Construction and Building Materials 156 (2017) 891–901

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier.com/locate/conbuildmat

Flexural strength of weathered granites: Influence of freeze and thaw

cycles
Sarkar Noor-E-Khuda a,⇑, Faris Albermani b, Martin Veidt c
a
School of Engineering and Technology, Central Queensland University, Perth, WA 6000, Australia
b
School of Engineering and Technology, Central Queensland University, Mackay, QLD 4740, Australia
c
School of Mechanical and Mining Engineering, The University of Queensland, Brisbane 4072, Australia

h i g h l i g h t s

Non-destructive testing of Australian Granite’s.

Effect of freeze-thaw cycles on the flexural strength of granite.

Ultrasonic pulse velocity of unweathered and weathered granite.

Correlation between flexural strength and UPV.

a r t i c l e i n f o a b s t r a c t

Article history: The effect of freeze-thaw cycles (FTC) on the flexural strength of granite panels is investigated in this
Received 23 May 2017 paper. Specimens from three different types of Australian granite were sampled. Fifteen specimens were
Received in revised form 11 September used as control (unweathered) specimens while another 36 specimens were subjected to accelerated
2017
weathering consisting of 100 FTC over a period of 34 days. A controlled freeze-thaw chamber and a tem-
Accepted 11 September 2017
perature range between 10 °C and +70 °C was used for this purpose. Ultrasonic pulse velocity (UPV) of
the specimens was measured before and after FTC testing. Flexural strength tests were conducted after
the FTC were completed. Results showed that thin granite veneers suffered physical degradation accom-
Keywords:
Ultrasonic pulse velocity
panied by a noticeable reduction in UPV, density and flexural strength following FTC. Correlation between
Granite flexural strength and UPV and between unweathered and weathered UPV is proposed for the three types
Flexural tensile strength of Australian granite used in this paper.
Freeze-thaw cycles Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction
Failure of cladding panels can lead to expensive restoration
works as well as serious risks from falling debris. Numerous inci-
dents of stone cladding failures have been reported around the
world [1]. A notable example is the Amoco building in Chigaco that
cost around US $75 million to replace the façade stone work. In
Australia, stone cladding failure was reported in the Melbourne
Metropolitan Board building, 210 George St building in Sydney
and 344 Queen St building in Brisbane, among others. There is a
notable increase in the number of new high-rise buildings around
Australia. With the recent safety concerns regarding aluminium
panels cladding, it is expected that the use of thin granite veneers
as a façade solution on high-rise buildings will increase.
⇑ Corresponding author.
E-mail addresses: s.noorekhuda@cqu.edu.au (S. Noor-E-Khuda), f.albermani@
cqu.edu.au (F. Albermani).
Over the past few decades, an increasing demand for aestheti-
cally pleasing, structurally adequate thin structural stone veneer
as cladding panels on buildings, particularly high-rise, has been
observed. Among thin structural cladding veneers, Granite is the
most commonly used (Pires et al. [2]). Being superior to other thin
structural stone veneers in terms of strength and durability, gran-
ite has received less research attention. In particular, a limited
research is available on the effect of long-term freeze-thaw cycles
(FTC) on the flexural strength of thin granite veneer [3]. It is gener-
ally accepted that weathering has the propensity to degrade the
strength of building stones and thin cladding panels over time
and reduce their capacity to perform under seismic and cyclonic
loadings during their service life [3,4].
Although destructive testing methods (such as four point bend-
ing test) are still preferred in conventional engineering practice,
recent developments in non-destructive testing (NDT) methods
such as acoustic emission, rebound hammer, laser scanning and
ultrasonic pulse velocity (UPV) have facilitated detailed structural

http://dx.doi.org/10.1016/j.conbuildmat.2017.09.049
0950-0618/Ó 2017 Elsevier Ltd. All rights reserved.
892 S. Noor-E-Khuda et al. / Construction and Building Materials 156 (2017) 891–901

investigation and retrofitting of modern structures [5]. Several Engineering laboratory of the University of Queensland. Details of
researchers adopted NDT techniques in micro level material charac- the experimental program are discussed in this section.
terisation and damage detection [5,6]. The UPV method has been
successfully employed for assessing the mechanical behaviour of
2.1. Material and test specimens’ description
concrete [7], masonry [6] and various thin structural stone veneers
[8,9]. High frequency sound waves above 20 kHz are referred to as
Three common Australian granite types were considered in this
ultrasound. These waves are transmitted through the material
study. These are; Calca, Imperial Black (or Adelaide Black) and
under test and the velocity of the wave passing through is termed
Grandee. Calca and Imperial Black stones are quarried in the Eyre
as the ultrasonic pulse velocity. The measured UPV is a function of
Peninsula and Black Hill regions of South Australia while Grandee
the inherited mineral composition, material properties and defects.
in the Mulliandry district of central New South Wales. These three
Vasconcelos et al. [10] used UPV to assess the strength, stiffness and
stone types are characterised by high compressive strength, hard-
fracture energy of granite. Furthermore, Chen et al. [11] successfully
ness, durability and visible crystalline structure with interlocking
estimated crack depth and direction and determined the thickness
crystals of medium to coarse grain. The three stones are distin-
of damaged surface layers using UPV. This paper refers to work con-
guishable by their appearances as shown in Fig. 1; with Calca hav-
ducted on the structural properties evaluation of various granite
ing pink/red clusters on black appearance, Imperial Black with dark
specimens [9–11] using the ultrasonic NDT method.
black appearance and Grandee with dark grey appearance. The
It is well known that thin structural stone veneers possess low
specimens were prepared from 32 mm granite slabs using laser
tensile strength, usually <10% of its compressive counterpart.
guided saw.
While more research is focused on the compressive strength of thin
Sample dimensions were selected based on ASTM C880/C880M
structural stone veneers [3,11], the flexural tensile behaviour has
specifications [21] and were approximately L400  W100  t
received fewer attention. The direct tensile and compressive
32 mm, with the exact dimension of each sample being specified
strength of granite specimens were reported in [10] from series
in Table 2. A total of 36 specimens (12 specimens of each stone
of destructive/non-destructive tests. A correlation between the
type, S1–S12, Table 2) were subjected to freeze-thaw cycling
strength and UPV was established [10] which showed that the
before conducting flexural test. Another 15 control specimens (5
direct tensile strength increased exponentially while the compres-
sive strength increased linearly with UPV. A number of studies
have reported on the influence of weathering due to salt crystalli-
sation [8], wetting/flooding [6,12] and heating [13] on the mechan-
ical properties of thin structural stone veneers. Franzoni et al. [13]
tested several stones including limestone and sandstone under
cycling heating between 100  400 °C, showed that the mechanical
properties of limestone degraded with temperature, while the
sandstone specimens showed the opposite characteristics, which
triggered the need to investigate the influence of FTC on the mate-
rial characteristics of granite. While FTC testing procedure has
been standardized, the guidelines provided in ASTM D5312/
D5312M-12 [14] and EN 12371 [15] reflect different climates and
practices and hence vary in-terms of the number of applied cycles,
heating/cooling duration and the required temperature range. In-
fact FTC for different rocks has been simulated using 50, 200 and
1400 FTC cycles in [16,3,17], respectively. Furthermore, various
temperature ranges were adopted; for example [3,18,19,20] used
temperatures range of (40 °C to 40 °C), (40 °C to 180 °C),
(20 °C to 20 °C) and (7 °C to 14 °C), respectively. Most of the
Australian cities experience a wide range of temperature from
above +50 °C in summer to 10 °C in winter. Furthermore, a sharp
temperature difference between daytime and nighttime tempera-
tures is common within the 24hrs cycle. For these reasons, a tem-
perature range of 10 °C to +70 °C is selected for the accelerated
FTC used in this work.
The effects of freeze-thaw cycling on the flexural strength of a
series of granite cladding panels using UPV and four point bending
tests are investigated in this paper. Specimens from three different
types of Australian granite, namely; Calca, Imperial Black and
Grandee were considered in this study. The specimens were par-
tially immersed in water and placed in a controlled freeze-thaw
chamber to simulate accelerated weathering through FTC. Based
on test results, correlations between UPV of weathered and
unweathered specimens and between UPV and flexural strength
are established.

2. Experimental program

Ultrasonic NDT, freeze-thaw cycling and flexural testing

of the granite cladding specimens were conducted at the Civil Fig. 1. Granite types (a) Calca; (b) Imperial Black; (c) Granadee.
 S. Noor-E-Khuda et al. / Construction and Building Materials 156 (2017) 891–901 893

specimens of each stone type, C1–C5, Table 2) were subjected to Table 1

flexural testing to ascertain the unweathered flexural strength. Petrographic description.

The pulse velocity test set-up depends on: the natural reso-
nance frequency of the used transducers (f), the minimum dimen-
sion of the test specimen (t) and the average grain size of the rock
used (d). The thickness (t) of the test specimens was checked
against the UPV test guidelines of ASTM D2845-05 [22] using Eq.
(1)
t P 5ðv =f Þ P 5d
Since the pulse propagation velocity ðv Þ and the grain size (d)
are intrinsic material properties and the coupon’s nominal thick-
ness, t = 32 mm, was set to meet the flexural test requirements; tri-
als were used to select suitable pulse frequencies for each test run.
Furthermore, since veneer claddings are commercially produced in
the form of thin slabs, it is expected that freeze-thaw cycling is
likely to have more impact on their mechanical properties. This
negates the use of thicker specimens in the experimental program.
Moreover, similar sized granite specimens used in direct tensile
and UPV testing were reported to give acceptable results [10]. Pet-
rographic description of the three types of granite used in this
study is summarized in Table 1.
2.2. Accelerated Freeze-Thaw cycles (FTC)
FTC were performed considering the ASTM D5312/D5312M-12
and EN 12371 [14,15] standard specifications. All the 36 granite
specimens were subjected to 100 freeze-thaw cycles over a period
of 34 days.
The accelerated FTC were conducted using a large freeze-thaw
chamber shown in Fig. 2(a). A control panel next to the chamber
is used to set the temperature range (10 °C to +70 °C) and freez-
ing/heating durations. In each of the four trays nine specimens
were placed; three each of Calca, Imperial Black and Grandee,
shown in Fig. 2(b), (c) and (d), respectively. Water is poured in
the tray such that the bottom face of the specimens (the polished
face) is immersed in water. The trays were designed to support
the specimens above the base of the tray to ensure adequate satu-
ration of the polished face of the specimens.
The specimens were subjected to eight hours freeze-thaw
cycles. Each cycle comprising five hours of freezing and three hours
of heating between 10 °C and +70 °C. Although the temperature
range is different from ASTM D5312/D5312M-12 [14] they were
considered an acceptable range that reflected Australian climate.
Granite Petrographic Description
Type
Calca Coarse grained granite of light reddish color, high orthoclase
feldspar and quartz content, and contains minor weakened
cleavage fractures; felsic rock
Imperial Dark black color Monzo gabbro with high plagioclase feldspar
Black content, mafic rock, hard, holocrystalline and with interlocking
ð1Þ crystals of medium to coarse grain
Grandee Dark colour quartz Monzo gabbro, high plagioclase feldspar and
pyroxene, and low quartz content, hard, holocrystalline and
with interlocking crystals of medium grain size, mafic rock
Fig. 3 (a) shows the FTC during one day period while Fig. 3(b)
shows FTC over the duration of the test (34 days). Each of the pos-
itive/negative temperature peaks were maintained for at least half
an hour throughout the freeze-thaw process. The trays were toped
up with water on a daily basis to compensate for evaporation
losses and to maintain the polished face of the specimens under
water throughout the test. This is usually done during the heating
cycle and when the temperature within the chamber is close to
ambient air temperature. The use of 100 FTC deemed adequate
to cause notable degradation in the mechanical properties of build-
ing stones [3,16,19]. A calibrated ohm reader placed inside the
freeze-thaw chamber was used to continuously record the temper-
ature inside the chamber throughout the test.
Fig. 2(e) shows a granite specimen after the completion of FTC
test. It is apparent that the specimen underwent physical degrada-
tion with salt crystallization (efflorescence) as white powdery
deposits could be seen all over the specimen. Moreover, visible fine
cracks on the surface of the specimens were observed as evidence
of differential thermal expansion and contraction and frost action
during FTC.
2.3. Ultrasonic pulse velocity (UPV) testing
UPV measurement of all granite specimens before and after the
FTC testing was conducted using a transient wave test method. The
test instrumentation and UPV measurements along the width and
length of the specimen are shown in Fig. 4. The test setup com-
prised of a Tektronix TDS210 LCD oscilloscope, a Panametrics
5077 PR pulser-receiver and two contact transducers. The UPV
was obtained by direct transmission, by placing the transducers

Table 2
Specimens used in the UPV and flexural testing program.

Specimen ID Calca Imperial Black Grandee

Dimension (mm)
Length Width Thickness Length Width Thickness Length Width Thickness
S1 401 97.89 33.39 400 97.07 30.24 399 96.96 30.33
S2 401 96.46 29.35 399 97.87 30.14 400 97.44 30.12
S3 400 97.19 31.19 401 97.01 29.92 399 96.80 30.03
S4 400 98.20 31.79 400 97.68 30.76 401 97.19 30.51
S5 401 97.44 33.82 400 97.13 30.59 400 97.36 31.39
S6 401 96.68 33.11 400 97.35 30.02 400 97.40 31.29
S7 400 97.21 31.35 400 96.92 30.35 400 97.36 30.94
S8 400 97.49 32.25 401 98.80 29.82 401 97.94 30.56
S9 401 98.43 32.59 400 97.31 30.53 400 97.60 30.87
S10 401 97.05 34.24 400 97.00 31.71 400 97.82 31.27
S11 401 97.30 31.52 399 97.65 29.77 399 97.90 30.61
S12 400 97.43 32.74 401 97.04 29.33 401 97.78 30.03
C1 400 98.70 31.72 400 97.48 30.82 400 98.18 30.47
C2 400 98.05 31.93 400 97.15 31.38 400 97.51 30.27
C3 400 97.40 32.98 400 97.20 30.49 400 98.56 31.19
C4 400 99.12 32.10 400 97.30 30.80 400 97.49 30.79
C5 400 97.51 30.91 400 97.05 29.49 400 98.05 31.27
894 S. Noor-E-Khuda et al. / Construction and Building Materials 156 (2017) 891–901

Time (Hours)
0 2 4 6 8 10 12 14 16 18 20 22 24
80.00
70.00 30mins
Measured
peak
60.00
8hours cycle

Te mpe rature o C
50.00 Simulated
40.00
30.00
5hours
20.00 freezing
10.00
0.00
-10.00 3hours
heating 30mins peak
-20.00
31/05 0:00 31/05 9:36 31/05 19:12
Time (Date/Hours/Mins)
(a)
100.00

80.00

60.00
Te mpe rature o C

40.00

20.00

0.00

-20.00

-40.00
8/05 15/05 22/05 29/05 5/06 12/06
Time (Date)
(b)
Fig. 3. Temperature cycles in the freeze/thaw chamber (a) Cycles in one day; (b)
Cycles over the testing period.

specimen were taken and the average of these in each direction

was recorded. The ratio of the distance travelled by the pulse
(Dwave) and the time interval ðDtÞ gave the UPV.

2.4. Flexural strength testing

The flexural strength testing of the granite specimens were car-

Fig. 2. Granite specimens under freeze and thaw cycling (a) Freeze-thaw chamber;
(b) Calca specimen; (c) Imperial Black specimen; (d) Grandee specimen; (e)
Efflorescence on weathered specimen.
on the opposite parallel faces of the specimen. The pulser-receiver
transmitted and received the ultrasonic waves by assigning the
transducers one as the emitter and the other one as the receiver.
The transducers were coated with suitable ultrasonic couplant
gel (Sonotech Ultragel II) for improved connectivity with the test
specimen. The ultrasonic waves propagating through the specimen
was monitored using the oscilloscope and the time interval
between transmission and receipt was recorded. Three readings
across the width and another three along the length of each
ried out in accordance with the ASTM C880/C880M [21] specifica-
tions, the test arrangement is shown in Fig. 5. All the 36 weathered
specimens from the FTC testing and the 15 control unweathered
specimens underwent flexural testing.
The 4-point bending test was performed on the specimens with
a span length (L) of 320 mm, which is 10 times the specimen’s
thickness. The overall length of each specimen was limited to
400 mm. A rigid steel loading frame was used to facilitate the roller
support arrangement: the two point loads (W/2 each) were placed
80 mm (L/4) inward from each support and 160 mm (L/2) apart
from each other. The granite specimens were mounted on the load
frame ensuring that the finished polished surface is on the tension
side during loading. An Instron 5500 R/6027 testing machine was
used to conduct the tests. The load was monotonically applied,
until failure of the specimen is obtained. A load rate of 1.4 kN/
min consistent with ASTM recommendation of 4.14 MPa/min is
 S. Noor-E-Khuda et al. / Construction and Building Materials 156 (2017) 891–901 895

Ge l O scilloscope
Pulser/Receiver
Transducers 80mm 160mm 80mm

(a) Fig. 5. Four point bending test of granite specimen.

The flexural tensile strength (ft) of granite specimen was com-

puted using Eq. (2b) rather than Eq. (2a) [21]. Eq. (2b) accounts
for the fact that the failure plane is usually not perpendicular to
the span, as was observed in most of the tested specimens.
3WL
ft ¼ 2
ð2aÞ
4bt

Dwave 3WLb
ft ¼ 2
ð2bÞ
4l t 2
where, W is the maximum applied load on the specimen, b and t are
the width and thickness dimensions, respectively and l is the actual
length of the inclined failure plane across the tension face. The
actual dimensions of the specimens (Table 2) rather than the nom-
inal dimensions were used in calculating the flexural strength (Eq.
(2b)). When the failure plane is perpendicular to the span, Eq.
(b) (2b) gives the same results as Eq. (2a). However, for inclined failure
plane, the flexural strength from Eq. (2b) is up to 4% lower than that
from Eq. (2a).

3. Test results and discussions

The ultrasonic and 4-point bending test results of the unweath-

ered and the weathered (FTC) specimens are discussed in this
section.

Dwave 3.1. Ultrasonic pulse velocity (UPV)

(c)
Fig. 4. Ultrasound testing of granite specimens (a) Test arrangement; (b) Measure-
ment along width; (c) Measurement along length.
used. The load-deflection response was recorded from the load cell
and the crosshead displacement data. Length and orientation of the
failure plane and the average thickness at the failure plane were
measured at the completion of each test.
Fig. 6 shows the UPV values for the Calca (Fig. 6a), Imperial
Black (Fig. 6b) and Grandee (Fig. 6c) specimens before FTC
(unweathered state). Three individual readings along the width
of each specimen are shown in Fig. 6.
The results shown in Fig. 6 are consistent and fall within Class 5
of UPV classification [23] (Table 3) for the three types of granite
used in this study. The average UPV of unweathered Imperial Black,
Grandee and Calca specimens were 6389 m/s, 5704 m/s and
5166 m/s, respectively.
Fig. 7 shows the average UPV values of the unweathered granite
specimens measured along the width and the length of the speci-
men. It was found that the UPV measurements along the length
of specimen were generally higher than those along the width.
The variation in UPV values along the length and width directions
corresponds to planar anisotropy of the granite material [22,23].
The difference is more consistent for Imperial Black and Grandee
than that for Calca.
896 S. Noor-E-Khuda et al. / Construction and Building Materials 156 (2017) 891–901

5500 Table 3
Classifications of UPV [23].
5400 Class UPV (m/s) Definition

5300 1 <2500 Very low velocity

2 2500–3500 Low velocity
5200 3 3500–4000 Medium velocity
4 4000–5000 High velocity
UPV [m/s]

5100 5 >5000 Very high velocity

5000
Table 4 presents the average UPV of the unweathered and
4900
weathered specimens measured along the length and the width
Reading 1
4800 of the specimen. Also shown in this table is the percentage reduc-
Reading 2 tion of pulse velocity due to weathering. The results show that
4700 Imperial Black was the worst affected by FTC with an average
Reading 3
reduction in UPV of 17% and 21% in the width and length direc-
4600
0 1 2 3 4 5 6 7 8 9 10 11 12 tions, respectively. Percent reduction of UPV for the Calca and
Specimen Grandee specimens were 12–13% and 8–12%, respectively.
(a)
3.2. Flexural strength
6550
Figs. 9 (Calca), 10 (Imperial Black) and 11 (Grandee) show the
6500 load-deflection curves obtained from the 4-point bending test of
all 36 weathered and 15 unweathered (control) specimens. The
6450 average load resisted by Calca, Imperial Black, and Grandee control
specimens were 4.36 ± 0.5 kN, 6.45 ± 0.58 kN and 6.36 ± 0.21 kN,
and the same of the weathered specimens were 3.38 ± 0.54 kN,
UPV [m/s]

6400
5.14 ± 0.38 kN and 4.49 ± 0.4 kN, respectively. A notable decrease
in the load capacity of all the weathered granite specimens is
6350 observed that indicates degradation of the structural properties
due to FTC. It is evident that Imperial Black is the strongest among
6300 Reading 1 the three granite types considered in this paper. This is in-line with
Reading 2 the FTC test observations discussed in Section 2.3.
6250 As expected, the obtained load-deflection curves confirm the
Reading 3
brittle nature of granite material; the load increased linearly until
6200 the ultimate capacity followed by a sudden sharp drop correspond-
0 1 2 3 4 5 6 7 8 9 10 11 12 ing to failure. These plots will be particularly useful in modelling
Specimen and validating the flexural response of weathered and unweath-
(b) ered granite, and veneer walls comprising of granite claddings.
Fig. 12 shows the average flexural tensile strength, ft, of the
6200 unweathered and weathered granite specimens. The results shown
in Fig. 12 demonstrate that ft of the unweathered (control set)
6000 Calca, Imperial Black and Grandee granite specimens is
10.25 MPa, 16.77 MPa and 15.98 MPa; respectively, and that of
the same weathered specimens is 7.74 MPa, 11.98 MPa and
5800 13.32 MPa, respectively, corresponding to a reduction in flexural
UPV [m/s]

strength due to FTC of 24.5%, 28.5% and 16.7%, respectively. These

5600 values are quite high which clearly indicate that thin veneer gran-
ites suffer notable loss in flexural strength due to weathering (FTC).
The results further reveals that Imperial Black despite being the
5400 strongest was the most affected by freeze-thaw cycling. The FTC
Reading 1 of Calca resulted in a reduction of flexural strength below the rec-
5200 ommended 8.27 MPa for thin structural stone veneers [21].
Reading 2
Reading 3 3.3. Density
5000
0 1 2 3 4 5 6 7 8 9 10 11 12
The mass of the specimens before and after FTC testing was
Specimen
recorded. Similarly, the volume of each specimen was calculated
(c) using the measured dimensions of the specimen. Then the density
of each specimen is obtained using the measured mass and vol-
Fig. 6. UPV measurement along the width of unweathered specimens (a) Calca; (b)
Imperial Black; (c) Grandee. ume. The mean density for each granite type is calculated using
the density of all the specimens within that particular granite type.
Fig. 8 shows the UPV (width direction) of unweathered (before The mean density of the unweathered and weathered granite spec-
FTC) and weathered (after FTC) granite specimens. A noticeable imens are shown in Fig. 13 together with the mean UPV. The values
and consistent decrease in UPV is seen due to FTC weathering of shown in Fig. 13 reveals that all three Australian granite types used
the three granite types. in the experimental program are of comparatively higher density
 S. Noor-E-Khuda et al. / Construction and Building Materials 156 (2017) 891–901 897

5600 5600

5500 5400

5400 5200

5300 5000

UPV [m/s]
UPV [m/s]

5200 4800

5100 4600

5000 4400

4900 4200
Unweathered
4800 Width 4000
Weathered
Length
4700 3800
0 1 2 3 4 5 6 7 8 9 10 11 12 0 2 4 6 8 10 12
Specimen Specimen
(a) (a)
7000 6800

6800 6300

6600
UPV [m/s] 5800
UPV [m/s]

6400 5300

6200 4800

6000 4300 Unweathered

Width Weathered
Length 3800
5800 0 2 4 6 8 10 12
0 1 2 3 4 5 6 7 8 9 10 11 12
Specimen Specimen
(b) (b)
6400 6300

6200
5800
6000
UPV [m/s]

5300
UPV [m/s]

5800

5600 4800

5400
4300
5200 Width Unweathered
Length Weathered
5000 3800
0 1 2 3 4 5 6 7 8 9 10 11 12 0 1 2 3 4 5 6 7 8 9 10 11 12
Specimen Specimen
(c) (c)
Fig. 7. Variation of UPV along the length and width of unweathered specimens (a) Fig. 8. Variation of UPV due to weathering (a) Calca; (b) Imperial Black; (c)
Calca; (b) Imperial Black; (c) Grandee. Grandee.

than other thin structural stone veneers reported in the literature (Table 4) and flexural strength obtained in the experimental
[9,16,18]. The mean density of unweathered Calca, Imperial Black program.
and Grandee specimens is 2644.97 kg/m3, 2980.58 kg/m3 and Furthermore, loss in density of the granite specimens due to FTC
2912.94 kg/m3, respectively. This explains the high UPV values is apparent in Fig. 13; Calca, Imperial Black and Grandee specimens
898 S. Noor-E-Khuda et al. / Construction and Building Materials 156 (2017) 891–901

Table 4
Average UPV of unweathered and weathered specimens and % loss.

Granite Type UPVUnweathered (m/s) UPVWeathered (m/s) % loss of UPV

Width Length Width Length Width Length
Calca 5166 ± 206 5243 ± 203 4486 ± 225 4595 ± 76 13.2 12.4
Imperial Black 6389 ± 68 6686 ± 306 5302 ± 462 5263 ± 406 17.0 21.3
Grandee 5704 ± 279 5917 ± 284 5264 ± 350 5195 ± 256 7.7 12.2

6 9
C1 C2 C3 C1 C2 C3
C4 C5 8 C4 C5
5
7

4 6
Load (kN)

Load (kN)
5
3
4
2
3

2
1
1
0
0 0.1 0.2 0.3 0.4 0.5 0
Deflection (mm) 0.00000 0.10000 0.20000 0.30000 0.40000 0.50000 0.60000
Deflection (mm)
(a)
(a)
6 7
S1 S2 S3 S4 S1 S2 S3 S4
S5 S6 S7 S8 S6 S7 S8 S9
5 6
S9 S10 S11 S12 S10 S11 S12
5
4
Load (kN)

Load (kN)

4
3

3
2
2
1
1

0
0 0.1 0.2 0.3 0.4 0.5 0
0 0.1 0.2 0.3 0.4 0.5 0.6
Deflection (mm)
(b) Deflection (mm)
(b)
Fig. 9. Load vs deflection plot of Calca granite (a) Control unweathered specimens;
(b) Weathered specimens. Fig. 10. Load vs deflection plot of Imperial Black granite (a) Control unweathered
specimens; (b) Weathered specimens.

experienced an average drop in density by 2.7%, 2.6% and 2.3%,

respectively. This is consistent with the reduction in UPV and flex- 3.4. Anisotropy
ural strength obtained from testing following the FTC regime. Flak-
ing, spalling and salt decay due to FTC contributed to reduction in Fig. 14. shows the planar anisotropy of the granite specimens;
density. anisotropy is calculated using Eq. (3) [23].
 S. Noor-E-Khuda et al. / Construction and Building Materials 156 (2017) 891–901 899

8 3100 7000
Unweathered
C1 C2 C3 Weathered
3000
7 C4 C5 UPV 6500
2900

De nsitymean (kg/m3 )
6 6000

UPVmean [m/s]
2800

5 2700 5500
Load (kN)

2600
4 5000
2500
3 4500
2400

2 2300 4000
Calca Imperial Grandee
Black
1
Fig. 13. Density of weathered and unweathered specimens.

0  
0 0.1 0.2 0.3 0.4 0.5 UPV max  UPV min
Deflection (mm) Anisotropy% ¼  100 ð3Þ
UPV max
(a) where, UPVmax and UPVmin represents the maximum and the mini-
mum recorded UPV values irrespective of the measurement direc-
8
tion. Planar anisotropy of granite can be attributed to the
S1 S2 S3 S4
alignment of microcracks with the rift plane [10], resulting in low
7 S5 S6 S7 S10 to moderate ansiotropic behaviour. From Fig. 14, it is obvious that
S11 S12 the Calca specimens showed higher anisotropy among the three
6 granite types both in the unweathered and weathered state. Aniso-
tropy of Imperial Black and Grandee specimens were comparatively
low, within 4%. Notable changes to anisotropy of the weathered
5
Calca and Imperial Black specimens was observed, which indicates
Load (kN)

possible microstructural changes due to differential thermal expan-

4 sion and contraction and frost action during the FTC regime. On the
other hand, anisotropy of the Grandee specimens was hardly
3 affected by weathering.

2 3.5. Corelations between UPV and weathereing

A correlation between UPV of unweathered and weathered

1 granite specimens for the three stone types is shown in Fig. 15a.
The correlation is shown by the straight line in Fig. 15 and is given
0 by Eq (4a) with R2 = 0.67. From this, it is possible to predict the UPV
0.00000 0.10000 0.20000 0.30000 0.40000 0.50000 0.60000 of weathered granite during its service life based on the UPV of
Deflection (mm) unweathered specimens. Hence the strength of weathered granite
(b) during its service life can be estimated based on the measured UPV
values.
Fig. 11. Load vs deflection plot of Grandee granite (a) Control unweathered
specimens; (b) Weathered specimens.
UPV weathered ¼ 0:07438UPV unweathered þ 778:04 ð4aÞ

UPV Length ¼ 0:987UPV width þ 177:12 ð4bÞ

20 Unweathered
Fle xural Strength, ft (MPa)

18 Weathered 7
Unweathered
16 6 Weathered
14
Anisotropy (%)

5
12
10 4
8 3
6
2
4
2 1
0 0
Calca Imperial Grandee Calca Imperial Grandee
Black Black

Fig. 12. Flexural strength of granite specimens (unweathered and weathered). Fig. 14. Anisotropy of granite specimens.
900 S. Noor-E-Khuda et al. / Construction and Building Materials 156 (2017) 891–901

6500 19
UPVweathered = 0.7438UPVunweathered + 778.04
6000 R² = 0.67 17

Fle xural Strength, ft (MPa)

5500
15
UPVweathered [m/s]

5000

4500 13

4000 11
3500
9 ft = 0.0041UPVmean - 9.6345
3000 R² = 0.72
7
2500

2000 5
4500 4700 4900 5100 5300 5500 5700 5900 6100 6300 6500 4000 4500 5000 5500 6000 6500 7000
UPVunweathred [m/s] UPVmean [m/s]
(a)
Fig. 16. Correlation between UPV and flexural strength.
8000

7000 rc (Eq. (7)). A range of UPV between 3000 m/s to 6500 m/s was
selected to represent various types of old and quarried granites,
6000
reported in the literature [4,9–11].
UPVLength [m/s]

5000
rt ¼ 0:701eð0:00052UPVÞ ð6Þ
4000 y = 0.9876x + 177.12
R² = 0.802 rc ¼ 0:0407UPV  36:31 ð7Þ
3000
Fig. 17(a) shows the predicted tensile strength (ft and rt ) using
2000 Eqs. (5) and (6) against the corresponding UPV values. As seen in
Fig. 17a, rt is less than ft for UPV below 6000 m/s. This is expected
1000 assuming evenly distributed defect density and uniform stress dis-
tribution under direct tension. However, for higher UPV values, Eq.
0 (6) seems to over-predict the direct tensile strength compared to
4000 4500 5000 5500 6000 6500 7000 flexural strength.
UPVWidth [m/s] Fig. 17(b) shows the ratio of flexural strength (Eq. (5)) to com-
(b) pressive strength (Eq. (7)) of granite plotted against the corre-
sponding UPV values. As expected, the flexural tensile strength of
Fig. 15. Linear correlation of UPV; (a) UPV in unweathered and weathered granite is under 10% of the corresponding compressive strength.
conditions; (b) UPV along the length and width of the specimens. Based on Fig. 17a and b, the compressive strength for Calca, Impe-
rial Black and Grandee specimens is calculated as 156 MPa,
185 MPa and 171 MPa, respectively. This is in reasonable agree-
Linear correlation between UPV values measured along the
ment with results obtained from compression test of
length and width dimension of the granite specimens is shown
50  50  50 mm cubes of Calca and Imperial Black granite which
in Fig. 15(b) and given by Eq. (4b) with R2 = 0.802. From practical
was reported as 148 and 192 MPa [1].
point of view, the developed relation can be particularly useful
where there is a limited access to take full measurements of
installed granite panels. 4. Conclusion

This paper investigated the effect of accelerated weathering

3.6. Correlations between UPV and material strength
As already mentioned in the previous sections, the flexural ten-
sile strength ft of the granite cladding panels can be estimated by
means of the pulse velocity. The relation between the flexural ten-
sile strength ft obtained from the 4-point bending tests and the
measured UPV along the width of the 36 weathered and the 15
unweathered Australian granite specimens is shown in Fig. 16. A
linear correlation is obtained as shown in Fig. 16 and Eq. (5) with
R2 = 0.72.
F t ¼ 0:0041UPV mean  9:6345
The proposed Eq. (5) for the flexural tensile strength ft is com-
pared against equations proposed by other researchers [10] for
the direct tensile strength rt (Eq. (6)) and the compressive strength
using FTC on thin granite panels used in veneer claddings. A total
of 51 specimens from three types of Australian granite (Calca,
Imperial Black and Grandee) were used in the experimental pro-
gram. UPV of the specimens were measured before and after FTC
testing. A controlled freeze-thaw chamber was used for the weath-
ering process and the specimens were subjected to 100 cycles of
freezing and thawing for a period of 34 days. Flexural strength test-
ing of weathered and unweathered control specimens was per-
formed to determine the flexural tensile strength.
Results from the FTC test showed noticeable evidence of phys-
ical degradation of the granite specimens. The UPV results clearly
ð5Þ
showed a reduction in UPV following FTC, which correlates to
strength reduction. This was validated by flexural strength testing.
Flexural strength reduction due to FTC of Calca, Imperial Black and
Grandee were 24.5%, 28.5% and 16.7%, respectively. Noticeable
 S. Noor-E-Khuda et al. / Construction and Building Materials 156 (2017) 891–901 901

25 Acknowledgements

The authors would like to acknowledge the assistance of Lucy

20 Cuddihy and Royce Erwin in conducting the experimental work
Tensile Strength (MPa)

and Dr R. Blackwell for his comments and discussion.

15 References

[1] R. Blackwell, No Stone Unturned: Towards the Design of Thin Structural Stone
Veneers [PhD Thesis], University of Queensland, 2005.
10 [2] V. Pires, L.G. Rosa, A. Dionisio, Implications of exposure to high temperatures
for stone cladding requirements of three Portuguese granites regarding the use
of dowel-hole anchoring systems, Constr. Build. Mater. 64 (2014) 440–450.
5 ft [3] X. Tan, W. Chen, J. Yang, J. Cao, Laboratory investigations on the mechanical
properties degradation of granite under freeze-thaw cycles, Cold Reg. Sci.
σt Technol. 68 (2011) 130–138.
[4] M. Uranjek, V. Bokan-Bosiljkov, Influence of freeze-thaw cycles on mechanical
properties of historical brick masonry, Constr. Build. Mater. 84 (2015) 416–
0
428.
2000 3000 4000 5000 6000 7000 [5] I.S. Colombo, I. Main, M. Forde, Assessing damage of reinforced concrete beam
UPV [m/s] using ‘‘b-value” analysis of acoustic emission signals, J. Mater. Civil Eng.
(ASCE). 15 (2003) 280–286.
(a) [6] G. Cardani, L. Cantini, S. Munda, L. Zanzi, L. Binda, Non invasive measurements
of moisture in full-scale stone and brick masonry models after simulated
8 flooding: effectiveness of GPR, in: Nondestructive Testing of Materials and
Structures, Springer, 2013, pp. 1143–1149.
[7] G. Trtnik, F. Kavčič, G. Turk, Prediction of concrete strength using ultrasonic
7 pulse velocity and artificial neural networks, Ultrasonics 49 (2009) 53–60.
[8] G. Barbera, G. Barone, P. Mazzoleni, A. Scandurra, Laboratory measurement of
6 ultrasound velocity during accelerated aging tests: Implication for the
determination of limestone durability, Constr. Build. Mater. 36 (2012) 977–
983.
5 [9] C. Cerrillo, A. Jiménez, M. Rufo, J. Paniagua, F. Pachón, New contributions to
ft /σc (% )

granite characterization by ultrasonic testing, Ultrasonics 54 (2014) 156–167.

4 [10] G. Vasconcelos, P. Lourenço, C. Alves, J. Pamplona, Ultrasonic evaluation of the
3 [11] J. Chen, Z. Xu, Y. Yu, Y. Yao, Experimental characterization of granite damage
2 [12] L. Binda, G. Cardani, L. Zanzi, Nondestructive testing evaluation of drying
1
0 [14] ASTM.D5312/D5312M-12. Standard test method for evaluation of durability of
2000 3000 4000 5000 6000
UPV [m/s]
(b)
Fig. 17. Correlation between UPV and strength (a) Flexural and direct tensile
strength; (b) Flexural and compressive strength.
changes in density and anisotropy were also observed following
FTC.
The UPV of the Australian granite specimens falls at the higher
end of the UPV spectrum for thin structural stone veneers. This is
consistent with their higher density and strength. A correlation
between UPV and flexural strength of weathered and unweathered
granite was established using the 51 test results reported in this
work. Furthermore, a correlation between UPV of unweathered
and weathered granite using 36 test results was also presented.
This can help in predicting the UPV, and hence strength, of granite
during its service life. Further research on the influence of other
weathering conditions such as cyclic wetting and drying on the
structural properties of Australian granites worth investigating.
physical and mechanical properties of granites, Ultrasonics 48 (2008) 453–
466.
using nonlinear ultrasonic techniques, NDT E Int. 67 (2014) 10–16.
process in flooded full-scale masonry walls, J. Perform. Constr. Facil. (ASCE). 24
(2010) 473–483.
[13] E. Franzoni, E. Sassoni, G.W. Scherer, S. Naidu, Artificial weathering of stone by
heating, J. Cult. Heritage 14 (2013) 85–93.
7000 rock erosion control under freezing and thawing conditions. American Society
for Testing and Materials, 2016.
[15] EN.12371. Natural stone test methods – Determination of frost resistance.
European Standard, 2010.
[16] J.P. Ingham, Predicting the frost resistance of building stone, Q. J. Eng. Geol.
Hydrogeol. 38 (2005) 387–399.
[17] J. Ruedrich, D. Kirchner, S. Siegesmund, Physical weathering of building stones
induced by freeze–thaw action: a laboratory long-term study, Environ. Earth
Sci. 63 (2011) 1573–1586.
[18] Y. Ozcelik, N. Careddu, E. Yilmazkaya, The effects of freeze–thaw cycles on the
gloss values of polished stone surfaces, Cold Reg. Sci. Technol. 82 (2012) 49–
55.
[19] M. Takarli, W. Prince, R. Siddique, Damage in granite under heating/cooling
cycles and water freeze–thaw condition, Int. J. Rock Mech. Min. Sci. 45 (2008)
1164–1175.
[20] D.-Y. Wang, W. Ma, Y.-H. Niu, X.-X. Chang, Z. Wen, Effects of cyclic freezing and
thawing on mechanical properties of Qinghai-Tibet clay, Cold Reg. Sci. Technol.
48 (2007) 34–43.
[21] ASTM.C880/C880M-15. Standard test method for flexural strength of
dimension stone. American Society for Testing and Materials, 2015.
[22] ASTM.D2845-05. Standard test method for laboratory determination of pulse
velocities and ultrasonic elastic constants of rock. American Society for Testing
and Materials, 2005.
[23] E. Vasanelli, D. Colangiuli, A. Calia, M. Sileo, M.A. Aiello, Ultrasonic pulse
velocity for the evaluation of physical and mechanical properties of a highly
porous building limestone, Ultrasonics 60 (2015) 33–40.