ORIGINAL ARTICLE
Direct observation of microcrack development in marble
caused by thermal weathering
A. Luque • E. Ruiz-Agudo • G. Cultrone •
E. Sebastia´n • S. Siegesmund
Received: 18 November 2009 / Accepted: 12 June 2010 / Published online: 26 June 2010
 Springer-Verlag 2010
Abstract One of the properties that makes marble such
an excellent construction and ornamental material is its low
porosity. It is very difficult for water or decay agents to
penetrate the internal structure of materials with no or few
pores, so enhancing the durability of these materials.
However, environmental temperature fluctuations bring
about significant physical changes in marbles that result in
an increase in porosity, due to the appearance of new
microcracks and the expansion of existing ones. These
cracks offer new paths into the marble which make it easier
for solutions containing pollutants to penetrate the material.
Thermal expansion tests were performed on three different
types of marble known as White, Tranco, and Yellow
Macael (Almeria, Spain), after which an increase in
porosity (from 17 to 73% depending on marble type) was
observed, mainly due to crack formation. The structural
changes occurring during thermal expansion tests were
more significant in the case of White Macael samples, a
fact that is not only related to its mineralogical composition
but also to the morphology of the grains, grain boundaries
and crystal size. Our research suggests that thermally
weathered White Macael marble could be more susceptible
to decay by other contaminant agents than Tranco or
Yellow Macael. The use of hot-stage environmental scan-
ning electron microscopy is proposed as a valid tool for
observing, both in situ and at high magnification, changes
in the fracture system of building stones induced by ther-
mal stress.
Keywords Marble  Microcracks 
Thermal expansion anisotropy  Grain boundaries
Introduction
All building stones are exposed to weathering from the
moment they are extracted from the quarry and used in the
construction of a building. They undergo a series of
structural and compositional changes in order to reach a
new thermodynamic equilibrium (Mingarro 1996; Aires-
Barros 2000). Apart from these natural changes, building
stones are also subject to different physical, chemical and
biological weathering processes (Ku¨hnel 2000) that may
affect their durability as structural and ornamental mate-
rials (Mingarro 1996; Doehne 2002).
Because of its low porosity, marble has historically been
considered a high quality material, and has been used in
many important civil and religious buildings. Unfortu-
nately, today there are numerous examples of historic
marble buildings and sculptures which show weathering
phenomena caused by thermal decay and the following
action of soluble salts (e.g. the churches of San Marco,
Santa Maria del Giglio and Santa Maria dei Miracoli
in Venice, Michelangelo’s David in Florence and the
Courtyard of the Lions in the Alhambra of Granada).
A. Luque (&)  E. Ruiz-Agudo  G. Cultrone  E. Sebastia´n
Department of Mineralogy and Petrology,
Faculty of Sciences, University of Granada,
Avenida Fuentenueva s/n, 18002 Granada, Spain
e-mail: analuque@ugr.es
E. Ruiz-Agudo
Institut fu¨r Mineralogie, University of Mu¨nster,
Corrensstr. 24, 48149 Mu¨nster, Germany
S. Siegesmund
Department of Structural Geology and Geodynamics,
Geoscience Centre, University of Go¨ttingen,
Goldschmidtstr. 3, 37077 Go¨ttingen, Germany
123
Environ Earth Sci (2011) 62:1375–1386
DOI 10.1007/s12665-010-0624-1
Environmental temperature fluctuations produce a series of
initial physical and mechanical changes in marble stones
(i.e. the first stage of weathering) that later enhance the
effect of other weathering mechanisms (Battaglia et al.
1993; Siegesmund et al. 2000, 2007). Granular decohesion
and bowing of marble due to temperature fluctuations have
been reported in some cases, particularly when the stone is
used in fac¸ades such as in Alvar Aalto’s Finland Hall in
Helsinki (Royer-Carfagni 1999), the Grande Arche de la
Defense in Paris, the Lincoln Tower in Rochester (Cohen
and Montiero 1991) and the Amoco building in Chicago
(Logan et al. 1993). Kessler (1919) found that repeated
heating may lead to permanent dilatation in marbles due to
the formation of microcracks. Other authors (Bortz et al.
1988; Thomasen and Ewart 1984; Winkler 1996) have
claimed that changes in moisture content may be respon-
sible for the deformation of marbles. More recently, Koch
and Siegesmund (2004) and Siegesmund et al. (2008)
discovered that the bowing that occurs in some marbles is
controlled by a combined effect of thermal cycling and the
presence of moisture. However, it seems that the response
of marble to temperature oscillations is mainly due to the
thermal anisotropy of its mineralogical components: calcite
and/or dolomite (Kleber 1959). The thermal expansion
coefficient, a, for these two minerals shows an extreme
directional dependence, as a result of their different crys-
tallographic directions. Parallel to the c-axis, both minerals
have an a value of about 26 9 10-6 K-1. However, par-
allel to the a-axis, dolomite shows a positive a value of
about 6 9 10-6 K-1, whereas calcite has a negative a
value (-6 9 10-6 K-1) (Grimm 1999; Weiss et al. 1999).
For instance, in experiments carried out on marbles with
different degrees of deterioration, porosity increased by 50%
or more compared to the original material, when samples
were subjected to thermal cycles with temperatures of over
50C (Koch and Siegesmund 2004; Malaga-Starzec et al.
2002). The increase in porosity is the consequence of the
effect produced by the marked anisotropy of calcite crystals.
When temperature rises, the crystal expands in one direction
(i.e. along the c-axis) and contracts perpendicularly to that
direction. Such movements cause internal cleavages and
the separation of crystals from their borders (Siegesmund
et al. 2000).
Even though the thermal decay process affects dolomite
and calcite marbles quite differently, the residual strain
does not seem to be controlled exclusively by the compo-
sition, as there are other intrinsic factors that also deter-
mine their behaviour when subject to thermal expansion
(Zeisig et al. 2002; Siegesmund et al. 2009). The rock
fabric, which includes grain size, grain aspect ratio, grain-
shape preferred orientation, lattice preferred orientation
(texture) and microcrack populations, plays an important
role in how the marble behaves when subjected to thermal
stress (Siegesmund et al. 2000; Royer-Carfagni 1999;
A˚kesson et al. 2006).
The main physical change produced by thermal oscil-
lations is the change in the pore size distribution, even
when the porosity is low (around 2%) (Ruedrich et al.
2001; Siegesmund et al. 2008). The opening of new
cleavage cracks between grain boundaries in marbles due
to thermal changes increases the porosity of the stone and
in most cases increases the number of large pores, and as a
consequence, facilitates the penetration of water and
solutions containing soluble salts or other pollutant agents
into intergranular spaces (Zeisig et al. 2002; Ruiz-Agudo
et al. 2008; Luque et al. 2009). This then causes different
weathering phenomena such as salt crystallization, car-
bonate dissolution and/or the formation of calcium sul-
phate, which occur not only on the surface, but also inside
the marble (Fassina et al. 2002; Simon and Snethlage
1993).
The aim of this paper is to characterize the changes in
the porous system of three different types of marble (two
calcitic marbles and one dolomitic) during thermal/
humidity tests. Textural modification will be monitored
using ultrasounds, mercury intrusion porosimetry, Ar
adsorption and hot-stage environmental scanning electron
microscopy (ESEM). The use of this last technique is
proposed as a novel approach to study, both in situ and at
high magnification, how grain boundaries are affected by
the residual strain generated by one thermal cycle. This
tool can be used to evaluate textural modifications caused
by thermal dilatation in dry conditions.
Materials and methods
Marbles
Three different varieties of marble were used, two of which
were calcitic, White Macael (WM) and Tranco Macael
(TM), and one dolomitic, Yellow Triana Macael (YM).
These types of rocks are widely used as building materials,
mainly for cladding, flooring and paving, and sometimes
show signs of decay. All these marbles are quarried in the
same geographic area, the ‘‘Comarca del Ma´rmol’’ (the
Marble County) in Almeria (Spain), although they are
extracted from different quarries. The WM marble is
quarried in Macael, while TM is quarried in Lubrı´n-
Zurgena and YM in Codbar. In geological terms, these
three marbles are Late Triassic and belong to the Nevado-
Filabride Complex in the Sierra de los Filabres (Betic
Internal Zone), which is the lowest tectonic unit of the
Alboran Domain (Balanya´ and Garcı´a-Duen˜as 1986): TM
and YM are Nevado-Lubrı´n units and WM is a Be´dar-
Macael unit (Weijermars 1991).
1376 Environ Earth Sci (2011) 62:1375–1386
123
Methodology
Textural analysis of selected marbles was performed using
an Olympus BX-60 polarized optical microscope (OM)
coupled with digital microphotography (Olympus DP-10).
The spatial and geometrical configuration of all the com-
ponents of the three marbles in terms of fabric and
microstructure was determined using the methodology
proposed by Passichier and Trouw (1996), where normally
the Z-axis is perpendicular to the foliation (Fig. 1). In order
to obtain more information about the type of grain
boundaries in each marble, we treated the negatives of
optical micrographs using Photoshop Elements 2.0 to
provide the same brightness, contrast and gamma values
for all the samples.
Due to their non-destructive nature, ultrasounds (US) are
particularly useful for determining the physical properties
of building stone. Measurements were performed using the
transmission method and three measurements were taken
for each spatial direction (X, Y and Z). These data were
used to infer information on the degree of compactness of
marbles (a decrease in the velocity suggests the develop-
ment of fissures) as well as on the textural anisotropy of
marbles (DM, in %), the value of which can be calculated
as follows:
DM ¼ 1  ð2  VpminÞ
Vpmax þ Vpmid
 100
where Vpmax is the maximum, Vpmin is the minimum and
Vpmid the medium value for ultrasonic wave velocity
(Guyader and Denis 1986; Weiss et al. 2002b; Sa´ez-Pe´rez
and Rodrı´guez-Gordillo 2009).
The flow of water into the stone pore system was
determined by carrying out a water absorption test (WA).
Real and apparent density and open porosity were mea-
sured by forced WA according to the UNE-EN 1936 (2007)
standard. The modifications in the distribution of the pore
access size and the pore/fissure volume of the marbles
before and after the thermal stress test were determined
using a Micromeritics Autopore III 9410 mercury intrusion
porosimeter (MIP), which is able to exert a maximum
injection pressure of 414 MPa. Ar-sorption isotherms (GS)
of sample fragments before and after the thermal salt tests
were obtained at 77 K using a Micromeritics Tristar 3000
under continuous adsorption conditions. In samples with
less than 5 m2/g surface area, Ar-sorption measurements
are more realistic than N2 measurements that usually yield
excessively high values and BET analysis was used to
determine the total specific surface area (Brunauer et al.
1938). The BJH method (Barrett et al. 1951) was used to
obtain pore size distribution curves, the pore volume and
the mean pore size of the samples. The surface fractal
dimension, Ds, was used to characterize surface roughness.
The analysis of the gas sorption isotherm using a modified
Frenkel–Halsey–Hill theory (Tang et al. 2003) allows the
determination of surface fractal dimension from the slope
(A) of the plot of ln(V) versus ln[ln(P/P0)], where V is the
adsorbed volume of gas, and P and P0 are the actual and
the condensation gas pressure. When surface tension (or
capillary condensation) effects are important, the relation-
ship between A and Ds is A = Ds - 3. Capillary conden-
sation is significant if d = 3(1 ? A) - 2 \ 0. The pressure
range and hence the thickness range of the adsorbed layer
being studied was only around monolayer (n = 1–2) cov-
erage to ensure that the determination of Ds was reliable
(Tang et al. 2003).
The degree of thermal anisotropy of the marbles was
evaluated by performing a thermal expansion test with
respect to specific orientations (X, Y, and Z), according to
pre-established axes. The test was carried out following the
methodology proposed by Koch and Siegesmund (2004) in
a chamber which allows the simultaneous analysis of six
samples. 10 cycles were performed: 3 in dry conditions and
7 under wet conditions. In order to simulate temperature
changes similar to those observed in buildings, each cycle
follows the same temperature sequence of 20–90C and
back down to 20C again over 15 h in dry conditions, and
17 h in wet conditions. The heating rate was 1/min to
ensure the thermal equilibration of the specimens.
This technique allows us to calculate the thermal
expansion coefficient (a):
a ¼ Dl
l þ DT
which expresses the relative change in length (l) or volume
of the sample due to temperature changes (DT), as well as
the thermal expansion ðers ¼ Dlrt=lrÞ; which is the ratio of
the change in length of the sample after cooling down to
room temperature (Dlrt) to the original sample length (lr).
The residual strain (r) generated by the thermal expansion
Fig. 1 Schematic representation of the marble samples with the
reference axes positioned according to the foliation planes
Environ Earth Sci (2011) 62:1375–1386 1377
123
can also be obtained using this test. This parameter is a
measurement of the irreversible damage that takes place in
the sample once it returns to its initial (environmental)
temperature (Kessler 1919).
An ESEM equipped with a heating stage was used to
observe the formation of new cracks and the widening or
closure of pre-existing ones during the following thermal
cycle: 20–45–90–20C. The images were obtained on a FEI
Quanta 400 ESEM, which operates at an accelerating
voltage of 20 kV. During heating, the detector-sample
distance was set to *12 mm and the ESEM chamber
pressure was set at *2 Torr water vapour. This water
vapour pressure is equivalent to that of environmental air at
20C and 15% RH. Each sample was heated at an average
heating rate of between 3 and 5C/min. A constant tem-
perature was maintained during image acquisition, after
15 min as equilibration time.
The stone pore system (pore volume, pore size distri-
bution, surface area and fractal dimension) was character-
ized using WA, MIP and GS (Xie et al. 1996; Pe´rez Bernal
and Bello Lo´pez 2000, 2001). Surface area and fractal
dimension are important as they frequently indicate the
presence of surface rugosity due to chemical weathering
(Ruiz-Agudo et al. 2008).
Results
Characterization of marbles
Although samples are quarried in the same area, OM
analysis showed important differences in the petrography
of the three varieties of marble we tested (Table 1). White
Macael, a calcitic marble, has a granoblastic micro-fabric
with polygonal shapes, a grain size between 0.1 and
3 mm and straight to slightly curved grain boundaries
(Fig. 2a). Microcracks and open cleavage planes are
straight and intra-particular. The micro-fabric in XY-plane
indicates a static recrystallization, in which grain bound-
aries become straighter and grains increase in size
becoming hexagonal in shape (Luque et al. 2009). Tranco
Macael, a white calcitic marble with a notable presence
of irregular grey bands, shows a granoblastic micro-fabric
with irregular shapes and grain sizes between 0.2 and
1 mm (up to 1.5 mm). Grain boundaries are mainly
‘‘interlobate’’, although occasionally pseudo-linear unions
were observed (Fig. 2b). Yellow Macael is a yellowish
dolomitic marble which includes disperse calcite grains of
residual origin and with preferred orientation and sec-
ondary calcite filling cracks and fractures. It shows a
granoblastic micro-fabric with sizes of 0.05–0.8 mm for
dolomite and 0.5–1 mm for calcite grains and it has
interlobate grain boundaries. Fe-oxides, quartz, pyrite,
muscovite and feldspars are also found as accessory
minerals (Fig. 2c).
The main petrophysical properties of the quarried mar-
bles are shown in Table 2. According to Weiss et al.
(2002b), the propagation velocity of ultrasonic waves (Vp)
within the stone matrix can determine the intrinsic and
extrinsic properties of marbles, and therefore, the structural
anisotropy of these crystalline materials, as well as the
existence of microcracks when the Vp values are measured
in dry and saturated conditions (Siegesmund et al. 1999,
2009). Taking into account the Vp values measured by
Dandekar (1968) in a single calcite crystal (Vpmax =
7,730 m/s and Vpmin = 5,710 m/s) and a single dolomite
crystal (Vpmax = 8,450 m/s and Vpmin = 6,280 m/s), we
can see that all three marbles types always have lower
values, but the same tendency of single calcite and dolo-
mite crystals. Table 2 shows the dry and water-saturated
values, and the anisotropy in the three cases.
This suggests that the three types of marble show some
degree of textural preferential orientation, when the c-axis
(minimum values in the three marbles) is parallel to the
z-axis established in our coordinate system. Moreover,
microcracks, whose existence was confirmed by comparing
dry and water-saturated values, have certain directionality,
perpendicular to the c-axis (maximum difference between
Vpdry and Vpsat).
The existence of pores (or fissures), of which there are
generally very few, is confirmed by the porosity values
obtained by water absorption, mercury intrusion porosi-
metry and gas adsorption tests in fresh marbles (Table 2).
Df and Ds parameters are consistent with the fissure
morphology and the degree of fissure surface roughness
observed under OM. Therefore, we can affirm that WM
has a fissure system with rectilinear trend morphology
(Df = 2.77) and high surface roughness (Ds = 2.88),
YM has more irregular morphology fissures (Df = 2.79)
and low surface roughness (Ds = 2.68) and TM has
Table 1 Mineralogical and petrographic features of the three marbles
we tested
White Macael Tranco Macael Yellow Macael
Mineralogical composition (%)
Cc 99 ± 1 98 ± 2 4 ± 1
Dol – – 95 ± 3
Others 1 ± 0.01 1 ± 0.01 1 ± 0.1
Grain size (mm)
Cc 0.1–3 0.2–1.5 0.5–1
Dol – – 0.05–0.8
Others B0.01 B0.01 B0.01
Texture Granoblastic Granoblastic Granoblastic-seriate
Cc calcite, Dol dolomite
1378 Environ Earth Sci (2011) 62:1375–1386
123
interlobate fissures (Df = 2.87) and high surface rough-
ness (Ds = 2.82).
These values give us some idea of the pore structure at
different scales; in particular, Ds can be considered as an
index of the pore structure at the nanoscale, i.e. the rugosity
of the pore surface.
Thermal expansion tests
Table 3 shows the thermal expansion coefficient (a, in
10-6 K-1), thermal expansion (ers, in mm/m) and residual
strain (r, in mm/m) values obtained along the X, Y and Z
perpendicular directions for each marble during the first
thermal dry cycle (20–90–20C), as well as the degree of
anisotropy of the marbles. WM and YM samples show the
highest a and ers values. However, in the three marbles, the
highest value for these two parameters is observed along
the Z-axis, which may indicate a preferred orientation of
carbonates along this direction. Moreover, if we take into
account the anisotropy values of each parameter (Da and
De) in the three marbles, we can see that the anisotropy
in WM (Da = 53.15%; De = 60.37%) and TM (Da =
62.13%; De = 71.43%) marbles is much higher than in
YM marble (Da = 13.67%; De = 13.73%), which indi-
cates that calcite causes a stronger anisotropy than dolo-
mite. These values only reflect the behaviour of the
marbles during the first dry cycle and they tend to remain
constant during the following two dry cycles (Fig. 3).
However, when these cycles are performed in wet condi-
tions, a and ers values show a significant increase. This is
mainly due to the effect of water and the pressure it exerts
at high temperature (90C), when it is retained in the grain
boundaries of the marble samples (Winkler 1994). Finally,
after 10 thermal cycles (3 dry cycles and 7 wet cycles) have
Table 2 Petrophysical characterization of White Macael, Tranco Macael and Yellow Macael (fresh samples)
White Macael Tranco Macael Yellow Macael
x y z DM (%) x y z DM (%) x y z DM (%)
Ultrasound tests, Vp (m/s)
Dry samples 5.885 5.756 5.058 13.10 6.210 5.678 5.387 9.37 6.597 6.573 5.165 21.56
Water-saturated samples 6.405 6.386 6.246 2.34 6.589 6.253 6.073 5.42 7.452 7.304 6.816 7.62
(a)
Water-access porosity, A (vol.%) 0.41 0.35 0.94
Real density by water absorption,
qrock (g/cm
3)
2.70 2.73 2.92
(b)
Porosity, P (%) 1.76 0.75 2.42
Real density, qrock (g/cm
3) 2.72 2.75 2.92
Surface area, SA (m2/g) 0.185 0.258 0.754
Fractal dimension calculated
using MIP data, Df
2.77 2.87 2.79
(c)
Pore volume, P vol. (cm3/g) 0.00006 0.00021 0.00022
Surface area, SA BET (m2/g) 0.198 0.316 0.347
Fractal dimension calculated
using GS data, Ds
2.88 2.82 2.68
Table 3 Thermal parameters of White Macael, Tranco Macael and Yellow Macael marbles
White Macael Tranco Macael Yellow Macael
X Y Z D (%) X Y Z D (%) X Y Z D (%)
a (10-6 K-1) 16.67 9.52 23.97 53.15 4.66 10.69 13.92 62.13 12.91 13.34 16.57 13.67
ers (mm/m) 0.86 0.43 1.31 60.37 0.18 0.54 0.72 71.43 0.66 0.68 0.85 13.73
r (mm/m) 0.31 0.09 0.29 69.80 0.08 0.06 0.13 42.57 0.01 0.01 0.07 74.97
r* (mm/m) 0.6423 0.2789 0.7723 60.57 0.3229 0.4843 0.5101 35.05 0.1695 0.1222 0.4677 61.63
a thermal dilatation coefficient, ers thermal expansion, r residual strain after first thermal dry cycle, r* residual strain after ten thermal cycles (3 in
dry conditions and 7 in wet conditions) (mm/m), D anisotropy determined in each marble for each parameter
Environ Earth Sci (2011) 62:1375–1386 1379
123
been carried out, the residual strain (r) is the main
parameter that evaluates the damage induced in the three
types of marble. Although r values are very different in
the three marbles, all of them increase as the number of
cycles increases. The increase follows this order: YM
(X = 0.01 mm/m; Y = 0.01 mm/m; Z = 0.07 mm/m)\
TM (X = 0.08 mm/m; Y = 0.06 mm/m; Z = 0.13 mm/m)
\WM (X = 0.31 mm/m; Y = 0.09 mm/m; Z = 0.29 mm/
m) after the three first dry cycles; YM (X = 0.17 mm/m;
Y = 0.12 mm/m; Z = 0.47 mm/m)\TM (X = 0.32 mm/m;
Y = 0.48 mm/m; Z = 0.51 mm/m)\WM (X = 0.64 mm/m;
Y = 0.28 mm/m; Z = 0.77 mm/m) after the following
seven wet cycles (Fig. 4).
To quantify the damage induced by the thermal expan-
sion test in each marble, the main petrophysical properties
(compactness and porosity) were measured again in marble
core samples. The new values are shown in Table 4. The
decrease in Vp values was higher in WM marble (between
37 and 55%) compared to YM (between 29 and 45%) and
TM (between 19 and 25%), and this may be related to the
damage induced by thermal changes, leading to an
important loss of compactness (Ko¨hler 1991; Siegesmund
et al. 2000; Weiss et al. 2002a). By the end of the tests,
porosity had increased in all three types of marble. The
pore size distribution had also changed, as was confirmed
by the values for fractal dimension, Df and Ds.
Although an increase in porosity was observed in all
three marble types, they each underwent different changes
in pore size distribution. Figure 5 shows the pore size
distribution of the samples before and after thermal
expansion tests. The number of pores of less than 1 lm
increased in TM and YM marbles; however, WM marble
experienced a substantial increase in the number of pores
of around 1 lm and above 30 lm, with no significant
change in the pore volume below 1 lm. This coincides
with the observed magnitude of Ds for the three marbles.
The main changes in the WM pore system occur in pores of
[1 lm, which cannot be quantified by GS; and this is why
the Ds value remains relatively unchanged compared to that
for the fresh sample (DDs = 1.4%). On the contrary, TM
and YM samples show significant modifications in the pore
system below 1 lm, which is reflected by a higher relative
decrease of the fractal dimension determined using GS
(DDs is 5.3 and 4.5% for TM and YM, respectively).
Hot-stage ESEM
The use of hot-stage ESEM allowed a direct observation of
the evolution of the marble microcracks system during
thermal cycles.
ESEM images were obtained using the same tempera-
ture range used in the thermal expansion test (from 20C to
45, 90 and again to 20C). At 20C (Fig. 6a), no modifi-
cations were observed, but when the temperature rose to
45C the samples started to suffer textural modifications,
Fig. 2 Optical microscopy image of a White Macael, b Tranco
Macael and c Yellow Macael marble micro-fabric
Fig. 3 Maximum elongation of marbles heated up to 90C during the
first three consecutive dry cycles along the three perpendicular
directions (X, Y and Z). WM White Macael, TM Tranco Macael, YM
Yellow Macael
1380 Environ Earth Sci (2011) 62:1375–1386
123
which were further enhanced when the temperature
reached 90C. Nevertheless, although grain boundaries in
the three marbles were observed to seal when the temper-
ature was reduced back to 20C, none of them returned to
their initial state. In the case of the WM marble, slight
textural changes were detected at 45C (Fig. 6b); at this
temperature, the space between the calcite grains began to
widen. This change was better observed when the tem-
perature was increased to 90C (Fig. 6c), and microcracks
that were just a few microns wide (*1 lm) but over
50 lm in length appeared. When the marble returned to its
starting temperature (20C), the separation between the
crystals decreased (*0.5 lm) but the length remained
the same, allowing connectivity between microcracks
(Fig. 6d). The behaviour of TM marble during the thermal
cycle in the ESEM chamber was considerably different.
Slight changes were detected in the marble crack and fis-
sure system when the temperature rose to 90C; at the end
of the first thermal cycle, when the marble went back to
20C, cracks and fissures had sealed and were almost
invisible (Fig. 7a–d) and connection along the grain
boundaries was lost. Finally, in spite of the different min-
eralogical composition of YM marble, significant altera-
tions also occurred in its fissure system when samples were
subjected to temperature fluctuations. This marble behaved
in the opposite manner to the TM marble, as crack opening
reached a maximum when the sample temperature fell back
to 20C after reaching 90C. At this temperature (20C),
Table 4 Petrophysical characterization of White Macael, Tranco Macael and Yellow Macael after the thermal expansion tests
White Maacel Tranco Macael Yellow Macael
x y z DM (%) x y z DM (%) x y z DM (%)
Ultrasound test, Vp (m/s)
Dry samples 2.647 3.625 2.794 17.53 5.010 4.550 4.057 15.13 4.714 4.567 2.866 38.24
Water-saturated samples 5.985 6.214 5.964 2.22 6.148 6.023 5.577 8.36 6.824 6.965 6.367 7.65
Difference between fresh and altered
dry samples (%)
55 37 45 26 19 20 25 13 29 31 45 24
(a)
Water-access porosity, A (vol.%) 0.67 0.48 0.92
Real density by water absorption,
qrock (g/cm
3)
2.71 2.74 2.93
(b)
Porosity, P (%) 6.63 1.21 2.91
Real density, qrock (g/cm
3) 2.72 2.70 2.92
Surface area, SA (m2/g) 1.342 0.573 0.449
Fractal dimension calculated using
MIP data, Df
2.69 2.92 2.94
(c)
Pore volume, P vol. (cm3/g) 0.00005 0.000111 0.00020
Surface area, SA BET (m2/g) 0.116 0.143 0.205
Fractal dimension calculated using
GS data, Ds
2.84 2.67 2.56
Fig. 4 Residual strain versus number of cycles for White Macael (WM), Tranco Macael (TM) and Yellow Macael (YM) marbles during thermal
expansion tests (3 dry cycles and 7 wet cycles). The dotted line divides dry (left) from wet (right) cycles
Environ Earth Sci (2011) 62:1375–1386 1381
123
the development of microcracks was evident (Fig. 8a–d).
These cracks were large (*2 lm 9 10 lm), but there was
less connection between the grains than with WM.
Discussion and conclusions
The results of our research have shown that numerous
factors contribute to marble weathering due to thermal
changes. All of these results indicate that marble mineral-
ogical composition, fabric, grain size and shape, and the
type of union between crystals are the main factors influ-
encing marble behaviour towards thermal changes.
Four aspects of the hot-stage ESEM test should be
considered. The first two are based on the technique we
have used; the other two depend on the finite element
models applied to the expansion and development of
microcracks in marbles with the increase of temperature
(Weiss et al. 2002a, 2003).
1. This test confirms that the main decay agent in marble
is thermal change, due to the anisotropic expansion of
Fig. 5 Pore size distribution plots for the three marbles tested, in fresh and altered samples, after the thermal expansion tests
Fig. 6 ESEM images of White
Macael marble surfaces during
thermal cycles in the
microscope chamber. Cracks are
observed to widen as
temperature is increased
(a 20C, b 45C, c 90C,
d 20C), and are still visible
once the sample returns to 20C
(d). Black arrows indicate the
position of microcracks
1382 Environ Earth Sci (2011) 62:1375–1386
123
calcite and dolomite crystals. Thermal expansion
harms calcite marbles more than dolomite marbles
and this effect is even more dangerous if the marbles
have straight grain boundaries.
2. The ESEM technique allows us to view directly the
formation and propagation of microcracks generated
by thermal stress in marbles, and, thus gain a better
knowledge of the kinematics displayed by the crystals
and the grain boundaries during heating.
3. From the images obtained during the thermal test in all
marbles, the greatest expansion and the largest sepa-
ration between grain boundaries occur at the highest
temperature selected in this work (90C). The micro-
cracks are intergranular and/or transgranular.
4. It is important to bear in mind that, once microcracks
have developed, the elastic energy is mitigated and the
largest concentration of residual elastic energy moves
toward the boundaries between the grains. When the
marble returns to its initial temperature at the end of a
heating cycle (20C), most of the edges of the
carbonate grains along the xz-plane are brighter (see
Figs. 6, 7, 8), which may indicate a higher concentra-
tion of electrical charge produced in this area by
increased residual energy.
5. The development of microcracks obtained with this
experiment is consistent with the porosity observed
using MIP. In the case of WM, the microcracks are
bigger than in TM and YM. This justifies our MIP
results for this marble, which showed an excessive
pore volume (over 6%) and a range of pores of over
10 lm. We can, therefore, also conclude that a large
grain size with straight grain boundaries helps the
formation of microcracks of great length and
connectivity.
These four aspects show that the hot-stage ESEM test is
an effective technique for directly observing the different
mechanisms of expansion–contraction and the distribution
of microcracks generated during the thermal change in
different types of marble.
From the results of the thermal expansion tests and the
in situ observations during hot-stage ESEM simulation, it is
clear that the three selected marbles, regardless of their
mineralogy, fabric or texture, undergo significant changes
in their crack and fissure systems.
If we compare the petrographic properties of fresh and
weathered samples, it can be inferred that well-developed
crystal shapes, larger grain size and linear grain
Fig. 7 ESEM images of Tranco
Macael marble surfaces during
thermal cycles in the
microscope chamber. Slight
changes in the crack system are
observed during the temperature
rise (a 20C, b 45C, c 90C,
d 20C); however, when the
sample returns to 20C, the
fractures seal up (d). Black
arrows indicate the position of
microcracks
Environ Earth Sci (2011) 62:1375–1386 1383
123
boundaries are the main properties responsible for the
dramatic effects of environmental thermal oscillations on
the pore system of the marbles and, as a consequence, in
their potential susceptibility to weathering. In the case of
the WM marble, its strong textural anisotropy and high
degree of thermal expansion are the main parameters that
determine its response to thermal changes. Its bigger
crystal size (compared to TM and YM samples) and its
simple, almost linear grain boundaries may result in a
high degree of thermal expansion and the development of
microcracks. As thermal expansion is a linear property
and the induced elongation is going to be proportional to
the initial length of the axes under consideration, it seems
reasonable to assume that the highest elongation is going
to take place in the marble with the biggest grain size
(WM) and, vice versa, the lowest elongation will occur in
marbles with smaller grain sizes (TM and YM). This may
result in higher impact energies when bigger crystals
interact than when smaller grains do. On the other hand,
as has already been mentioned, the type of grain bound-
aries is an important factor which determines the behav-
iour of marbles during thermal tests. Simple grain
boundaries indicate lower binding energy between them
and interlobate-type unions reflect higher binding energy
between grains; thus, crystal separation will occur more
easily in WM marble than in TM and YM marbles, which
show tortuous boundaries and a higher binding energy.
This is important when considering the results of the
study of the porous system by MIP and GS, as these tests
help to predict the behaviour of marbles that are thermally
altered when they come into contact with soluble salts or
other contaminant agents. The most pronounced modifi-
cation of the stone porous system after the thermal
expansion tests [in terms of porosity and pore size dis-
tribution, as well as other parameters such as compactness
(inferred from the value of the propagation velocity of
ultrasound waves), surface area or fractal dimension] was
observed for WM marble. These changes can be
explained by the formation of new linear fractures
(*1 lm in size) or the widening of pre-existing ones
(*10–100 lm). As explained earlier, even though these
openings are large, they are consistent with those pro-
duced (during a single cycle) in the hot-stage attached to
the ESEM. The opening of new linear fractures exposes
new, flat surfaces that result in lower fractal dimension
compared to fresh samples. In general, these modifica-
tions may enhance the weathering action of dissolved
contaminants such as soluble salts or sulphuric acid,
Fig. 8 ESEM images of
Yellow Macael marble surfaces
during thermal cycles in the
microscope chamber. Widening
of cracks is observed as the
temperature (a 20C, b 45C,
c 90C, d 20C) is increased,
and is still visible once the
sample returns to 20C (d).
Black arrows indicate the
position of microcracks
1384 Environ Earth Sci (2011) 62:1375–1386
123
mainly due to both increased accessibility of such pollu-
tants to the stone matrix and the increase in the volume of
material affected by decay agents.
Another interesting aspect is that the marbles we studied
exhibit different mechanisms of grain rearrangement when
they return to 20C after a thermal cycle. ESEM image
sequences confirm such differences. Whereas in the case of
TM marble calcite grains reorganize themselves resulting
in the closure of cracks, in WM and YM marbles the cracks
that open or widen during thermal cycles remain visible
after the test has finished. Differences in crystal shape and
size, and overall, grain size distribution may explain the
mechanisms observed in each marble. Equidimensional
and homogeneously sized crystals (such as those observed
in WM and YM samples) cannot easily rearrange, which
means that cracks remain open. Non-equidimensional
crystals with a polydisperse size distribution can be reor-
ganized in a more compact way, thus resulting in the clo-
sure of cracks and fissures. This may also help to interpret
the residual strain values, which are higher in the case of
TM marble and the changes in the rate of ultrasound wave
propagation and porosity after thermal tests, both of which
indicate that TM samples show the smallest degree of
alteration when subjected to thermal oscillations. We can,
therefore, deduce that in TM marble a temperature rise
results in length changes, but once the temperature falls
back to its original level, grain reorganization leads to
crack closing, which is in turn reflected in the small change
in porosity and pore size distribution.
We are aware that the duration of the test is limited (10
cycles in total) and that the test does not exactly reproduce
the real weathering process that occurs when a material is
exposed to temperature changes for decades or centuries.
However, it is likely that the structural modifications will
be more pronounced as the number of cycles is increased.
This work offers an accurate representation of the physical
processes taking place during exposure to environmental
temperature oscillations, and also allows us to compare
different materials in terms of their response to thermal
stress. The results of this work may contribute to a better
understanding of the processes that cause the weathering of
marbles used as building or ornamental materials. In par-
ticular, the environmental temperature oscillations that
generally affect marbles result in changes in the structure
of the pore system (i.e. a first stage of decay) of the stones
that enhances or facilitates the potential action of other
contaminant agents.
Acknowledgments This research was financed by Research Project
FQM 1635, the Integrated Action HA 2007-0012, the European
Commission VIth Framework Program (Contract no. SSP1-CT-2003-
501571) and Research Group RNM-179 (Junta de Andalucı´a, Spain).
We thank I. Sanchez-Almazo (CEAMA, Junta de Andalucı´a-
Universidad de Granada) for her assistance with ESEM analysis.
References
Aires-Barros L (2000) Modes and mechanisms of rock weathering.
Protection and conservation of the cultural heritage of the
Mediterranean cities. In: Proceedings of the 5th international
symposium on the conservation of monuments in the Mediter-
ranean Basin, Seville, Spain, pp 3–6
A˚kesson U, Lindqvist JE, Schouenborg B, Grelk B (2006) Relation-
ship between microstructure and bowing properties of calcite
marble claddings. Bull Eng Geol Environ 65:73–79
Balanya´ JC, Garcı´a-Duen˜as V (1986) Grandes fallas de contraccio´n y
de extensio´n implicadas en el contacto entre los dominios de
Albora´n y Sudibe´rico en el Arco de Gibraltar. Geogaceta 1:19–21
Barrett EP, Joyner LS, Halenda PPJ (1951) The determination of pore
volume and area distributions in porous substances. I. Compu-
tations from nitrogen isotherms. J Am Chem Soc 73:373–380
Battaglia S, Franzini M, Mango F (1993) High sensitivity apparatus
for measuring linear thermal expansion: preliminary results on
the response of marbles. Il Nuovo Cimiento 16:453–461
Bortz SA, Erlin B, Monk CB (1988) Some field problems with thin
veneer building stones, new stone technology, design and
construction for exterior wall systems. American Society for
Testing and Materials, Philadelphia, pp 11–31
Brunauer S, Emmett PH, Teller EJ (1938) Adsorption of gases in
multimolecular layers. J Am Chem Soc 60:309–319
Cohen JM, Montiero JM (1991) Durability and integrity of marbles
cladding: a state of the art review. J Perform Constr Facil
5(2):113–124
Dandekar DP (1968) Variation in the elastic constants of calcite with
pressure. AGU Trans (Washington, DC) 49(1):323
Doehne E (2002) Salt weathering: a selective review. In: Siegesmund
S, Weiss T, Vollbrecht A (eds) Natural stone, weathering
phenomena, conservation strategies and case studies. Special
Publications, vol 205. Geological Society, London, pp 51–64
Fassina V, Favaro M, Naccari A (2002) Principal decay patterns on
Venetian monuments. In: Siegesmund S, Weiss T, Vollbrecht A
(eds) Natural stone, weathering phenomena, conservation strat-
egies and case studies. Special Publications, vol 205. Geological
Society, London, pp 381–391
Grimm W (1999) Beobachtungen und U¨berlegengen zur Verformung
von Marmorprojekten durch Gefu¨geauflockerung. Zeitschrift der
Deutschen Geologischen Gesellschaft 150(2):195–236
Guyader J, Denis A (1986) Propagation des ondes dans les roches
anisotropes sous constrainte evaluation de la qualite des schistes
ardoisiers. Bull Eng Geol 33:49–55
Kessler DW (1919) Physical and chemical test of the commercial
marbles of the United States. Technologic Papers of the Bureau
of Standards, 123. Government Printing Office, Washington, DC
Kleber W (1959) Einfu¨hrung in die Kristallographie. VEB Verlag
Technik, Berlin
Koch A, Siegesmund S (2004) The combined effect of moisture and
temperature on the anomalous expansion behaviour of marble.
Environ Geol 46:350–363
Ko¨hler W (1991) Untersuchungen zu Verwitterungsvorga¨ngen an
Carrara-Marmor in Potsdam-Sanssouci. In: Berichte zu Forschung
und Praxis der Denkmalpflege in Deutschland, Steinscha¨den-
Steinkonservierung 2, pp 50–53
Ku¨hnel RA (2000) Driving forces of rock degradation. Protection and
conservation of the cultural heritage of the Mediterranean cities.
In: Proceedings of the 5th international symposium on the
conservation of monuments in the Mediterranean Basin, Seville,
Spain, pp 7–9
Logan JM, Hadedt M, Lehnert D, Denton M (1993) A case if study of
properties of marble as building veneer. Int J Rock Mech Min
Sci Geomech 30:1531–1537
Environ Earth Sci (2011) 62:1375–1386 1385
123
Luque A, Cultrone G, Mosch S, Siegesmund S, Sebastian E, Leiss B
(2009) Anisotropic behaviour of White Macael marble used in
the Alhambra of Granada (Spain). The role of thermohydric
expansion in stone durability. Eng Geol (available on line, 6 July
2009)
Malaga-Starzec K, Lindqvist JE, Schouenborg B (2002) Experimental
study on the variation in porosity of marbles as a function of
temperature. In: Siegesmund S, Weiss T, Vollbrecht A (eds)
Natural stone, weathering phenomena, conservation strategies
and case studies. Special Publications, vol 205. Geological
Society, London, pp 81–88
Mingarro F (1996) Procesos de alteracio´n. Degradacio´n y conserva-
cio´n del Patrimonio Arquitecto´nico. Cursos de verano del
Escorial, Complutense, pp 387–398
Passichier CW, Trouw RAJ (1996) Microtectonics. Springer, Berlin
Pe´rez Bernal JL, Bello Lo´pez MA (2000) The fractal dimension of
stone pore surface as weathering descriptor. Appl Surf Sci
161:47–53
Pe´rez Bernal JL, Bello Lo´pez MA (2001) Fractal geometry and
mercury porosimetry. Comparison and application of proposed
models on buildings stones. Appl Surf Sci 185:99–107
Royer-Carfagni G (1999) Some considerations on the warping of
marble facades: the example of Alvar Alto’s Finlandia Hall in
Helsinki. Constr Build Mater 13:449–457
Ruedrich J, Siegesmund S, Richter D (2001) Marble columns and
their state of weathering: structural evidences and ultrasonic
tomography. Z Dtsch Geol Ges 152:665–680
Ruiz-Agudo E, Luque A, Sebastia´n E, Rodrı´guez-Navarro C (2008)
Changes in the pore structure of marble after salt decay tests. In:
Libro de Actas del IX Congreso Internacional de Rehabilitacio´n
del Patrimonio Arquitecto´nico y Edificaciones, pp 147–150
Sa´ez-Pe´rez MP, Rodrı´guez-Gordillo J (2009) Structural and compo-
sitional anisotropy in Macael marble (Spain) by ultrasonic, XRD
and optical microscopy methods. Constr Build Mater 23:2121–
2126
Siegesmund S, Weiss T, Vollbrecht A, Ullemeyer K (1999) Marbles
as a natural building stone: rock fabrics, physical and mechanical
properties. Z dt Geol Ges 150/2 Stuttgart (Schweizerbart):237–
257
Siegesmund S, Ullermeyer K, Weiss T, Tschegg EK (2000) Physical
weathering of marbles caused by anisotropic thermal expansion.
Int J Earth Sci 89:170–182
Siegesmund S, Koch A, Ruedrich J (2007) Ursachen mangelnder
Formbesta¨ndigkeit von Fassadenplatten: Fallstudie Universita¨ts-
bibliothek der Universita¨t Go¨ttingen. In: Siegesmund S, Ehling
A (eds) Rohstoff Naturstein Teil 1. Z dt Ges Geowiss 158:630–
648
Siegesmund S, Ruedrich J, Koch A (2008) Marble bowing: compar-
ative studies of three different public building facades. Environ
Geol 56:473–494
Siegesmund S, Kracke T, Ruedrich J, Schwarzburg R (2009) Jewish
cemetery in Hamburg Altona (Germany): state of marble
deterioration and provenance. Eng Geol (available on line, 22
July 2009)
Simon S, Snethlage R (1993) The first stages of marbles weathering,
preliminary results after short-term exposure of nine month.
Forschungs-Bericht (Eurocare-Euromarble EU 496), Mu¨nchen
(Bayrisches Landesamt fu¨r Denkmalpflege) 11:37–44
Tang P, Chew NYK, Chan H-K, Raper JA (2003) Limitation of
determination of surface fractal dimension using N2 adsorption
isotherms and modified Frenkel–Halsey–Hill theory. Langmuir
19:2632–2638
Thomasen SE, Ewart CS (1984) Durability of the thin-set marble. In:
Third international conference on durability of building mate-
rials and components, pp 313–323
UNE-EN 1936:2007, Me´todos de ensayo para piedra natural,
Determinacio´n de la densidad real, aparente, la porosidad abierta
y total. AENOR
Weijermars R (1991) Geology and tectonics of the Betic Zone, SE
Spain. Earth Sci Rev 31:153–236
Weiss T, Leiss B, Oppermann H, Siegesmund S (1999) Microfabric
of fresh and weathered marbles: Implications and consequences
for the reconstruction of the Marmopalais Postdam. Zeitschrift
der Deutschen Geologischen Gesellschaft 152(2):313–332
Weiss T, Siegesmund S, Fuller ER (2002a) Thermal stresses and
microcracking in calcite and dolomite marbles via finite element
modelling. In: Siegesmund S, Weiss T, Vollbrecht A (eds)
Natural stone, weathering phenomena, conservation strategies
and case studies. Special Publications, vol 205. Geological
Society, London, pp 65–80
Weiss T, Rasolofosaon PNJ, Siegesmund S (2002b) Ultrasonic wave
velocities as a diagnostic tool for the quality assessment of
marble. In: Siegesmund S, Weiss T, Vollbrecht A (eds) Natural
stone, weathering phenomena, conservation strategies and case
studies. Special Publications, vol 205. Geological Society,
London, pp 149–164
Weiss T, Siegesmund S, Fuller ER (2003) Thermal degradation
of marbles: indications from finite-element modelling. Build
Environ 38:1251–1260
Winkler EM (1994) Stone in architecture. New York
Winkler EM (1996) Technical note: properties of marble as building
veneer. Int J Rock Mech Min Sci Geomech 33:215–218
Xie H, Wang J, Qan P (1996) Fractal characters of micropore
evolution in marbles. Phys Lett A 218:275–280
Zeisig A, Siegesmund S, Weiss T (2002) Thermal expansion and its
control on the durability of marbles. In: Siegesmund S, Weiss T,
Vollbrecht A (eds) Natural stone, weathering phenomena,
conservation strategies and case studies. Special Publications,
vol 205. Geological Society, London, pp 65–80
1386 Environ Earth Sci (2011) 62:1375–1386
123