Upload
unibo
View
5
Download
0
Embed Size (px)
Citation preview
Franzoni E., Sassoni E., Scherer G.W., Naidu S., Artificial weathering of stone by heat-
ing, Journal of Cultural Heritage 14S (2013) e85-e93, DOI:10.1016/j.culher.2012.11.026
ARTIFICIAL WEATHERING OF STONE BY HEATING
Elisa Franzoni1, Enrico Sassoni2,*, George W. Scherer3, Sonia Naidu4
1 Dipartimento di Ingegneria Civile, Ambientale e dei Materiali, Università di Bologna, via Terracini 28, 40131 Bologna, Italy,
Tel: +39 051 2090329, Fax: +39 051 2090322, e-mail: [email protected] Assistant Professor
2 Dipartimento di Ingegneria Civile, Ambientale e dei Materiali, Università di Bologna, via Terracini 28, 40131 Bologna, Italy,
Tel: +39 051 2090363, Fax: +39 051 2090322, e-mail: [email protected] Ph.D.
* Corresponding Author
3 Department of Civil and Environmental Engineering, Princeton University, Eng. Quad. E-319, Princeton, NJ 08544, USA,
Tel: +1 609 258 5680, Fax: +1 609 258 1563, e-mail: [email protected] Full Professor
4 Department of Chemical and Biological Engineering,
Princeton University, Eng. Quad. E-226, Princeton, NJ 08544, USA, Tel: +1 609 258 9089, Fax: +1 609 258 1563, e-mail: [email protected]
Ph.D.
Franzoni E., Sassoni E., Scherer G.W., Naidu S., Artificial weathering of stone by heat-
ing, Journal of Cultural Heritage 14S (2013) e85-e93, DOI:10.1016/j.culher.2012.11.026
1
ABSTRACT
Since the effectiveness of stone consolidants significantly depends on the weather-
ing level of the stone samples on which they are tested, in this study the suitability
of heating stone to high temperature, as an artificial weathering method to induce
controllable microstructural, physical and mechanical alterations, was investigated.
Three lithotypes with different characteristics were used: Giallo Terra di Siena (GS,
a highly porous calcareous sandstone), Globigerina limestone (GL, a highly porous
limestone) and Pietra Serena (PS, a porous quartzitic sandstone with low porosity).
The lithotypes were characterized in terms of mineralogical composition, pore size
distribution and water absorption, as well as dynamic modulus, static modulus,
compressive and tensile strength. They were then heated for 1 hour, in different
conditions: (i) dry samples were heated to 100, 200, 300 and 400°C; (ii) water-
saturated samples were heated to 200°C; (iii) water–saturated samples were heated
to 200°C and, after cooling to room temperature, re-heated to 400°C. After heating,
all the lithotypes experienced an increase in open porosity and water absorption, as
a consequence of the anisotropic thermal deformation of calcite crystals. Corre-
spondingly, GS and GL exhibited an increasing reduction in mechanical properties
for increasing heating temperature. PS, on the contrary, exhibited an increase in
compressive and tensile strength, which was attributed to chemical-physical trans-
formations undergone by secondary mineralogical fractions (clay minerals, etc.) at
high temperature. All things considered, heating proved to be a fairly effective and
reproducible method to cause artificial weathering in stone samples for the testing
of consolidants. However, depending on the microstructural characteristics of the
lithotypes, the effectiveness of heating may vary significantly, which requires a
case-by-case adjustment of the most suitable heating procedure and the develop-
ment of complementary methods for artificial weathering.
KEYWORDS
Artificial weathering; Heating; Microstructure; Mechanical properties; Thermal
deformation; Stone consolidation; Calcite
Franzoni E., Sassoni E., Scherer G.W., Naidu S., Artificial weathering of stone by heat-
ing, Journal of Cultural Heritage 14S (2013) e85-e93, DOI:10.1016/j.culher.2012.11.026
2
1. INTRODUCTION
Stones used in sculpture and architecture, as well as rocks in their original loca-
tion, are exposed to environmental weathering, which is responsible for alterations
in their microstructure (e.g. open porosity, pore size distribution, chemical-
mineralogical composition of the phases, etc.), that, in turn, result in a modifica-
tion of physical and mechanical properties.
The open porosity and effective porosity (i.e. the interconnected open po-
rosity between two opposite sides of a specimen) generally increase as a conse-
quence of weathering, responsible for the opening of new micro-cracks, the wid-
ening of existing micro-fractures and the dissolution of the most soluble fractions
[1, 2]. For instance, a sensible increase in effective porosity (+4% in the first 5 cm
from the surface, compared to the underlying unweathered part) was detected in
the case of a tuff-made stonework exposed to environmental weathering for 150
years [3]; more dramatic increases (up to 10 times) were found in the case of
rocks weathered to various degrees [1, 4]. Alongside the modification in porosity,
pore size distribution is generally affected by weathering as well. On the one
hand, the average pore size increases when dissolution of the most soluble frac-
tions occurs, so that smaller pores are internally dissolved and/or collapsed, thus
forming larger pores [1, 5]; on the other hand, the average pore size decreases
when clay minerals, formed through weathering, and/or crystallized soluble salts
partially fill the pores, thus reducing the pore size [1]. As a consequence of the
modification in open porosity and/or pore size distribution, also specific surface
area is affected by weathering, generally increasing with increasing weathering
degree [1]. Due to these modifications in the pore system, the physical properties
of weathered stones are altered as well. In particular, water absorption, responsi-
ble for all the weathering mechanisms related to water (e.g., freeze-thaw cycles,
crystallization of soluble salts, swelling of clays, etc.), was found to increase by
up to 5 times in weathered stones, compared to unweathered ones [1]. Moreover,
the increase in pore amount and size is responsible for significant decreases in
mechanical properties of weathered stones. Ultrasonic pulse velocity and dynamic
Franzoni E., Sassoni E., Scherer G.W., Naidu S., Artificial weathering of stone by heat-
ing, Journal of Cultural Heritage 14S (2013) e85-e93, DOI:10.1016/j.culher.2012.11.026
3
elastic modulus were reduced by about 40% in weathered rocks and stoneworks
[3, 4]. Moreover, weathered rocks exhibit a strong decrease in static elastic modu-
lus (-30-60%), compressive strength (-30-80%) and tensile strength (-20-60%), in
comparison with the same unweathered lithotypes [1, 4].
To restore cohesion between grains and mechanical properties of weath-
ered architectural stones, consolidating treatments are usually applied. The effec-
tiveness of such consolidants significantly depends on the weathering level of the
tested stone [6], so the actual performance of new consolidating materials should
preferably be evaluated on weathered specimens. Nevertheless, environmentally
weathered stone samples are rarely available in sufficient quantity and with suita-
bly constant characteristics, hence the importance of developing methods for arti-
ficial weathering of stone, able to provide altered specimens with uniform and re-
producible characteristics, is evident.
Extensive experimental studies on artificial weathering of stone have been
performed with the aim of reproducing and accelerating environmental weathering
processes occurring in the field, i.e. dissolution in clean and acid rain and for-
mation and dissolution of soluble salts, owing to pollutant dry deposition (see re-
view in [7]). These studies, performed on different types of stone (mainly car-
bonate stones), involved a wide range of environmental conditions, in terms of
temperature, relative humidity, presence of atmospheric gaseous pollutants (such
as SO2, NOx, O3) and particulate matter, as well as presence of acidic solutions
(such as H2SO4, HNO3, H2CO3, HCl) (see, among others, [8-16]). However, those
experimental investigations generally aimed at evaluating the rate of stone surface
recession (determined by measuring stone weight loss and/or the amount of Ca2+
ions dissolved from the stone). Therefore, they provide little guidance regarding
the effects of the artificially induced decay on stone microstructure and physical-
mechanical properties that would be of use in producing weathered stone samples
for testing of consolidants.
A more promising method for obtaining artificially damaged stone speci-
mens could be to adopt some methods described in international standards for as-
sessing stone resistance to various weathering mechanisms (such as ageing by
Franzoni E., Sassoni E., Scherer G.W., Naidu S., Artificial weathering of stone by heat-
ing, Journal of Cultural Heritage 14S (2013) e85-e93, DOI:10.1016/j.culher.2012.11.026
4
SO2 action [17], salt crystallization [18], freeze-thaw [19] or thermal shock [20]).
Such methods can be used to intentionally induce damage in the stone [2, 21, 22];
however, these methods exhibit some drawbacks:
(i) the duration of the weathering procedures: according to the cited European
Standards, a prolonged exposure period is required for determining the re-
sistance to SO2 (21 days, [17]), salt crystallization (up to 15 cycles [18]),
freeze-thaw (up to 168 cycles [19]) and thermal shock (up to 20 cycles, unless
failure happens earlier [20]). Accordingly, in the cited studies 5-32 cycles of
salt crystallization [2, 22], 50-75 cycles of freeze-thaw [2, 21] and 32-50 cy-
cles of thermal shock [21, 22] were repeated to achieve significant weather-
ing;
(ii) the conditions of stone samples at the end of the weathering procedure: artifi-
cial weathering by freeze-thaw, salt crystallization and thermal shock is hard-
ly controllable and capable of being modulated to a desired weathering level,
so that weathered samples often exhibit intense cracking and huge material
loss [2, 22]. In addition, samples subjected to salt crystallization are usually
contaminated by high amounts of salts. Such opening of cracks, detachment
of parts and contamination by salts, although being somehow representative
of decay conditions that may be experienced in the field, complicate the eval-
uation of consolidant performance (in terms of chemical reactivity, ability to
bond grains, etc.).
Therefore, in this study, as an alternative artificial weathering method, the
effectiveness of heating stone to high temperature for obtaining damaged speci-
mens with uniform and reproducible characteristics, but still uncontaminated by
salts, was investigated. Heating was selected as a promising damaging method be-
cause:
(i) crystals of calcite, present in most architectural stones, such as limestones,
marbles and calcareous sandstones, are known to undergo a marked aniso-
tropic thermal expansion, expanding parallel and contracting perpendicular to
crystallographic c-axis (respectively, about 0.19% and 0.04% for a thermal
Franzoni E., Sassoni E., Scherer G.W., Naidu S., Artificial weathering of stone by heat-
ing, Journal of Cultural Heritage 14S (2013) e85-e93, DOI:10.1016/j.culher.2012.11.026
5
variation from 20 to 100°C) [23, 24]; such anisotropic deformations are re-
sponsible of stress that can lead to the opening of micro-cracks at grain
boundaries. As a matter of fact, grain decohesion and intergranular cracking
owing to natural heating-cooling cycles have been recognized as a main cause
of weathering processes that typically affect marbles, e.g. bowing and sugar-
ing [24-28];
(ii) crystals of different minerals, which may be present in heterogeneous stones
such as granite and sandstone, have different thermal expansion coefficients
and hence undergo different deformations, which may result in additional mi-
cro-cracks [29];
(iii) the possible presence of water inside the pores, when stone is heated to high
temperature, may be responsible of further damaging, as a consequence of the
pressure exerted on pore walls [30] and of the chemical reaction between wa-
ter and Si-O bonds on the surface of the newly formed micro-cracks [31].
Given the frequent problems of thermal weathering of marbles, many experi-
mental studies have been carried out to analyze the effects of heating-cooling cy-
cles on marble porosity and/or mechanical properties (see for instance [24, 30, 32,
33]). However, these studies were basically aimed at imitating the thermal varia-
tions in the field, hence they generally involved heating-cooling cycles with a rel-
atively low maximum temperature (rarely exceeding 100 °C). The effects of heat-
ing stone samples to higher temperatures – not representative of realistic thermal
excursions, but useful to induce more dramatic alterations in stone microstructure
– have been considered only in few cases and only for marble and granite [29,
32].
The effect of heating as an artificial weathering method for stone was in-
vestigated for the first time in a previous study [6], where Indiana limestone (a
medium porous limestone) was heated at 100, 200, 300, 400 and 500 °C for 1, 4
and 16 hours. For the heated samples, a decrease in dynamic elastic modulus line-
arly proportional to the heating temperature was found. Samples heated to 300 °C
for 1 hour exhibited a decrease of -43.4% in dynamic modulus and of -27.1% in
Franzoni E., Sassoni E., Scherer G.W., Naidu S., Artificial weathering of stone by heat-
ing, Journal of Cultural Heritage 14S (2013) e85-e93, DOI:10.1016/j.culher.2012.11.026
6
tensile strength. Heating duration longer than 1 hour proved to have no further ef-
fect on the dynamic modulus reduction.
In this study, for three lithotypes with different microstructural characteris-
tics, the effectiveness of heating to high temperature (up to 400°C) to induce con-
trollable microstructural, physical and mechanical alterations, has been systemati-
cally investigated. In particular, the effects of heating have been evaluated in
terms of modifications in porosity and pore size distribution, water absorption and
sorptivity, dynamic and static modulus, compressive and tensile strength.
2. EXPERIMENTAL PROCEDURE
2.1 Stones
Three lithotypes with significantly different microstructural characteristics were
selected:
(i) Giallo Terra di Siena (GS), a highly porous calcareous sandstone traditionally
used in architecture in Tuscany (provided by Il Casone S.p.A., Italy). It is
mainly composed of carbonatic granules of algal fragments, shells, fragments
of bivalves and micro-crystalline calcite and may contain low amounts of
quartz, feldspars and fragments of magmatic and metamorphic rocks; it has
fine to very fine grains, partially oriented, bonded by carbonatic cement that
does not completely fill the voids between the grains;
(ii) Globigerina Limestone (GL), a very highly porous limestone, typically used
in modern and historical architecture in Malta (provided by Xelini Skip Hire
and High-Up Service, Malta): it is mainly composed of calcite crystals and
fossils, including globigerinae (from which the name derives), shells and sea
urchins; it may contain low amounts of quartz, feldspars and clays; it has fine
grains, bonded by carbonatic cement that does not completely fill the pores;
(iii) Pietra Serena (PS), a quartzitic sandstone with low porosity, typically used in
historical architecture in Florence (provided by Il Casone S.p.A., Italy): it is
mainly composed of quartz, feldspars, micas and fragments of silicate and
carbonatic rocks; it has medium grains, oriented parallel to the bedding planes
Franzoni E., Sassoni E., Scherer G.W., Naidu S., Artificial weathering of stone by heat-
ing, Journal of Cultural Heritage 14S (2013) e85-e93, DOI:10.1016/j.culher.2012.11.026
7
and bonded by carbonatic cement that almost completely fills the voids be-
tween the grains.
2.2 Stone characterization
2.2.1 Microstructural characteristics
The mineralogical composition was determined by powder X-ray diffraction
(XRD, Philips Diffractometer PW 1840, 40kV/20mA, Cu Kα radiation) and the
carbonate content was assessed on duplicate samples by the Dietrich-Frühling gas
volumetric method (based on the measurement of the CO2 volume released by re-
acting the powdered sample with HCl). The morphological characteristics were
evaluated by observation on a scanning electron microscope (SEM, Philips
XL20). The total open porosity (OP), the pore size distribution and the specific
surface area (SSA) were determined by mercury intrusion porosimetry (MIP,
Fisons Macropore Unit 120 and Porosimeter 2000 Carlo Erba).
2.2.2 Physical properties
The rate of water sorption by capillarity (sorptivity) was measured according to
EN 15801 [34] using 5 cm cubes. To avoid any alterations due to sample drying in
an oven, even at relatively low temperature, the samples were preliminarily dried
at room temperature using a fan until constant weight. To evaluate the effect of
heating on stone sorptivity, the same samples used for the unheated condition
were dried again at the end of the test and then heated to different temperatures,
then the sorptivity of the heated samples was determined (§ 2.4): in this way any
possible uncertainty owing to stone microstructural variability was avoided. Since
all the three lithotypes exhibited some anisotropy owing to bedding planes (§ 3.1),
the sorptivity was measured with the water flux in the direction parallel to the
bedding planes. This orientation was chosen because anisotropic stones are usual-
ly placed in the building with the bedding planes horizontally (i.e. perpendicular
to the loads, so that the strongest direction is stressed), hence, the penetration of
water inside the stone by capillarity (for instance in case of a rain event) happens
in the direction parallel to the bedding planes.
Franzoni E., Sassoni E., Scherer G.W., Naidu S., Artificial weathering of stone by heat-
ing, Journal of Cultural Heritage 14S (2013) e85-e93, DOI:10.1016/j.culher.2012.11.026
8
The water absorption by capillarity (WA) was determined on triplicate
samples (5 cm side cubes) as the total amount of water penetrated after 24 hours,
in the same conditions as the sorptivity test.
2.2.3 Mechanical properties
Dynamic elastic modulus (Ed) was determined on 2 cubes (5 cm side) for each
lithotype and calculated according to the formula Ed = ρV2, where ρ is the bulk
density and V the ultrasonic pulse velocity, measured by means of a commercial
instrument (Matest) with 55 kHz transducers (transmission method), using a rub-
ber couplant between the transducers and the sample. Given the anisotropy of the
lithotypes, the Ed was calculated as the average for the values measured in the
three directions.
Static elastic modulus (E) and compressive strength (σc) were determined
by axial compressive tests perpendicular to the bedding planes on 2 cubes (5 cm
side), using an Amsler loading machine (maximum load 600 kN) at a constant
loading rate of 2.5 kN/s. Strains were measured by two LVDT coupled to the
plates of the loading machine. The reliability of this system had been preliminari-
ly validated by comparing to results obtained in parallel measurement with strain
gauges and an LVDT. The static modulus was calculated as the slope of the
straight part of the stress-strain curve.
Tensile strength (σt) was measured by performing the Brazilian test on 6
cylinders (5 cm height and 2 cm diameter, core-drilled perpendicular to the bed-
ding planes), using an Amsler-Wolpert loading machine (maximum load 100 kN)
at a constant displacement rate of 4 mm/min.
2.3 Artificial weathering by heating
Cubic and cylindrical samples of each lithotype were artificially weathered
according to the following procedures:
(i) heating in a furnace at 100, 200, 300 and 400 °C for 1 hour in dry conditions;
Franzoni E., Sassoni E., Scherer G.W., Naidu S., Artificial weathering of stone by heat-
ing, Journal of Cultural Heritage 14S (2013) e85-e93, DOI:10.1016/j.culher.2012.11.026
9
(ii) heating in a furnace at 200 °C for 1 hour after water saturation, reached by
partially immerging the samples into deionized water and letting them soak
for 24 hours (hereinafter, this condition is referred to as WS+200°C) ;
(iii) heating in a furnace at 200 °C for 1 hour after water saturation (reached as
above) and then, after cooling to room temperature, re-heating to 400 °C in
dry conditions (hereinafter referred to as WS+200+400°C).
All the samples were inserted into the furnace once it had reached the desired
temperature.
2.4 Evaluation of heating effects
The modifications in porosity, pore size distribution and water absorption were
evaluated for samples heated in the WS+200°C condition, at 400°C and in the
WS+200+400°C condition. To prevent variability in stone microstructure from
confounding the evaluation of the heating effects, the samples for the MIP test af-
ter heating were taken very close to the sampling point used for MIP test on the
unheated stone. As for the water absorption test, the same samples used for char-
acterizing the unheated stone were used, after they were dried and then heated in
the three conditions reported above.
Finally, the modification in sorptivity was evaluated by repeating the sorp-
tivity test on the same sample that had been used to characterize the unheated
stone, after it was re-dried at room temperature and heated to 400°C.
3. RESULTS AND DISCUSSION
3.1 Stone characterization
The mineralogical composition and the carbonate content of the three lithotypes
are reported in Table 1. Since both calcite and dolomite are present in sample PS,
the carbonate content, measured by the Dietrich-Frühling method, accounts for
both carbonate minerals and is hence indicated as equivalent calcium carbonate
content in Table 1. The SEM images illustrating the lithotype morphology are re-
ported in Figure 1. The pore size distributions and the results of water absorption
Franzoni E., Sassoni E., Scherer G.W., Naidu S., Artificial weathering of stone by heat-
ing, Journal of Cultural Heritage 14S (2013) e85-e93, DOI:10.1016/j.culher.2012.11.026
10
and sorptivity tests are reported in Figure 2, while the values of OP, average pore
radius, SSA and WA are summarized in Table 2. The stress-strain curves obtained
from the axial compressive test are reported in Figure 3 and the average values of
Ed, E, σc and σt are summarized in Table 4.
GS is a highly calcareous sandstone (CaCO3 = 84.3%), also containing
quartz and Na-feldspar. Its highly porous microstructure can be noted in the SEM
image and from MIP results (OP = 20.12%). It exhibits a quite coarse average
pore radius (rav = 3.5 μm), resulting in moderately high sorptivity and a total water
absorption of 4.0%. GS exhibits a dynamic elastic modulus Ed = 35.3 GPa with
some anisotropy (Ed measured perpendicular to the bedding planes is about 10%
smaller than that parallel to the bedding planes). The static elastic modulus, E, is
sensibly lower, since it amounts to 11.8 GPa. It is typical for Ed to be higher than
E, which has been attributed to the difference in strain amplitude between dynam-
ic and static measurements, the possible presence of fluid inside pores, and the
presence of cracks, clays, and planes of weakness and foliation [35, 36]. Com-
pressive and tensile strength of GS amount to 43.6 MPa and 6.2 MPa, respective-
ly.
GL is a limestone mainly containing calcite (CaCO3 = 91.4%) and small
amounts of quartz. The calcium carbonate is largely owing to fossil micro-
organisms, as can be seen in the SEM image. It exhibits a highly porous micro-
structure (OP = 40.06%, rav = 3.0 μm) resulting in a very high sorptivity and high
water absorption (WA = 14.6%). As a consequence of the very high porosity, GL
exhibits low values of both dynamic and static modulus (Ed = 15.2 GPa and E =
6.5 GPa), E and Ed being quite different for the reasons discussed above. GL
shows a slight anisotropy as well, the Ed perpendicular to the bedding planes be-
ing about 6% higher than Ed measured parallel to them.† Compressive and tensile
strength show pretty small values (σc = 27.5 MPa and σt = 3.1 MPa).
† This is opposite to the typical anisotropy, where the direction normal to the bedding is less stiff; however, the result is reproducible: parallel to the bedding planes Ed = 14.8 and 14.7 GPa, while in the perpendicular direction Ed = 15.6 GPa. After water saturation, Ed decreases in all directions (Ed = 13.1 and 13.0 in the two parallel directions, and Ed = 14.2 in the perpendicular direction), but the sense of the anisotropy is preserved.
Franzoni E., Sassoni E., Scherer G.W., Naidu S., Artificial weathering of stone by heat-
ing, Journal of Cultural Heritage 14S (2013) e85-e93, DOI:10.1016/j.culher.2012.11.026
11
PS is a highly quartzitic sandstone, mainly containing quartz and feldspars,
alongside calcite and dolomite (CaCO3 = 12.8%) and small amounts of micas and
pyroxenes. It exhibits a very dense microstructure (OP = 6.05%) and a small av-
erage pore radius (rav = 0.029 μm) and consequently the sorptivity is very slow
and the total water absorption limited to 1.3%. Consistently, mechanical proper-
ties are high (Ed = 32.2 GPa, E = 11.8 GPa, σc = 109.4 MPa, σt = 8.0 MPa). A
marked anisotropy was detected: Ed measured perpendicular to the bedding planes
is about 35% lower than that measured parallel.
The three lithotypes selected in this study hence show a significant varia-
bility in mineralogical composition and porosity, which was chosen to evaluate
the heating effects on stones with different microstructural characteristics.
3.2 Heating effects
The effects of heating on the porosity, water absorption and sorptivity of the three
lithotypes are illustrated in Figure 2 and summarized in Table 2. The changes in
Ed, E, σc and σt after heating are reported in Table 3 and the modifications in com-
pressive stress-strain curves are illustrated in Figure 3.
In the case of lithotype GS, heating proved to have a significant effect on
both microstructural-physical and mechanical properties. Indeed, an increasing
reduction in mechanical properties for increasing temperatures was experienced,
as suggested by the results reported in Figure 3 (top) and Table 3. In particular,
both dynamic and static modulus exhibited a maximum decrease after heating in
the WS+200+400°C condition (ΔEd = -32.5% and ΔE = -23.5%). Compressive
strength also decreased for increasing temperatures, reaching a minimum after
heating at 400 °C (Δσc = -14.0%), while tensile strength experienced an equal re-
duction (Δσt = -14.3%) after heating at 400 °C and in the WS+200+400°C condi-
tion.
The decrease in mechanical properties can be attributed to the opening of
micro-cracks consequent to the stress induced by calcite anisotropic thermal de-
formation during heating [24, 33]. This is confirmed by the increase in porosity
detected for increasing heating temperatures (Figure 2). The maximum increase in
Franzoni E., Sassoni E., Scherer G.W., Naidu S., Artificial weathering of stone by heat-
ing, Journal of Cultural Heritage 14S (2013) e85-e93, DOI:10.1016/j.culher.2012.11.026
12
porosity (ΔOP = +15.7%) was found after heating at 400 °C, i.e. in the condition
that caused the greatest decrease in compressive and tensile strength. After heat-
ing at 400 °C, the amount of small pores increased in absolute terms (pores having
radius r < 1 μm increased, for instance, from 18.1 mm3/g to 21.2 mm3/g after heat-
ing to 400 °C).
The increases in porosity and amount of small pores led to increases in wa-
ter absorption and sorptivity (Figure 2). Samples heated at 400 °C experienced the
highest WA increase (ΔWA = +20%). Correspondingly, after heating at 400 °C the
results of the sorptivity test exhibited basically no modification in the initial part
of the graph (whose slope depends on the amount and size of the coarser pores,
which were not substantially altered by heating) and an increase in the slope of the
second part of the graph (which is more influenced by the smaller pores, which
increased in number after heating), leading to a higher amount of water absorbed
after 24 hours.
In addition to the effect of calcite crystal anisotropic deformation, some
contribution to stone damage from the presence of water during heating (and dur-
ing microcrack opening) seems present [30, 31], since both Ed and E exhibited an
additional decrease after heating in the WS+200°C condition, compared to 200°C
in dry conditions, and in the WS+200+400°C condition, compared to 400°C in dry
conditions (Table 3).
GL exhibited a decrease in mechanical properties after heating to high
temperature, as shown in Figure 3 (medium) and Table 3. In this case, the maxi-
mum decreases in dynamic and static elastic modulus were found after heating in
the WS+200+400°C condition (ΔEd = -28.2% and ΔE = -14.8%). However, the
reduction in compressive strength in the same heating condition is smaller (Δσc =
-9.8%) and, more significantly, a definite trend of reduction for increasing tem-
peratures cannot be clearly recognized (Table 3). In the case of tensile strength, a
reduction Δσt = -13.7% was found after heating both at 400 °C in dry conditions
and in the WS+200+400°C condition.
The reduction in mechanical properties of GL seems overall to be less pro-
nounced than that experienced by GS, which may be explained considering that the
Franzoni E., Sassoni E., Scherer G.W., Naidu S., Artificial weathering of stone by heat-
ing, Journal of Cultural Heritage 14S (2013) e85-e93, DOI:10.1016/j.culher.2012.11.026
13
very high open porosity of GL (OP = 40.06%, basically double that of GS) can al-
low some deformation of calcite crystals without causing stress in the stone and mi-
cro-cracks opening. Consistently, no significant systematic modification in pore
size distribution was observed (Figure 2), the difference in total open porosity be-
tween unheated and heated samples being more likely attributable to the intrinsic
stone variability than to modifications occurring after heating. However, new pores
and cracks at the nano-scale, not detectable by MIP measurement, may have devel-
oped after heating and be responsible for the decrease in mechanical properties.
According to the small variation in the pore system, the variations in
sorptivity and water absorption are quite small (Figure 2). The highest increase in
WA, experienced after heating in the WS+200+400°C condition, is +7.5% (Table
2), while only small changes in sorptivity were found after heating at 400°C.
In the case of lithotype PS, some significant and partly unexpected modifi-
cations in mechanical properties were found (Figure 3, bottom, and Table 3). Dy-
namic elastic modulus steadily decreased for increasing heating temperatures,
reaching a minimum for the WS+200+400°C condition (ΔEd = -21.3%). However,
for the same heating condition, the decrease in the static elastic modulus only
amounts to 3.4% and no definite trend for increasing temperatures exists (actually,
E value after heating to 400°C is slightly higher than that of unheated samples).
The fact that, for increasing heating temperatures, Ed steadily decreases while E on-
ly shows little changes can be explained considering that, as a consequence of heat-
ing, some micro-cracks opened in the stone (OP increased from 6.05% to 8.58% af-
ter heating at 400°C, Table 2). Such micro-cracks, although causing an increase in
the time required for the ultrasonic wave propagation (hence a decrease in the dy-
namic modulus), apparently had a limited effect on the static modulus, which also
depends on the relative position and orientation of the cracks [36].
Moreover, in spite of the decrease in Ed and insignificant change in E, a
sharp increase in both compressive and tensile strength was measured (Table 3).
Although unexpected, such increase in stone strength after heating is however
consistent with some experimental findings reported in the literature [37, 38].
Those studies report that compressive strength, tensile strength and elastic modu-
Franzoni E., Sassoni E., Scherer G.W., Naidu S., Artificial weathering of stone by heat-
ing, Journal of Cultural Heritage 14S (2013) e85-e93, DOI:10.1016/j.culher.2012.11.026
14
lus of sandstone at high temperature (300 °C in [37] and 400 °C in [38]) exhibit an
increase that was attributed to the “hot-melt effect” of some minerals [38]. To val-
idate the hypothesis that the increase in compressive and tensile strength observed
in this study could be a consequence of some chemical-physical transformation
undergone by secondary mineralogical fractions (e.g., clay minerals), the possible
presence of clays, although not detected by XRD, was further investigated. The
morphology of some fractions observed by SEM suggested that clay minerals
could actually be present (Figure 4, left) but, to have a further confirmation, com-
pressive strength test was performed on dry and water-saturated samples. As a
matter of fact, clay bearing stones are known to undergo a dramatic decrease in
static modulus when they are tested in water-saturated conditions [39], hence a
comparison of the static elastic modulus of dry and water-saturated PS triplicate
samples was performed. Considering that (i) a decrease ΔE = -17% was found for
the water-saturated PS samples (Figure 4, right); (ii) stones containing high
amounts of clays are reported to undergo reductions in ΔE from -45% to -120%
[39]; (iii) fired clay bricks, not containing any clay mineral, underwent reductions
in ΔE of 3-5%, the presence of small amounts of clay minerals seems confirmed.
The increase in compressive and tensile strength found in PS samples after heat-
ing at high temperature could therefore be attributed to some chemical-physical
transformation undergone by clay-type minerals, which led to a more effective
bonding of stone grains.
As a consequence of heating, PS underwent a progressive increase in po-
rosity and amount of smaller pores with increasing temperature. After heating to
400 °C, OP increased by +41.8% and the average pore radius decreased from
0.029 μm in the unheated sample to 0.008 μm. This increase in porosity is due to
the opening of new micro-cracks, owing to the anisotropic thermal deformation of
the calcite crystals contained in the stone cementing fraction and to the differential
thermal deformation of the different minerals. The opening of micro-cracks is also
favored by the very dense microstructure of the stone, which (contrary to the case
of GL) does not allow any deformation of grains without causing stress at the
grain boundaries.
Franzoni E., Sassoni E., Scherer G.W., Naidu S., Artificial weathering of stone by heat-
ing, Journal of Cultural Heritage 14S (2013) e85-e93, DOI:10.1016/j.culher.2012.11.026
15
As a consequence of the increase in porosity, sorptivity and water absorp-
tion increase for samples heated to high temperature (for instance, ΔWA = +23.1%
after heating to 400 °C).
All things considered, heating to high temperature caused significant
changes in microstructural, physical and mechanical properties of the studied
lithotypes: the alterations in OP, WA and mechanical properties after heating at
400 °C (which caused the most significant changes in most of the cases) are sum-
marized in Table 4. Such alterations can suitably resemble alterations owing to
environmental weathering, which generally causes increases in OP and WA and
decreases in mechanical properties. However, in the case of PS, in spite of an in-
crease in porosity and water absorption, heated samples exhibited an increase in
strength, which suggests that preliminary tests should be carried out on each spe-
cific lithotype to set-up the most effective heating procedure.
4. CONCLUSIONS
The results of the present study indicate that heating is a fairly effective and reproduc-
ible method to cause artificial weathering in stone samples that can later be used for
the testing of stone consolidants. Indeed, by using artificially weathered stone speci-
mens, where micro-cracks and grain decohesion have been induced, a more signifi-
cant evaluation of a consolidating treatment’s effectiveness, in terms of binding ac-
tion and strengthening effect, can be achieved, compared to the use of unweathered
stone specimens.
In particular, heating at 400 °C proved to be the most effective weathering
condition in most of the cases. In fact, especially for lithotypes containing high
amounts of calcite (such as GS and GL), the increases in porosity, sorptivity and wa-
ter absorption and the decrease in mechanical properties after heating at 400 °C suita-
bly resemble those experienced by stone subjected to environmental weathering.
Moreover, heating has the advantage of being a controllable method, which provides
samples “weathered” to a desired degree, by choosing the appropriate heating condi-
tion.
Franzoni E., Sassoni E., Scherer G.W., Naidu S., Artificial weathering of stone by heat-
ing, Journal of Cultural Heritage 14S (2013) e85-e93, DOI:10.1016/j.culher.2012.11.026
16
However, depending on the nature and microstructure of the stone subject-
ed to artificial weathering by heating, the effectiveness of the treatment may vary
significantly. In fact, in spite of the satisfactory artificial damage induced in cal-
careous lithotypes (GS and GL) by heating at 400 °C, the same weathering condi-
tion proved to cause no mechanical damaging in the quartzitic sandstone analyzed
(PS). Consequently, the effects of heating on the stone type to be used for testing
of consolidants should be assessed by means of preliminary tests and the most ap-
propriate heating conditions hence tailored to the specific stone type. Further
complementary methods for artificial weathering are anyway worth of investiga-
tion.
ACKNOWLEDGEMENTS
Dr. Gabriela Graziani and Dr. Enzo Padula are gratefully acknowledged for their
collaboration on stone characterization. Dr. Francesco Fusi is gratefully acknowl-
edged for the SEM observations.
REFERENCES
[1] A. Tuğrul, The effect of weathering on pore geometry and compressive strength of selected rock types from Turkey, Eng Geol 75 (2004) 215-227
[2] D.T. Nicholson, Pore properties as indicators of breakdown meachanisms in experimentally weathered limestones, Earth Surf Process Landforms 26 (2001) 819-838
[3] T. Esaki, K. Jiang, Comprehensive study of the weathered condition of welded tuff from a his-toric bridge in Kagoshima, Japan, Eng Geol 55 (1999) 121-130
[4] G. Vasconcelos, P.B. Lourenço, C.A.S. Alves, J. Pamplona, Ultrasonic evaluation of the physical and mechanical properties of granites, Ultrasonics 48 (2008) 453-466
[5] M.G. Orkoula, P.G. Kousoukos, Dissolution effects on specific surface area, particle size and porosity of pentelic marble, J Colloid Interf Sci 239 (2001) 483-8
[6] E. Sassoni, S. Naidu, G.W. Scherer, The use of hydroxyapatite as a new inorganic consolidant for damaged carbonate stones, J Cult Herit 12 (2011) 346-355
[7] E. Franzoni, E. Sassoni, Correlation between microstructural characteristics and weight loss of natural stones exposed to simulated acid rain, Sci Total Environ, 412-413 (2011) 278-285
[8] K.M. Neal, S.H. Newnam, L.M. Pokorney, J.P. Rybarczyk, Elemental analysis of simulated acid rain stripping of Indiana limestone, marble and bronze, In: R Bobain (Ed), Materials deg-radation caused by acid rain, ACS Symposium Series; Washington, DC, 1986. p. 285-300
Franzoni E., Sassoni E., Scherer G.W., Naidu S., Artificial weathering of stone by heat-
ing, Journal of Cultural Heritage 14S (2013) e85-e93, DOI:10.1016/j.culher.2012.11.026
17
[9] R.J. Cheng, J.R. Hwu, J.T. Kim, S.M. Leu, Deterioration of marble structures – The role of acid rain, Anal Chem 59 (1987) 104A-106A
[10] J.B. Johnson, S.J. Haneef, B.J. Hepburn, A.J. Hutchinson, G.E Thompson, G.C. Wood, Labo-ratory exposure systems to simulate atmospheric degradation of building stone under dry and wet deposition conditions, Atmospheric Environment, 24A, 10 (1990) 2285-2592
[11] J.B. Johnson, M. Montgomery, G.E. Thompson, G.C. Wood, P.W. Sage, M.J. Cooke, The in-fluence of combustion-derived pollutants on limestone deterioration: 1. The dry deposition of pollutants gases, Corros Sci 38 (1996) 105-131
[12] J.B. Johnson, M.,,Montgomery, G.E. Thompson, G.C. Wood, P.W. Sage, M.J. Cooke. The in-fluence of combustion-derived pollutants on limestone deterioration: 2. The wet deposition of pollutants species, Corros Sci 38 (1996) 267-278
[13] E. Zendri, G. Biscontin, A. Bakolas, G. Finotto. Simulated study on the chemical and physical decay of the acid rain on carbonate stone, In: J. Riederer (Ed), Proceedings of the 8th Interna-tional Congress on Deterioration and Conservation of Stone, Berlin, 1996, p. 273-279
[14] L.Tecer, Laboratory experiments on the investigation of the effects of sulphuric acid on the deterioration of carbonate stone and surface corrosion, Water Air Soil Poll 114 (1999) 1-12
[15] C.M. Grossi, M. Murray, Characteristics of carbonate building stones that influence the dry deposition of acidic gases, Constr Build Mater 13 (1999) 101-108
[16] M.J. Thornbush, H.A. Viles, Simulation of the dissolution of weathered versus unweathered limestone in carbonic acid solutions of varying strength, Earth Surf Proc Land 32 (2007) 841-852
[17] European Standard EN 13919, Natural stone test methods - Determination of resistance to ageing by SO2 action in the presence of humidity, 2004
[18] European Standard EN 12370, Natural stone test methods - Determination of resistance to salt crystallization, 2001
[19] European Standard EN 12371, Natural stone test methods - Determination of frost resistance, 2010
[20] European Standard EN 14066, Natural stone test methods - Determination of resistance to ageing by thermal shock, 2004
[21] H. Yavuz, Effect of freeze-thaw and thermal shock weathering on the physical and mechani-cal properties of an andesite stone, Bull Eng Geol Environ 70 (2011) 187-192
[22] J. Martìnez-Martìnez, D. Benavente, M.A. Garcìa-del-Cura, Spatial attenuation: The most sensitive ultrasonic parameter for detecting petrographic features and decay processes in car-bonate rocks, Eng Geol 119 (2011) 84-95
[23] S. Siegesmund, K. Ullemeyer, T. Weiss, E.K. Tschegg, Physical weathering of marbles caused by anisotropic thermal expansion, Int J Earth Sci 89 (2000) 170-182
[24] J.M. Logan, Laboratory and case studies of thermal cycling and stored strain on the stability of selected marbles, Env Geol 46 (2004) 456-467
[25] G. Royer-Carfagni, Some considerations on the warping of marble façades: the example of Alvar Aalto’s Finland Hall in Helsinki, Const Build Mater 13 (1999) 449-457
Franzoni E., Sassoni E., Scherer G.W., Naidu S., Artificial weathering of stone by heat-
ing, Journal of Cultural Heritage 14S (2013) e85-e93, DOI:10.1016/j.culher.2012.11.026
18
[26] A.M. Ferrero, M. Migliazza, A. Spagnoli, Theoretical modelling of bowing in cracked marble slabs under cyclic thermal loading, Constr Build Mater 23 (2009) 2151-2159
[27] S. Chatterji, Aspects of generation of destructive crystal growth pressure, J Cryst Growth 277 (2005) 566-577
[28] V. Shushakova, E.R. Fuller Jr, S. Siegesmund, Influence of shape fabric and crystal texture on marble degradation phenomena: simulations, Environ Earth Sci 63 (2011) 1587–1601
[29] S. Chaki, M. Takarli, W.P. Agbodjan, Influence of thermal damage on physical properties of a granite rock: porosity, permeability and ultrasonic wave evolutions, Constr Build Mater 22 (2008) 1456-1461
[30] A. Luque, E. Ruiz-Agudo, G. Cultrone, E. Sebastiàn, S. Siegesmund, Direct observation of microcrack development in marble caused by thermal weathering, Environ Earth Sci 62 (2011) 1375-1386
[31] S.M. Wiederhorn, S.W. Freiman, E.R. Fuller Jr, C.J. Simmons, Effects of water and other di-electrics on crack growth, J. Mater. Sci. 17 (1982) 3460-3478
[32] K. Malaga-Starzec, J.E. Lindqvist, B. Schouenborg, Experimental study on the variation in porosity of marble as a function of temperature. In: Siegesmund S, Weiss T, Vollbrecht A (eds) Natural stone, weathering phenomena, conservation strategies and case studies. Geo-logical Society Special Publication 205 (2002) 81-88
[33] K. Malaga-Starzec, U. Åkesson, J.E. Lindqvist, B. Schouenborg, Microscopic and macro-scopic characterization of the porosity of marble as a function of temperature and impregni-ation, Constr Build Mater 20 (2006) 939-947
[34] European Standard 15801, Conservation of cultural property - Test methods - Determination of water absorption by capillarity, 2010
[35] E. Fjær, Static and dynamic moduli of a weak sandstone, Geophysics 74, 2 (2009) WA102-WA112
[36] J. Martìnez-Martìnez, D. Benavente, M.A. Garcìa-del-Cura, Comparison of the static and dy-namic elastic modulus in carbonate rocks, Bull Eng Geol Environ (2011) 1-6, DOI: 10.1007/s10064-011-0399-y
[37] Rao Qiu-hua, Wang Zhi, Xie Hai-feng, Xie Qiang, Experimental study of the mechanical properties of sandstone at high temperature, J Cent South Univ Technol (2007) s1-0478-06
[38] Zhang LianYing, Mao XianBiao, Lu AiHong, Experimental study on the mechanical proper-ties of rocks at high temperature, Sci China Ser E-Tech Sci 52 (2009) 641-646
[39] I. Jimenez Gonzalez, G.W. Scherer, Effect of swelling inhibitors on the swelling and stress re-laxation of clay bearing stones, Env Geol 46 (2004) 364-377
Franzoni E., Sassoni E., Scherer G.W., Naidu S., Artificial weathering of stone by heat-
ing, Journal of Cultural Heritage 14S (2013) e85-e93, DOI:10.1016/j.culher.2012.11.026
19
Table 1. Mineralogical composition of the lithotypes: C = Calcite; D = Dolomite; Q = Quartz; A = Albite (Na-feldspar); S = Sanidine ((K,Na)-feldspar); M = Mi-crocline (K-feldspar); P = Phlogopite (Mica); B = Bustamite (Pyroxene); CaCO3 = equivalent calcium carbonate content (+ + + = dominantly present; + + = present; + = traces; - = not present)
C D Q A S M P B CaCO3 wt.%
GS + + + - + + - - - - 84.3 (±0.6)
GL + + + - + - - - - - 91.4 (±1.6)
PS + + + + + + + + + + + - + - 12.8 (±1.0)
Franzoni E., Sassoni E., Scherer G.W., Naidu S., Artificial weathering of stone by heat-
ing, Journal of Cultural Heritage 14S (2013) e85-e93, DOI:10.1016/j.culher.2012.11.026
20
Table 2. Microstructural and physical properties of the lithotypes: OP = open po-
rosity; rav = average pore radius; SSA = specific surface area; WA = water absorp-
tion
OP [%] rav [μm] SSA [m2/g] WA [wt. %]
GS
Unheated 20.12 3.516 1.51 4.0
WS + 200 °C 20.81 3.508 1.33 4.2
WS + 200 + 400 °C 22.83 3.832 1.74 4.5
400 °C 23.28 4.563 1.76 4.8
GL
Unheated 40.06 3.087 1.77 14.6
WS + 200 °C 37.41 3.607 1.40 15.1
WS + 200 + 400 °C 39.89 3.055 1.57 15.7
400 °C 36.50 3.104 1.18 15.6
PS
Unheated 6.05 0.029 1.54 1.3
WS + 200 °C 6.52 0.030 1.65 1.3
WS + 200 + 400 °C 6.63 0.022 1.60 1.7
400 °C 8.58 0.008 2.73 1.6
Franzoni E., Sassoni E., Scherer G.W., Naidu S., Artificial weathering of stone by heat-
ing, Journal of Cultural Heritage 14S (2013) e85-e93, DOI:10.1016/j.culher.2012.11.026
21
Table 3. Mechanical properties of the lithotypes: Ed = dynamic elastic modulus; E
= static elastic modulus; σc = compressive strength; σt = tensile strength (Ed, E and
σc values are averages for 2 samples; σt values are averages for 6 samples; stand-
ard deviations in brackets)
Ed [GPa] E [GPa] σc [MPa] σt [MPa]
GS
Unheated 35.3 (±1.0) 11.8 (±0.1) 43.6 (±0.2) 6.2 (±0.5)
100 °C 33.8 (±1.1) 10.7 (±0.2) 42.9 (±3.1) -
200 °C 32.8 (±1.2) 10.7 (±0.6) 40.3 (±0.0) -
WS + 200 °C 31.0 (±0.1) 10.6 (±0.3) 41.4 (±2.8) 5.6 (±0.5)
300 °C 27.8 (±0.0) 10.6 (±0.1) 38.2 (±0.9) -
400 °C 24.9 (±0.7) 9.4 (±0.1) 37.5 (±0.0) 5.3 (±0.8)
WS + 200 + 400 °C 22.7 (±1.6) 9.0 (±0.0) 40.1 (±0.9) 5.3 (±0.3)
GL
Unheated 15.2 (±0.1) 6.5 (±0.2) 27.5 (±0.3) 3.1 (±0.2)
100 °C 14.5 (±0.0) 6.8 (±0.1) 25.3 (±1.6) -
200 °C 14.2 (±0.1) 6.5 (±0.0) 24.4 (±0.2) -
WS + 200 °C 13.6 (±0.4) 6.2 (±0.0) 26.0 (±1.6) 2.9 (±0.3)
300 °C 13.2 (±0.1) 5.8 (±0.1) 23.7 (±4.1) -
400 °C 11.3 (±0.5) 5.7 (±0.1) 26.6 (±1.2) 2.7 (±0.2)
WS + 200 + 400 °C 11.0 (±0.0) 5.5 (±0.1) 24.8 (±0.5) 2.7 (±0.2)
PS
Unheated 32.2 (±0.3) 11.8 (±0.0) 109.4 (±0.0) 8.0 (±0.4)
100 °C 30.6 (±0.0) 10.6 (±0.0) 102.1(±0.0) -
200 °C 30.6 (±0.7) 11.2 (±0.3) 102.4 (±6.6) -
WS + 200 °C 27.7 (±0.2) 11.5 (±0.1) 112.1 (±3.1) 8.0 (±1.1)
300 °C 28.3 (±0.4) 11.2 (±0.6) 106.1 (±10.6) -
400 °C 27.2 (±0.4) 12.0 (±0.0) 122.6 (±2.3) 8.8 (±0.9)
WS + 200 + 400 °C 25.0 (±0.2) 11.4 (±0.5) 114.8 (±0.6) 8.6 (±0.4)
Franzoni E., Sassoni E., Scherer G.W., Naidu S., Artificial weathering of stone by heat-
ing, Journal of Cultural Heritage 14S (2013) e85-e93, DOI:10.1016/j.culher.2012.11.026
22
Table 4. Changes in microstructural, physical and mechanical properties of the
lithotypes after heating to 400°C, compared to unheated condition (OP = open po-
rosity; WA = water absorption; Ed = dynamic elastic modulus; E = static elastic
modulus; σc = compressive strength; σt = tensile strength)
ΔOP [%] ΔWA [%] ΔEd [%] ΔE [%] Δσc [%] Δσt [%]
GS +15.7 +20.0% -29.5 -20.3 -14.0 -14.5
GL -8.9 +6.8% -25.7 -12.3 -3.3 -12.9
PS +41.8 +23.1% -15.5 +1.7 +12.1 +10.0
Franzoni E., Sassoni E., Scherer G.W., Naidu S., Artificial weathering of stone by heat-
ing, Journal of Cultural Heritage 14S (2013) e85-e93, DOI:10.1016/j.culher.2012.11.026
23
Figure 1. SEM images of the three lithotypes GS (top), GL (medium) and PS (bottom)
Franzoni E., Sassoni E., Scherer G.W., Naidu S., Artificial weathering of stone by heat-
ing, Journal of Cultural Heritage 14S (2013) e85-e93, DOI:10.1016/j.culher.2012.11.026
24
Figure 2. Pore size distribution, water absorption and sorptivity of the three lithotypes GS (top),
GL (medium) and PS (bottom) (the graph y-axis has a different scale for different lithotypes to al-
low a better readability)
Franzoni E., Sassoni E., Scherer G.W., Naidu S., Artificial weathering of stone by heat-
ing, Journal of Cultural Heritage 14S (2013) e85-e93, DOI:10.1016/j.culher.2012.11.026
25
Figure 3. Compressive stress-strain curves for the three lithotypes GS (top), GL (medium) and PS
(bottom) (the graph axes have a different scale for different lithotypes to allow a better readability)
Franzoni E., Sassoni E., Scherer G.W., Naidu S., Artificial weathering of stone by heat-
ing, Journal of Cultural Heritage 14S (2013) e85-e93, DOI:10.1016/j.culher.2012.11.026
26
Figure 4. SEM image of sample PS (left) and comparison between static elastic modulus of PS in
dry conditions and after water saturation (right)