polymers

Article

Moisture Absorption/Desorption Effects on Flexural

Property of Glass-Fiber-Reinforced Polyester
Laminates: Three-Point Bending Test and Coupled
Hygro-Mechanical Finite Element Analysis
Xu Jiang 1, , Jie Song 2, , Xuhong Qiang 3, *, Henk Kolstein 4, and Frans Bijlaard 4,
1
2
3
4

Department of Bridge Engineering, College of Civil Engineering, Tongji University, Shanghai 200092, China;
jiangxu@tongji.edu.cn
Shandong Provincial Academy of Building Research, Jinan 250031, China; shipgoing@126.com
Department of Structural Engineering, College of Civil Engineering, Tongji University,
Shanghai 200092, China
Faculty of Civil Engineering and Geosciences, Delft University of Technology, Delft 2628CN,
The Netherlands; M.H.Kolstein@tudelft.nl (H.K.); F.S.K.Bijlaard@tudelft.nl (F.B.)
Correspondence: qiangxuhong@tongji.edu.cn; Tel./Fax: +86-21-6598-0531
These authors contributed equally to this work.

Academic Editors: Alper Ilki and Masoud Motavalli

Received: 30 May 2016; Accepted: 3 August 2016; Published: 10 August 2016

Abstract: Influence of moisture absorption/desorption on the flexural properties of Glass-fibre-reinforced

polymer (GFRP) laminates was experimentally investigated under hot/wet aging environments.
To characterize mechanical degradation, three-point bending tests were performed following the
ASTM test standard (ASTM D790-10A). The flexural properties of dry (0% Mt /M8 ), moisture
unsaturated (30% Mt /M8 and 50% Mt /M8 ) and moisture saturated (100% Mt /M8 ) specimens at
both 20 and 40 C test temperatures were compared. One cycle of moisture absorption-desorption
process was considered in this study to investigate the mechanical degradation scale and the
permanent damage of GFRP laminates induced by moisture diffusion. Experimental results confirm
that the combination of moisture and temperature effects sincerely deteriorates the flexural properties
of GFRP laminates, on both strength and stiffness. Furthermore, the reducing percentage of
flexural strength is found much larger than that of E-modulus. Unrecoverable losses of E-modulus
(15.0%) and flexural strength (16.4%) for the GFRP laminates experiencing one cycle of moisture
absorption/desorption process are evident at the test temperature of 40 C, but not for the case of 20 C
test temperature. Moreover, a coupled hygro-mechanical Finite Element (FE) model was developed to
characterize the mechanical behaviors of GFRP laminates at different moisture absorption/desorption
stages, and the modeling method was subsequently validated with flexural test results.
Keywords: glass fibre reinforced polymer; bridge deck; environmental degradation; mechanical
property; coupled hygro-mechanical numerical analysis

1. Introduction
During the last two decades, fibre-reinforced polymer (FRP) bridge decks are increasingly being
used for the rehabilitation of old concrete-steel composite bridges and new construction of pedestrian
and highway bridges [17], due to their various advantages [8]: high strength to weight ratio, corrosion
resistance, controllable quality, rapid installation with minimum traffic disruption and low maintenance
cost. Although FRP decks are increasingly being employed in civil infrastructure applications,
their durability and long-term performance are still not comprehensively understood. In the normal

Polymers 2016, 8, 290; doi:10.3390/polym8080290

www.mdpi.com/journal/polymers

Polymers 2016, 8, 290

2 of 15

service life of such infrastructures, FRP decks are usually exposed to harsh and changing environments,
involving large variations in temperature and humidity. The hot/wet exposure is supposed to be
the severest environmental condition to degrade the performance of polymeric materials [917],
which decreases the service-life of FRP composite bridges. The influence of moisture absorption on
mechanical properties of FRP composites is well documented in literatures [914,1626], regarding
the tensile, interlaminar shear and flexural properties. Shao et al.s research [9] found that the tensile
strength of the aged composites decreased with the increase of percent absorption. In their test
results, the reduction was estimated at about 60% for both the web and the flange pultruded sheet pile
sections. In Nogueira et al.s research [11], the effects of absorbed water on the mechanical properties
of an amine-cured epoxy resin system were studied. The overall reductions of elastic modulus,
elongation at break, tensile strength, tensile toughness, Rockwell hardness and glass transition
temperature were recorded in their test results, which were attributed to the plasticizing effect of water.
From the SEM image investigation of E-glass fiber reinforced vinyl ester composite after hygrothermal
aging, the fiber-matrix interfacial bonding and consequently the stability of the composite were found
highly sensitive to moisture absorption [24]. In DellAnno et al.s research [17], the interlaminar shear
and flexural performance of the carbon fibre composites with isophthalic polyester, vinyl ester and
urethane acrylate matrices were investigated under the aging condition of 40 C-deionised water
and compared to those of equivalent composites impregnated with three grades of epoxy resin.
Their results demonstrated that, although the epoxy systems perform equally or better than the
alternative resins in the dry state, they are also more sensitive to property degradation due to
water ingress.
However, most of available researches were conducted in the aerospace engineering field
and automotive industry. For FRP composites used in the civil engineering field, the available
research results are very limited. Although test data collected from other research fields can be
suggestive, but it is not convincing to directly extend the experimental data and results to the civil
engineering field. Since FRP composites used for civil infrastructures have essential differences,
which include geometries, types of fibres and matrix, fabrication methods, curing process and service
environmental conditions as well. In order to keep pace with the application of FRP composites in Civil
Engineering field, this research was undertaken to reveal more knowledge about the environmental
degradation of Glass-fiber-reinforced Polymer (GFRP) composite bridge under hot/wet environment.
In GFRP bridge decks, the GFRP laminates are mainly loaded by the wheel load in the
through-thickness direction. Therefore, the flexural properties of GFRP materials need to be carefully
studied. Thus, in this paper, the influence of moisture and temperature on the flexural properties of
GFRP laminates was experimentally investigated. Three-point bending tests were performed following
the ASTM test standard (ASTM D790-10A). The flexural properties of dry (0% Mt /M8 ), moisture
unsaturated (30% Mt /M8 and 50% Mt /M8 ) and moisture saturated (100% Mt /M8 ) specimens at the
test temperatures of both 20 and 40 C were compared. One cycle of moisture absorption-desorption
process was also included in this study to investigate how the residual damage induced by the moisture
diffusion degrades the mechanical properties of GFRP laminates. Furthermore, to better understand
the environment-dependent mechanical performance of GFRP laminates, a coupled hygro-mechanical
FE model was developed by writing a specific post-progressing subroutine to work together with the
FEM software ABAQUS, and subsequently validated with the flexural test results.
2. Experiment
To investigate the flexural property of GFRP laminates, three-point bending tests were employed.
The whole test procedure follows the test code ASTM D790-10 [27]. GFRP laminates studied in this
paper were manufactured by resin vacuum infusion (Infra Composite BV, Delft, the Netherlands) using
polyester, and then cut into specific dimensions with the principal axis parallel to the warp direction of
the woven roving (as shown in Figure 1). The glass transition temperature (Tg ) of GFRP laminates
is 78 C. The 5.64 mm thick specimen is composed of six layers of standard 0.94 mm EQX1200.

Polymers 2016, 8, 290

Polymers 2016, 8, 290

3 of 15
3 of 15

0.94 mm EQX1200. The layup configuration of each piece of the standard 0.94 mm EQX1200 is
The layup configuration of each piece of the standard 0.94 mm EQX1200 is illustrated in Table 1.
illustrated in Table 1. The mechanical properties of GFRP laminates supplied by the manufacturer
The mechanical properties of GFRP laminates supplied by the manufacturer are shown in Table 2.
are shown in Table 2. The nominal length and width of the specimens were selected to be 150 and 20
The nominal length and width of the specimens were selected to be 150 and 20 mm respectively.
mm respectively. According to the test code ASTM D790-10 [27], the specimen length shall be
According to the test code ASTM D790-10 [27], the specimen length shall be sufficient to allow for
sufficient to allow for overhanging on each end of at least 10% of the support span. The specimen
overhanging on each end of at least 10% of the support span. The specimen width shall not exceed
width shall not exceed one fourth of the support span for specimens greater than 3.2 mm in depth.
one fourth of the support span for specimens greater than 3.2 mm in depth.

Figure
1. 1.
Fibre
-re inforce d polyme
r (FRP)
cime nfor
forflexural
fle xuraltests.
te sts.
Figure
Fibre-reinforced
polymer
(FRP)laminate
laminate spe
specimen
Table
Glass-fibre -re inforce
d polymer
rties of(supplied
EQX1200
(supplie d by
Table 1.1.Glass-fibre-reinforced
polymer
(GFRP)(GFRP)
laminatelaminate
propertiesprope
of EQX1200
by Owenscorning,
Owe
nscorning,
Tole
do,
OH,
USA).
Toledo, OH, USA).
2)
Weight uniformity (g/m
Weight uniformity (g/m2 )
Product name Total weight (g/m )
Yarn roving
Knit yarn
Total weight (g/m2 )
Product name
0(Warp) Yarn
+45roving
90(Weft) 45
Knit yarn

90 300
(Weft)
EQX 1200
1,193 0 (Warp) 283 +45 300
30045
10
2

EQX 1200

1,193

283

300

300

300

10

Table 2. Me chanical prope rties of GFRP laminate s (supplie d by Owe nscorning, Tole do, OH, USA).
Table 2. Mechanical properties of GFRP laminates (supplied by Owenscorning, Toledo, OH, USA).
Property
Tensile (ISO 527-4)
Compression (ISO 8515)
Flexural (ISO 14.125)
Mean value
Warp
Weft
Warp
Weft
Warp
Weft
Property
Tensile (ISO 527-4)
Compression (ISO 8515)
Flexural (ISO 14.125)
Stre ngth
331 MPa
314 MPa
220 MPa
200 MPa
473 MPa
433 MPa
Mean
value
Warp
Weft
Warp
Weft
Modulus
18 GPa
17 GPa
14
GPa
14
GPa
13Warp
GPa
11 Weft
GPa
Strength
331 MPa
314 MPa
220 MPa
200 MPa
473 MPa
433 MPa
Modulus
18
GPa
17
GPa
14
GPa
14
GPa
13
GPa
11 content,
GPa
The numbering of specimens is given in Table 3, with regard to the moisture uptake

test temperature, absorption/desorption process and replicated number. Two test temperatures 20
and 40
were selected
for the three
pointinbending
which to
were
by the content,
climate
TheCnumbering
of specimens
is given
Table 3, tests,
with regard
the controlled
moisture uptake
chamber
with theabsorption/desorption
tolerance of 2 C, as shown
in Figure
2.
test temperature,
process
and replicated
number. Two test temperatures 20

and 40 C were selected for the three point bending tests, which were controlled by the climate chamber
with the tolerance of 2 C, as shown in Figure 2.

F-50%-20 C-D
F-50%-40 C-D
F-30%-20 C-D
Se t-6
F-30%-40 C-D
C-D
Polymers 2016, F-0%-20
8, 290
Se t-7
F-0%-40 C-D
Se t-5

fle xural
fle xural
fle xural
fle xural
fle xural
fle xural

50%
50%
30%
30%
0
0

20 C
40 C
20 C
40 C
20 C
40 C

ye s
ye s
ye s
ye s
ye s
ye s

5
5
5
5
5 4 of 15
5

Figure 2. Climate chambe r.

Figure 2. Climate chamber.

As listed in Table 3, the test in each condition was repeated five times to investigate the spread
Table 3. GFRP laminate specimens for flexural tests.
of test results. The hydrothermal aging condition was 40 C-water, which is supposed to be a severe
hot/wet condition for GFRP laminates, as discussed in previous
70 pieces
Test research [28].
AfterIn total,Numder
of of
Specimen identification
Test
M t /M 8
temperature
desorption
specimens
specimens were prepared. During the hydrothermal aging process, all the specimens were
C
C
F-0%-20at
0
no
immersed in the water
40 C, flexural
except the F-0-20
C and20F-0-40
C specimens,
which5 are the
Set-1
F-0%-40 C
flexural
0
40 C
no
5
as-received reference specimens (Set-1 in Table 3). The as-received specimens
were stored
in the

F-30%-20
C
flexural
30%
20
C
no
5
laboratory
moisture content of them is very
low, and thus can be ignored. As
Set-2 environment. The
F-30%-40 C
flexural
30%
40 C
no
5
illustrated in Table 3, the Set-2
specimens
(F-30%-20
C
and
F-30%-40
C) were tested at 30% relative

F-50%-20 C
flexural
50%
20 C
no
5
Set-3uptake content. The
moisture
specimens (F-50%-20
C40
and
C) were tested5 at 50%
C F-50%-40 no
F-50%-40 C Set-3
flexural
50%
relative moisture uptake
content.
The
Set-4 specimens
C) were
C
F-100%-20
flexural
100% (F-100%-20
20 C C and F-100%-40
no
5 tested
F-100%-40 level
C
flexural
40 Ccontent). Until
no this time point,
5
at theSet-4
moisture saturation
(100%
relative 100%
moisture uptake
the

above test process

was C-D
considered
as the moisture
absorption
process.yesThen, the remaining
F-50%-20
flexural
50%
20 C
5
Set-5
C-D
F-50%-40
50%
40 C
yes into an oven5at 42 C
specimens were all
taken out
of theflexural
hydrothermal
aging environment,
and put

F-30%-20
flexural
30%
20 C process. yes
5
to drySet-6
them, which
was C-D
considered
as the moisture
desorption
In this way, the
Set-5
F-30%-40 C-D
flexural
30%
40 C
yes
5
specimens (F-50%-20 C-D and F-50%-40 C-D) were tested at 50% relative moisture uptake content
F-0%-20 C-D
flexural
0
20 C
yes
5
after aSet-7
certain time
of moisture desorption.
Subsequently,
the
Set-6 specimens
(F-30%-20 C-D
and
F-0%-40 C-D
flexural
0
40 C
yes
5
F-30%-40 C-D) were tested at 30% relative moisture uptake content after the moisture desorption.

As listed in Table 3, the test in each condition was repeated five times to investigate the spread
of test results. The hydrothermal aging condition was 40 C-water, which is supposed to be a severe
hot/wet condition for GFRP laminates, as discussed in previous research [28]. In total, 70 pieces of
specimens were prepared. During the hydrothermal aging process, all the specimens were immersed
in the water at 40 C, except the F-0-20 C and F-0-40 C specimens, which are the as-received reference
specimens (Set-1 in Table 3). The as-received specimens were stored in the laboratory environment.
The moisture content of them is very low, and thus can be ignored. As illustrated in Table 3,
the Set-2 specimens (F-30%-20 C and F-30%-40 C) were tested at 30% relative moisture uptake content.
The Set-3 specimens (F-50%-20 C and F-50%-40 C) were tested at 50% relative moisture uptake
content. The Set-4 specimens (F-100%-20 C and F-100%-40 C) were tested at the moisture saturation
level (100% relative moisture uptake content). Until this time point, the above test process was
considered as the moisture absorption process. Then, the remaining specimens were all taken out
of the hydrothermal aging environment, and put into an oven at 42 C to dry them, which was
considered as the moisture desorption process. In this way, the Set-5 specimens (F-50%-20 C-D and

Polymers 2016, 8, 290

Polymers 2016, 8, 290

5 of 15
5 of 15

The
Set-7
specimens
The final
final
Set-7
specimens (F-0-20
(F-0-20 C-D
C-Dand
and F-0-40
F-0-40 C-D)
C-D) were
were the
the fully
fully dry
dry specimens
specimens after
after one
one5cycle
cycle
Polymers
2016,
8, 290
of 15
of
moisture
absorption-desorption
process.
Herein,
the
symbol
D
indicates
the
moisture
desorption.
of moisture absorption-desorption process. Herein, the symbol D indicates the moisture desorption.
As
As shown
shown in
in Figure
Figure 3,
3, the
the whole
whole three-point
three-point bending
bending test
test set-up
set-up was
was put
put into
into the
the chamber.
chamber.
C-D)
F-50%-40
were
tested
at
50%
relative
moisture
uptake
content
after
a
certain
time
of moisture
According
to
ASTM
D790-10
[27],
the
support
span-to-depth
ratio
is
16:1.
Thus,
the
support
span
According to ASTM D790-10 [27], the support span-to-depth ratio is 16:1. Thus, the support
span
C-D
C-D)
desorption.
Subsequently,
the
Set-6
specimens
(F-30%-20
and
F-30%-40
were
tested
was
proposed
to
be
90.24
mm,
but
it
varied
between
different
groups
of
specimens,
since
the
value
was proposed to be 90.24 mm, but it varied between different groups of specimens, since the value of
of
at
relative
moisture
uptakecalculated
content after
theon
moisture
desorption.
The finalof
the
support
span
was
based
the
real
each
group
the30%
support
span
was exactly
exactly
calculated
based
on
the average
average
real thickness
thickness
ofSet-7
eachspecimens
group of
of
C-D
C-D) were
(F-0-20
F-0-40
dry specimens
after
oneThe
cycle
of moisture
specimens.
The
radii
of
and
supports
were
specimens
were
specimens.
Theand
radii
of the
the loading
loading nnose
osethe
andfully
supports
were 5.0
5.0 0.1
0.1 mm.
mm.
The
specimens
were
absorption-desorption
process.
Herein,
the
symbol
D
indicates
the
moisture
desorption.
loaded
at
the
strain
rate
0.01
mm/mm/min.
Correspondingly,
the
rate
of
crosshead
motion
was
loaded at the strain rate 0.01 mm/mm/min. Correspondingly, the rate of crosshead motion was 2.4
2.4
As shown
Figure
3, the
whole [27]:
three-point
bending test set-up was put into the chamber.
mm/min,
which
is
as
mm/min,
whichin
is calculated
calculated
as follows
follows
[27]:
According to ASTM D790-10 [27], the support span-to-depth
ratio is 16:1. Thus, the support span was
2
(1)
2 / 6d
R

ZL
(1)
proposed to be 90.24 mm, but it varied between
R different
ZL / 6d groups of specimens, since the value of the
support
span
was exactly
calculated
based
the average
ofof
each group
oftested,
specimens.
R
of crosshead
motion,
mm/min,
LL ==on
support
span,real
mm,thickness
d == depth
beam
R == rate
rate
motion,
mm/min,
support
depth
of FRP
FRPwere
beam
tested,atmm,
mm,
The
radiiofofcrosshead
the loading
nose and
supports
were 5.0 span,
0.1 mm,
mm. dThe
specimens
loaded
the
Z
=
rate
of
straining
of
the
outer
fiber,
mm/mm/min.
Z = rate
of straining
of the outer fiber,
mm/mm/min.
strain
rate
0.01 mm/mm/min.
Correspondingly,
the rate of crosshead motion was 2.4 mm/min,
The specimen
was
deflected
until the load dropped to 30% of the maximum load or the
specimen as
was
deflected
whichThe
is calculated
follows
[27]: until the load dropped to 30% of the maximum load or the
maximum
10
whichever occurred first (as illustrated in
maximum displacement
displacement of
of mid-span
mid-span reached
reached
10 2mm,
mm,
in
R ZL
{ 6d whichever occurred first (as illustrated(1)
Figure
4).
The
experimental
data
was
recorded
per
second.
To
track
the
moisture
absorption
process
Figure 4). The experimental data was recorded per second. To track the moisture absorption process
R =GFRP
rate of crosshead
motion, mm/min, L tests
= support span,
mm, d = depthdetailed
of FRP beamprocedure
tested, mm,
in
in GFRP laminate
laminate specimens,
specimens, gravimetric
gravimetric tests were
were also
also conducted,
conducted, the
the detailed test
test procedure of
of
Z = rate
of be
straining
of previous
the outer research
fiber, mm/mm/min.
which
can
found
in
[28].
which can be found in previous research [28].

Figure 3. Fle xural te st de vice .

Figure
xural te
st device.
de vice .
Figure 3.
3. Fle
Flexural
test
Load
Load

Load
Load

FMax
FMax

FMax
FMax

70%FMax
70%FMax

Displacement
Displacement

(a)
(a)

10mm
10mm

Displacement
Displacement

(b)
(b)

Figure 4. Te rmination rule of the fle xural te st. (a) Drop to 30% of the maximum load; (b) Maximum
Figure 4.
4. Termination
Te rmination rule
rule of
ofthe
the flexural
fle xuraltest.
te st. (a)
(a) Drop
Drop to
to 30%
30% of
of the
the maximum
maximumload;
load; (b)
(b) Maximum
Maximum
Figure
displace ment of 10 mm.
displace
ment
of
10
mm.
displacement of 10 mm.

Polymers 2016, 8, 290

6 of 15

The specimen was deflected until the load dropped to 30% of the maximum load or the maximum
displacement of mid-span reached 10 mm, whichever occurred first (as illustrated in Figure 4).
The experimental data was recorded per second. To track the moisture absorption process in GFRP
Polymers 2016, 8, 290
6 of 15
laminate specimens, gravimetric tests were also conducted, the detailed test procedure of which can be
found
in previous
research [28].
3. Results
and Discussion
3. Results
Discussion
Figureand
5 shows
the moisture absorption process of GFRP laminate flexural test specimens
immersed in water of 40 C, compared with the FE moisture diffusion analysis. The moisture uptake
Figure 5 shows the moisture absorption process of GFRP laminate flexural test specimens
content (Mt ) absorbed by each specimen was calculated according to its weight before exposure (w 0)
immersed in water of 40 C, compared with the FE moisture diffusion analysis. The moisture uptake
and after exposure (w t) as follows:
content (Mt ) absorbed by each specimen was calculated according to its weight before exposure (w0 )
and after exposure (wt ) as follows:
w w0
M t 100 t
(2)

w0w0
wt

Mt 100
(2)
w0
Moisture content is established as the function of the square root of time. It can be found that
the moisture
level (M) isas
about
0.77%, which
in line
with
the gravimetric
experimental
Moisturesaturation
content is established
the function
of the is
square
root
of time.
It can be found
that the
results obtained
in the
research
[28].which
The specimens
developed
a similarexperimental
moisture diffusion
moisture
saturation
levelprevious
(M8 ) is about
0.77%,
is in line with
the gravimetric
results
curve
as
that
simulated
by
the
FE
analysis
[28],
which
verifies
the
accuracy
of
the
FE
moisture
obtained in the previous research [28]. The specimens developed a similar moisture diffusion curve as
diffusion
model.
that
simulated
by the FE analysis [28], which verifies the accuracy of the FE moisture diffusion model.
1 month

1 day

3 months

1 year

6 months

110
0.8

100

0.7

90

Mt (%)

70
0.5
60
0.4

50

0.3

40

Mt/M (%)

80

0.6

30
0.2

FE analysis
test data

0.1
0.0

20
10
0

1000

2000

3000
1/2

4000

5000

6000

1/2

t (S )

Figure 5.5.Comparison
Comparison
moisture
uptake
be twee
n test and
results
and FEon
analysis
on FRP
Figure
of of
moisture
uptake
curvecurve
between
test results
FE analysis
FRP specimens
speflexural
cimens for
for
test.fle xural te st.

The typical failure mode of specimens under flexural tests is shown in Figure 6, where rupture
The typical failure mode of specimens under flexural tests is shown in Figure 6, where rupture
occurs in the outer surface of the test specimen. In order to obtain the E-modulus of GFRP laminates,
occurs in the outer surface of the test specimen. In order to obtain the E-modulus of GFRP laminates,
the stress and strain at the mid-span of GFRP specimens were calculated as follows.
the stress and strain at the mid-span of GFRP specimens were calculated as follows.
According to ASTM D790-10 [27], the flexural stress in the outer surface of the specimen at
According to ASTM D790-10 [27], the flexural stress in the outer surface of the specimen at
midpoint was calculated by means of the following equation:
midpoint was calculated by means of the following equation:
(3)
3PL / 2bd 2
3PL { 2bd2
(3)
where: = stress in the outer fibers at midpoint, MPa, P = load at the mid-span on the
load-deflection
b = width
ofmidpoint,
FRP beamMPa,
tested,
where:
= stresscurve,
in theN,
outer
fibers at
P =mm.
load at the mid-span on the load-deflection
The
flexural
strain,
that
nominal
fractional
change
in the length of an element of the outer
curve, N, b = width of FRP beam tested, mm.
surface of the test specimen at mid-span, was calculated for any deflection using Equation (4):

6Dmax d / L2

(4)

where: = strain in the outer fibers at midpoint, MPa, Dmax = maximum deflection of the center of
the beam, mm.

Polymers 2016, 8, 290

7 of 15

The flexural strain, that nominal fractional change in the length of an element of the outer surface
of the test specimen at mid-span, was calculated for any deflection using Equation (4):
6Dmax d { L2
Polymers
8, 290
where: 2016,
= strain

(4)

7 ofthe
15
in the outer fibers at midpoint, MPa, Dmax = maximum deflection of the center of

beam,
Polymersmm.
2016, 8, 290
7 of 15
The
stress-strain curves
curvesare
arepresented
presented
Figure
7. make
To make
the comparison
the
The stress-strain
in in
Figure
7. To
the comparison
moremore
clear, clear,
the curve
curve
ofone
only
one specimen
under
each
test condition
is presented,
which
was
selected
visually
as
The
stress-strain
curves
aretest
presented
inis Figure
7.
To
make
the
comparison
more
clear,
the
of only
specimen
under
each
condition
presented,
which
was
selected
visually
as the
average
the
average
curve
of
the
five
specimens.
curve of
of the
only
one
specimen under each test condition is presented, which was selected visually as
curve
five
specimens.
the average curve of the five specimens.

Figure6.6.Failure
Failuremode
modeof
ofthe
thefle
flexural
test
Figure
xural te
st specimen.
spe cime n.

Figure 6. Failure mode of the fle xural te st spe cime n.

400

stress
[MPa]
stress
[MPa]

400
300
300
200
200
100
100
0
0.000
0
0.000

0.020

F-0%-20C
F-0%-40C
F-0%-20C
F-30%-20C
F-0%-40C
F-30%-40C
F-30%-20C
F-50%-20C
F-30%-40C
F-50%-40C
F-50%-20C
F-100%-20C
F-50%-40C
F-100%-40C
F-100%-20C
F-50%-20C-D
F-100%-40C
F-50%-40C-D
F-50%-20C-D
F-30%-20C-D
F-50%-40C-D
F-30%-40C-D
F-30%-20C-D
F-0%-20C-D
F-30%-40C-D
F-0%-40C-D
F-0%-20C-D
F-0%-40C-D

0.005

0.010

0.015

0.025

0.030

0.035

0.040

0.005

0.010

strain 0.020
[mm/mm]
0.015
0.025

0.030

0.035

0.040

[mm/mm]
Figure 7. Stre ss-strain curve strain
s of FRP
spe cime ns unde r fle xural te sts.

Figure
7. 7.
Stre
ss-strain curve
s of
FRP specimens
spe cime nsunder
unde rflexural
fle xuraltests.
te sts.
Figure
Stress-strain
curves
of FRP

The E-modulus is represented by the chord modulus, which was calculated from two discrete
points
on E-modulus
the load-deflection
curve, using
(5) [27]: which was calculated from two discrete
The
is represented
by theEquation
chord modulus,
The E-modulus is represented by the chord modulus, which was calculated from two discrete
points on the load-deflection curve, using Equation (5) [27]:
E Equation
( 2 1 )(5)/ ([27]:
2 1 )
(5)
points on the load-deflection curve, using
E ( 2 1 ) / ( 2 1 )
(5)
1 , 1 and 2 , 2 are the flexural
stress
E p
q { pstrain
(5)
2 1and
2 1 qselected at two points of stress-strain

1(see
, 1Figure
and
, the
thestable
flexural
stress and strain selected at two points of stress-strain
curves
7)2in
linear
range.
2 are
The
flexural
strength
is
the
maximum
flexural
stress sustained by the test specimen during the
curves (see Figure 7) in the linear stable range.
flexural
The strength
environment-dependent
properties
(including
andduring
strength)
Thetest.
flexural
is the maximum flexural
flexural stress
sustained
by theE-modulus
test specimen
the
are
shown
in The
Figure
8 and Table 4. The R-square
for each
curve is E-modulus
also presentand
in Figure
8,
flexural
test.
environment-dependent
flexuralvalue
properties
(including
strength)
which
indicates
how
well
the
curve
fits
test
data
points.
are shown in Figure 8 and Table 4. The R-square value for each curve is also present in Figure 8,

Polymers 2016, 8, 290

8 of 15

1 , 1 and 2 , 2 are the flexural stress and strain selected at two points of stress-strain curves
(see Figure 7) in the linear stable range.
The flexural strength is the maximum flexural stress sustained by the test specimen during the
flexural test. The environment-dependent flexural properties (including E-modulus and strength)
are shown in Figure 8 and Table 4. The R-square value for each curve is also present in Figure 8,
which
indicates
how well the curve fits test data points.
Polymers
2016, 8, 290
8 of 15
18000

450
2

R =0.6784

16000

400
350

R =0.6329

Flexural strength (MPa)

12000
10000
8000
6000

absorption
absorption-desorption

4000

R =0.8402

300
250

R =0.9570
200
150

absorption
absorption-desorption

100
50

2000

0
0.1

0.2

0.3

0.4

0.5

Mt/M

0.6

0.7

0.8

0.9

0.0

1.0

0.1

0.2

0.3

0.4

0.5

Mt/M

0.0

(a)

0.6

0.7

0.8

0.9

1.0

E-modulus (MPa)

14000

(b)

18000

450

16000

400

R =0.6326
350

12000

Flexural strength (MPa)

R =0.0632
10000
8000
6000

absorption
absorption-desorption

4000

300
2

R =0.8760

250

200

R =0.9403

150

absorption
absorption-desorption

100

2000

50

0.1

0.2

0.3

0.4

0.5

Mt/M

(c)

0.6

0.0

0.7

0.8

0.9

1.0

0.0

0.1

0.2

0.3

0.4

0.5

Mt/M

0.6

0.7

0.8

0.9

1.0

E-modulus (MPa)

14000

(d)

Figure 8. Env ironme nt-de pe nde nt fle xural prope rty degradation of FRP laminates. (a) E-modulus,
Figure 8. Environment-dependent flexural property degradation of FRP laminates. (a) E-modulus,
20 C; (b) Stre ngth, 20 C; (c) E-modulus, 40 C; (d) Stre ngth, 40 C.
20 C; (b) Strength, 20 C; (c) E-modulus, 40 C; (d) Strength, 40 C.

Specimen
identification
Specimen
identification
F-0%-20
C
C
F-0%-40
F-0%-20C
C
F-0%-40C
F-30%-20

F-30%-20
F-30%-40 CC
C
F-30%-40
F-50%-20 C
F-50%-20 C
F-50%-40
C
F-50%-40C
C
F-100%-20
F-100%-20C
C
F-100%-40C
F-100%-40
C-D
F-50%-20
F-50%-20 C-D
F-50%-40 C-D
F-50%-40 C-D
F-30%-20 C-D
F-30%-20
C-D
F-30%-40C-D
C-D
F-30%-40
F-0%-20C-D
C-D
F-0%-40C-D
F-0%-20
F-0%-40 C-D

Table 4. Fle xural prope rty de gradation of GFRP laminate s.

Table 4. Flexural property degradation of GFRP laminates.
E-Modulus *
Standard deviation
Flexural strength *
Standard
(MPa)
(MPa)
(MPa) *
deviation
(MPa)
E-Modulus
*
Standard
deviation
Flexural strength
Standard
deviation
(MPa)
(MPa)
(MPa)
(MPa)
16,609
386
411
6.89
15,409
852
8.49
16,609
386
411375
6.89
15,409
852
375299
8.49
15,873
867
8.88
15,873
867
299249
8.88
13,874
500
11.42
13,874
500
249269
11.42
15,038
710
17.03
15,038
710
269
17.03
13,870
535
252
17.57
13,870
535
252
17.57
14,022
514
265
13.74
14,022
514
265
13.74
12,780
538
214214
6.08
12,780
538
6.08
14,408
551
260260
15.46
14,408
551
15.46
13,019
832
239
17.85
13,019
832
239
17.85
15,059
740
324
37.04
15,059
740
324
37.04
12,336
166
277
9.85
12,336
166
9.85
16,333
204
410277
10.80
13,095
698
314410
13.92
16,333
204
10.80
13,095
314
13.92
* mean value698
of five specimens.
* me an value of five spe cime ns.

To be simplified, the predictive equation for the E-modulus degradation as the function of
moisture content was curve fitted by the linear interpolant function, while for the flexural strength
the exponential function was used. All the curve fitting processes were conducted by MATLAB
R2011b, employing the least square method. The obtained predictive equation is as follows:
E-modulus, 20 C, absorption process:

Polymers 2016, 8, 290

9 of 15

To be simplified, the predictive equation for the E-modulus degradation as the function of
moisture content was curve fitted by the linear interpolant function, while for the flexural strength the
exponential function was used. All the curve fitting processes were conducted by MATLAB R2011b,
employing the least square method. The obtained predictive equation is as follows:
E-modulus, 20 C, absorption process:
E 2725 Mt {M8 ` 16, 609

(6)

E-modulus, 20 C, absorption-desorption process:

E 2428 Mt {M8 ` 16, 333

(7)

Flexural strength, 20 C, absorption process:

Mt

S 103p M8 1.1q ` 257

(8)

Flexural strength, 20 C, absorption-desorption process:

Mt

S 15.5p M8 1.87q ` 242

(9)

E-modulus, 40 C, absorption process:

E 2795 Mt {M8 ` 15, 409

(10)

E-modulus, 40 C, absorption-desorption process:

E 628 Mt {M8 ` 13, 095

(11)

Flexural strength, 40 C, absorption process:

Mt

S 123p M8 1.04q ` 221

(12)

Flexural strength, 40 C, absorption-desorption process:

Mt

S 3.23p M8 4.25q ` 166

(13)

All the predictive curves are illustrated in Figure 8 for comparison with the experimental results.
For the specimens tested at 20 C (Figure 8a), the E-modulus of GFRP laminates decreases
gradually as the moisture content increases from fully dry to fully saturated. The E-modulus of the
moisture saturated specimen is 14,022 MPa (as shown in Table 4), which is 15.6% lower than that of
the unconditioned dry specimen. For the specimens with the moisture contents 30% and 50% of the
saturation level, the loss of E-modulus is 4.4% and 9.5% respectively. With regard to specimens in
the moisture desorption process, the E-modulus does not decrease significantly as compared to the
specimens at the same moisture uptake level. Accordingly, the slight loss of E-modulus is 1.7%, 5.1%
and 4.2% at the moisture uptake levels 0%, 30% and 50% respectively. In terms of flexural strength
(see Figure 8b), there is a general exacerbation of decrease between the fully dry specimens and 30%
moisture content specimens, regardless of the moisture absorption/desorption process. More than
20% loss of flexural strength is evident. After this stage, as the moisture uptake content increases,
the flexural strength of specimens slightly decreases, until reaching 265 MPa for the moisture fully
saturated specimens. At the end, the total drop of flexural strength is 35.4% of the fully dry specimens.
Similar to the E-modulus degradation, the difference of flexural strength between absorption process
and desorption process is very limited.

Polymers 2016, 8, 290

10 of 15

With regard to the environment-dependent flexural properties of GFRP laminates at 40 C,

the E-modulus of GFRP laminates regularly decreases from 15,409 MPa (dry) to 12,780 MPa (moisture
fully saturated), and then slightly increases to 13,095 MPa after being fully dried in the moisture
desorption process. Different from that at 20 C, there is a significant unrecoverable loss (15.0%)
of the E-modulus for the dry specimens tested at 40 C. Accordingly, the loss of 11.1% and 6.1%
of the E-modulus is evident for the specimens with 30% and 50% moisture content respectively in
the desorption process. As to the flexural strength of specimens tested at 40 C, a rapid decrease
is observed at a moisture content of about 30% of moisture saturation level, and then the flexural
strength slightly decreases to 214 MPa of the moisture fully saturated specimens, which is the lowest
value among the whole series of tests. It is 42.9% lower than that of the unconditioned dry specimens
(375 MPa, as listed in Table 4) tested at 40 C, and 47.9% lower than that of the unconditioned dry
specimens (410 MPa) tested at 20 C. The most severe loss of flexural strength indicates that the
combination of moisture and temperature effects can significantly influence the mechanical properties
of GFRP laminates. Comparing the unconditioned dry specimens with the dry specimens after the
moisture desorption, a decrease of 16.4% of the flexural strength is observed for the specimens tested
at 40 C. However, for the specimens with moisture contents 30% and 50%, the loss of flexural strength
is not obvious. Even a slight increase is observed for the specimens with 30% moisture content in the
moisture desorption process.
It is clear that the hot and wet environment seriously degrades the flexural properties of GFRP
laminates, and in turn influences the durability of FRP composite structures. For the moisture effects,
as stated in the researches [21,29,30], the absorbed moisture can cause both reversible and irreversible
changes to the FRP composites, such as water plasticization and the disruption of hydrogen bonds
between the molecular chains in the polymer. It is because the matrix often has hydrophilic groups
that attract water molecules, which can form hydrogen bonds with water molecules via the hydroxyl
groups [21]. Plasticization is a physical mechanism, which degrade the stiffness and strength of
the polymer. It can be reversible at the initial stages of aging when the absorbed water is removed
and no chemical reaction occurs. Furthermore, prolonged environmental exposure often leads to
irreversible changes that induce permanent property alterations within the matrix, the fibre surfaces
and the fibre/matrix interface. For instance, the polymer relaxation process is related to the polymer
chain rearranging or redistribution of free-volume elements to provide additional sites of suitable size
and accessibility to accommodate more penetrant molecules. It is a long-term and complicated
process, which is a combination of chemical and physical reaction. Moreover, the fibre/matrix
debonding, matrix cracking due to moisture/thermal cycles and ultimately fiber breaking would
also contribute to the mechanical degradation. The fibre/matrix interface and the interphase region
around fibres, are susceptible to deterioration due to moisture absorption. With the fibre/matrix
debonding [31], there is enhanced interfacial capillary action along the fibres that can offer more space
for moisture penetration and further promote their degradation. With regard to the temperature effects,
when the test temperature is approaching the Tg of FRP composite materials, the mechanical
performance such as E-modulus, strength and fatigue resistance significantly decreases [32,33].
Sometimes, the mass loss of FRP composites occurs following high-temperature exposure [28,34].
The recommended working temperature for FRP composite structures should be at least 20 C lower
than the Tg of FRP composite materials. Furthermore, the researches [33,35] confirmed that a decrease
of Tg was evident when the moisture uptake content increased in the FRP composite materials.
Meanwhile, as already proven in previous research [28], the high temperature can speed up the
moisture diffusion process. Thus, the interaction between moisture and temperature effects accelerates
the environmental degradation process on the FRP composite material, which in turn explains why
the combination of moisture and temperature seriously deteriorates the mechanical properties of
FRP materials.

Polymers 2016, 8, 290

11 of 15

4. Coupled Hygro-Mechanical FE Analysis

Generally, the environmental degradation experiments (mainly concerning moisture and
temperature effects) for FRP composite materials are limited to a fairly short time, normally
no more than 5 years. However, the expected service life of infrastructures such as bridges
exceeds 50 years. Thus, the short-term experimental investigations are not sufficient to estimate
the long-term performance of FRP structures. To achieve this aim, some accelerated experimental
methods were developed by the researches [10,3638], in which the temperature or atmospheric
pressure of the environmental aging conditions were raised beyond the normal service conditions to
accelerate the moisture diffusion and degradation process. These accelerating experimental methods
were confirmed useful and time-effective to investigate the durability of FRP composites and adhesive
Polymers
2016,But,
8, 290researches indicated that the high aging temperatures approaching glass transition
11 of 15
materials.
temperature of specimens would improve the mechanical performance by post-cure or deteriorate the
the
materials
by inducing
thermal
notinoccur
in use
the of
real
use
of FRP composite
materials
by inducing
thermal
cracks,cracks,
which which
do not do
occur
the real
FRP
composite
structures.
structures.
Another
method
for
studying
the
environmental
degradation
of
mechanical
behaviour
of
Another method for studying the environmental degradation of mechanical behaviour of FRP
materials
FRP
materials is
and
the
predictive
FEThe
modeling
[39].
The first moisture
step is modeling
and structures
thestructures
predictiveisFE
modeling
[39].
first step
is modeling
transportmoisture
through
transport
through
FRP
structures
in
order
to
determine
the
moisture
concentration
distribution
FRP structures in order to determine the moisture concentration distribution through the cross-sections
through
the cross-sections
a function
of time.required
The material
parameters
transient
as a function
of time. The as
material
parameters
for the
transient required
diffusionfor
FE the
analysis
are
diffusion
FE
analysis
are
the
diffusion
coefficient
and
the
solubility
coefficient,
which
can
be
the diffusion coefficient and the solubility coefficient, which can be obtained from the short term
obtained
from
the
short
term
gravimetric
experiments,
as
stated
in
previous
research
[28].
And
then,
gravimetric experiments, as stated in previous research [28]. And then, based on the obtained moisture
based
on the obtained
moisture
concentration distribution,
the environment-dependent
mechanical
concentration
distribution,
the environment-dependent
mechanical
behaviour of FRP structures
can be
behaviour
of
FRP
structures
can
be
investigated
using
the
coupled
hygro-mechanical
FE
investigated using the coupled hygro-mechanical FE analysis. The input moisture-dependentanalysis.
material
The
input moisture-dependent
properties
of FRP tests
composites
obtained
the material
properties
of FRP composites arematerial
obtained
by the material
(such asare
flexural
test, by
tensile
test and
tests
(such
as
flexural
test,
tensile
test
and
short-beam
shear
test).
short-beam shear test).
To
FE modeling
odel of
To develop
develop the
the coupled
coupled hygro-mechanical
hygro-mechanical FE
modeling method,
method, the
the FE
FE m
model
of the
the GFRP
GFRP
flexural
test
specimen
was
calibrated
(as
shown
in
Figure
9).
The
moisture
diffusion
process
in the
flexural test specimen was calibrated (as shown in Figure 9). The moisture diffusion process
in
FRP
specimen
was modeled
by theby
transient
diffusion
FE analysis
(thoroughly
describeddescribed
in previous
the FRP
specimen
was modeled
the transient
diffusion
FE analysis
(thoroughly
in
research
and validated
comparing
with the gravimetric
results, as shown
in
previous [28])
research
[28]) and by
validated
by comparing
with the experimental
gravimetric experimental
results,
Figure
5. From
the moisture
diffusion
analysis,
theanalysis,
moisture
distributiondistribution
across the
as shown
in Figure
5. From the
moisture
diffusion
theconcentration
moisture concentration
FRP
specimen
section
was
obtained
as
a
function
of
time,
which
can
be
read
into
the
stress
analysis
across the FRP specimen section was obtained as a function of time, which can be read into the
stress
as
a
predefined
field
variable.
The
predictive
equations
(Equations
(6),
(7),
(10)
and
analysis as a predefined field variable. The predictive equations (Equations (6), (7), (10) and(11))
(11)) of
of
environment-dependent
material
properties
served
as
the
input
of
field-dependent
material
environment-dependent material properties served as the input of field-dependent material properties
properties
of FRP composites,
which was
by tests.
the flexural
Hence, theofE-modulus
of
of FRP composites,
which was obtained
by obtained
the flexural
Hence, tests.
the E-modulus
each element
each
element
was
determined
by
the
local
moisture
concentration.
Thus,
the
coupled
was determined by the local moisture concentration. Thus, the coupled hygro-mechanical FE modeling
hygro-mechanical
m odeling
was realized
on the flexural
specimen.toSubsequently,
can be
was realized on theFE
flexural
test specimen.
Subsequently,
it can test
be employed
simulate otheritmaterial
employed
to
simulate
other
material
tests
(such
as
short-beam
shear
tests)
and
FRP
structures.
tests (such as short-beam shear tests) and FRP structures.

Figure
l ofofthe
xural te
st spe
cime n.
Figure9.9.FE
FEmode
model
thefle
flexural
test
specimen.

To validate this coupled hygro-mechanical FE modeling method, two groups of FRP specimens
To validate this coupled hygro-mechanical FE modeling method, two groups of FRP specimens
were selected: F-50%-20 C
and F-30%-40 C-D. Firstly, the moisture concentration distributions
were selected: F-50%-20 C and F-30%-40 C-D. Firstly, the moisture concentration distributions
throughout the mid-section of these two specimens are shown in Figures 10 and 11. Based on these
throughout the mid-section of these two specimens are shown in Figures 10 and 11. Based on these
two predefined fields, the coupled hygro-mechanical FE analysis was executed. The FE stress
two predefined fields, the coupled hygro-mechanical FE analysis was executed. The FE stress analysis
analysis was linear-elastic. The element used was C3D8R, which is an 8-node linear brick element,
was linear-elastic. The element used was C3D8R, which is an 8-node linear brick element, with the
with the capacity of reduced integration and hourglass control. In total 2000N load was applied at
capacity of reduced integration and hourglass control. In total 2000N load was applied at mid-span of
mid-span of the specimen.
the specimen.

were selected: F-50%-20 C and F-30%-40 C-D. Firstly, the moisture concentration distributions
throughout the mid-section of these two specimens are shown in Figures 10 and 11. Based on these
two predefined fields, the coupled hygro-mechanical FE analysis was executed. The FE stress
analysis was linear-elastic. The element used was C3D8R, which is an 8-node linear brick element,
with
the capacity of reduced integration and hourglass control. In total 2000N load was applied
at
Polymers 2016, 8, 290
12 of 15
mid-span of the specimen.

Polymers
2016, 8,
290Nominal
12 of 15
Figure
10.
ntration distribution
Figure
10.
Nominalmoisture
moistureconce
concentration
distributionacross
across the
the mid-plane
mid-plane of
of the
the FRP
FRP

specimenwith
30%moisture
moisture uptake content (time =
24h).
h).
spe
cimenwith 30%
= 24

Polymers 2016, 8, 290

12 of 15

Figure
n
Figure 11.
11. Nominal
Nominal moisture
moisture concentration
concentration distribution
distributionacross
across the
the mid-plane
mid-plane of
of the
the FRP specime
specimen
Figure
11.
Nominal
moisture
concentration
across the mid-plane of the FRP specime n
with
50%
content
h).
with
50% moisture
moisture
uptake
content
(time = 229distribution
h).
with 50% moisture uptake content (time = 229 h).

The
The comparisons
comparisons of
of experimental
experimental results
results and
and FE
FE results
results on
on the
the load-deflection
load-deflection curves
curves of
of
The
comparisons
of
experimental
results
and
FE
results
on
the
load-deflection
curves
of13
F-50%-20
C
specimens
and
F-30%-40
C-D
specimens
are
illustrated
in
Figures
12
and

F-50%-20
C specimens
and and
F-30%-40
C-D C-D
specimens
are illustrated
in Figures
12 and 1312respectively.
F-50%-20 CGood
specimens
F-30%-40
specimens
illustrated
and 13
respectively.
agreement
is obtained
in means
the arelinear
stage,in Figures
which means
the
Good
agreement
is
obtained
in
the
linear
stage,
which
the
environment-dependent
stiffness
respectively.
Good
agreement
is
obtained
in
the
linear
stage,
which
means
thebe
environment-dependent
stiffness
of
GFRP
specimens
with
different
moisture
uptake
contents
can
of environment-dependent
GFRP specimens withstiffness
different
moisture
uptakewith
contents
can
be predicted
by the can
coupled
of GFRP
specimens
different
moisture
uptake contents
be
predicted
by
the
coupled
hygro-mechanical
FE
model.
Hence,
the
accuracy
of
the
FE
model
is
hygro-mechanical
FE
model.
Hence,
the
accuracy
of
the
FE
model
is
validated,
and
the
methodology
predicted by the coupled hygro-mechanical FE model. Hence, the accuracy of the FE model is
validated,
and
the
methodology
of
the
hygro-mechanical
numerical
analysis
is
proven.
of validated,
the hygro-mechanical
numerical
is proven.
and the methodology
ofanalysis
the hygro-mechanical
numerical analysis is proven.
1400
1400

1000
1000

Load [N]

Load [N]

1200
1200

800
800
600600

F-50%-20C-01
F-50%-20C-01
F-50%-20C-02
F-50%-20C-02
F-50%-20C-03
F-50%-20C-03
F-50%-20C-04
F-50%-20C-04
FEM
FEM

400400
200200
0 0
0 0

11

22

33

44

55

66

77

88

Deflection [mm]
[mm]
Deflection

Figure 12.Comparison
Comparison of e xperimental
ntal and FE
FE results on
the load-de
flection
of
Figure
flection curve
curve
Figure 12.
12.
Comparisonofofe xperime
experimentaland
and FE results
results on
on the
the load-de
load-deflection
curve of
of
F-50%-20
spe
cimens.
CC
F-50%-20
spe
cimens.
F-50%-20 C
specimens.
1400
1400
1200
1200

[N]

ad [N]

1000
1000
800

F-50%-20C-04
FEM

200
0
0

Deflection [mm]

Figure
Polymers
2016, 8,12.
290 Comparison of e xperime ntal and FE results on the load-de flection curve of
13 of 15
F-50%-20 C spe cimens.
1400
1200

Load [N]

1000
800
600

F-30%-40C-D-01
F-30%-40C-D-02
F-30%-40C-D-03
F-30%-40C-D-04
F-30%-40C-D-05
FEM

400
200
0
0

Deflection [mm]

Figure 13. Comparison of e xperime ntal and FE results on the load-de flection curve of
Figure 13.
Comparison of experimental and FE results on the load-deflection curve of
F-30%-40 C-D spe cime ns.
F-30%-40 C-D specimens.

5. Conclusions
This paper describes the investigation on the environment-dependent flexural properties of the
FRP laminate material, which is achieved by the three-point bending tests according to the ASTM test
code D790-10. The hydrothermal aging condition is a typical hot/wet aging environment (40 C-water)
for the application of FRP bridge decks. The test conditions vary in terms of test temperature, moisture
uptake content and absorption/desorption process. The following conclusions can be drawn:
(1)

(2)

(3)
(4)

(5)

For GFRP laminates tested at 20 C, the E-modulus and flexural strength of moisture saturated
GFRP laminates dropped 15.6% and 35.4% respectively, comparing with the fully dry materials.
While, the difference of E-modulus and flexural strength between absorption process and
desorption process was very limited.
For GFRP laminates tested at 40 C, the E-modulus and flexural strength of moisture saturated
GFRP laminates dropped 17.1% and 42.9% respectively, comparing with the fully dry materials.
There were significant unrecoverable losses of E-modulus (15.0%) and flexural strength (16.4%)
for the GFRP laminates experienced both moisture absorption and desorption process.
Experimental results confirmed that the combination of moisture and temperature seriously
deteriorated the flexural properties of GFRP laminates, on both strength and stiffness.
Predictive equations for environment-dependent flexural properties of FRP laminates were
obtained by using the least square method for the curve fitting. These predictive equations
can be employed as the input parameters for a coupled hygro-mechanical FE model, as well
as a contribution to the design code as far as the long-term performance of FRP structures
is concerned.
A coupled hygro-mechanical FE modeling method was developed to analyze the
environment-dependent mechanical behaviours of FRP laminates, and was validated by the
experimental results of flexural tests. In the following research works, it could be employed in
the parametric FE analysis, and predict the environment-dependent behaviours of other material
tests or FRP structures.

Acknowledgments: The authors would like to thank the National Natural Science Foundation of China
(Program No. 51408424), the Fundamental Research Funds for the Central Universities and Infra Composites B.V.
for financial support and supplying GFRP laminates of this research.
Author Contributions: Xu Jiang conceived and designed the test setup and experiments. Xu Jiang performed the
experiments. Xu Jiang and Xuhong Qiang analyzed the test data. Xuhong Qiang did the finite element analysis.
Xu Jiang and Xuhong Qiang co-wrote the paper. Jie Song behaved as research consultant and offered valuable
knowledge to guide the research. Jie Song and Xu Jiang revised the paper. Henk Kolstein and Frans Bijlaard
supervised the whole research process, and organized the works related to the experiments.

Polymers 2016, 8, 290

14 of 15

Conflicts of Interest: The authors declare no conflict of interest.

References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.

22.
23.

Knippers, J.; Pelke, E.; Gabler, M.; Berger, D. Bridges with Glass Fibre-Reinforced Polymer Decks:
The Road Bridge in Friedberg, Germany. Struct. Eng. Int. 2010, 20, 400404. [CrossRef]
Luke, S.; Canning, L.; Collins, S.; Knudsen, E.; Brown, P.; Taljsten, B.; Olofsson, I. Advanced composite bridge
decking systemproject ASSET. Struct. Eng. Int. 2002, 12, 7679. [CrossRef]
Alampalli, S.; Kunin, J. Rehabilitation and field testing of an FRP bridge deck on a truss bridge.
Compos. Struct. 2002, 57, 373375. [CrossRef]
Schollmayer, M. Through-Thickness Performance of Adhesive Connections Between FRP Bridge Decks and Steel
Main Girders; Ecole Polytechnique Federale de Lausanne: Lausanne, Switzerland, 2009.
Gurtler, H.W. Composite Action of FRP Bridge Decks Adhesively Bonded to Steel Main Girders Lausanne;
Ecole Polytechnique Federale de Lausanne: Lausanne, Switzerland, 2004.
Keller, T.; Bai, Y.; Vallee, T. Long-term performance of a glass fiber-reinforced polymer truss bridge.
J. Compos. Constr. 2007, 11, 99108. [CrossRef]
Chen, A.; Davalos, J.F. Strength evaluations of sinusoidal core for FRP sandwich bridge deck panels.
Compos. Struct. 2010, 92, 15611573. [CrossRef]
Hollaway, L.C. A review of the present and future utilisation of FRP composites in the civil infrastructure
with reference to their important in-service properties. Constr. Build. Mater. 2010, 24, 24192445. [CrossRef]
Shao, Y.X.; Kouadi, S. Durability of fiberglass composite sheet piles in water. J. Compos. Constr. 2002, 6,
280287. [CrossRef]
Bank, L.C.; Gentry, T.R.; Barkatt, A. Accelerated Test Methods to Determine the Long-Term Behavior of Frp
Composite StructuresEnvironmental-Effects. J. Reinf. Plast. Compos. 1995, 14, 559587.
Nogueira, P.; Ramirez, C.; Torres, A.; Abad, M.J.; Cano, J.; Lopez, J.; Barral, L. Effect of water sorption on the
structure and mechanical properties of an epoxy resin system. J. Appl. Polym. Sci. 2001, 80, 7180. [CrossRef]
Earl, J.S.; Shenoi, R.A. Hygrothermal ageing effects on FRP laminate and structural foam materials.
Compos. Part A-Appl. Sci. Manuf. 2004, 35, 12371247. [CrossRef]
Robert, M.; Roy, R.; Benmokrane, B. Environmental Effects on Glass Fiber Reinforced Polypropylene
Thermoplastic Composite Laminate for Structural Applications. Polym. Compos. 2010, 31, 604611. [CrossRef]
Karbhari, V.M. E-glass/vinylester composites in aqueous environments: Effects on short-beam shear strength.
J. Compos. Constr. 2004, 8, 148156. [CrossRef]
Daly, H.B.; Brahim, H.B.; Hfaied, N.; Harchay, M.; Boukhili, R. Investigation of water absorption in pultruded
composites containing fillers and low profile additives. Polym. Compos. 2007, 28, 355364. [CrossRef]
Karbhari, V.M.; Xian, G.J. Hygrothermal effects on high V(F) pultruded unidirectional carbon/epoxy
composites: Moisture uptake. Compos. Part B-Eng. 2009, 40, 4149. [CrossRef]
DellAnno, G.; Lees, R. Effect of water immersion on the interlaminar and flexural performance of low cost
liquid resin infused carbon fabric composites. Compos. Part B-Eng. 2012, 43, 13681373. [CrossRef]
Springer, G.S. Enivormental Effects on Composite Materials, Vol.1; Technomic Pubnishing Company: Lancaster,
PA, USA, 1981.
Springer, G.S. Enivormental Effects on Composite Materials, Vol.2; Technomic Pubnishing Company: Lancaster,
PA, USA, 1984.
Springer, G.S. Enivormental Effects on Composite Materials, Vol.3; Technomic Pubnishing Company: Lancaster,
PA, USA, 1987.
Loh, W.K.; Crocombe, A.D.; Wahab, M.M.A.; Ashcroft, I.A. Modelling anomalous moisture uptake, swelling
and thermal characteristics of a rubber toughened epoxy adhesive. Int. J. Adhes. Adhes. 2005, 25, 112.
[CrossRef]
Haque, A.; Mahmood, S.; Walker, L.; Jeelani, S. Moisture and Temperature Induced Degradation in Tensile
Properties of Kevlar Graphite Epoxy Hybrid Composites. J. Reinf. Plast. Compos. 1991, 10, 132145. [CrossRef]
Akay, M.; Mun, S.K.A.; Stanley, A. Influence of moisture on the thermal and mechanical properties of
autoclaved and oven-cured Kevlar-49/epoxy laminates. Compos. Sci. Technol. 1997, 57, 565571. [CrossRef]

Polymers 2016, 8, 290

24.

25.
26.
27.
28.

29.
30.

31.
32.
33.

34.
35.
36.
37.
38.
39.

15 of 15

Eslami, S.; Honarbakhsh-Raouf, A.; Eslami, S. Effects of moisture absorption on degradation of E-glass fiber
reinforced Vinyl Ester composite pipes and modelling of transient moisture diffusion using finite element
analysis. Corros. Sci. 2015, 90, 168175. [CrossRef]
Carra, G.; Carvelli, V. Long-term bending performance and service life prediction of pultruded Glass Fibre
Reinforced Polymer composites. Compos. Struct. 2015, 127, 308315. [CrossRef]
Silva, M.A.G.; da Fonseca, B.S.; Biscaia, H. On estimates of durability of FRP based on accelerated tests.
Compos. Struct. 2014, 116, 377387. [CrossRef]
Standard Test Methods for Flexural Properties of Unreinforced and Reinforced Plastics and Electrical Insulating
Materials; D79010 A; ASTM International: Philadelphia, PA, USA, 2010.
Jiang, X.; Kolstein, H.; Bijlaard, F.; Qiang, X.H. Effects of hygrothermal aging on glass-fibre reinforced
polymer laminates and adhesive of FRP composite bridge: Moisture diffusion characteristics. Compos. Part
A-Appl. Sci. Manuf. 2014, 57, 4958. [CrossRef]
Gellert, E.P.; Turley, D.M. Seawater immersion ageing of glass-fibre reinforced polymer laminates for marine
applications. Compos. Part A-Appl. Sci. Manuf. 1999, 30, 12591265. [CrossRef]
Grammatikos, S.A.; Evernden, M.; Mitchels, J.; Zafari, B.; Mottramd, J.T.; Papanicolaou, G.C. On the response
to hygrothermal aging of pultruded FRPs used in the civil engineering sector. Mater. Des. 2016, 96, 283295.
[CrossRef]
Joliff, Y.; Belec, L.; Heman, M.B.; Chailan, J.F. Experimental, analytical and numerical study of water diffusion
in unidirectional composite materialsinterphase impact. Comput. Mater. Sci. 2012, 64, 141145. [CrossRef]
Ashcroft, I.A.; Hughes, D.J.; Shaw, S.J.; Wahab, M.A.; Crocombe, A. Effect of temperature on the quasi-static
strength and fatigue resistance of bonded composite double lap joints. J. Adhes. 2001, 75, 6188. [CrossRef]
Ashcroft, I.A.; Wahab, M.M.A.; Crocombe, A.D.; Hughes, D.T.; Shaw, S.J. The effect of environment on the
fatigue of bonded composite joints. Part 1: Testing and fractography. Compos. Part A-Appl. Sci. Manuf. 2001,
32, 4558. [CrossRef]
Zhou, J.M.; Lucas, J.P. The effects of a water environment on anomalous absorption behavior in
graphite-epoxy composites. Compos. Sci. Technol. 1995, 53, 5764. [CrossRef]
Barjastech, E.; Nutt, S.R. Moisture absorption of unidirectional hybrid composites. Compos. Part A-Appl. Sci
Manuf. 2012, 43, 158164. [CrossRef]
Muliana, A.; Nair, A.; Khan, K.A.; Wagner, S. Characterization of thermo-mechanical and long-term behaviors
of multi-layered composite materials. Compos. Sci. Technol. 2006, 66, 29072924. [CrossRef]
Davies, P.; Evrard, G. Accelerated ageing of polyurethanes for marine applications. Polym. Degrad. Stab.
2007, 92, 14551464. [CrossRef]
Moulzakis, D.E.; Zoga, H.; Galiotis, C. Accelerated environmental ageing study of polyester/glass fiber
reinforced composites (GFRPCs). Compos. Part B-Eng. 2008, 39, 467475. [CrossRef]
Crocombe, A.D.; Ashcroft, I.A. Abdel Wahab MM. Chapter 8 Environmental Degradation. Modeling of Adhesively
Bonded Joints; Springer-Verlag: Berlin, Germany, 2008.
2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC-BY) license (http://creativecommons.org/licenses/by/4.0/).