Professional Documents
Culture Documents
discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/316447627
CITATIONS READS
2 141
4 authors, including:
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Yiming Yao on 25 October 2017.
Abstract Fiber reinforced concrete (FRC) can be designed textile reinforced concrete, and ultra-high performance
to exhibit pronounced ductility, energy absorption capac- fiber reinforced concrete.
ity, post-cracking strength depending on the fiber type and
volume fraction. FRC have been classified into two cate- Keywords Cementitious composites
gories, namely, strain softening and strain hardening Distributed cracking Textile reinforced concrete
cement composites (SSCC and SHCC). SSCC and SHCC Strain hardening Fiber-reinforced concrete Toughness
are ultra-ductile class of materials developed for applica- Design
tions in the large material volume usage in the cost sensi-
tive construction industry. Strain hardening behavior can
be obtained by adding relatively low volume (typically 1 Introduction
\2%) of randomly distributed fibers and demonstrates a
well formed distributed crack system. Mechanical proper- The high level of strength, ductility, and versatility attained
ties under uniaxial tensile, flexural, and shear tests indicate in the general field of strain hardening cement composites
superior performance such as tensile strength as high as (SHCC) has been observed for a broad range of composites
25 MPa, and strain capacity of 1–8%. Development of such as: textile reinforced concrete (TRC), high perfor-
proper design and analysis tools are very essential to fully mance fiber reinforced concrete (HPFRC), and ultra-high
utilize these materials. Several approaches are presented in performance fiber reinforced concrete (UHPFRC). This
this paper for analysis, simulation, back-calculation, and strain hardening behavior improves the durability of con-
design of strain softening and strain hardening cement crete structures due to the ability to limit the width of
composite systems, and are applicable to all classes of cracks and carry tensile stresses due to bridging effect of
SSCC and SHCC such as steel fiber reinforced concrete, fibers [1]. A variety of fiber and fabric systems such as
Alkali resistant (AR) glass fibers, polypropylene (PP),
& Barzin Mobasher
polyethylene (PE), and polyvinyl alcohol (PVA) have been
barzin@asu.edu utilized for the manufacture of SHCC [2–4]. While these
Yiming Yao
composites are manufactured differently, their mechanical
yiming.yao@asu.edu responses are under the general category of strain harden-
Xinmeng Wang
ing or deflection hardening.
xinmeng.wang@asu.edu The general strain hardening behavior of cement com-
Karan Aswani
posites is illustrated in Fig. 1, representing the tensile stress–
kaswani1@asu.edu strain and crack spacing-strain responses. The initial
response is linear elastic up to about 3 MPa when the first of a
1
School of Sustainable Engineering and Built Environment, series of parallel cracks takes place. The mechanism of
Arizona State University, Tempe, AZ 85287-8706, USA
multiple cracking is due to load transfer to the fibers or fabric
2
DiGioia Gray & Associates, Tempe, AZ 85282, USA layers across the crack [5]. As multiple cracking takes place,
3
PK Associates, Scottsdale, AZ 85250, USA the stiffness of the sample significantly drops while the crack
123
Int J Adv Eng Sci Appl Math
123
Int J Adv Eng Sci Appl Math
Table 2 Neutral axis depth ratio k, normalized moment mi and curvature ji for each stage i of normalized tensile strain at bottom fiber (b)
Stage ki mi and ji
2b½ðc1Þk13 þ3k12 3k1 þ1
1 k1 ¼ 12 for c ¼ 1. m1 ¼ 1k1
pffiffi
0\b B 1 1þ c
k1 ¼ 1þc for c \1 or c [ 1
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
b2 cþD21 c2 b4 þD21 cb2 3
ð2bcþC21 Þk21 2
3C21 k21 þ3C21 k21 C21
2.1 k21 ¼ D21 m21 ¼ 1k21
1\b B a C21 ¼ 2gb3 þ3gb2 3b2 gþ1
D21 ¼ g b2 2b þ 1 þ 2b b2 c 1 b2
0\k B x
2.2 k22 ¼ D22 Dþ2xcb
22 2
m22 ¼ ð3xc þ C22 Þk22 2C22 k22 þ C22
2 3
3gb2 þ3b2 x3 cþg1
1\b B a D22 ¼ g b 2b þ 1 þ 2b þ x2 c 1 C22 ¼ 2gb b2
x \ k B kcu pffiffiffiffiffiffiffiffiffiffi
D31 cb2 D31 3
ðC31 2bcÞk31 2
3C31 k31 þ3C31 k31 C31
3.1 k31 ¼ D31 b2 c
m31 ¼ k31 1
2 3ðlb2 la2 ga2 þa2 Þþ2ga3 þg1
a \ b B btu D31 ¼ gða 2a þ 1Þ C31 ¼ b2
0\k B x þ2lðb aÞ þ 2a 1
3.2 k32 ¼ D32 Dþ2xcb
32
m32 ¼ ðC32 þ 3xcÞk32 2
2C32 k32 þ C32
2 2 3ðlb2 la2 ga2 þa2 Þþ2ga3 x3 cþg1
a \ b B btu D32 ¼ x c þ ga C32 ¼ b 2
x \ k B kcu þ2ðlb ga la þ aÞ þ g 1
[2]. An example of design procedure using simplified (b) cross sectional elastic-softening/hardening analysis, and
solutions is illustrated in this paper. (c) yield line analysis.
123
Int J Adv Eng Sci Appl Math
123
Int J Adv Eng Sci Appl Math
3.3 Parameterized moment–curvature relationship The mid-span deflection is obtained directly using the
double integration of curvature distribution while the
Parametrized bilinear moment–curvature model as a constant terms are evaluated by imposing boundary
simplification of the closed-form solutions has been conditions and ensuring the continuity of displacement
derived and proposed as a practical design tool [18]. fields [16]. Full range deflection distributions along the
Figure 4 presents the idealized parametric moment–cur- beams were derived for various configurations including
vature response as a bilinear function that includes an simply support and cantilever beams and loading types
elastic and a post-cracking stages, which are referred to as such as point load/moment, distributed load etc. In the
Stages 1 and 2. The bilinear model is defined by two present study, the analytical expressions of the mid-span
control points (ucr, Mcr) and (up, Mp). The linear elastic deflections for 3 PB and 4 PB are presented for elastic
range is characterized by the slope of EIg and extends up (Stage 1) and cracked (Stage 2) stages:
to the first flexural cracking (ucr, Mcr), as defined in Eq. 2. j
d1 ¼ u L2
The post-crack region has a reduced stiffness EIcr and 12 cr
3PB : u L ðj 1ÞL ð2Le LÞj L2e ðL 2jLe Þ
2
extends to the ultimate flexural capacity (up, Mp), beyond d2 ¼ cr þ þ
which it is modeled using the limit state plastic hinge L 2Le 24 8 6L2
formation. ð5Þ
One can terminate the response at specified maximum, 2
a L2
curvature, tensile or compressive strain or any limit state d1 ¼ jucr
criterion. The fully plastic stage is defined by a constant 6 8
123
Int J Adv Eng Sci Appl Math
energy dissipation within the yield lines, both of which are considered. It is further assumed that the curvature–de-
computed independently. The work done by the internal flection relation is independent of end conditions of the
moment term causes the localization and rotation of the rigid slab and the fixity of the supports only increases the overall
sections about the yield lines. Computations are simplified length of yield lines thus the required load to cause the
by making assumptions regarding constant magnitudes of collapse mechanism.
resisting moment and rotation along the hinge. Rotations at the hinges and the supports as a function of
The slab is divided into rigid sections that rotate about imposed deformations were calculated using kinematics and
their respective axes along the support lines. Energy dissi- extended to obtain curvature–deflection relationship by
pation is quantified by projecting all the yield lines around a assuming a linear rotation–curvature relationship using a
region onto, and at right angles to, that region’s axis of constant localization hinge width L*, which defines a nomi-
rotation. These projected lengths are multiplied by the nal curvature across the localization width. Parametric
moment and the angle of rotation of each length segment. analysis of the effect of hinge width have been addressed by
Given the small angles considered, the angle of rotation is Bakhshi et al. [20]. Case studies are addressed to obtain the
equated to the tangent of the angle produced by the deflection load–deflection of a slab using three complementary proce-
of the region. By assigning the maximum deflection a value dures of (a) load-internal moment, (b) curvature–deflection
of unity, the vertical displacement of any point is thereby kinematics, and (c) moment–curvature relationship.
defined. The work done by the external loads is evaluated by
considering the center of gravity of the load applied in each 4.2 Square slabs
region and multiplying it by the distance it travels.
Two most common methods of approach based on limit Square slab of edge length ‘L’ subjected to a distributed load
analysis using upper and lower bound theories of limit ‘q’ is considered. It is assumed that yield lines are at 45 to
analysis are presented. As seen in the simple case of the the sides and each point on the yield line is consistent in
statically determinate beam, results are identical since they carrying a fixed moment magnitude. Hogging and sagging
match the exact solution. moments about the yield lines and the clamped supports
respectively are also assumed to be equal in magnitude.
4.1 Virtual work method (upper bound approach) The yield pattern is used to define the potential collapse
mechanism. If the panel has fixed boundaries, then the
Using the principal of virtual work, the work by the yielding along the edge is included in the calculations as
external forces on the virtual displacement is equal to the well. Energy equilibrium is ensured by equating the work
work done by the internal stresses on the virtual strains for done by loads and the internal energy dissipated by rota-
a kinematically admissible virtual displacement, if the tions about yield line which is expressed as:
system is in equilibrium. The virtual work expression can X X
L2 dmax dmax
be written as a work balance equation using the virtual and Fd ¼ m0 l h ) 4 q ¼4 mL
4 3 0:5L
real parameters for the external and internal work mea-
sures. Owing to its ease of use and applicability to various ð7Þ
situations, the yield line method is highly effective in the In left hand side of Eq. 7, q is the uniformly distributed
design of 2-D plane members subjected to flexure and can load and L2/4 is the area of each wedge (for an equivalent
provide a good estimate of load bearing capacity after the point load of qL2/4) and dmax/3 is the deflection of the
peak load. The approach proposed by Johansen [13] is centroid of each triangle. On the right-hand side, as the
implemented with the objective of fitting the results from rotations are projected onto the sides, the rotation angle h is
full scale tests. calculated from geometry of Fig. 5b as dmax/0.5L. Solving
The moment–load-geometry relationship for basic con- for moment per unit length m0, one obtains:
figurations of square, rectangular, and circular slabs with
qL2
different support conditions are used. Only the concen- m0 ¼ ð8Þ
trated and uniformly distributed loads are considered in the 24
present case. It is assumed that moment about each point For clamped slabs, the yield line moment will be higher
on yield line is consistent and sagging moment is equal to due to rotation of the corner supports. All the four wedges
hogging moment. Yield line formation in square and have projection of their positive (sagging) yield line of a
rectangular slabs is assumed to be at 45 to the edges. value m0 surrounding its axis of rotation with length L.
However, other angles and patterns can also be addressed Sides also have continuous supports (clamped) and
using a minimization algorithm. General cases for yield negative (hogging) yield line causing moment of value
lines not at 45 are also calculated and reported in detail m0’, forms along the support. External work due to energy
[16, 19]. Both simple support and clamped edges are expended is independent of the support conditions
123
Int J Adv Eng Sci Appl Math
therefore it is the same as the expression obtained for For plane #1 (DAO), the normal !
n1 is the cross product
! !
simply supported slab: of vectors DA and AO and in plane #3 (ABO), normal ! n2 is
! !
L2 dmax dmax dmax obtained from vectors AB and BO . The rotation h and u
0
4 q ¼ 4 m0 L þ m0 L ð9Þ are obtained as:
4 3 0:5L 0:5L
( )
! ! L2
By assuming the positive and negative moments (equal 1 1 n1 n1 1 1
u ¼ cos ! ¼ cos ð14Þ
sagging and hogging moment) are the same, i.e. m0 = m0’, L n1 !
n2 L 4d2 þ L2
one would obtain:
Deflection–curvature relationship is therefore obtained
qL2
m0 ¼ ð10Þ by solving Eq. 14:
48
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Similarly, the load and moment relationship can be L cos uL ð1 cos uL Þ
d¼ ð15Þ
obtained for the case of concentarted load applied at the 2 cos uL
center:
4.3 Rectangular slabs
P dmax
Simply supported : 4 dmax ¼ 4 m0 L ; P ¼ 8m0
4 0:5L
The uniformly loaded slab is divided into rigid regions
P dmax dmax
Clamped : 4 dmax ¼ 4 m0 L þ m00 L ; P ¼ 4m0 that rotate about their respective axes of rotation along
4 0:5L 0:5L
the support lines, as shown in Fig. 7. The work of
ð11Þ external loads is evaluated by multiplying the resultant
Figure 6 shows a square slab with length of L and yield load in each region by the deformation of its center of
lines at 45 to the sides, where the maximum deflection, d gravity. Two groups of triangles and trapezoidal sections
occurs at the center. By defining parameter u as the mean are considered. Following the similar procedure, the
curvature in the yield hings and L* as the hinge width, uL* force–moment and curvature–deflection relationships are
represents the hinge rotation. obtained [18]:
The angle of rotation, h between two adjacent surfaces is qb2 ð3a bÞ
computed using the unit vectors. The curvature u and angle m0 ¼ ð16Þ
36ða þ bÞ
between the planes h can then be related as: sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
uL ¼ h ð12Þ a 1 cos 2uL
d¼ ð17Þ
! ! 2 1 þ cos 2uL
DA ¼ ðL; 0; 0Þ AO ¼ ð0:5L; 0:5L; dÞ
! ! 4.4 Round panels
AB ¼ ð0; L; 0Þ BO ¼ ð0:5L; 0:5L; dÞ
! ! ð13Þ
!n 1 ¼ DA AO ¼ ðLdÞ^j þ 0:5L2 k^ A round panel of radius R with a point load acting at center
! ! ! is considered and the yield lines are assumed in the form a
n ¼ AB BO ¼ ðLdÞ^i þ 0:5L2 k^
2
fan shape. Hogging moment along the yield lines and
123
Int J Adv Eng Sci Appl Math
Fig. 6 Rotation–deflection
relationship for a square slab
dα
R
δ
Fig. 7 Rotation–deflection relationship for a simply supported rect-
angular slab [18, 19]
θ
Section A-A
sagging moment along the supports are also assumed to be 2R
equal in magnitude although finite element solutions have
shown a range of moment distributed along the length [2].
Fig. 8 Principal of virtual work to determine the ultimate load
Two end conditions are considered which consist of simply carrying capacity of a round panel test simply supported in its contour
supported and clamped support. and subjected to center point load [18, 19]
Flexural capacity of a simply supported round slab
subjected to a center-point loading is shown in Fig. 8.
Slabs subjected to a large concentrated load may failed as observed in experiments of SFRC with relatively high fiber
‘‘fan cracking pattern’’ with positive yield lines radiating dosages [22]. By assuming rigid body rotation, yield line
from the load and a negative circular yield line centered approach may underestimate the strain energy dissipated
under the point load [15, 21]. On the other hand, the fan- due to pronounced plastic deformations. A model param-
mechanism formed by fine distributed cracks is also eter ‘‘n’’ as the number of yield lines/fan cracks is
123
Int J Adv Eng Sci Appl Math
introduced to account for the changes in internal energy Other cases studies involving square, rectangular, and
dissipation. The round panel with three supports as speci- circular slabs with simple, clamped and mix mode supports
fied by ASTM C1550 were dervied elsewehere where have been dervied for wider design applications [18, 19].
n was asumed to be 3 [18, 19].
The applied load and yield line moment is related as 4.5 Use of nonlinear moment–curvature
[21]: relationship
P
Simply supported : m0 ¼ ð18Þ The algorithm to calculate load–deflection response is
2p
based on merging curvature–deflection relationship with
P
Clamped : m0 ¼ ð19Þ applied load-yield line moment into the moment–curvature
4p
relationship. As shown in Fig. 10 the algorithm is defined
A round slab with radius R and yield lines originating at for a step wise approach in the following manner:
center and extending to the boundary are shown in Fig. 9.
(1) Generate M–u using Table 2 with cross sectional and
With the maximum deflection at center, the angle of
material data. The intersections of discrete interacting
rotation between two adjacent cracked segments OAB and
curves determine the transfer from one mode to
OBC are analyzed for the curvature–deflection
another.
relationship.
(2) The slab is segmented into finite sections and for a
By assuming rigid crack segments and rotations repre-
! ! ! given curvature both the moment and deflection are
sented by vectors OC , OD and OE as shown in Fig. 9. The
calculated using the curvature–deflection relationship
vectors normal to crack segment OAB and OBC are found
as derived in Sects. 4.2 to 4.4. The load is calculated
by cross product of two adjacent normal
! ! from moment using yield line Eqs.
planes. OA ¼ ðR cosðan Þ; R sinðan Þ; dÞ, OB ¼ ðR; 0; dÞ,
! (3) The load and deflection response at each curvature
OC ¼ ðR cosðan Þ; R sinðan Þ; dÞ, an ¼ 2p
n
step.
! ! !
n1 ¼ OA OB
¼ Rd sinðan Þ; Rdð1 cosðan ÞÞ; R2 sinðan Þ ð20Þ
! ! ! 5 Model simulation
n2 ¼ OB OC
¼ Rd sinðan Þ; Rdð1 cosðan ÞÞ; R2 sinðan Þ ð21Þ
5.1 Modelling of TRC in 1-D flexure
The angle of curvature u between segments OAB and
OBC is the angle between these two normal vectors, In order to demonstrate the analytical methods for 1-D
determined by the dot product. problem, flexural results of laminated Textile Reinforced
( ) Concrete (TRC) beams studied by Mobasher et al. [4] are
! !
1 1 n1 n2 simulated. Flexural TRC beam samples with a rectangular
u ¼ cos !
L n1 !
n2 cross section of 30 9 9 mm2 and a span of 220 mm were
1 cosðan ÞR2 þ 2 cosðan Þd2 þ R2 tested under three-point bending (3 PB). The experimental
¼ cos1 ð22Þ
L cosðan ÞR2 þ R2 þ 2d2 program covered various textiles including carbon (C),
123
Int J Adv Eng Sci Appl Math
aramid (A), glass (G), polypropylene (PP), as well as E = 20 MPa result in cracking moment and curvature of
hybrid reinforcements. Results of a hybrid (25G:75PP) Mcr = 810 N-mm and ucr = 22.2(10-6) m-1. The
with 25:75 composition of glass and polypropylene textile dimensionless moment–curvature in Fig. 11a is applied to
yarn ratios are used. Load–deflection responses of five construct the model for the deflection equations. The
replicate 3 PB specimens per batch were used. applied load is obtained from static equilibrium of a 3 PB
Bilinear moment–curvature model is used to simulate beam while the mid-span deflections are obtained from
the results and the normalized model is constructed using curvature used in Eq. 5. The simulated load–deflection
the two values of mp = 3.8 and jp = 37.5, see Fig. 11a. A response is compared with experimental results in Fig. 11b.
cracking strength of rcr = 2.0 MPa, and modulus of Due to distributed cracking, TRC samples exhibit pro-
nounced deflection hardening behavior. The analytical
model agrees well with the experimental data up to max-
imum load.
(a) (b)
Fig. 11 a Normalized moment–curvature model, and b comparison between model simulation and experimental results of TRC beam specimens
123
Int J Adv Eng Sci Appl Math
(a) (b)
Fig. 12 a Tension models, and b comparison of experimental results with the simulations for SFRC square slabs with volume fractions of 1.0
and 1.5%
The simulated parameters used in the model show that obtained for the polynomial curve fit approach. By sub-
as the fiber volume fraction vf increases from 1.0 to 1.5%, stituting the model parameters into the equations for the
both maximum flexural loads and residual loads increase. three stages in Table 2, normalized moment (m) and nor-
The improvement in flexural properties are addressed by malized curvature (j) can be obtained:
increasing tensile strength rcr and residual strength lrcr,
which have been shown to positively affect the moment Stage 1 : m ¼ j
1:882
capacity in parametric studies [8]. On the other hand, the Stage 2:1 : m ¼ 2:805 pffiffiffi ðascendingÞ
deflection and crack opening observed in flexural slabs are j
generally higher than those obtained in beam tests. This m ¼ ð104 Þj2 7:966ð103 Þj þ 2:45 ðdescendingÞ
3949
may be attributed to the failure mechanism pattern in slab Stage 3:1 : m ¼ 0:592 þ 2
j
specimens with larger number of major cracks (3 or 4)
ð23Þ
compared to the beam specimens. Moreover, if the fibers
are well distributed both in beam and slab specimens, the Step 2 The equations obtained in Step 1 represent the
contribution of fibers that are oriented parallel to the numerical values of moment–curvature by factoring in
cracked plane is negligible. However, in the slab specimens the cracking moment and curvature and cross section
subjected to biaxial tensile stress and cracking in multiple sizes:
directions which results in extra load required to impose
rcr bh2
the kinematic deformation, more fibers are involved in M ¼ mMcr ¼ ½m
6 ð24Þ
toughening and bridging processes [22]. Therefore, to 2ecr
accommodate the large crack opening and deflection of the u ¼ jucr ¼ ½j
d
slabs, large tensile strain capacity up to 6% in post-
cracking stages are used, see Fig. 12a. For clear span of 680 mm, thickness 80 mm, cracking
Alternatively, direct use of the nonlinear curve fitting to stress as 1.5 MPa, moment–curvature relationship of Stage
the entire moment curvature relationship in the calculations 2.1 can be shown as:
!
(Eqs. 17 and 21) is also demonstrated, which may be more
1:882 1:5 1 802
applicable in the practical design. The design procedure is M ¼ 2:806 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u=ð1:5 106 Þ 6
illustrated in the following steps:
Step 1 Determine the full range moment–curvature 3:570
¼ 4489:6 pffiffiffiffi N mm ðascendingÞ ð25Þ
relationship using the closed-form equations summarized u
in Table 2. Figure 13 shows the moment–curvature
M ¼ 7:111ð107 Þu2 8497:1u þ 3:920 ð103 Þ
responses generated using the same model parameters
N mm ðdescendingÞ
shown in Fig. 12a. The following simplified relations are
123
Int J Adv Eng Sci Appl Math
Similar expression can also be calculated for Stage 3.1: and given by Eq. 14. While the load and deflection
responses are directly related using elastic solutions of
1:422ð105 Þ
M ¼ 946:72 þ N mm ð26Þ plate theory based on Kirchoff’s assumptions and
u2
derivations by Lagrange in 1811. For a homogeneous and
Step 3 The boundary conditions are simply supported on an isotropic plate, the bending stiffness, D will be given as
all sides. From Sect. 4.2, the moment–load relationship for EI/(1-t2) and the deflection is computed as [18]:
a point load for (Case 1.1) is as follows: 0:00406PL2 Et3
Elastic stage : d ¼ ; D¼
P ¼ 8m0 ð27Þ D 12ð1 t2 Þ
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð28Þ
Step 4 The curvature–deflection relationship for a square L cos uL ð1 cos uL Þ
Cracked stage : d ¼
slab in cracked stages (2.1 and 3.1) are derived in Sect. 4.2 2 cos uL
Step 5 Calculate the load–deflection result using the
results from Steps 3 and 4. The comparison between
experimental results and simulations of different stages are
shown in Fig. 14. It is noted that the curve fitting approach
is not limited to the present case. Mobasher [2] obtained
the simplified relations of moment–curvature, moment-
natural axis depth, as well as moment-tensile strain for
varying a and g values that may apply to common SHCC
materials. The curve fitting approach can be extended to
even wider range of material properties in terms of design
charts and tables for flexural design of SFRC, HPFRC,
TRC, and UHPC.
(a) (b)
Fig. 14 Model simulations using nonlinear curve fitting equations for square slabs with different fiber volume fractions a vf = 1.0%, and
b vf = 1.5%
123
Int J Adv Eng Sci Appl Math
thickness. The specimens were supported by three supports 1.0% and 1.25%. Figure 16b compares the experimental
at 120 and subjected to the point load at the center. The and predicted load–deflection curves. Similar to the pre-
responses of load–deflection for each specimen were sim- vious case study, as the fiber dosage increases, tensile
ulated by the proposed model. The hinge length was strength increases from 2.5 to 3.0 MPa, and the normalized
assumed as 65 mm for all samples. Figure 15 shows the residual strength increases from 0.2 to 0.25.
effect of number of cracks n, on the simulation results. It is
clearly seen that as the number of cracks increases, the
deflection after peak increases for the same level of load, 6 Conclusions
indicating higher energy dissipated during the test. The
simulated load–deflection curve fits well with the experi- Serviceability limits based analytical procedures for anal-
mental responses when n = 24. Figure 16a illustrates the ysis and design of SSCC and SHCC were presented in this
tension models used for different fiber volume fractions of paper. Several different approaches were proposed to
address the flexural behaviour of 1-D (beam) and 2-D (slab
and panel) structural members. The methods cover typical
softening and hardening cement composites such as ECC,
FRC, SFRC, TRC, and UHPFRC. Linearized compression
and tension materials models were used to represent the
constitutive behaviour of SSCC and SHCC materials. The
softening/hardening branch was addressed by the post-
cracking stiffness parameter that is either negative (soft-
ening) or positive (hardening). Closed-form solutions for
moment–curvature responses were subsequently derived
through cross sectional analysis that connects the load and
deflection responses.
In order to obtain closed-form solutions to the load–
deflection, various strategies were presented including
moment-area method, parameterized moment–curvature
relationship and principle of virtual work or Yield Line
Theory. In addition, nonlinear curve fitting of moment–
curvature responses in different stages were obtained for
Fig. 15 Effects of n in post cracking load–deflection behavior varying hardening/softening parameters. The fitted
(a) (b)
Fig. 16 a Tension models, and b comparison of experimental results with the simulations for SFRC round panels with volume fractions of 1.0
and 1.25%
123
Int J Adv Eng Sci Appl Math
equations can be extended to generate a design chart and 11. Mobasher, B., Yao, Y., Soranakom, C.: Analytical solutions for
table for practical design purposes. The proposed models flexural design of hybrid steel fiber reinforced concrete beams.
Eng. Struct. 100, 164–177 (2015)
were verified by several experimental studies including 12. ACI Committee 318, American Concrete Institute: building code
TRC beams, SFRC square slabs and round panels from requirements for structural concrete (ACI 318-14): an ACI stan-
literature. The tension softening/hardening and multiple dard: commentary on building code requirements for structural
cracking behavior were addressed by the models to account concrete (ACI 318R-14), an ACI report. (2014)
13. Johansen, K.W.: Yield-Line Theory. Cement and Concrete
for the deflection hardening mechanisms and energy dis- Association, London (1962)
sipation in flexural members under 1-D and 2-D loading 14. Johansen, K.W.: Yield-Line Formulae for Slabs. Cement and
conditions. Concrete Association, London (1972)
15. Kennedy, G., Goodchild, C.: Practical Yield Line Design. The
Concrete Centre, Riverside house (2004)
16. Wang, X., Subramaniam: Analytical load-deflection equations for
beam and 2-D panel with a bilinear moment–curvature model. In:
References ASU Electronic Dissertations and Theses. Arizona State
University (2015)
1. Destrée, X., Yao, Y., Mobasher, B.: Sequential cracking and their 17. Yao, Y., Aswani, K., Wang, X., Mobasher, B.: Analytical dis-
openings in steel-fiber-reinforced joint-free concrete slabs. placement solutions for statically determinate beams based on a
J. Mater. Civ. Eng. 28(4), 04015158 (2015). doi:10.1061/ trilinear moment–curvature model. J. Struct. Eng. (Under review)
(ASCE)MT.1943-5533.0001377 18. Wang, X.: Analytical Load-deflection equations for beam and
2. Mobasher, B.: Mechanics of Fiber and Textile Reinforced 2-D panel with a bilinear moment–curvature model.
Cement Composites. CRC Press, Boca Raton (2011) http://gradworks.umi.com/15/95/1595325.html, (2015)
3. Yao, Y., Silva, F.A., Butler, M., Mechtcherine, V., Mobasher, B.: 19. Aswani, K.: Design procedures for strain hardening cement
Tension stiffening in textile-reinforced concrete under high speed composites (SHCC) and measurement of their shear properties by
tensile loads. Cem. Concr. Compos. 64, 49–61 (2015) mechanical and 2-D digital image correlation (DIC) method.
4. Mobasher, B., Dey, V., Cohen, Z., Peled, A.: Correlation of http://adsabs.harvard.edu/abs/2014PhDT……..23A, (2014)
constitutive response of hybrid textile reinforced concrete from 20. Bakhshi, M., Barsby, C., Mobasher, B.: Comparative evaluation
tensile and flexural tests. Cem. Concr. Compos. 53, 148–161 of early age toughness parameters in fiber reinforced concrete.
(2014) Mater. Struct. 47, 853–872 (2014)
5. Mobasher, B., Peled, A., Pahilajani, J.: Distributed cracking and 21. Darwin, D., Dolan, C., Nilson, A.: Design of concrete structures.
stiffness degradation in fabric-cement composites. Mater. Struct. McGraw-Hill Education, New York (2015)
39, 317–331 (2006) 22. Kim, J., Kim, D.J., Park, S.H., Zi, G.: Investigating the flexural
6. Mobasher, B., Pahilajani, J., Peled, A.: Analytical simulation of resistance of fiber reinforced cementitious composites under
tensile response of fabric reinforced cement based composites. biaxial condition. Compos. Struct. 122, 198–208 (2015)
Cem. Concr. Compos. 28, 77–89 (2006) 23. Khaloo, A.R., Afshari, M.: Flexural behaviour of small steel fibre
7. Soranakom, C., Mobasher, B.: Flexural analysis and design of reinforced concrete slabs. Cem. Concr. Compos. 27, 141–149
textile reinforced concrete. In: 4th Colloquium on Textile Rein- (2005)
forced Structures (CTRS4), Dresden, Germany June 3–5 (2009) 24. Sawyer Jr., H.A.: Design of concrete frames for two failure
8. Soranakom, C., Mobasher, B.: Correlation of tensile and flexural stages. Spec. Publ. 12, 405–437 (1965)
responses of strain softening and strain hardening cement com- 25. de Montaignac, R., Massicotte, B., Charron, J.-P., Nour, A.:
posites. Cem. Concr. Compos. 30, 465–477 (2008) Design of SFRC structural elements: post-cracking tensile
9. Soranakom, C., Mobasher, B.: Closed-Form moment–curvature strength measurement. Mater. Struct. 45, 609–622 (2012)
expressions for homogenized fiber-reinforced concrete. Mater. J.
104, 351–359 (2007)
10. Soranakom, C., Mobasher, B.: Closed-form solutions for flexural
response of fiber-reinforced concrete beams. J. Eng. Mech. 133,
933–941 (2007)
123