe passiveresistivityofthe surroundingcement paste, allows
water and air from the atmosphere to permeate through to the reinforcement where the corrosion
mechanismis initiated. Under normal conditionsthe pH ofthe concrete matrix pore fluids is in excess of
12.5, but with affectivecarbonationthisfigure willdrop to around pH 8.0. Carbonation therefore is only
the catalystfor the corrosion of embedded steel and takes no active part in the process. The time taken
for the carbonation front to reach the reinforcement and the subsequent corrosion is a function of
numerous elementsnotably cement content ofthe original mix, depth ofsteel cover, permeability of the
mix, prevailing atmosphericconditions,etc.
Dl
The greatest influencing factor on steel corrosion is the presenceof free chloride ions within the pore
fluids. These ions can occur in a concretemix from a numberofsources, firstly by direct addition to the
mix(i.e. calcium chloride accelerators or contaminated aggregates), or secondly as a result ofexposure to
de-icing salts, marine environments or other chloride bearing liquids. Chlorides in concrete can be in
several forms such as free ions, weakly bound hysi/chemi-absorbed)and strongly chemicallybound. A
percentage ofthe chlorides that exist in the mix at the time ofhydration will participate in the hydration
process, forming strongly bound chioroaluminates which are fixed and are believed not to initiate
corrosion. The remaining chlorides that do not form these strongly bound compoundswill remain as free
ions in the pore fluidsofthe cementmatrix. Chlorides that are introducedto the concreteafter hydration
willnot form any degree ofchemical alliance within the matrix and will remain as freeions andit is these
free ions that are believed to representthe greatest threat to the reinforcement. When sufficient chloride
are present,corrosion ofthe steel willbe initiated and does not relyon the depthofcarbonation whichwill
onlyinfluencethe rate ofcorrosion. This corrosion initiation maybe related to the size ofthe chloride ion,
whichbeingsmall maypermitready diffusion throughthe passivating layer aroundthe steel in a manner
that is not possible for the largeroxygenmolecules. Thebasicchemical reactions for this oxidation ofthe
iron in a chloride ion environment, are:
i) the availability, concentration and the difflisibility ofchloride ions at the anode,
ii) the availability of hydroxyl ions as influenced by the extent of neutralisation, the rate of
hydrolysis, cementcontentand composition,
iii) the moisture content of the concrete which influences the oxygen difiuision and concrete
resistivity.
02
Investigations And Results
Ourinvestigation took two forms, the first was a physicalinvestigation to assess the prevailing conditions
within the structureand the second an analytical check to attempt to allay fears we had about possible
punching shear problems aroundthecast-incollar column connections.
Thephysicalinvestigationincluded:
Visual inspection to logcracking patterns within the slabs. Withthiscar park we were fortuitous that
the waterproof membrane over the roof area had recently been removed as part of a replacement
Contract, this allowed a closer inspection at this level than wouldhave otherwise been possible.
Our observations of the entire structure revealed that the major of columns throughout the structure
displayed radial cracking patternsin the adjoining slabs. The most prevalent type ofcrack radiated from
the column at approx. 45 from the re-entrant corners of the column penetrations. Crack widths were
estimated to be between1 & 3 mm. On each all levels, exceptthe roof, there are numerouslocation were
spalling concrete on the top surface ofthe slabs indicate active corrosion is occurring and numerousother
locations wherepatch repairs have been employed in the past to rectii,'concretedamage.
A sophisticated covermeter survey to map out reinforcement locations, assess concretecovers and to
gainan indication as to the sizes ofbars present. This was generally locatedaroundthe column locations
that displayed the severest cracking patterns. Results showeda good correlation with the construction
drawings for size, location and concretecovers. The cover being between20 & 30 mm to the top layerof
reinforcement.
Carbonation tests in a limited number of locations which indicated lower than expected rates of
carbonation, these rangingfrom 3 to 22 mmwithin the first floor slab, with the majority being less than 12
mmand 2 to 3 mm in the roofslab. This suggeststhat this is not the root cause ofthe observed corrosion
in theslabs.
Our observations showthat corrosion was on all floor levels (excluding the roof), evenwhen chloride ion
concentrations were as low as .13%. This wouldappearto contradict the suggestions ofthe BRE Digest
N 246 that suggests that the risksofthe initiationofcorrosion dueto chlorides are:
c) 1.0% <highrisk.
I would postulate therefore that we, as engineers, are not comparing like with like. I believe that the
investigations employed for this BRE document were to assess the implications of chlorides occurring
within a mix prior to setting, historically this would most likely be in the form of calcium chloride,
traditionally used as an accelerator. When included in a mix in this form a percentage ofthe chlorides
react with compounds in the mixto form complex chloride compounds duringhydration and which do not
03
influence corrosion as the chloride ion is too tightly bound, these are generally refereed to as "fixed"
chlorides.
Chlorides that enterthe concretematrix after setting will remain as free ions andare known to affectthe
corrosion rate of steel. The methods used to assess chloride ion concentrations in concrete do not
differentiate between "free" and "fixed" chlorides and therefore do not accurately assess corrosion
potentials. I believe that anyconcretecontaining chlorides at percentages above 0.1% should, in the first
instance, be regardedwith suspicion. This is not to say that the BRE recommendations are inaccurate, it
is that we are using infonnation from a related series of tests and applying these results to a similar
problemand aretherefore trulyrepresentative. This is not an indictment ofour profession, but is intended
that as atimely warning that we must be morecautiousin applying investigative resultswith whichwe are
not ifilly conversant.
Withthe lift slab car park in question the extensive cracking patterns around the more severely loaded
columns will have allowed a more rapid ingress ofchlorides, moisture and air through to the
reinforcing
steel, all necessary to promote corrosion ofthis steel. The majority ofall the historic patch repairs to the
slabs are adjacent to columns.
Compressive testing of concrete cores removed from throughout the structure to establish insitu
strengths. The mean ofthe resultswas 33.4N/mm2 with a standarddeviation of4.9 N/mm2, higher than
the specified strength of20 N/mm2.
Finally two petrography examinations were conductedto ascertain the original concrete constituent
parameters. The results showed nothing untoward except for high water/cement ratios in the concrete
mix. Therefore, from the results of the compressive tests and the petrography examination there is no
problem with the strength of concrete used, but as with a majority of structures of this era, concrete
robustness was not fullyunderstood or appreciated resulting in low concrete covers, low cementcontents,
highwater/cement contentsetc. and this structureis no different.
Design Appraisal
As part of our commission a structural design assessment was completed, using a plane grillage elastic
analysis, to ascertain whether the structure conformed to the prevailing Code ofPractice at design and
whetherit conforms to the presentBS 8110. Calculations were completed usingthe simplifiedpermissible
stress designprocedure ofthe load factormethodofCP 114 and the ultimate limit state designprocedure
ofBS8110. For the design to this later code the material partial factorofsafety was maintained at 1.15 to
reflect thevariance in steelquality at the time ofconstruction. Our resultswere:
For compliance with CP 114, the code ofpractice to whichwe have beenled to believe this car
park was designed; the general bendingmoment capacity ofthe slab adjacent to the columns was
slightly less than Code requirements. This is not necessarily an indictment ofthe designers but
possibly a reflection betweenthe variances in resultsbetweenanalyticalmethodsemployed.
for compliance with BS 8110,the ultimate capacity ofthe slab in bending was slightly less than
coderequirements. The structuredoes not meet the codes requirements for robustness nor does it
comply to the requirements for lateral stability. The codes requirements for punching shear can
only be adequately resisted should it be assumed that the shear perimeter is measured from the
exterior of the cast-in collar system. The validity ofthis assumption must be contentious is as
muchthat these codes ofpractice applyto normal construction techniques, with CP 114 specially
excluding '...reinforcedconcreteconstructionofa specialistcharacter....". We consider that the
cast-in collarsupport system is not adequately addressed in eithercode, andtherefore such a novel
system should been reasonably expected to have been subjected to a suitable testingprogramme to
04
confirm its suitability, and not rely on assumptions regarding the interpretations of a Code of
Practice. We do not want to judge this innovative construction method but would recommend
that a morerigoroustest/design regimeshould havelbe applied to the system. Without having this
necessary information to hand we could not verifjthe capacity ofthe slab underpunching shear.
In applying both Codes,there were not indications on the construction drawings or in the structureas to
how the structural resisted lateral wind loading except through a reliance on unreinforced, site mixed
concrete usedto infihl the voidsbetweencolumns and the slabsand betweenthe stair structures locatedat
eitherendofthe car park.
Recommendations
The two optionsavailableto ensurethe continuing safe operation ofthe car park are:
Repair/strengthen the existing structure; whereby the structure is efficiently back propped to
allow the removal ofcover concreteto exposeall thenegative reinforcementto allow an assessment ofthe
remaining steel area to establish if replacement steel needs to be added to achieve the minimum areas
required, will allow the existing steel to be thoroughly cleaned and the cover concrete to be replaced.
Steps wouldthen be required to be taken to ensurethat chloride ion initiated corrosion is arrested in the
future. As the chlorides have penetrated to well below the steel depth on the first floor, removing all
contaminated concrete is not a viable option. This then leaves two recognised methods to control this
form ofcorrosion, these are:
o Cathodic Protection: where,by applying and electrical chargeto the affected reinforcement
willreverse its potential in sucha manner that it will promotethe active migration ofchloride
ions away from the affected steel. Caution must be taken to ensurethat chlorides that have
penetrated deeperthan the treated steelwillbe encouraged to migrateto the opposite surface
and any steel there that is unprotectedwill be susceptibleto corrosion.
05
o
Total rebuild;This solution is by and far the most expensive option availablebut also offers
the widest scopeforchangesuchas improved capacity, layout, security, aesthetics etc.
Conclusions
In summing up our investigations high lighted that we must be more suspicious of reinforced concrete
with chloride ion contentsbelowthe 0.4% we havetraditionally assumed to be the necessary threshold for
the onset of chloride initiated corrosion ofreinforcement. When results of ongoing investigations are
published, that the applicable parameters used to assess the situation are clearly defined to avoid
misunderstandingthe application ofthese results.
Our findings haveraised concerns about the early lift slab designs, thesewe hope are not misconstrued as
a denunciation of the construction technique but as a recommendation that certain areas should be re-
examined to ensure that the structures will provide the desired levels ofperformance with the necessary
levels ofsafety. This methodofconstruction means that the last pouredslab in the firstto be loaded and if
time restraints are tight thenthere will always be the temptationto lift earlywillalways be present. At the
end ofthe day it is ourresponsibilityto ensurepublic safety whenusing/occupying ourbuildings.
One remaining area ofconcernwasthe ability ofthe precast cladding unitsto withstand vehicular impacts.
These precast concrete units span from floor to floor with a connection at each floor that in theory
transfers both vertical and horizontal loads to the associated slabs. From the construction drawings there
was no evidence that any form ofrestraining reinforcement is in place to preventthe dislodgement ofthis
fixing under a horizontal loading. If the present car park is to be retained in its present form and
refrbished this wouldrequire attention.
06
Column-slab connections
and developments in shear design
Dr Pal Chana
ConcreteResearch and Innovation Centre, Imperial College
1. Introduction
This paper is a mixture oflookingat the past and also lookinginto the future. It concentrates
specifically on column-slab connections and aspects relatingto punching shear.
Typically a slab can be tested as shown in Figure 1, supported on a column with the loads
appliedon a circularperimeterat some distance from the column. In this test rig, whichwas
at the BCA's research station at WexhamSprings,the load reacts againsta large concrete
block.
As the load is applied, radial cracks form from the column locationat loads approaching the
service load. Theseradial cracks extend towards the edge ofthe slab. At loads close to the
calculated designload some circumferential cracks form in successive perimetersaround the
column. The slab fails soon after the formationofthese cracks. Failure is generally brittle in
nature. However, this depends upon the layout ofthe reinforcement.
CP 114:1957Punchingshear
El
Shear reinforcement not consideredto be necessary.
Moment transfernot considered.
CP 110:1972Punching shear
4. Membrane effects
Codes ofpractice rules are based on cantilevertype test specimensand ignorethe effect of
continuity. There has been test workdone to investigatethe effect ofcontinuityor so called
membrane effects which increase the shear resistanceofthe connection. In these teststhe slab
was extendedto the point ofcontraflexure (Figure2).
5 Shearheads
One solutionfor strengthening the column slab connection is to use a steel shearhead within
the slab depth. A shearhead is a prefabricated steel sectionassemblynormallyplaced within
the slab depth over the column.
E2
Negative slab reinforcement can be reduced
Rulesare given in ACT code. CRIC hasjustcompleteda project (with financial support
from the RCC and DETR) to develop rules for UK practice
Various prefabricated shear reinforcements systems are now availablewhich ease the placing
of shear reinforcement in the slab.
Shearhoops are one such system which have now been used effectively for over 5 years.
Thereare other systems also includingthe stud rail system. This is a system whichhas been
used in Europeand America.
Consists of a line ofsteel studs fixedto a steelplate
Radial arrangement ofstud rails is preferred.
This systemgives excellentperformance and CRIC is developing designguidelinesfor UK
practice
Perhaps codes lay too much emphasis on strengthand less on ductility and robustness.
E3
8. Coventry Car Park collapse
A four-storey car-park structure collapsedin 1988 duringdemolitionofthe top floor. The 200
mm concrete slab dropped 3metres onto the floor belowcausinga progressive collapse.
There were concerns about existingcar parks in Coventiywhich were showing signsof
crackingaround the columns. BCAwere asked by Coventry City Council to carry out tests in
order to assessthe load capacity.
Four tests carriedout in total, includingsome with a service hole adjacentto the column. We
used shearhead collarsrecoveredfrom the collapse. Steeldetail was reproducedin the test
specimen. No top or bottom steel in the immediate vicinityofthe collar for a distance of 1.5d.
This represents more than adequate margin ofsafetyand demonstratesthat, when properly
designed to moderncodes,flat slabsare a robust and efficient form ofconstruction.
10. References
E4
inp:intS
I 22i
EII
0- ooo
PIcnof stab
Bendun moment
I
041
041
fI
1St.
(b) Fullpanelspecimen
E5
Punchingshear failure
E6
A typical shear head
E7
/&/,
E8
Modern lift slab construction
OsamaEl Garf
Tilbury Douglas ConstructionLtd
Introduction
Lift Slabwas developedto providean economicalmethod of constructiOn. Thesystemhas been used
in wide variety of applicationsin theconstruction industry. The mainadvantageof thesystemis that it
can simplify or eliminate complicated falsework by casting concrete at ground level then lifting or
lowering it to its final position.
Lift SlabApplications
Thesystemis mostefficientifthe useofthetraditionalmethodsof construction require
a lot offalseworktoachieve a tight programme
examples
Multi-storey buildings with large footprints (e.g. Multi-storey Car Parks)
Steelor concreteroofs or slabsat height
complicated falsework
examples
Pile capsdeep underwater (e.g. bridgefoundations)
Multi-storey buildings with unusual footprints
Elevatedwatertowers
Conicalor othercomplicatedroofs (e.g. large diametersilo roofs)
Column Inserts
When pre-cast concrete columnsare adoptedinsertsfabricatedfromsteel sections are cast into each
floor level. Intermediate inserts may be used to park the slabs temporarilyfor multistage lifting on
tall buildings. Withsteel columnsa bearingmemberis weldedbetween flanges.
Fl
Steel Collar
Collarsfabricatedfrom steel sections are cast into each slab around the columns. These collars are
used as shear heads which also serve as the attachment point for the lifting jack rods during lifting.
The collar incorporates twe shear plates.
Steel Connectors
The steel connectors are usedto transmit thevertical load from theslab into the columns. In thefinal
connection the connectors engagewith the shear plates in the collar, and bear on the column insert
forminga permanent mechanical connection.
After the load Is transferred successfully to the columns the lifting jacks can be removed and the
remaining voids can be filled with concrete encasing the mechanical shear joint, and providing
resistance to lateral loading, fire and corrosion.
The use offlat slab design forthe Lift Slab multi-storeycar parks permitsthe reduction of floor to floor
height(typically2.550m), subsequently theoverall heightofthecar park can be reduced
to save on claddingcost
meet planning requirement
to fit an extra floor and increase the car park capacity
First Generation
In the early 60's thetypical structural layout adopted had four columnsacross the bay width of 16.0
metres. The typical reinforced concrete floor slab thickness was 225mm, and the average area of
floor slab per column was 25 squaremetres.
Second Generation
Thedevelopment ofthe structural layout in the late 60's increased the averagearea of floor slab per
columnto 50 squaremetres with consequent reductionof cost. The layout used twe columnsacross
the 16.0 metres bay width, spaced at 9.80 metres with cantilevers of 3.10 metres at each side. The
reinforced slab was typically280mm thick.
Third Generation
In the early 80's post-tensioned concrete floor slabswere introduced for all Lift Slab car parks. The
samestructural layoutas thesecond generation was used, butwith slab thicknessof250mm.
The useof post-tensioned concreteslabs
What next
As a result of post-tensioning the slab a more flexible structural layout can be used. A 13.6 metres
span prototypehas been constructed with a slabthicknessof 320mm.
F2
1)
U)
iO)
/
{; -:'--
iooi
Cl)
LI
k.
;
System Overview I
Pre-cast Concrete Columns
Column Grouting
Ground Floor Slab
Foundation
Backfill and Ground Slab Formation
System Overview 2
Lifting Jacks in Position
and Rods Connected to The Top Slab
1 -
E
0. 1
w Ii
w
I
0
I
L.
(I)
.2
w
b
Steel Connector
N
a used to transmit vertical load
from the slab into the column
W75 & W150 W75 WISO
C G)
E
o -
I C)
-I
:4 p.'
8
: q q
o ..
(a)
.)I. w
.
I I I
/ :'
0
C)
0.
0
C.)
r '
.
OWU)
(I)
00
I
L.
0 0
U)
C1
WC
W V
I- 1 o
Cl)
w
:4 '
.0
w
E oI .0
E
C) I
5 w 2
C
-J
0.
w 4
C
Ui
U)
I- C
0Cl)C
U)
I- Cl) 0
._
0 C)
C)
Insitu concrete fill
Collar's Lifting Member
Post-tensioned Collar's Side Member
lined slab Collar's End Member
I
.k
Connector / Shear Plate
Column's Insert
V ...Pre-Cast Concrete Column
U U
1'
. N. '.. . "
-a ?t1d exit route
.
i'rizontal-SIabedges;externaflyWhich
. - . .
ensure
.
Cl)
0)
C.)
0
0
0 Cl)
0 a)
0.
-c
.0 .0
I
\
Cu
Cl)
\ : a)
0
t
-
-
Cl)
0
;I
C.)
Il/s
0* 0001.
I.
1
4- -
c
.1
- .
; -I'
' ,- I.
-C. -
II-
I-
'I co 0098
I ,' OSLC' 'OSLC
L
/ _______
00091.
4-
-k--
Fe withthe t
I I I I I I
warpedi.s$ab Iayo t) ' I
0
1, StructuraI layout 0
0
0,
satisfy the lStructEn9
'o:
rcommditinsi 984
Using Lift Slab
Eliminates 90% of the formwork required
for cast in place concrete
General
Atmospherically exposed steel will corrode in the presence of oxygen and water. Steel
reinforcement which is embedded in concrete is protected from corrosion by the chemical
characteristics of the cement and the physical barrier provided by the concrete cover.
Hydrated cement is highly alkaline with pH values in the region 1 2.6 to 1 3.5. It is well
establishedthat when steel is in contactwith alkalinesolutions of pH values in this range,
the metal is normally passivated by a thin layer of oxide film.
Steel reinforcement in concrete will therefore only corrode when the protection afforded
by this passive layer is affected in some way. The two common causes of depassivation
are carbonationand chloridecontamination. Unfortunatelythere are now many structures,
including car parks, which are suffering from deterioration of the reinforced concrete
elements. In dealing with such deterioration, there are three questions that should be
asked:
The principal options for the repair of concrete damaged by reinforced corrosion are
coatings
concrete repair
replacement
electrochemicaltechniques - cathodic protection
- chloride removal (for chloride contaminated
concrete)
- realkalisation(for carbonated concrete)
corrosion inhibitors
The selection of the appropriatetechnique will depend upon a variety of factors including
the extent of deterioration, the degree of carbonation or chloride contamination, the
characteristicsof the concrete, accessibility, aestheticconsiderations,the future use of the
structure and of course the availability of funding.
Various features of the different techniques are discussed in the following sections.
GI
Coatings
Research into protective coatings for concrete has been extensive. Industry has moved
from the belief that concrete was a durable material that needed no additional protection.
A variety of paints, sealersand water repellenttreatments are availableto provide first line
defence against deterioration of new structures, and to renovate and extend the life of
structures suffering from reinforcement corrosion. Working Group WG1 of TC1O4/SCBof
the European Committee for Standardisation(CEN) is responsible for establishing unified
standardsfor the surface protection of concrete.
Where reinforcement has already started to corrode, coatings on their own may not have
a significant effect. As long as there is moisture and oxygen present in the concrete the
corrosion process will continue. It may be possible to reduce the rate of corrosion by
keeping the concrete dry and a vapour permeable coating may go some way towards
achieving this. However it may be more successfully achieved by some form of cladding
or guttering if the geometry of the structure permits. If this solution is chosen, a
monitoring regimeshould be establishedto ensure that the deteriorationhas beenarrested.
Concrete Repairs
For concrete repairsto be successful, it is important that all areas of concrete affected by
carbonation or by chloride contamination are removed and replaced. If only visually
defective concrete is repaired (ie spalled, delaminated or cracked areas), there is a high
probability that the reinforcement adjacent to a repair will rapidly become anodic and
corrode if carbonated or chloride contaminated concrete is left behind.
G2
Repair Materials
The repair materialmust properlyfill the void, carry the loadingappliedto the member, and
provide protection to the reinforcement. This is normallyachievedby using a cementitious
material which might be a hand-applied mortar, spray applied gunite or a high flow
concrete, and may contain cement replacements(eg PFA or silica fume) or additives (eg
polymers or plasticisers). Non cementitious materials are only likely to be used for the
repair of small, non-structural defects.
Replacement
Where a reinforced concrete memberis so badly damaged that it is beyond repair (eg the
reinforcement is so badly corroded that it cannot be satisfactorily reinstated), or where
deterioration is so extensivethat repair becomes very expensive,then replacementof the
whole element may be a cost effective option. This may however entail temporary
proppingto the structure, depending on the residualstrength and whether the loadingcan
be temporarily reduced (eg can heavy equipmenton a suspended slab be relocated whilst
a damaged column is replaced).
A significant advantageof complete replacementis that the new memberwill be of better
quality than one which has been patched up in a piecemealfashion, enabling its life to be
predicted with much more confidence.
Cathodic Protection
Only cathodic reactions should now occur at the steel, producing hydroxyl ions, and
corrosionshould cease. Negativelycharged ions such as hydroxyls and chlorides will tend
to move towards the positive anode. The reactions occurring at the anode will tend to
yield acidic products. This may be an important factor in consideringthe durability of the
anode.
It has not been conclusively demonstrated that cathodic protection will stop corrosion
completely. However in trials at the Transport and Road Research Laboratory during the
late 1 980's, cathodic protection was shown to significantly reduce the corrosion that
developed in test specimens.
The principal elements of a cathodic protection system are the power supply, the anode
and the monitoring system.
G3
The power for a cathodic protection system is usually derived from an alternating current
source via a transformer-rectifier. A typical zone will probably only require 50 to 1 50
watts which can be readily obtained from a street lighting circuit or the domestic supply
to a building. The output from the transformer-rectifier is likely to be between 2 and 1 2
volts giving a current density at the surface of the concrete of between 5 and 2OmA/m2.
Metal oxide coated titanium mesh - pinned to the surface of the concrete and
covered with a cementitious overlay.
Conductivecoatings - proprietary paints with a conductive graphite filler in a binder
and carrier mix, usually covered with a decorative/protective topcoat.
Various criteria for ensuring satisfactory protection of the reinforcement have been
considered for reinforced concrete structures. The draft European standard for the
cathodic protection of steel in atmosphericallyexposed concrete proposes the following:
a) An instant off potential (measured between 0.1 s and 1 s after switching the d.c.
circuit open) more negative than -720mV with respect to Ag/AgCI/0.5M KCI.
The virtue of b) or C) is that the decay is independent of the nature of the reference
electrode used or any variation in its absolute potential in the long term. The decay period
used is typically 4 hours: However, the rate of depolarisationis likely to be variable and
may becomeslower the longer the structure has been protected.
G4
Chloride Removal
The Principles
Chloride removal, like cathodic protection, involves the passage of a direct current from
the surface of the concrete to the reinforcement, though currents and voltages are far
higher. The chloride ions are negatively charged and are therefore attracted towards the
anode on the surface of the concrete from where they can be removed. With the correct
operating conditions, sufficientchloridecan usually be extracted aftersome 6 to 8 weeks.
Since the current is only applied for a relatively short period, the anode can be much
simplerthan in cathodic protection systems. Typically a mesh of activated titanium is fixed
to the concrete using wooden battens to space it off the surface. Instead of using a
cementitious overlay, a saturated paper pulp is sprayedon to completely enclose the wire
mesh.
An alternative to the "papier mache" system has been developed which involves the
construction of a tank against the face of the concrete to retain water as the electrolyte.
This removesthe needto keep wetting the paper pulp requiring a continuous presenceon
site.
An ordinary steel mesh anode may be used in place of the activated titanium, althoughthe
steel will corrode quite rapidly and may stain the surface of the concrete. However in
many cases the concrete is grit blasted and coated on completion of the process, so any
staining is removedor hidden.
Operating Conditions
Current levels are usually maintained at around 0.5 to 1 .0 A/m2 (compared with some
lOmA/m2 for. cathodic protection). The duration required is typically 6 to 8 weeks but it
will depend upon the extent of chloride contamination, rebar configuration and the circuit
resistance of the system.
Limitations
The process will not remove all of the chlorides from the concrete. Although the chloride
levels can be considerably reduced in the vicinity of the reinforcement and in the cover
zone, less will be removed from those areas behind the steel. It is therefore likely that the
remaining chlorides will diffuse towards the reinforcement and the process may need to
be repeated at some stage. It is therefore also likely that chloride removal will not be
appropriatefor concrete with cast-in chlorides.
Realkalisation
High pH levels are restored by the production of hydroxyl ions at the reinforcement,
together with the movement of positively charged alkali ions from the electrolyte on the
surface into the concrete.
G5
Corrosion Inhibitors
Calcium nitrite has been used for someyears now as an additive to fresh concrete, to raise
its tolerance to chlorides ie greater concentrations of chloride can be present before the
reinforcement will start to corrode. However because of its molecular size, it will not
readily penetrate into hardened concrete from the surface if you are attempting to arrest
ongoing corrosion.
There is now a wide range of restoration options available to the owners of reinforced
concrete structures. The electrochemicaltechniquesmay seemrather complicated at first
sight, but there are cost savings to be had, so it is worth giving them proper consideration.
However, it is a specialist field and the systems must be selected, designed,operated and
monitored accordingly.
References
G6
Deterioration modelling to determine the residual life
of reinforced concrete structures
C P Atkins, Dr M B Roberts, I R Lane and Mike Gower
Maunsell Ltd, Birmingham
Synopsis
The deterioration ofreinforced concrete structures due to carbonation and chloride contaminationis
a widespreadproblem. Maunsell has developed a methodology for modelling the most common
deterioration processes for reinforced concrete. In particular, a procedure for quanilfying the
deterioration processes associated with chloride contamination of reinforced concrete has been
developedand this may be used to predict the loss ofarea ofreinforcementand the loss ofcover
concrete with time. A worked example of the deterioration model is presented. Conventional
structural analysis can then be used to determine the residual service life of the deteriorated
structure.
Introduction
All structuresdeteriorate with time - if the deterioration mechanisms are known,then structures can
be designedto last for their full service life. Unfortunately,the current evidence shows that many
deterioration mechanisms were not fully understood in the past and as a result many structures are
now deteriorating at an accelerated rate. Many structures, such as bridges, car parks and concrete
framed buildings are constructedofsteel reinforced concrete which under normal circumstances is a
durable constructionmaterial. However, circumstances can arise which lead to corrosion ofthe steel
reinforcementand a subsequent reduction in the strengthofa structure.
profile
Time Tinteen6on
HI
The problemthat engineersare now facedwith is how to determinethe residual strengthofa structure
in its present deterioratedcondition and to predict when the structurewill no longer be consideredto
be safe, as illustratedin Figure 1.
Maunsell has addressedthis problem by developinga methodology for modellingthe most common
deterioration processesfor reinforcedconcrete. In particular,Maunsell has developeda procedurefor
carbonation;
chloride contamination.
The corrosion products occupy a volume significantly greater than the original steel and this in turn
results in a build-upof stresseswithin the protectivecover concrete. These stresses are subsequently
relieved by cracking and delamination of the concrete and will eventually result in the concrete
breakingaway from the structure. The exposedsteel reinforcementwill continueto corroderapidly.
The lossof area ofthe steel reinforcement, as a result ofcorrosion,reducesthe abilityofthe structure
to support the applied loading. Delamination of the concrete reduces the bond forces required to
anchorthe reinforcement, furtherreducingits structural effectiveness.
Carbonation
Carbon dioxide in the atmosphere dissolves in the moisture present in the concrete and forms an
acidic solution.This reduces the alkalinityofthe concrete so that the passive layer protectingthe steel
reinforcementbecomesineffective. This is known as carbonation. The rate of progressof carbonation
in concreteis knownto be non-linearandreduceswith time.
H2
To identify carbonation,concreteis freshly broken away from the surfaceanda pH indicatorspray is
applied. Typicallythe indicator will turn pink in strongly alkaline environments but remains clear in
less alkalineconditions. The pink/clearboundary may be measured from the surface ofthe concreteto
indicate the depth of carbonation. Ifthe date of construction and the depth ofcarbonation at the time
of inspection is known, then the rate of carbonation can be calculated and the time at which the
carbonation front will reach the level of the steel reinforcement and hence the time to onset of
corrosion can be determined.
ChlorideInducedCorrosion
The major causeofreinforcedconcretedeterioration in the UK is the presenceof chloride ions. Once
the chloridesreach the level ofthe steel reinforcement in sufficient quantitiesthey disrupt the passive
protective film andcorrosion commences.
The first potential source of salt is found in the constituents of concrete. Concrete is made up of
cement,aggregatesand water. If the aggregates are taken from the sea they will contain quantities of
salt, unless they are washed thoroughly. It is also possible that chlorideshave been deliberatelyadded
to the concreteto make it set quicker. Calcium chloride was used extensivelyfor this purpose, until its
detrimentaleffects were recognised and its use in reinforcedconcretewas subsequently banned.
Salt may also enter the concrete from external sources. Every winter thousandsof tonnes of salt are
liberally scattered on the roads to reduce accidents. Concrete structures located in marine
environments are also exposed to the chloridesin sea water.
Figure2 TypicalChlorideProfiles
To identify chloride contamination it is possible to sample the concrete at different depths to
determine a profile of chloride concentration with depth, as shown in Figure 2. If the amount of
chloridesremains relatively constantor increases with depth, the chlorideshave probablybeen added
to the concrete at the time of construction. If the chloride levels in the concrete are higher at the
surfacethan inside, then the chlorideshaveprobablyentered the concretefrom externalsources.
H3
Deterioration Modelling
The first step in the production of a deterioration model for reinforced concrete is to identif' the
primary cause of deteriorationand to establishthe mechanism(s)by which the structural integrity is
affected.
structure
arrival of chlorides onset of onsetof failure
considered
at concrete surface corrosion delamination unsafe
0 : corrosion &
:
chloride active
U)
ingress corrosion : delamination :
.4
0 4...
C-) .4
rate of corrosion i
.4
delamination threshold
T0 T3 T4
Time
As.bullt
section
Apply deterioration
model
Modified
section
Reduced
diameter and bond
H4
delaminate. These parameters can then be used to predict a modified cross-section as illustrated in
Figure 4. This modified section can then be subjected to conventional structural analysisto derive
the residual strengthofthe deteriorated section.
Non-destructivesurvey techniques, such as half-cellpotential surveys, can indicate the risk of active
corrosion by measuring the local electrical potentials arising from active corrosion of the
reinforcement. Half-cell potential surveys may be presented as contour plots ofthe concrete surface,
with areas of active corrosion generally being indicated by potentials more negative than -200mV.
The more negativethe potential,the more likely that activecorrosion is taking place. Such a survey
can then be used to identify representative areas of active corrosion. By removingsome ofthe cover
concrete, to expose the steel reinforcementin representativeareas, the amount of steel lost due to
corrosion can be measured.Since the time that corrosion started has been calculated and the loss of
section is known, if a constant rate of corrosion is assumed its value can be calculated. Finally, a
relationshipbetween corrosion rate and the non-destructive half-cell measurements can be derived.
This relationshipwill enable the likely corrosion rate I at anypointto be determinedfrom local half-
cell potentials; a measurementwhich can be taken with ease at any point on the structure. Therefore,
the likely corrosion loss ofsteel reinforcementmay be calculated at any pointin time.
By comparing areas of corrosion with areas of delamination the final piece of the jigsaw can be
produced, i.e. how much loss of reinforcementsection causes delamination. In this way a full model
of the deterioration process can be produced for the structure.
Having calibrated the deterioration
model, computersoftware can be developedto enable its application on similar areas ofthe structure
to be simplyperformed. Maunsellhas developedin-house deteriorationmodellingcomputer software
for this very purpose.
Summary
Maunsell has established a general methodology for the development of quantitative deterioration
modelling for reinforced concrete structures. In particular, Maunsell has developed a quantitative
model for a large group of structures subjected to chloride induced corrosion and has successfully
H5
used this model to determinethe residual service life of the structures. For the firsttime the risk of
continuing deterioration can be quantified and this is enabling a rational repair strategy to be
developed. High priority repairs can be identified, while enabling a rational maintenanceprogramme
to be establishedfor structuresdeterminedto be at low risk.
While the equations developed by Maunsell are specific to one particular form of structures, the
general methodology is equally applicable to any deteriorating reinforced concrete structure, from
carbonated buildings to chloride contaminated bridge decks and car parks. Maunsell has the
capability to develop and calibrate these deterioration models and to analyse the resultant structures
with either a conventional code based approach or by using more advanced probability-based
techniques.
A worked example ofa quantitativedeterioration model for chloride contamination from an external
source is attached.
H6
Worked Example (based on chlorides diffusing from outside the concrete)
N
E
E
Step 2: determine T2 C
0
- -1 corrosionratei U
0
assumed constant
0
(0
(0
Corrosion 0
-J
initiation
Construction
To TI T2 T0 Time
timecorroding
Full half-cellsurvey
Selectedchlorideprofiles taken in areasofactivecorrosion
Full coversurvey
Step 1
Table 1 Example ofa Chloride Profile Taken from a Structure
Depth (mm) Mid Point (mm) % Chloride
5-50 27.5 3.00
50-75 62.5 1.12
75-100 87.5 0.75
100-125 112.5 0.20
125-150 137.5 0.20
Comparethe chloride profilesobtained from the structurewith those shown in Figure 6, and decide
on the lengthof time that the structurehas beenexposed to chlorides.From thisresult derive the time
ofarrival ofchloridesat the surfaceofthe beam T1.
H7
For example, if a chloride profile was as shown in the Table 1, the length of time of exposure to
chlorideswouldbe 12 years:
T1 = 1997-l2years1985.
12
0
0 2 4 8 8 10 12 14 16
Depth (cm)
Step 2
300
200
E
E
a
100
0 .t I t
0 10 20 30 40 50
lime afterexposure (yrs)
From Figure 7 and the cover data findthe length of time that the depth of steel has been exposed to
the 0.4% threshold chloride value, 2- T1. If the cover was 40mm, then the chlorides would have
arrived at the level of the steel after 4 years. In the aboveexample,this would mean that the steel had
been exposedto 0.4% chloridesfor 8 years, i.e. T2 = 1989, T,r= 8yrs.
H8
as so
a
A103. be
and
-100
marked procedure
either
generated.
section, be
be The
of concrete can
can
and
loss should
this
that method.
-200
steel
total
estimated.
the
8 the
of
concrete
forces
capacity, based
the
Rate
structure
obtain area
Figure between
to hence
the
that
-300
structural
probability of
bond
using and its
Corrosion years) life
vs rate (8
threshold,
H9
effective
residual
determine sophisticated
the
-400 Half.celI(mVvsCSE)
Half-cell
corrosion
corrosion reinforcement to the
3.8mm2yr.
of
8 the more
the
1= a and
delamination
reduce
of or
time
mV, the will analysed
future
the
Figure calculate
2 be the
-500 -460
by than
anchorage
value of rate
may
approach
into
30mm
the
= greater
is delamination on based
times
half-cell half-cell sections
at
loss This code
corrosion
effect
8 7 6 2 0
f5
600
local an
the
S
local
a
3 C)
the
for
example
5 section
have modified
repeated
be
this the
Step From e.g., Step4 Multiply In Step If delaminated. will The conventional can
Design and maintenance for serviceability
and durability
Peter Matthew
Matthew Consultants
Car parks have too long been the poorrelations of structures, -not obtaining the acclaim of
say a Norman Foster Office development or a Richard Rogers Millennium dome. Yet, no
other structure has such demands put on it by the users or can evoke more anger in its
critics.
Car parks are often called ugly, tatty, intrusive, useless, user-unfriendly or a blot on the
landscape and at best a necessary evil or plain serviceable. Why have they become such a
target for ridicule, when so manyofus depend on them so much?
It might be that none of us really like having to park cars, but we don't like going to
hospital either and hospital buildings have acclaim and we don't like going to court but
courthouses have acclaim. So what is it?
Let's consider the usual criteria a car parkgets built by - costper space. A developer or his
Architect will plough plenty into the 'main' building facade services and fit out but the car
park can look like a halffinished Spanish hotel, as long as the cars can park. But cost per
space will limit stall sizes, headroom, ramp widths, stairs and lift access, lighting levels,
waterproofing and certainly exterior appearance.
I have no gripewith a car park looking like a car park, although sometimes valiant efforts
are made to conceal them as if they are a social pariah, but they can actually be quite
attractive ifthe effort and budget is adequate. The U.S.A probably leads the way with the
best and sometimes the worst car parks, but those at the top of the scale , employing
creative precast or pigmented and sculptured insitu concreteor cleverly worked brick and
metal can be as stunning as your best office, hospital or courthouse.
Most complaints are not from those lookingfrom outside, however, they come from the
users who we all know are different creaturesin cars than they are when standing on two
legs. Theyexpect to drive straight to a spacewithout delay, have no problem with turning
II
or manoeuvring, be in a clean and bright environment and have a pleasant and quick exit
on foot. On return they do not want to queue, have payment problems or travel
unnecessary tedious routes. Unfortunately the cost per space often kills these aspirations
and 'carparkrage' creeps in.
Car park operators can differ in their approach to these problems and there is no doubt a
vast difference in attitude to the customeraccordingto the car park's purpose. The best
car parks are usually those provided for shoppers , particularly at large out of town
centres which are keen to satisf' the clients. Municipal car parks, airport and station car
parks, office car parks and residential car parks all vary in standards according to the
importance the developerrates the quality ofthe facility in his need to impress.
Those items which create the best image and custom satisfaction apart from the exterior
dispersion of cars into the facility. It may also include guidance to what level
parkingis available to avoid tiresome searching.
2. Good signage.
Just putting arrows on the floor is not enough. The signs should be plentiful,
carefully planned to avoidconfusionand easily visible.
12
Ramp widths should be generous and comfortably exceed the recommended
minimum where at all possible. Fitting in a couple of extra spaces by saving on
Parkingbays are usually 2.4 metres wide but are sometimes reducedto 2.3 metres
at considerable discomfort to users. Some shopping centres now advocate 2.5
metre bays.
shopping facilities this should be at least 100 lux and preferably 200 lux. Higher
levels are oftenused in turningand ramp areas.
6. Gooddrainage.
All floors should have adequate falls to avoid water ponding and well designed
collection by cross falls or channels. Intermediate floors drainage must be to the
foul system via an interceptor since these floors are more likely to have oil
contamination.
Where possible the outer spaces of the structure should fall to the perimeter to
avoid wingblown rain from traversing the slab to drain.
13
A fall of 1 in 65 should be targetedto allow for toleranceeffects reducing this to 1
in 80 in places.
Above these user friendly measures to promote a quality car park the durability question
must be answered, and this too can effect the price per space.
Firstly I believe that all car parks in use should be designed for 'very severe' rating, sinceit
is almost guaranteed that they will be subjected to dc-icing salts, either directly or from car
tyres. The design of adequate falls and good maintenance will help but resistance to
penetrationby salt-water must be resistedby the concreteitself.
BS 8110 creates a dilemma for designers of very severe resistance slabs, unless the
concrete used is C50. The use of C40 or C45 requires air-entraining, but unfortunately
few suppliers ofair entrained concretewill guaranteea strengthof4ON/mm2.
I believe that air entraining is a great advantage, but perhaps the strengthofthe concrete
should not be such a limiting factor if the design can cope and the other parameters of
cement content and water: cementratio are met.
More effort should be made to improve the concrete's resistance by use of pore blockers
such as silane. Care is also needed here since there is a range of products all sounding
similar (e.g. Silane, siloxane, Isobutyl Silane, etc.). They may be permanent protection
and require periodicre-applications.
14
The movement of car park decks due to thermal effects are more severe than in other
building types because oftheir expensive condition. For this reason careftil consideration
and analysisofjointing and restraintsare needed. Too often, expansion /contraction joints
are cheap and cheerftil, failing withinone year ofinstallation.
Use of proprietary mechanicaljoints are well designed nosings plus inserts are especially
required at roof levels where movement is greatest. These joints should preferably be
used at all levels.
In conjunction with good joints the vertical structure must be detailed to allow the
movement. I am not a fan ofhalvingjoints in the structure,having seen too many fail, but
believe that double columns present the best alternative where possible. Stainless steel
dowels can often suffer the same fate as halvingjoints unless their installation is perfect.
When designing the vertical structure due allowance must be made for horizontal
movement effects in the flexural design. The connection between the vertical and
horizontal structure (as well all now know) is also paramount. In short, thin about how
the structure can movewithout distress.
Finally, I would like to call on all designers to be responsible for providing a maintenance
manual for car parksjust as they would expectwhen buying a car. At least this puts some
onus on the owner to do something and not expect the structure to withstand all abuse
withoutany distress.
The PCI of the United States have produced such a manual which is simple and
straightforward. I commend it as a good step in the right direction.
15
Precast multi-storey car parks
Paul Bracey
Composite Structures Ltd
INTRODUCTION
Prefabrication of structures is as old as civilisation itself. Some of the best known ancient structures surviving
today were made "off site", and erected on location. They were generally prefabricated because elements of
repetition are present. This meant that the work necessary in shaping the units could more easily be undertaken
off site.
Modern car park structures similarly contain a large repetitious element. Precast concrete can be used to
advantage, utilising the flexibility of form of reinforced concrete, but in an engineered off site production mode.
Precast concretecan be moulded to an infinite variety of shapes, and be texturedand colouredto a vastarray of
finishes.
Worldwide, precast concrete forms an important part of the construction techniques portfolio, with the USA in
particular producing highly engineered and efficientstructureson a scale rarely seen in the UK. Resistance to
prefabrication in the UK means that maybe only 3% of British multi-storey frame projects are built in precast
concrete. However, within the multi-storey car parking sectorin Britain, arguably25% of structures are being built
in precast concrete. This is indicative of the advantages that prefabricated concrete construction can bring to
parking structures.
JI
SPECIALREQUIREMENTS OF ANY PARKING STUCTURE
A car parking structure, in addition to being cheap and ideally self finished, must resist these actions during its
design life :-
Rain and the contaminants that can turn It Into weak acid
Wind forces
Cyclical daily
Fire
A typical 50N/mm2 (70001b/in2) mix for precasting has 400kg/rn3 minimum cement content and 0.45 maximum
water cement ratio. This mix is the highest grade available under BS8I 10 1997 table 3.4, and offers excellent
durability. A high cement content offers good resistance to carbonation. In simple terms, carbonation is
acidification of the "covercrete". High cement content ensures the protective alkalinity of this important layer.
However, too alkali amix can cause alkali!aggregatereaction problems, so it is important to ensure that alkalis in
the final mix are within allowable maximum limits. Using the C50 mix described, 30mm of cover reinforcement will
meet the "very severe" durabilityclassification of BS6IIO 1997, affording reasonable protectionagainst occasional
de-icing salts.
J2
Table 3.3 Nominalcovertoall reInforcement(including links) to meetdurability requirements
(SEE NOTE 1)
Conditionsof exposure Nominalcover
(see 3.3.4) Dimensionsin millimetres
Mild 25 20 20" 20" 201J
Moderate 35 30 25 20
Severe - - 40 30 25
Verysevere - - 50" 40" 30
Most severe - - - - 50
Abrasive - - - See NOTE 3 See NOTE3
Maximum free water/cement 0.65 0.60 0.55 0.50 0.45
ratio
Minimum cement content 275 300 325 350 400
(kg/rn3)
Lowest grade ofconcrete C30 C35 C40 C45 C50
1) nese covers may De reauce to imm proviaed thatthe nominal maximum size of aggregatedoes not
exceed 15mm.
2) Where concrete is subject to freezing whilst wet, air entrainment should be used (see 5.3.3 ofBS.5328:
Part 1:1987)and the strength grade may bereduced by 5.
NOTE 1. Thistable relates to normal-weightaggregate of20mm nominal size. Adjustments to minimum
cement contents foraggregates otherthan 20mm nominal maximum size are detailed intable8 of
BS.5328: Part 1:1987.
NOTE 2. Use ofsulphate resisting cement conforming to BS.4027. These cements havelowerresistance
to chlonde ion migration. Iftheyare used in reinforced concrete in very severe ormostsevere exposure
conditions, the covers intable 3.3 should be increased by 10mm.
NOTE 3. Covershould be not less than the normal value correspondingtothe relevant environmental
category plus any allowance for loss ofcoverduetoabrasion.
Precasting also offers advantages in terms of factory control of placing and curing. With most production under
cover, issues such as frost and ice on shutters, freezing of newly placed concrete, difficulties of compaction,
difficult conditions for workers etc. are minimised. All of these can have significanteffects on the quality of site
placed concrete. Precasting also allows units to shrinkbeforethey are placed intothe structure, reducing durability
reducing insitu shrinkage cracking.
Generally then precast concreteframes are and always have been made of significantly stronger concrete than
insitu frames, and offer betterdurability as a consequence ofthis.
DRAINAGE ____________________________
All frames must be configured with a minimum 1:60 or 10 fall. This is required on all decks, to provide for wash
down of lowerdecks as well as rainfallon upperdecks. Fallsmust be planned in from the outset many schemes
have been produced where drainageis ignored. Falls have significant effects on elevations and building height
(often limited by planning restrictions) that must be catered for right from the outset. Water falling on a deck must
havea quick, effectiverouteto the drainage system. On top decks at least, this usually meansformingscreeds to
counterfalls in orderthat localised ponding is avoided. In particular, a definitepath for waterfalling on exposed top
rampsmust be defined.
FLOORING SYSTEMS
Flooring systems used with precast frame construction include composite and unscreeded systems. Composite
systems includeHollowcore, DoubleTees, and Solid Slabs. Unscreeded systems include thick table DoubleTees
and bespoke units.
In the UK, all top decks using these systems should be waterproofed with a good quality membrane. This is
because it is not possible to build these systems to water retaining structure standards. Generally though, lower
decksare leftwith an exposed concretefinish, due to the weatherprotection afforded by the upperdecks.
These flooring systems have the same high durability as frame components for the same reasons. They also
benefitby having their strandslocated at the bottomofthe units, away fromthedeck level and its pollutants.
Elsewhere in Europe, unscreeded bespoke precast floor systems are used for top deck construction. Units are
mastic pointed, and the relatively high maintenance requirements of the mastic are accepted.
J3
WELD CONNECTORS
Unscreeded flooring systems mustbe tied togetherto satisfy robustness requirements. This is usuallyachievedby
means of stainless steel weld connectors, comprising a cast in insertin the floor units joined together by means of
a site weldedconnectorplate. The connectoris usually coveredby groutor mastic.
SCREEDS
Lower deck structuralscreeds require careful detailing if adequate durability and performance is to be achieved.
Concrete is supplied by standard readymix contractors, Ideally, a C50 mix. with 30mm cover is used. Typically
though, C50 readymix is not available due to limitations on local aggregate strength. Composite Structures
therefore favours a 10mm aggregateC40 mix adjusted to meet the water cement ratio and cement content of a
C45 mix and meetingthe "very severe"exposurecategoryof BS8I10 with 40mm of cover.
Ifadequatespace is to be provided for mesh laps, this means that 75mm minimum screed is required at thecrown
of the units. Attempts to use less are likely to be unsuccessful I Curing of thin toppings is notoriously difficult in
summerconditions, and polypropylene fibres can helpto prevent drying shrinkage.
The screed is used to tie the structuretogether, both to provide robustness and resist vehicle impact. It is also
used as a horizontal diaphragm to spread wind forces. It is therefore most important that adequateconsideration
is givento durability.
FIRE
If boundary wall conditionsexist or the parking structure forms part of a basement, then significant fire resistance
willbe required from the structural frame. Precast concrete structures offer excellent fireresistanceup to at least4
hours rating.
TEMPERATURE
It is important to allow for temperature movement in design. Indicative figures for planning purposes show
0.3mm/metre run for lower decks and 0.6mm/metre run for top decks. Movement at top decks is typically twice
that at lower decks, due to exposure. Top decks can be seriously affected by dark membrane systems, due to
surface temperature gain.
J4
The required spacingof movement joints dependslargely on the elevational cladding system used and its ability to
absorb movement. The need to avoid constrained expansion between stair towers, ramp sets and lateral vertical
stability systems is also an important factor. The figures suggestedindicate 30mm of top deck movement for a
50m length of structure.
Some Engineers takethe view that thermal movement is not a "real" issue, and can be ignored. Practice indicates
otherwise. Stair cores located at gable ends can be particularly prone to thermal movement, and require careful
detailing.
MOVEMENT JOINTS
Movement joints are best formed by means of double beams and columnsratherthan slidingjoints. Sliding joints
lead to the need for excessivecare during construction and maintenance thereafter. They really require bridge
construction levels of QA to be successful, a luxury not available on most car park sites. It is worth pointing out
that doublecolumnshave architectural implications that require early planning consideration.
The movement joint seals are themselves often uunder specified. This leads to ingress of water and high
maintenance costs in service.
200
30
MovementJoint Detail
LATERAL STABILITY
Lateral stability of precast structures is usually provided either by wall or vertical steel bracing systems. Usually
small areas of ramp or core area wall can be utilised if other walls are not available. This provides a very direct
route for wind and lateral vehicle forces to the foundations. Constrained thermal movement should be designed
for. Lateral load is distributed to the verticalstability sets by diaphragm action of the floors.
J5
SPANDREL/FLOOR CONNECTION
This spandrelto floor connection has a number of important functions to fulfil. As well as supportingthe main floor
planks, the edge spandrel also resists the current BS6399 15 tonne design vehicle impact force. Bars projecting
from the precast spandrelare cast into the screed, providing a load path back to the mainverticalstability systems.
119
'I'
Appmx. 600
[Sc,eed andmesh
Reinfamementbarftom
upstand
/ /pandm!
to ff8 in
Spandrej,/
panel
if Floor
plank.
\_Drip groove
The floor plank seating must allow for fit up tolerances without causing spalling of seatings. A temperature
differential of up to 300 may exist over the depth of a top deck floor construction on a hot day. Daily changes in
precast plank levels of up to 6mm due to thermal bowing have been measured. It is therefore important that the
connection details are able to accommodate all these conflicting requirements. As mentioned before, it is
particularly important that the insitu work is of good quality.
WALL PANEL/FLOORCONNECTION
Wall connections are similar to spandrel connections. Hollowcore floor planks should be provided with weepholes
to prevent build up of water in the cores both during construction and in the event of waterproofing breakdown.
Whilstweepholesare only required at the low end of the units, it is usual to provide them at both ends, since one
doesn'tknowwhich way round the floor units will be erected.
J6
SPANDRELBEAM! COLUMN CONNECTION
Beam/column connections are formed most economically with reinforced concrete corbels. Various other
connections are possible, including recessed steel brackets and billets. Concrete corbels have advantages with
regard to durability, speed, cost and simplicity.
Spandrel beams
GaWaniedpin
intcj
castIn socket1th
6mm Thkk lead
beang pock
Colum,. SECT1ONA-A
Tvoica!
_ri SDandrel Beam To Column Corbel Detail
Beams are usually dowelled for robustness, and seated on galvanised steel, lead or rubber bearing pads to ensure
fit up. As with floor plank connections, there is often considerable thermal movement on these connections. For
this reason, some flexibility is desirable.
J7
COLUMN/FOUND CONNECTION
BS8I1O 1997 robustness clauses requires that all columns and walls are positively tied down to the foundation
structure. This is most easily achieved by means of starter bars fixed with high strength non-shrink grout.
Baseplates and holding down bolts are possible with precast construction, but expensive and usually unnecessary.
Al
castin regashefiths
150x150x10Lcatfo,7' I
bars to
L- projectout offoundations
WALLS
Reinforced precast concrete walls can be cast in many different forms. Casting of walls into tall panels eliminates
expensive and slowsite joints, and assists with robustness.
Walls with slendeness ratios up to 60 can be handled using tilt lift techniques as shown here on a car park
structure. Thiswall is 11 m tall by 175mm thick.
Voided walls may require "in plane" analysis and additional reinforcement to deal with thermalstresses.
J8
PARCON & SPECIALSSYSTEMS
There are a number of special precast frame systems available for parking structures. This is Composite
Structures Parcon cantilever portal frame system. Frames are cast as 16m x storey height units, and transported
to site upside down. Once on site, work progresses quickly, since all beam/column joints are integral with the
precast frame. The systemproduces a particularly robustframewithoutstability walls.
padfoundaon3
Typical Section Through Split Level Car Park
+62 +204
FINISHES
The use of self finished structuralprecast units can eliminate the maintenance problems associated with secondary
cladding systems and speed up on site programmes. A very wide range of exposed aggregate, cast in brick, stone
and otherfinishesare possible.
Farnborough CarPark
PLANNING
it is helpful, if the use of precast concrete is to be maximised, to plan early. In this way maximum benefitcan be
obtained from repetition, and the use of the material and moulds can be optimised. A simple structurewill usually
have less joints, and potentially less maintenance.
The optimal car park frame in terms of best value, short programme, good durability, and low lifetimemaintenance
will often be selffinished precast.
J9
CONCLUSIONS
This brief paper highlights some of the aspects of precast concretecar park construction important with regard to
durability. Precast concrete has many in-built strengths to offer the designer, contractorand client when used in
this context. In summary, they include :-
JI 0
A critical review of
precast car parking systems
Dr Howard Taylor
TarmacPrecast Concrete Ltd
INTRODUCTION
Precastconcretecar parks have been erected in significantnumbers over the last thirty
years. A review of the performanceof our building stock is always of value. This
conference gives a timely opportunity to learn and improve all forms of car park
construction.
The writer has been actively involved with the design and construction of precast
concrete car parks of many forms of layout for the last twenty years. The further,
older stock ofcar parks executed by businessesthat have operatedfrom precast plants,
some now closed and some now owned by Tarmac Precast Concrete Ltd. have also
provided a useful extended population of which records are known and recollections
are strong. A further useful reference is a summary of the experience of car parks in
the USA 1,2.
Car parks are difficultstructures: planning requirements and the necessary exposureto
the elements give their structure a testing time. Equally, it is not every owner's desire
to spend money on quality either in terms of build or maintenance. It is of course the
designer's duty to consider safety and robustness.
Structural form is based on planning requirements and these usually relate to the type
of site and usage. The site will determine whether a flat floor, split level, or helical
ramp system is appropriate and the application, long or short stay, will determine the
circulation, bay layout and access locations. The application with respect to the
identityofthe car park in its environment, as an introduction to a shoppingcentre for
example, may define the amount of money that may be spent on waterproofing.
drainage, air changesand finishes.
This review considers a range of systems with respect to some of these points and
concentrates on the related structural issues.
CLEARSPAN SYSTEMS
Clear span systems have the obvious benefit of columns at the ends of parking bins
only, and from the user point of view there are no columns to avoid during parking
and getting out the car. The clear span ofapproximately I6m brings greater structural
depths than short span systems.
KI
probably be completedwith an insitu topping. Difficulties are the depth of the cross
beam whichis required to pick up the load for possibly as much as 6m of floor and the
beam to column connection which is usually substantial and, ifcorbeled,may impinge
on head room, albeit in a non critical area. Edge beams would probablybe connected
to both the columns and the insitu topping and would be short in the longitudinal
direction. In the gableelevation,a full span gable beam may be difficultto design and
intermediate gable columnsmay be preferred.
Double tees of 500 or 600mm depth can carry car park loadings to 1 6m and span to
load bearing edge beams. These beams have a structural depth that allows them
conveniently to cope with impact and pedestrian loading. Double tee car parks are
conventionally finished with a concrete topping although in some countries "pre-
topped" DTS are used. Double Tee cranked ramp units can be made which remove
columns at the ends ofsplit level ramps.
Some systems can have a precast portal spanning 16m with provision for flooring
spanning on to it in the longitudinal direction. As a frame, the system provides its
own stability in one direction. Longitudinal beams are required to trim the ends of
ramps unless the frame has a cranked or inclinedbearing for the longitudinal spanning
slab.
Hollowcore 400mm deep with insitu topping can provide a clear span solution at a
minimum floor depth and with what some Architects perceive as an advantage of a
flat soffit. The units do have the difficulty of the closed voids which must always be
drained. Build up of water in the cores has caused problems in winter with freezing
in structures during construction. The writer is unaware of any difficulty in a
completed structure. Hollowcore planks are heavier than other systems such as
double tees or beam and longitudinal slab systems and thereforealso require a heavier
supporting structure.
Single Tee
Single Tee systems existing in the USA where 3 metre single tees are supported by
columns at similar centres or by columns spaced further apart with an insitu
intermediate slab/topping concrete to provide continuity. This slab is sometimes
prestressed in the direction transverse to the single tee span.
Tableclear span
This system which is used in the USA has a reinforced concrete table, consisting of
four columns, a longitudinaland transverse frame and a floor all cast together. The
tables are erected on top of each other to producean instantlystable structure. Tables
can also be spreadapart just as single tees and have drop in slabs and infill concrete if
required. The writer has seen a car park ofthis form with a helical floor layout with
K2
the table units being made with skewed tops. The great stabilityof the system make
this excellentin seismic zones.
Durability
Car parks may be waterproofed on all decks, on the top deck only, not waterproofed at
all and in some cases known by the writer, neither waterproofed nortopped with insitu
concrete. The US experienceis that in all cases, frame or shallow floor systems have
less durability than deep simply supported systems as in the latter case, critical
spanningsteel is always at the soffit, well away for the top surface which is first to be
attacked by water and chlorides. At stiff points, colunins and re-entrant corners, top
cracking in a radial direction is common and this is also allows water access to top
steel that would naturally be provided in a framed system, Drainage is a must, floors
can have generous slopes, span to alternatecolumns lines and have local thickeningof
screeds which then channel water to down pipes. For neatness sake, drains or
downpipes are best at columns, although this can present difficulty with beam and
framed systems. The final essentialfor good durability is for responsible management
and good maintenance. There is little point in having drains if plants are allowed to
grow out ofthem!
Cars can and do catch fire in car parks; the writer has had experienceofthis in a small
number from the car park stock that was described earlier. The drainage was
superficial in these cases and repairs were minimal and straightforward.
Movements
The large plan areas and exposed nature of car park floors make them respond to
temperature and moisture more than many other structures. We do not design car
parks like bridges and, in general, accept some movementdistress. Movementjoints
are used in long structures and in structures with L shaped or other inappropriate plan
forms. From bitter experience, there appears only one sure way to provide allowance
for movement and that is to have open or simply sealed joints with independent
structures on each side. Sliding joints simply do not work. Double columns at
movement joints, despite Architectural complaints, should be insistedupon.
Erection
The different systems have a range of maximum piece size and weight which is
sometimes a deciding factor in choice. Very busy town centre sites serviced by tower
cranes may require light units and may more appropriately be erected in a floor by
floor basis. Sites where a crawler crane can operate within the footprint of the car
park can have longer units and can be erected virtually to full height. This is often
quicker than floor by floor erectionand can be accommodated in town centre sites by
the last units being erected by a mobile crane after the crawler has been removed.
K3
SHORT SPAN SYSTEMS
These systems are less demanding structurally as spans may be alternately 2, 12, 4, 12
metres in an extended flat car park. Apart from the difficulty of the columns in
parkingareas variants ofthe floor systemspreviouslydescribed will all apply.
Beams may frame longitudinallyor laterally and opportunities exist for beams to span
across columns levels to give economy of depth through continuity as well as
simplifying the jointing.
Frameshort span
The systems are similar to the long span frame systems and have the advantage of the
moment being more easily carried into columns for frame action.
STRUCTURAL ISSUES
The different systems have structural design aspects defined by their geometry. The
main structural issues are considered below.
Stability
The stability of precast car parks with respect to lateral loads is usually provided by
bracing or frame action. Car parks require stair towers at centres such that they can
usually providesimple strong points to brace a non-framed structure. Stair towers are
usually provided at the edge or corners of car parks for planning reasons with respect
to access and maximising floor lengths between escape routes. This can raise
difficulties with respect to shrinkage and thermal movements of the floor plates and
crackingis sometimes observed. Ideally bracing in the long direction along the centre
line with walls at right angles to the braced axis provides stabilityand allows out of
plane movement from the braced grid line in each direction. In this way movement
crackingis minimised.
Robustness
Car parks present no difficulty in providing ties for robustness and critical elements
can be protected from impact.
Impact resistance
The requirementto protect the structure from impact by cars and to restrain cars at the
periphery is important and has just been highlighted by SCOSS. It is easier to deal
with edge beam restraint in layouts with edge columns and deep spanning elements
than with short span cantileversystems. The strength of opening corner frame joints
is notoriously difficultto guarantee.
K4
CONCLUSION
Precast concrete car parks are the same as other car park structure types in that they
can be provided to a range of durability and finish standards. Structural systems
depend upon site layout, planning,environmental and usage requirements and choice
of systems requires careful thought. Our built stock shows that the structures can be
durable and if properly managed and maintained can present a good and safe
environment for an economic and useful life. Precast concrete has an important role
in the building stock and can not only cope with all layout requirements but can also
be a vehicle for high quality concreteinternalfinishes and a free choice of elevational
treatment.
REFERENCES
K5
Summing up and the way forward
Prof. George Somerville
British Cement Association
I will make a distinction between assessing old structures and designing new
structures, because standards have changed. We are talking about building structures
that have existed only for some 40 to 50 years. They were built from the mid-50s
onwards and are functional, but uniquely, didn't feature in the structural codes ofthat
time. They still don't, but there is good guidance available.
Assessment
The priority needs to be given to the structures built in the 1 960s and 70s, when
standards were lower and less well defined. Perhaps we should also distinguish
between purelystructuralissues (models) and the influence on these of durability.
We need to bring in the technology used in assessing bridges. We are not well placed
to carry out a rigorous structural assessment of building structures. We rely on and
modify design models, and this does not always work, especially when deterioration is
involved. Possibly the best example of that is shear, where the underlying empiricism
is a handicap. The figures in Mike Gower's paper have to be developed realistically,
not just for car parks, but because future engineers will spend more time on assessing
and upgrading old structures, not on new build. At the moment, assessmentsare done
as a series of one-offs. Those involved need not take such a reticent or conservative
view. The process of assessment is as skilled as design, and needs to be developed.
The top priority is safety. My personal view is that where we do have a failure there
will always be a numberofunderlyingreasons. There may be a final trigger, but there
will be a combination ofcauses, as pointed out by SatishDesai.
The real strength is invariably higher than the design estimates. This is a another area
where more authoritative guidance is required. It is terribly easy to be conservative,
and to play safe. We must do better than that. Already, we have seen moves in that
direction in the bridge field, with the introduction of reliability-based methods. This
has to be done with care, since not all the factors are random, but it does add
perspective.
It is only a few years ago that owners were panickingthat their structures would 'die',
due to the dreaded concretecancer (ASR). The IStructE report corrected that situation,
and recent work at the BCA indicates that the same progressive approachcan be used
for corrosion.
LI
Car parks are an intermediate between buildings and civil engineering structures, and
there is a case for reviewing design and performance standards, especially for
durability, movementand protection. Emphasis needs to be put on how durability and
movement might reduce strength and serviceability. Neither of these are currently a
formal design issue. These need to be introduced as additional items, not just by
strengthening existingrules.
New build
New build presents a much healthier picture, as we have a wide range of concrete
options available, and real confidence in design methods with supporting
documentation The models are fine, but performance criteria, including loads, may
still require looking at. For example, handling moisture correctly can extend the life of
the structure.
The industry really has to adopt life-time performance as a design matter. The culture
is not yet universally right, as first cost is still often the main criterion. But owners are
becoming more discerning and demanding. We need to adopt whole-life costing
methods, risk analysisand value engineer.ingconcepts.
Finally, a personal request, not connected with detailed calculations or safety, but still
important - can we make car parks more attractive, inside and out?
L2
Concrete car parks - Cl/SIB
Design and maintenance issues
Conference papers UDC
BRITISH CEMENTASSOCIATION PUBLICATION M21 625.712.63-033.3
OBC