Sie sind auf Seite 1von 10

MA TE RI A L S CH A R A CT ER IZ A TI O N 60 ( 20 0 9 ) 7 0 0–7 0 9

Application of petrographic examination techniques to the


assessment of fire-damaged concrete and masonry structures

Jeremy P. Ingham⁎
Halcrow Group Limited, Vineyard House, 44 Brook Green, Hammersmith, London W6 7BY, United Kingdom

AR TIC LE D ATA ABSTR ACT

Article history: The number of building fires has doubled over the last 50 years. There has never been a
Received 16 March 2008 greater need for structures to be assessed for fire damage to ensure safety and enable
Received in revised form appropriate repairs to be planned. Fortunately, even after a severe fire, concrete and
3 August 2008 masonry structures are generally capable of being repaired rather than demolished.
Accepted 1 November 2008 By allowing direct examination of microcracking and mineralogical changes, petrographic
examination has become widely used to determine the depth of fire damage for reinforced
Keywords: concrete elements. Petrographic examination can also be applied to fire-damaged masonry
Concrete structures built of materials such as stone, brick and mortar. Petrography can ensure
Masonry accurate detection of damaged geomaterials, which provides cost savings during building
Fire repair and increased safety reassurance.
Petrography This paper comprises a review of the role of petrography in fire damage assessments,
Optical microscopy drawing on a range of actual fire damage investigations.
© 2008 Elsevier Inc. All rights reserved.

1. Introduction expenditure and also savings in consequential losses, by permit-


ting much earlier reoccupation of the structure.
This number of building fires has increased by more than a 100% Petrographic examination has become a well established
over the last 50 years [1] and currently the cost in the United method for determining the depth of fire damage for reinforced
Kingdom exceeds two million pounds per day. Consequently, concrete elements. The information gained from microscopical
there has never been a greater need for structures to be assessed examination of concrete samples is now regularly used to aid the
for fire damage to ensure safety and enable appropriate repairs to decision of whether to repair, or demolish fire-damaged
be planned. Concrete and masonry construction materials offer concrete structures. It is less well known that the same
good resistance to fire because they are incombustible (in petrographic examination techniques can be successfully
comparison to wood) and have low thermal conductivity (in applied to other fire-damaged building materials such as natural
comparison to steel). However, physiochemical changes and stone, clay bricks and mortars (both lime and cement based).
mechanical damage caused by heating will eventually compro- Petrographic examination of samples from fire-damaged stone
mise the load-bearing capacity of concrete and masonry ele- and brick masonry structures can be used to make informed
ments. In practice, the worst damage is usually confined to the decisions regarding the extent of repair required. In the case of
outer surface and even severe fires seldom cause total structural historic structures, petrographic examination can also be used to
collapse. Experience shows that following detailed appraisal, fire- determine the composition and identify the source of the
damaged structures can nearly always be reinstated using a materials, enabling a compatible match to be selected.
selection of repair techniques, sometimes combined with repla- The application of petrographic examination of fire-
cement of selected structural elements. As an alternative to large- damaged concrete and masonry structures (including case
scale demolition this provides very substantial savings in capital studies) will be discussed in detail in the following pages.

⁎ Tel.: +44 20 8233 3574.


E-mail address: inghamjp@halcrow.com.

1044-5803/$ – see front matter © 2008 Elsevier Inc. All rights reserved.
doi:10.1016/j.matchar.2008.11.003
M A TE RI A L S CH A RACT ER IZ A TI O N 60 ( 20 0 9 ) 7 0 0 –7 0 9 701

clay bricks, petrographic examination methods could be


2. Investigation Procedures for adapted from BS EN 12407.
Fire-damaged Structures Following arrival in the laboratory samples are first examined
in the as-received condition both with the unaided eye and using
A thorough appraisal is normally required after a fire as soon the a low-power binocular microscope at magnifications of up to
building can be entered and generally before the removal of x100. This initial examination is used to observe macroscopic
debris [2]. After a fire, an estimate of the severity of temperature features such as macrocracking and colour changes. It also
exposure is required in terms of an equivalent standard test [3]. A allows selection of the most appropriate location for thin-
visual examination and classification of damage for each sections (typically 75 ×50 mm area) to be taken for further,
structural member is conducted on-site. The maximum concrete more detailed high-power microscopical examination.
temperature profile during a fire can be estimated from post-fire Preparation of thin-section specimens involves production
assessment of the concrete. This is commonly achieved by a of typically 30 micron thick ground slices of the sample mounted
combination of on-site visual inspection and petrographic on glass slides, through which light will pass to allow micro-
examination of diamond-drilled core samples in the laboratory. scopical observation. A technique developed for the study of
The findings of the on-site and laboratory investigations are used rocks, the thin-section preparation of fire-damaged concrete and
to prepare key diagrams and schedules detailing the damage. A masonry materials presents considerable challenges for the
comparison between the cost of repair and the cost of removal technician. Samples may be relatively soft, friable and/or cracked
and replacement is then made for each damaged element, taking so that it is necessary to consolidate the material by impregna-
into account its function. Following this, a repair strategy may be tion with resin. Cementitious concrete and mortars are also heat
drawn up. and water sensitive so that the drying and resin curing stages of
2.1. On-site Investigation Techniques specimen preparation have to be conducted at low temperatures
(b60 °C). In addition, coolant/carrier liquids used during cutting
Prior to undertaking on-site inspection the investigator must be and grinding must comprise oil or alcohol, rather than being
satisfied that the structure is safe to enter. Temporary falsework water-based.
(props) may be required to secure individual members and Thin-sections are examined in plane-polarised or cross-
stabilise the structure as a whole. The primary on-site investiga- polarised transmitted light using a high quality, medium to
tion technique is the visual inspection, which records such high-power petrological microscope at magnifications typically
features as collapse, deflections, spalling and cracking. Impor- up to ×600. Fluorescent dyes added to the consolidating resin
tantly, certain colour changes can often be used to identify the during sample preparation may be used to aid the examination of
presence and extent of damaged material. Visual inspection may cracks and pore structures when the specimen is viewed in
be aided by the use of a magnifying hand lens (typically x10 combination with a strong light source and an excitation filter.
magnification) or field microscope (typically x50 magnification). The combination of visual/low-power examination in hand
A small hammer is commonly used to conduct a tapping survey specimen and more detailed high-power microscopical exam-
that will detect hollow sounding delaminated material. A ination in thin-section, is commonly termed ‘petrographic
number of complimentary non-destructive techniques, such as examination’.
Schmidt (rebound) hammer and ultrasonic pulse velocity (UPV)
can be used to assess material strength in-situ. Samples of 2.3. Other Laboratory-based Investigation Procedures
damaged material (and undamaged references) may be removed
for laboratory investigation, often by diamond drilling of cores or Thermoluminescence is a laboratory test that can be used to
by careful extraction of lump samples. determine if quartz aggregate particles within concrete have
been heated to temperatures exceeding 300-500 °C [7],
2.2. Petrographic Investigation Techniques although the usefulness of this technique is somewhat
reduced by its limited availability and relatively high cost. A
Petrographic examination of concrete, natural stone and number of other microscopical and chemical analysis meth-
masonry mortar are performed in accordance with methods ods have been used to investigate fire-damaged concrete.
given in ASTM C856 [4], BS EN 12407 [5] and ASTM C1324 [6] These include scanning electron microscopy and mineralogi-
respectively. In the absence of a specific standard procedure for cal analysis by X-ray diffraction (XRD) [8,9]. Thermal analytical

Table 1 – Simplified visual concrete fire damage classification (after [2] ).


Class of Features observed
damage
Finishes Colour Crazing Spalling Reinforcement Cracks Deflection
bars

0 (Decoration required) Unaffected Normal None None None exposed None None
1 (Superficial repair Some peeling Normal Slight Minor None exposed None None
required)
2 (General repair required) Sunstantial loss Pink/red Moderate Localised Up to 25% exposed None None
3 (Principal repair required) Total loss Pink/red or Extensive Considerable Up to 50% exposed Minor None
whitish grey
4 (Major repair required) Destroyed Whitish grey Surface lost Almost total Up to 50% exposed Major Distorted
702 MA TE RI A L S CH A R A CT ER IZ A TI O N 60 ( 20 0 9 ) 7 0 0–7 0 9

methods used include differential thermal analysis (DTA),


thermal gravimetric ananysis (TGA) and derivative thermo- 3. Concrete Structures
gravimetric analysis [8–10]. To date, these methods have been
used mainly for academic research and are not routinely used Concrete buildings most likely to be subjected to fire include
to investigate fire-damaged structures commercially. private and public buildings such as warehouses, offices and
In recent years, research has been conducted into the schools. Other common scenarios involve vehicle fires in car
application of image analysis techniques to assessment of the packs or concrete lined tunnels. Research into the effect of fire
colour changes caused to concrete by fire-damage [11,12]. on concrete and concrete structures has been conducted since
These methods involve using computer software to analyse as least 1922 [13] and a considerable quantity of information is
the colours of digital images captured from finely ground now available.
slices of concrete. Reliance on this method to detect fire- The strength of concrete after cooling varies depending on
damage has a number of drawbacks as different geomaterials temperature attained, the heating rate, mix proportions,
show differing colour changes on heating and some show no applied loading and any external sealing that may influence
changes at all. Unlike petrographic examination, colour image moisture loss from the surface [3]. For temperatures of up to
analysis does not cross-check with other features of damage 300 °C the residual strength of structural quality concrete is
such as microcracking or changes in the optical properties of not severely reduced [14]. Generally, between 300 °C and 500 °C
the cement paste. Image analysis techniques have their the compressive strength reduces rapidly and concrete that
greatest chance of success when calibrated against a set of has been heated in excess of 600 °C will be of no use
reference samples of the same mix, heated to a range of structurally. 300 °C is normally taken to be the critical
temperatures under controlled conditions. For assessment of temperature above which, concrete is deemed to have been
fire-damaged structures these references would need to be significantly damaged. Normally concrete exposed to tem-
cored from undamaged parts of the structure with exactly the peratures above 300 °C is replaced if possible. Otherwise the
same concrete mix. An experienced construction materials dimensions are increased (for example, by reinforced col-
petrographer should always be involved in the assessment umns), depending upon the design loads.
procedure to ensure that any perceived colour changes are not Visually apparent damage induced by heating include
actually an inherent feature of the material, or the result of a spalling, cracking, surface crazing, deflection, colour changes
different degenerative process. and smoke damage. Visual survey of reinforced concrete

Table 2 – Mineralogical and strength changes to concrete caused by heating (compiled from [10,17–19] ).
Heating Changes caused by heating
temperature
Mineralogical changes Strength changes

70–80 °C Dissociation of ettringite, Ca6Al2(SO4)3(OH)12·26H2O causing its depletion in the


cement matrix
105 °C Loss of physically bound water in aggregate and cement matrix commences Minor loss of strength possible (b 10%)
causing an increase in the capillary porosity and minor microcracking
120–163 °C Decomposition of gypsum, CaSO4·2H2O causing its depletion in the cement
matrix
250–350 °C Pink/red discoloration of aggregate caused by oxidation of iron compounds Significant loss of strength commences at
commences at around 300 °C. 300 °C
Loss of bound water in cement matrix and associated degradation becomes
more prominent
450–500 °C Dehydroxylation of portlandite, Ca(OH)2 causing its depletion in the cement
matrix.
Red discoloration of aggregate may deepen in colour up to 600 °C.
Flint aggregate calcines at 250-450 °C and will eventually (often at higher
temperatures) change colour to white/grey.
Normally isotropic cement matrix exhibits patchy yellow/beige colour in cross-
polarised light, often completely birefringent by 500 °C
573 °C Transition of α-to β-quartz, accompanied by an instantaneous increase in Concrete not structurally useful after
volume of quartz of about 5% in a radial cracking pattern around the quartz heating in temperatures in excess of 550–
grains in the aggregate 600 °C
600–800 °C Decarbonation of carbonates; depending on the content of carbonates in the
concrete, e.g. if the aggregate used is calcareous, this may cause a considerable
contraction of the concrete due to release of carbon dioxide, CO2; the volume
contraction will cause severe microcracking of the cement matrix
800–1200 °C Complete disintegration of calcareous constituents of the aggregate and cement
matrix due to both dissociation and extreme thermal stress, causing a whitish
grey coloration of the concrete and severe microcracking. Limestone aggregate
particles become white
1200 °C Concrete starts to melt
1300–1400 °C Concrete melted
M A TE RI A L S CH A RACT ER IZ A TI O N 60 ( 20 0 9 ) 7 0 0 –7 0 9 703

Fig. 1 – A photomicrograph of concrete in plane-polarised light.


Showing both reddened (red, marked R) and calcined (brown
mottled, marked C) flint aggregate particles. Numerous fine Fig. 3 – View of the interior of fire-damaged reinforced
cracks (white) within the cement matrix (dark), some of which concrete structure showing a spalled slab soffit and burnt
radiate from quartz fine aggregate (white) (from [16]). formwork debris.

structure is performed using a classification scheme from The colour of concrete can change as a result of heating [15]
Concrete Society Technical Report No. 33 [2]. This system uses and may be used to indicate the maximum temperature
visual indications of the degree of damage to assign each attained and the equivalent fire duration. In many cases, at
structural member a class of damage from 1 to 5. Each damage above 300 °C a red discoloration is important as it coincides
classification number has a corresponding category of repair, approximately with the onset of significant strength loss. Any
ranging from decoration to major repair. The Concrete Society pink/red discolored concrete should be regarded as being
classification system is summarised in Table 1. suspect [2]. Actual concrete colours observed depend on the
Spalling of the surface layers is a common effect of fires types of aggregate present in the concrete. Colour changes are
and may be grouped into two types. Explosive spalling is most pronounced for siliceous aggregates and less so for
erratic and generally occurs in the first thirty minutes of the limestone, granite and Lytag (shows very little colour change).
fire. A slower spalling (referred to as ‘sloughing off’) occurs as The red colour change is a function of (oxidizable) iron content
cracks form parallel to the fire-affected surfaces leading to a and it should be noted that as iron content varies, not all
gradual separation of concrete layers and detachment of a aggregates undergo colour changes on heating. Also, due
section of concrete along some plane of weakness, such as a consideration must always be given to the possibility that the
layer of reinforcement. Forms of cracking include those pink/red colour may be a natural feature of the aggregate
caused by differential thermal expansion, which often run rather than heat-induced.
perpendicular to the outer surface. Also, differential incom- Petrographic examination is invaluable in determining the
patability between aggregates and cement paste may cause heating history of concrete as it can determine whether
surface crazing. Thermal shock caused by rapid cooling from features observed visually are actually caused by heat rather
fire-fighting water may also cause cracking. than some other cause. In addition to colour changes of

Fig. 4 – View of a spall on a fire-damaged concrete slab soffit


Fig. 2 – View of the exterior of the fire-damaged reinforced associated with a melted embedded plastic reinforcement
concrete frame. bar spacer.
704 MA TE RI A L S CH A R A CT ER IZ A TI O N 60 ( 20 0 9 ) 7 0 0–7 0 9

Fig. 5 – View of a fire-damaged concrete column showing red Fig. 7 – A photomicrograph of fire-damaged concrete in cross-
discoloration of exposed fine aggregate particles (marked by polarised transmitted light. showing anisotropic properties
arrow). and yellow-beige colour of the cement matrix (marked C).
Indicating heating to 500 °C.
aggregate, the heating temperature can be cross-checked with
changes in the cement matrix and evidence of physical
distress such as cracking and microcracking. A compilation
of the changes undergone by concrete as it is heated is the approximately 600 °C in the area represented by the
presented in Table 2. Careful identification of microscopically sample.
observed features allows thermal contours (isograds) to be By determining the position of thermal contours through
plotted through the depth of individual concrete members. In the cross-section of a concrete element, an estimate can also
the most favourable situations contours can be plotted for be made of the likely condition of reinforcement bars. At 200-
105 °C (increased porosity of cement matrix), 300 °C (red 400 °C prestressed steel shows considerable loss of strength; at
discoloration of aggregate), 500 °C (cement matrix becomes N450 °C cold-worked steel losses residual strength and at
wholly isotropic), 600 °C (α-to β-quartz transition), 800 °C N600 °C hot-rolled steel losses residual strength.
(calcination of limestone) and 1200 °C (first signs of melting).
Fig. 1 shows some microscopical features that may be 3.1. Case Study of a Fire-damaged Concrete Structure
observed in fire-damaged concrete (example from [16]). Some
aggregate particles have been reddened indicating that the An investigation was commissioned to determine the extent
concrete has reached at least 300 °C at that point. Particles of of damage caused by a large fire to the reinforced concrete
flint have been calcined and so have been heated to 250- frame of a ten-storey building (Fig. 2). The fire started during
450 °C. The cement matrix is bisected by numerous fine construction and swept through three whole storeys, burning
cracks, some of which radiate from quartz grains in the fine the wooden formwork that was still in-situ after placement of
aggregate fraction. This deep cracking and cracking associated the upper three concrete floor slabs (Fig. 3).
with quartz suggest that the concrete has reached 550-575 °C. The investigation was divided into two phases. The first
Overall we can deduce that the concrete has been heated to phase consisted of a limited trial of on-site visual inspection

Fig. 6 – A photomicrograph in plane-polarised transmitted Fig. 8 – A photomicrograph of concrete showing a crack


light. Showing red discoloration of a flint fine aggregate running parallel to the outer surface and red discoloration
particle near to the spalled outer concrete surface (left, (marked R) of a flint fine aggregate particle (plane-polarised
marked R). Indicating heating to 300-600 °C (from [20]). transmitted light).
M A TE RI A L S CH A RACT ER IZ A TI O N 60 ( 20 0 9 ) 7 0 0 –7 0 9 705

Table 3 – Changes caused by heating of various types of natural stone that may be observed visually or microscopically
(compiled from [9,21,22,27] ).
Heating Stone type
Temperature
Limestone Sandstone Marble Granite

250 °C Pink or reddish-brown Red discoloration starts at Heating marble through a At less than 573 °C, if heating rate
300 °C discoloration starts at 250-300 °C but may not range of temperatures causes is less than 1 °C per minute the
250-300 °C but may not become visible until non-reversible expansion thermal expansion is fully reversible.
become visible until 400 °C known as thermal hysteresis If heating rate is greater than 5 °C per
400 °C minute the expansion is not totally
reversible
400 °C Discoloration becomes
more reddish at 400 °C
600 °C Calcination of calcium Heating above 573 °C Above 600 °C complete Develops cracks or shatter at 573 °C
carbonate commences at causes internal rupturing disruption due to differential due to quartz expansion
600 °C of quartz grains with expansion, becomes friable
associated weakening and and reduces to powder
friability
Clay minerals in the
cement disintegrate
(kaolinite up to 600 °C,
chlorite above 600 °C)
800 °C Calcium carbonate Red discoloration may Differential thermal expansions at
calcines to a grey-white persist until 1000 °C higher temperatures (900 °C) gives rise
powder at 800-1000 °C Any calcium carbonate to tensile and compressive stresses
with associated loss of cement calcines to powder causing permanent strain in the stone
strength at 800-1000 °C causing
disintegration
1000 °C + Melting starts Melting starts Melting starts Melting starts

and petrographic examination of twenty concrete core sam- granite coarse aggregate was naturally pink coloured and
ples in the laboratory. On-site inspection revealed that the fire exhibited no apparent heating-related colour changes. The
was unusual as the seat of fire was very extensive. The worst fine aggregate contained a proportion of flint particles that
damage comprised spalling associated with combustible exhibited well defined colour changes (reddening) as a result
plastic spacers for the reinforcement bars that were cast into of heating (Fig. 6). Other notable features included heat-
the concrete (Fig. 4). It was noted that certain fine aggregate induced mineralogical changes to the cement matrix (Fig. 7)
particles exhibited red discoloration (Fig. 5). and various types of cracking and microcracking (Fig. 8). The
Petrographic examination determined that the concrete results of the Phase 1 investigation indicated that the
comprised crushed granite coarse aggregate and natural sand structure could be economically repaired (representing con-
fine aggregate, bound by a matrix of hardened Portland siderable cost savings over demolition) and that colour
cement with an addition of pulverised-fuel ash (pfa). The changes in the concrete aggregate could be used to determine

Fig. 9 – A photomicrograph showing sandstone that has Fig. 10 – A photomicrograph showing clay brick that includes
been deliberately heated to redden its appearance. Iron flint particles exhibiting both red discoloration (right) and
compounds have been oxidised (black). Quartz grains appear calcination (lower left). The flints were heated during
white and pore spaces are shown yellow (plane-polarised manufacture of the brick (firing) and are therefore not
transmitted light). indicative of fire damage (plane-polarised transmitted light).
706 MA TE RI A L S CH A R A CT ER IZ A TI O N 60 ( 20 0 9 ) 7 0 0–7 0 9

Fig. 11 – A photomicrograph showing red discoloration Fig. 13 – A photomicrograph of fire-damaged mortar in


(marked R) of limestone from a fire-damaged masonry block fluorescent reflected light, showing near surface
(plane-polarised transmitted light). microcracking (yellow) caused by fire-damage (from [28]).

the position of the 300 °C isograd throughout each structural


element. A large scale programme of visual inspection, hammer action drill). The original depth of cover was then
concrete core sampling and reinforcement bar sampling reinstated with sprayed concrete (gunite).
(Phase 2) was then undertaken to determine the depth of fire
damage to every structural element. Phase 2 included the
petrographic examination of two hundred concrete samples 4. Stone and Brick Masonry Structures
which determined that none of the concrete had been heated
to temperatures of much higher than 600 °C. The types of masonry structures most likely to be subjected to
Overall, although the damage was widespread the fire fire include domestic and public buildings and notably, are
damage was generally confined to the outer 5-30 mm of the likely to include buildings of particular historic and cultural
slab soffits and some columns. Strength testing of steel value [21]. Stone and brick masonry can be seriously affected
reinforcement samples indicated that the steel had not been by building fires. The damage tends to be concentrated around
significantly affected by heating. It was concluded that despite window openings and doorways but may also affect structural
the fire being widespread the damage was eminently repair- masonry [22]. A loadbearing wall exposed to fire will suffer a
able. This had been aided by the presence of formwork that progressive reduction in strength due to deterioration of the
had afforded the slab soffits a degree of protection and the mortar in the same manner as concrete. Severe damage is
relatively short duration of fire due to a lack of combustible more likely to be caused by expansion or collapse of other
material. The combined findings of the visual inspections and structural members [23]. At high temperatures (600-800 °C) the
petrographic examinations were then used to draw up a repair strength of most natural stones and masonry mortars is
specification. seriously affected and if thermal shock occurs the stone can
The repairs comprised cutting away the damaged cover disintegrate [22]. Cracking can be caused by quenching
concrete on the floor slab soffits and columns, either by water masonry heated by fire with cold water [24].
jetting (hydro demolition) or manual breaker (usually a large

Fig. 12 – A photomicrograph showing microcracking of Fig. 14 – A photomicrograph of lime mortar in plane-polarised


limestone caused by heating near to the surface of transmitted light, showing red discoloration of chalk
fire-damaged masonry block (fluorescent reflected light). aggregate (left) caused by heating (from [28]).
M A TE RI A L S CH A RACT ER IZ A TI O N 60 ( 20 0 9 ) 7 0 0 –7 0 9 707

Clay bricks can withstand temperatures in the region of


1000 °C or more without damage, but under very severe and
prolonged heating the surface of the brick may fuse. Spalling
may occur with some types of brick particularly of the
perforated type [23]. At low temperatures (250-300 °C) damage
is usually restricted to colour changes, such as reddening of
iron-bearing stones and mortars. Although not structurally
significant, as the colour change is non-reversible it may be
significant for aesthetic reasons, especially in the case of
historic buildings. In some cases, masonry that is well away
from the fire can suffer from smoke staining (blackening).
Water used to fight the fire or water that has ingressed due to
compromised weatherproofing, may cause salt effluorescence
on surfaces and leaching of mortars.
The changes caused by heating of various types of natural Fig. 16 – View showing the edge of the fire-damaged doorway.
stone are shown in Table 3. The most important are the red Brickwork mortar has been discolored red by the fire.
discoloration caused by oxidation of iron compounds which
commences at around at 300 °C. The red discoloration
corresponds with the onset of significant strength loss and it changes in stone include cracking or shattering of quartz
can be used to detect the 300 °C thermal contour in a similar resulting from the α- to β-quartz phase transition at 573 °C and
manner to concrete. Care must be taken to ensure that the the calcination of limestone and marble at 800-1000 °C.
pink/red colour is attributable to heating, as some stones are When looking for fire damage in clay brick, it must be
naturally pink or red. In addition, the author is aware of two remembered that the bricks are typically fired at high
types of natural stone product (one sandstone and one temperature (900-1150 °C for modern brick) during manufac-
granite) that are deliberately heat-treated by the stone ture. The presence and condition of certain brick constituents
producer to change their colour, in order to increase sales. can be used to determine the firing temperature of the brick
Fig. 9 shows a microscopical view of sandstone that was kiln [25]. Fig. 10 shows a microscopical view of a brick from a
deliberately heat-treated to change its colour from light brown historic building containing discolored and calcined flints,
to a more marketable deep red. Other important heat-induced suggesting uneven firing at temperatures of up to 800 °C.
When repairing masonry structures it is desirable (and an
essential requirement when dealing with historic buildings) to
match the original materials to ensure compatibility. Petro-
graphic examination is routinely used to determine the source
of natural stone and the ingredients of mortar for matching in
historic masonry structures [26]. Repair techniques for fire-
damaged masonry structures start with the removal of loose
materials for safety reasons. With historic structures efforts are
usually made to retain as much of the original material as
possible by using various consolidation techniques. Severely
damaged masonry units and mortars are replaced in order to
maintain structural integrity. Damage to any embedded metal
cramps or fixings must also be considered in the repair scheme.

4.1. Case Study of a Fire-damaged Stone Masonry Structure

An investigation was conducted to determine the extent of


damage caused by a fire in a historic stone masonry barn. The
fire was relatively small with a well-defined seat where the
adjacent wall appeared to exhibit significant damage. The
construction comprised a thick masonry wall built from blocks
of limestone with a lime mortar. On-site visual inspection
revealed that the fire damage was localised and that it
consisted of red surface discoloration of a number of lime-
stone blocks and cracking/crazing of both blocks and lime
mortar.
Petrographic examination of core and lump samples
determined that the masonry blocks consisted of biosparite
limestone that exhibited red discoloration (Fig. 11) suggesting
Fig. 15 – Taking core samples at a fire-damaged brickwork that the outer surface had been heated to 300-500 °C. The outer
building. The damage is concentrated around the door. 40-50 mm exhibited a network of microcracks (Fig. 12) and the
708 MA TE RI A L S CH A R A CT ER IZ A TI O N 60 ( 20 0 9 ) 7 0 0–7 0 9

integrity of the stone appeared to have been compromised to a The assessment of fire damage depth on-site by visual
maximum depth of 50 mm from the outer surface. The mortar inspection of colour changes (such as red discoloration of
was found to comprise quartzitic natural sand fine aggregate, aggregate particles) should be performed with care. An
bound by a hardened matrix of non-hydraulic/feebly hydraulic experienced construction materials petrographer should
lime. The mortar sample exhibited a pattern of cracks between always be involved to ensure that features attributed to fire
the outer surface and 70 mm depth and the outer 5-6 mm damage are not in fact natural features, or that the material
exhibited abundant microcracking of the binder (Fig. 13), both has not been deliberately heat-treated for some aesthetic
apparently caused by heating. Chalk aggregate particles reason. If no discoloration is observed it does not necessarily
exhibited discoloration (Fig. 14) to a maximum depth of mean that there has been no fire damage. Microscopical
9 mm, suggesting a 300 °C thermal contour at approximately examination can detect a range of other features caused by
10 mm depth from the outer surface. The outer surface may fire damage thus ensuring that nothing is overlooked.
have been heated to temperatures of up to 500 °C. The Petrography is also the method of choice for the identification
integrity of the mortar appeared to have been compromised to of materials for the sourcing of matching materials for the
a maximum depth of 70 mm from the outer surface with the repair of fire-damaged historic structures.
most concentrated damage in the outer 5-6 mm.
It was concluded that the presence of heat-induced cracks/
microcracks would be expected to increase the susceptibility REFERENCES
of the masonry to weathering on external elevations. How-
ever, as the fire damage was confined to the surface and
not compromising the structural integrity a decision was [1] Office of the Deputy Prime Minister. Fire statistics, United
made to monitor the damage rather than make any immedi- Kingdom, 2004. ODPM Publications; 2006.
ate repairs. [2] Concrete Society. Assessment and repair of fire-damaged
structures; 1990. Concrete Society Technical Report No. 33.
[3] Khoury GA. Effect of fire on concrete and concrete structures.
4.2. Case Study of a Fire-damaged Brick Masonry Structure Prog Struct Eng Mater 2000;2:249–447 2000.
[4] ASTM International. ASTM C856-04, Standard practice for the
Following an arson attack, an investigation was undertaken to petrographic examination of hardened concrete; 2004.
determine the extent of fire damage at a school gymnasium. Philadelphia, USA.
The fire was caused by a car being crashed into the emergency [5] BS EN 12407: Natural stone test methods – petrographic
examination. London: British Standards Institution; 2000.
doors and then set alight (Fig. 15). The construction comprised
[6] ASTM International. ASTM C1324-05, Standard test method
a cavity wall with an inner leaf of concrete blockwork and an
for the examination and analysis of hardened masonry
outer leaf of clay brick masonry. mortar; 2005. Philadelphia, USA.
On-site visual inspection revealed that the brickwork [7] Placido F. Thermoluminescence test for fire-damaged
mortar around the door exhibited red discoloration (Fig. 16). concrete. Mag Concr Res 1980;32(11):112–6.
A number of core samples were diamond drilled from the [8] Handoo SK, Agarwal S, Agarwal SK. Physicochemical,
brickwork around the door. The core holes were also used mineralogical, and morphological characteristics of concrete
exposed to elevated temperatures. Cem Concr Res 2001;32
for a boroscope inspection of the cavity to look for masonry
(7):1009–18.
damage and check the condition of the wall ties. Petrographic
[9] Hajpál M, Török Á. Mineralogical and colour changes of quartz
examination of core samples in the laboratory determined sandstones by heat. Environ Geol 2004;46(3-4):303–532.
that the mortar comprised quartz sand fine aggregate, bound [10] Alarcon-Ruiz L, Platret G, Massieu E, Ehrlacher A. The use of
by a hardened Portland cement matrix that incorporated thermal analysis in assessing the effect of temperature on a
a yellow pigment. Red discoloration and microcracking cement paste. Cem Concr Res 2005;35(3):609–13.
were observed up to 20 mm from the outer surface. The [11] Lin DF, Wang HY, Luo HL. Assessment of fire-damaged mortar
using digital image process. J Mater Civ Eng 2004;16(4):383–6.
clay brick appeared to be undamaged by the heating in all
[12] Short NR, Purkiss JA, Guise SE. Assessment of fire damaged
cases. concrete using colour image analysis. Constr Build Mater
It was concluded that the masonry around the doorway 2001;15(1):9–15.
was significantly compromised and the area up to 1 m around [13] Lea F, Stradling R. The resistance to fire of concrete and
the door frame was dismantled and rebuilt with new reinforced concrete. Engineering 1922;114(2959).
materials. [14] Malhotra HL. The effect of temperature on the compressive
strength of concrete. Mag Concr Res 1956;8(23):85–94.
[15] Bessey GE. The visible changes in concrete or mortar exposed
to high temperatures. Investigations on building fires, Part 2,
5. Conclusions National Building Studies Technical Paper No. 4. London:
HMSO; 1950. p. 6–18.
Petrographic examination has an established track record for [16] Smart S. Concrete- the burning issue. Concrete July/August
accurately determining the depth of fire damage to structural 1999:30–4.
concrete. The technique has also been sucessfully applied to [17] Larbi JA, Nijland TG. Assessment of fire-damaged concrete:
combining petrology and concrete petrography. Proceedings
structures built from natural stone and brick masonry,
of the 8th Euroseminar on Microscopy Applied to Building
including historic buildings. The expense of conducting
Materials, Athens, 4-7 September; 2001. p. 191–8.
petrographic examination can be paid back many times over [18] Lin WM, Lin TD, Powers-Couche LJ. Microstructures of
in costs savings resulting from consequentially well informed fire-damaged concrete. Am Concr Inst Mater J 1996;93
repair decisions and accurate repair specicification. (3):199–205.
M A TE RI A L S CH A RACT ER IZ A TI O N 60 ( 20 0 9 ) 7 0 0 –7 0 9 709

[19] Riley MA. Possible new method for the assessment of [25] Dunham AC. Developments in industrial mineralogy: I. The
fire-damaged concrete. Mag Concr Res 1991;43(155):87–92. mineralogy of brick making. Proc Yorks Geol Soc 1992;49
[20] Ingham JP, Hamm T. Using microscopy and microanalysis to (2):95–104.
investigate concrete. Concrete March 2005:27–9. [26] Ingham JP. The role of light microscopy in the investigation of
[21] Sippel J, Siegesmund S, Weiss T, Nitsch KH, Korzen M. Decay historic masonry structures. Proc R Microsc Soc 2005;40
of natural stones caused by fire damage. In: Přikryl R, Smith BJ, (1):14–22.
editors. Building Stone Decay: From Diagnosis to [27] Koca MY, Ozden G, Yavuz AB, Kincal C, Onargan T, Kucuk K.
ConservationGeological Society Special Publication No. 271; Changes in the engineering properties of marble in fire-exposed
2007. p. 139–51. columns. Int J Rock Mech Min Sci 2006;43(4):520–30.
[22] Chakrabarti B, Yates T, Lewry A. Effect of fire damage on [28] Ingham JP. Investigation of traditional lime mortars - the role
natural stonework in buildings. Constr Build Mater 1996;10 of optical microscopy. Proceedings of the 10th Euroseminar
(7):539–44. on Microscopy Applied to Building Materials, Paisley, United
[23] Tucker DM, Read REH. Assessment of fire-damaged structures. Kingdom, 21-25 June; 2005.
BRE Information Paper IP 24/81. Watford: Building Research
Establishment; 1981.
[24] McLaren IA. Uppark: a study in fire damage to brickwork. In:
Baer NS, editor. Conservation of Historic Brick Structures.
Donhead Publishing Ltd; 1998. pp. 467–469.