Professional Documents
Culture Documents
h i g h l i g h t s
Effect of curing time on the triaxial mechanical behavior of concrete was studied.
Failure strength, plastic deformation is highly dependent on the confining pressure.
Present curing time dependent model can describe the main behavior of the concrete.
a r t i c l e i n f o a b s t r a c t
Article history: Due to its hydration process and applications under complex loadings, concrete is a live material and
Received 9 December 2016 might be subjected to triaxial loadings in short time after pouring. The aim of this study is to investigate
Received in revised form 23 April 2017 the main mechanical behavior of normal concrete at different curing time under different confining stress
Accepted 24 April 2017
state. A series of triaxial compression tests of a concrete are conducted firstly. Four curing times of 3, 7, 14
Available online 3 May 2017
and 28 days, as well as four confining pressures of 0, 5, 10 and 20 MPa, are considered at a C30 concrete
type. The interesting observation is that the failure strength and plastic deformation of concrete is not
Keywords:
only related to curing time, but also highly dependent on the confining pressure. With the increasing
Concrete
Curing time
of confinement, curing time dependence on deformation and failure strength is increasing and decreasing
Triaxial compression respectively. Based on experimental investigations, a curing time dependent constitutive model is pro-
Mechanical behavior posed. In this model, the effect of curing time is introduced to elastic characteristic, failure function
Constitutive law and plastic hardening law. Finally, calibration of the proposed model shows good agreement with the test
results.
2017 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.conbuildmat.2017.04.180
0950-0618/ 2017 Elsevier Ltd. All rights reserved.
D. Chen et al. / Construction and Building Materials 147 (2017) 488496 489
Based on the relevant experimental data, continuous researches mixture was put into the U 50 mm 100 mm cylindrical molds
have been concentrated on developing the constitutive models. by three times, and compacted by a vibrating table for 23 min at
The constitutive law is a general term and it may include yield a frequency of 50 Hz and vibration amplitude of 0.5 mm. After com-
criteria, failure criteria, hardening law and flow rule. After the pacting, the samples were cured at 20 C for 24 h. Then the demold-
well-known Mohr-Coulomb theory, many modelings have been ing was done and the specimens were cured in water saturated
established and modified by introducing the principal shear with lime at 20 C until the day of test in accordance with GB/
stresses, the principal normal stress and stress invariants. Tresca T50081-2002 [34]. Three cylindrical samples were used for each
criterion [11], von Mises criterion [12], Drucker-Prager criterion compression test, and there were totally 84 cylinders were tested
[13], Ottosen criterion [14], Willam-Warnke criterion [15], and in this study. Meanwhile, 12 100 mm 100 mm 100 mm cubic
their modified formulations, as well as others work [1618], have samples were made in the same conditions. Three cubic samples
been proposed to develop the modeling the mechanical behavior of were used in each splitting tensile test.
concrete. Meanwhile, some damage criteria established by Mazars The compression tests and splitting tensile tests were per-
[19] and Kitsutaka [20] can be introduced to describe its softening formed at 3, 7, 14 and 28 days, respectively. After being cured in
behavior. More detailed contents can be found in relevant review saturated lime water for 3, 7, 14 and 28 days, the concrete samples
articles and books [21,22]. As classical quasi-brittle material, e.g., were taken out from lime water, the surface of each specimen was
rock and concrete, the mechanical behavior of concrete is similar thoroughly washed and was wiped with soft cloth. The triaxial
to that of rock [23]. The further modeling research of concrete is compressive tests of cylindrical samples were performed in the tri-
analogue to that of rock [24,25]. axial cell V5 system (Fig. 1) of TOP INDUSTRY, France. Meanwhile,
As well-known, concrete is a live material and its strength the splitting tests with cubic samples were conducted in a 3000 kN
development is a result of the hydration process or curing time electro-hydraulic servo universal test machine.
[26]. Therefore, the concrete at early age might show time depen- In the present triaxial system, the maximum of confining
dent elastoplastic behavior with respect to different curing time. pressure and deviator stress can be up to 40 MPa and 150 MPa
However, it is found that the applied constitutive laws are gener- respectively. The axial displacement is measured by two LVDTs
ally established on uniaxial compression tests while the plastic in the cell, and the lateral deformation is measured with a special
evolution at early age under triaxial stress state is always neglected strain ring located at the middle of the sample. In our study, the tri-
[27]. Meanwhile, a series of researches on elastic characteristic, axial test was performed by two loading stages shown in Fig. 2. The
uniaxial compressive strength, flexural strength, as well as shrink- axial and confining pressures were simultaneously applied to tar-
age of concrete at early age have been conducted [2831]. We can geted values r3 in load control at first. Then the confining pressure
see clearly that there are few studies investigating the mechanical was kept constant at the targeted level, the control load mode was
behavior of early age concrete under multiaxial loading. Consider- transformed to axial displacement control with rate of 1 lm/s and
ing the above mentioned applications related to both early age and the deviatoric stress (r1r3) increased until the specimen reached
complex loading, it is necessary to carry out the investigation of fractured. In this study, four series of triaxial compression tests
concrete at different curing time under triaxial loading. were carried out at 3, 7, 14 and 28 days. In each series, four triaxial
In this paper, a C30 normal concrete is investigated systemati- compression tests with the confining pressures of 0, 5, 10 and
cally. A series of triaxial compression tests are performed firstly. 20 MPa were performed.
Four curing times of 3, 7, 14 and 28 days, as well as four confining The tensile strength was obtained by the splitting test and
pressures (Pc) of 0, 5, 10 and 20 MPa, are considered. Then a curing would be used as a reference for hydrostatic tension strength. In
time dependent constitutive law is proposed. Based on the exper- the following section, hydrostatic tension strength will be intro-
imental data and relevant calculation method, the models param- duced to establish the modeling. It is generally accepted that it is
eters are determined and the numerical simulation is calibrated hard to measure the hydrostatic tension strength. Meanwhile,
finally. hydrostatic tension strength might be some smaller than the ten-
sile strength. Thus, the value of the tensile strength can be used
2. Experimental research to limit the range of hydrostatic tension strength. In this study,
four splitting tests were performed at 3, 7, 14 and 28 days.
2.1. Experimental preparation
2.2. Experimental results and analysis
The Ordinary Portland cement (P.O. 42.5) produced by Anhui
Conch Cement Company, China was in use and its strength charac- Triaxial compression tests were carried out respectively to char-
teristics was confirmed to GB/T 17671-1999 [32]. The fine aggre- acterize basic mechanical behavior of concrete at different curing
gates with fineness modulus of 2.5 were used in this study. time. The stress-strain curves obtained from these triaxial com-
According to the dimension of the samples, the crushed limestones pression tests are shown in Fig. 3(ad).
with a maximum particle size of 10 mm were selected as the coarse The elastic characteristic is discussed firstly and Table 2 shows
aggregates. Mixture proportions were designed to achieve a mean the elastic modulus of the concrete at different curing time. The
value compressive strength of 30 MPa at 28 days. Table 1 shows elastic modulus is obtained at the initial linear elastic stage, and
the mix proportions of the mixture by weight. Furthermore, the defined as the slope of stress versus axial strain. The elastic modu-
slump test of fresh concrete was performed according to GB/T lus of the concrete enhances gradually as the curing time increases
50080-2002 [33]. The slump of the concrete was 10 cm, which indi- due to the continuous hydration. Such early-age concrete stiffening
cated that the concrete mixture had good workability. The fresh has also been reported by other authors [30,29,35], and is often
introduced as aging elasticity. In our study, there is no evident
evolution of the Poissons ratio (m) with respect to curing time. In
literature, the evolution of the Poissons ratio is neither not always
Table 1 taken into account. Truman and Oluokun conclude that m remains
Mix proportions of the concrete (kg/m3). constant during hydration [36,37], although some lower values at
Cement Sand Coarse aggregate Water very early age are mentioned [37].
Under low confining pressure, the curing time causes a signifi-
350 736.5 1104.7 210
cant increase of failure strength. The hydration between 3 days and
490 D. Chen et al. / Construction and Building Materials 147 (2017) 488496
Upper clamp
LVDT
Sample in Viton membrane
Ring for radial strain
Low clamp
Fig. 1. The device and the sample fabricated for triaxial compression test.
Pc = 20 MPa
80 80
Pc = 20 MPa
60 60
Pc = 10 MPa Pc = 10 MPa
40 40
Pc = 5 MPa Pc = 5 MPa
20 20
Pc = 0 MPa
3( ) Pc = 0 MPa 1( ) 3( ) 1( )
0 0
-4 -3 -2 -1 0 1 2 3 4 -4 -3 -2 -1 0 1 2 3 4
(a) 3 days (b) 7 days
1-3 (MPa) 1-3 (MPa)
100 100
Pc = 20 MPa Pc = 20 MPa
80 80
60 60
Pc = 10 MPa Pc = 10 MPa
40 40
Pc = 5 MPa Pc = 5 MPa
20 20
Pc = 0 MPa Pc = 0 MPa
3( ) 1( ) 3( ) 1( )
0 0
-4 -3 -2 -1 0 1 2 3 4 -4 -3 -2 -1 0 1 2 3 4
80 20
20
q (MPa) 3( ) 1( )
60 0
15 -4 -3 -2 -1 0 1 2 3 4
40 28 days
10
Fig. 5. Stress-strain curves of concrete at different days with Pc = 20 MPa.
28 days
7 days
14 days 5
7 days
20 3 days
3 days p (MPa)
0
14 days -4 0 4 Table 3
p (MPa) Splitting tensile strength of concrete at different curing time.
0
-10 0 10 20 30 40 50 60 Curing time (d) 3 7 14 28
Splitting tensile strength (MPa) 2.5 3.1 3.7 3.8
Fig. 4. Deviatoric failure surface of concrete at different curing time.
3.1. Framework of model where the dot represents the incremental form or time derivative of
variables.
Under the thermodynamical framework of continuum mechan- The general stress formulation can be expressed as
ics and elastoplastical theory, the assumption of small strain is
r C : ee 2
considered and the total strain e can be composed by an elastic part
ee and a plastic part ep. Thus, the rate form of the strain can be where C is the elastic stiffness tensor of concrete.
described as: The rate form of Eq. (2) can be rewritten as:
e_ e_ e e_ p 1 r_ C : e_ e_ p 3
492 D. Chen et al. / Construction and Building Materials 147 (2017) 488496
By using Hills notation [39], the effective elastic stiffness can be where vp is the regularizing coefficient to take into account the
rewritten in the following general form: effect of confinement. According to the experimental data, it can
be described as a function of minor principal stress:
C 2lK 3kJ 4
hr3 i jr3 j
where l is shear modulus and k is bulk modulus. They can be vp exp b 11
Pr
expressed as:
E E where b is a models parameter. hxi is Heaviside function with
l and k 5 hxi 1; x P 0 or hxi 0; x < 0.
21 t 31 2t
where E and t are the Youngs modulus and Poissons ratio. As 3.3. Characterization of plastic flow rule
above discussed, the Youngs modulus of young concrete is highly
depending on the curing time and it will be discussed in Part 3.4. To describe correctly the transition of volumetric deformation
The two fourth order tensors K and J are defined as: from compressibility to dilatancy, it is necessary to introduce a
plastic flow rule. Inspired by the work of Pietruszczak et al. [40],
1
J d d and K I J 6 the following non-associated flow rule is used directly:
3
where d is the second order unit tensor and is the symmetric pC
g q gp C ln 12
fourth order unit tensor with the component Iijkl 12 dik djl dil djk . I0
where the parameter g defines the transition boundary between
3.2. Characterization of failure function and plastic yield function compressibility and dilatancy. It can be simplified determined by
the stress states at the transition points in triaxial tests. I0 defines
According to our experimental research, it is mentioned that the the intersection point between flow rule and the axis p.
failure condition is highly depending on confining pressure. Thus, When the plastic mechanism is activated, the plastic strain can
the linear criteria, e.g. Mohr-Coulomb or Drucker-Prager criteria, be deduced by plastic flow rule as:
and quadratic function [40] are not suitable for present work.
@g
Inspired by our previous work [41], a high ordered nonlinear fail- e_ p k_ 13
ure function is proposed:
@r
q n where k_ is the plastic multiplier of the plastic mechanism. Consid-
pC
F A 7 ering the plastic consistency condition, the following expression
Pr Pr
can be obtained:
where q and p are deviatoric stress and mean stress with
q @f @f
q 3
K : r : K : r and p 13 trr respectively. The parameters f_ : r_ c_ p 0 14
2 @r @ cp
A and n define the sharp of failure surface. C is hydrostatic tension
strength of material while Pr is a given reference stress with Combining Eqs. (3), (10), (13) and (14), the plastic multiplier k_
Pr = 1 MPa. In fact, it is hard to determine C and its value can be con- can be deduced as:
sidered some less than tensile strength. The values of A, n and C can @f
: C : e_
be determined by back-analysis method with experimental data. k_ @r
p
@g @g
15
2K: :K:
@f
For young concrete, C is effected by the curing time obviously and @r
: C : @@gr @@fc 3 @r
vp
@r
p
its evolution will be discussed in Part.3.4.
According to the failure function Eq. (7), the plastic yield func-
tion can be defined with a plastic hardening law: 3.4. Introduction of curing time
q n
pC
f aA 8 The proposed general modeling contains 8 parameters includ-
Pr Pr ing two elastic parameters E and m, three parameters A, n and C
where a is the plastic hardening law. Neglecting the softening dealing with failure surface, and three plastic parameters B, b
behaviors, it can be generally described as: and g. Based on the following research in Section 4.1, it is found
that E, n, C and B are highly dependent on curing time while the
cp
a 9 other parameters present less dependence. Since concrete is sel-
B cp dom used for loadings within 3 days in construction, the loading
where B and cp are hardening parameter and generalized plastic capacity of concrete at 7, 14 and 28 days is more important for
distortion to control the plastic hardening process. The incremental structure design. The mechanical properties of the concrete at
form of cp can be defined as: 28 days are regarded to be characteristic. The empirical formulas
q based on the material parameters at 3 and 28 days may be more
2
3
K : e_ p : K : e_ p effective in practice. According to their values listed in Table 4,
c_ p 10 the following relations can be proposed as following:
vp
Table 4
Typical values of parameters for different curing time.
1.194 n 4.0
C (MPa)
1.192 3.5
1.190
3.0
1.188
1.186 2.5
1.184 2.0
1.182
1.5
1.180
1.0
1.178
t (d) t (d)
1.176 0.5
0 10 20 30 40 50 60 0 10 20 30 40 50 60
(a) n (b) C
26 9 B
E (GPa)
25 8
24
7
23
6
22
5
21
20 4
t (d) t (d)
19 3
0 10 20 30 40 50 60 0 10 20 30 40 50 60
(c) E (d) B
Fig. 6. Evolution of parameters with curing time t.
q
Et ht 3ibE3 E28 E3 1 h28 ti28 t2 =28 32 c the value of C should be considered with respect to the tensile
q strength. The parameter B controls the evolution of plastic harden-
nt ht 3ibn3 n28 n3 1 h28 ti28 t2 =28 32 c ing. This value can be determined according to Eq. (9) and volu-
q metric plastic deformation with respect to a from experimental
Ct ht 3ibC 3 C 28 C 3 1 h28 ti28 t2 =28 32 c data. Once B is determined, the hardening plastic law can be
q obtained through the determination of b. In order to determine
Bt ht 3ibB3 B28 B3 1 h28 ti28 t2 =28 32 c g, it is necessary to find the point of transition from volumetric
16 strain curve. We collect all transition point in pq plan and the
value of the slope can be used as g. Figs. 712 represent the
where t is the curing time and subscript denotes the corresponding influence of 6 parameters respectively.
time. Fig. 6 shows the curves according to Table 4 and Eq. (16). It is Using the presented method, all the parameters for four curing
clear that the proposed empirical function is suitable for the ten- times have been determined and listed in Table. 4.
dency of these parameters with respect to curing time. According to the relation between parameters and curing time
For elastic modulus, with 2 parameters E3 and E28, the elastic Eq. (16), the present curing time dependent modeling can be finally
modulus of concrete at any curing time can be predicted in Eq. described by E3, E28, t, A, n3, n28, C3, C28, B3, B28, b and g in which the
(16). The prediction of the other three parameters, n, C and B, is subscript denotes the corresponding curing time. With these 12
the same with the parameter E. Totally, 8 parameters are needed parameters listed in Table. 5, the mechanical behavior of this con-
for the evolution of main parameters influenced by the curing time. crete at any curing time under any confining stress can be easily
Also, there are another 4 parameters, m, A, b and g less dependent described by the following calculation method.
on the curing time, but necessary for the whole elastoplastic mod-
eling. Therefore, the present curing time dependent modeling can 4.2. Calibration of the calculation method
be described with 12 parameters. The elastic parameters are E3,
E28 and m. The parameters of failure surface are A, n3, n28, C3 and The proposed model is implemented into the finite element
C28. The plastic parameters are b, g, B3 and B28, respectively. program THMPASA developed at Lille Mechanics Laboratory. This
code has been developed for numerical modeling of mechanical
4. Calibration and numerical simulation evolution problems in engineering media.
In THMPASA program, the global governing equations are
4.1. Calibration of the parameters established in weak form. The loading strain path is divided into
many incremental steps. The core of THMPASA is the update of
The parameters of concrete at different curing time are deduced plastic strain, stress and internal variables at each loading step.
by above experimental tests. The two elastic parameters E and t The iteration process of any step k is summarized as follows:
have been obtained in Part 2. Putting the values of failure strength
of triaxial compression tests with different confining pressures in Step 1. The increment of total strain e_ k is provided as a known
pq plan, the parameters A, n and C are identified easily. Note that loading condition.
494 D. Chen et al. / Construction and Building Materials 147 (2017) 488496
120 60
q (MPa) 1- 3 (MPa)
100 50
B = 8 E-6
80 40
B = 8 E-5
B = 8 E-4
60 30
40 A = 4.5 20
A = 4.0
20 A = 3.5 10
Fig. 7. Influence of parameter A on failure surface. Fig. 10. Influence of parameter B with Pc = 5 MPa.
60
120 1- 3 (MPa)
q (MPa)
50
100
= 0.05
40 = 0.15
80
= 0.25
30
60 n = 1.19
n = 1.18
n = 1.20 20
40
10
20
3 (%) 1 (%)
p (MPa) 0
0 -2 -1 0 1 2
-10 0 10 20 30 40 50 60
Fig. 11. Influence of parameter b with Pc = 5 MPa.
Fig. 8. Influence of parameter n on failure surface.
60
1- 3 (MPa)
120
q (MPa)
50
100 = 0.6
40 = 1.0
80 = 1.6
30
60
20
40 C = 4
10
C=3
20
v (%) v (%)
C=2 0
p (MPa)
0 -2 -1 0 1 2
-10 0 10 20 30 40 50 60
Fig. 12. Influence of parameter g with Pc = 5 MPa.
Table 5
Typical values of parameters for curing time dependent modeling.
1- 3 (MPa) 1- 3 (MPa)
100 100
80 80 Pc = 20 MPa
Pc = 20 MPa
60 60
Pc = 10 MPa Pc = 10 MPa
40 40
Pc = 5 MPa Pc = 5 MPa
20 20 Pc = 0 MPa
Pc = 0 MPa
3 (%) 1 (%) 3 (%) 1 (%)
0 0
-4 -3 -2 -1 0 1 2 3 4 -4 -3 -2 -1 0 1 2 3 4
(a) 3 days (b) 7 days
1- 3 (MPa) 1- 3 (MPa)
100 100
Pc = 20 MPa Pc = 20 MPa
80 80
60 Pc = 10 MPa 60 Pc = 10 MPa
40 Pc = 5 MPa 40 Pc = 5 MPa
Pc = 0 MPa 20 Pc = 0 MPa
20
3 (%) 1 (%) 3 (%) 1 (%)
0 0
-4 -3 -2 -1 0 1 2 3 4 -4 -3 -2 -1 0 1 2 3 4
(c) 14 days (d) 28 days
Fig. 13. Simulations and experimental data of concrete at different curing time.
in Fig.13(a)(d). And the relevant experimental data is drawn as (2) The present curing time dependent modeling with 12
dotted lines at the same time. We can see that the curing time is parameters can describe the triaxial mechanical properties
a much important factor in the early curing time while its influence of the concrete at different curing time in a large range of
decreases with increasing of curing time. Furthermore, it is clear confining pressure. In this model, the effect of curing time
that there is a good agreement between simulations and experi- is introduced to elastic characteristic, failure function and
mental data. The proposed model can describe the main behavior plastic hardening law. Through finite element program
of this concrete under different confining stress state and different THMPASA, numerical simulation of triaxial compression
curing time condition. tests was performed and shows a good agreement with
experimental results.
5. Conclusions
[4] I. Imran, S.J. Pantazopoulou, Experimental study of plain concrete under [24] V. Lefort, G. Pijaudier-Cabot, D. Grgoire, Analysis by Ripleys function of the
triaxial stress, ACI Mater. J. 93 (6) (1996) 589601. correlations involved during failure in quasi-brittle materials: experimental
[5] J.C. Chern, H.J. Yang, H.W. Chen, Behavior of steel fiber reinforced concrete in and numerical investigations at the mesoscale, Eng. Fract. Mech. 147 (2015)
multiaxial loading, ACI Mater. J. 89 (1) (1992) 3240. 449467.
[6] Y. Farnam, M. Moosavi, M. Shekarchi, S.K. Babanajad, A. Bagherzadeh, [25] S.M. Vrech, G. Etse, Gradient and fracture energy-based plasticity theory for
Behaviour of slurry infiltrated fibre concrete (SIFCON) under triaxial quasi-brittle materials like concrete, Comput. Methods Appl. Mech. Eng. 199
compression, Cem. Concr. Res. 40 (11) (2010) 15711581. (1) (2009) 136147.
[7] X.B. Lu, C.T.T. Hsu, Behavior of high strength concrete with and without steel [26] M. AbdElaty, Compressive strength prediction of Portland cement concrete
fiber reinforcement in triaxial compression, Cem. Concr. Res. 36 (9) (2006) with age using a new model, HBRC J. 10 (2) (2014) 145155.
16791685. [27] B.F. Zhu, Thermal Stress and Temperature Control of Mass Concrete,
[8] D.C. Candappa, J.G. Sanjayan, S. Setunge, Complete triaxial stress-strain curves Butterworth-Heinemann, Oxford, 2013.
of high-strength concrete, J. Mater. Civil Eng. 13 (3) (2001) 209215. [28] D. Shen, X. Wang, D. Cheng, J. Zhang, G. Jiang, Effect of internal curing with
[9] H.S. Shang, Y.P. Song, Experimental study of strength and deformation of plain super absorbent polymers on autogenous shrinkage of concrete at early age,
concrete under biaxial compression after freezing and thawing cycles, Cem. Constr. Build. Mater. 106 (2016) 512522.
Concr. Res. 36 (10) (2006) 18571864. [29] O. Bernard, F.J. Ulm, E. Lemarchand, A multiscale micromechanics-hydration
[10] S.Y. Xie, J.F. Shao, N. Burlion, Experimental study of mechanical behaviour of model for the early-age elastic properties of cement-based materials, Cem.
cement paste under compressive stress and chemical degradation, Cem. Concr. Concr. Res. 33 (9) (2003) 12931309.
Res. 38 (12) (2008) 14161423. [30] J.K. Kim, S.H. Han, Y.C. Song, Effect of temperature and aging on the mechanical
[11] H. Tresca, Sur Iecoulement des corps solids soumis a de fortes pression, properties of concrete: part I. Experimental results, Cem. Concr. Res. 32 (7)
Comptes Rendus hebdomadaires des Sance de lAcademic des Scriences 70 (2002) 10871094.
(1870) 473480. [31] D. Shen, J. Jiang, J. Shen, P. Yao, G. Jiang, Influence of curing temperature on
[12] W.H. Yang, A generalized von Mises criterion for yield and fracture, J. Appl. autogenous shrinkage and cracking resistance of high-performance concrete at
Mech. 47 (2) (1980) 297300. an early age, Constr. Build. Mater. 103 (2016) 6776.
[13] D.C. Drucker, Limit analysis of two and three-dimensional soil mechanics [32] Standard for method of testing cements-determination of strength. GB/T
problems, J. Mech. Phys. Solids 1 (1953) 217226. 17671-1999, Beijing, China, 2000.
[14] N.S. Ottosen, A failure criterion for concrete, ASCE Eng. Mech. Div. J. 103 (4) [33] Standard for test method of performance on ordinary fresh concrete. GB/T
(1977) 527535. 500802002, Beijing, China, 2003.
[15] K.J. Willam, E.P. Warnke, Constitutive model for the triaxial behavior of [34] Standard for test method of mechanical properties on ordinary concrete. GB/
concrete, Int. Assoc. Bridge. Struct. Eng. Proc. 19 (1) (1975) 130. T50081-2002, Beijing, China, 2003.
[16] Z.H. Guo, C.Z. Wang, Investigation of strength and failure criterion of concrete [35] G. De Schutter, L. Tearwe, Degree of hydration-based description of
under multi-axial stresses, China Civ. Eng. J. 24 (3) (1991) 114. mechanical properties of early age concrete, Mater. Struct. 29 (7) (1996)
[17] K.H. Gerstle, Simple formulation of triaxial concrete behavior, ACI Mater. J. 78 335344.
(5) (1981) 382387. [36] K.Z. Truman, D.J. Petruska, C.D. Norman, Creep, shrinkage, and thermal effects
[18] P.R. Sheorey, A.K. Biswas, V.D. Choubey, An empirical failure criterion for rocks on mass concrete structure, J. Eng. Mech. 117 (6) (1991) 12741288.
and jointed rock masses, Eng. Geology 26 (2) (1989) 141159. [37] F.A. Oluokun, E.G. Burdette, J.H. Deatherage, Elastic modulus, Poissons ratio, and
[19] J. Mazars, G. Pijaudier-Cabot, Continuum damage theory-application to compressive strength relationships at early ages, Mater. J. 88 (1) (1991) 310.
concrete, J. Eng. Mech. 115 (2) (1989) 345365. [38] F.A. Oluokun, E.G. Burdette, J.H. Deatherage, Splitting tensile strength and
[20] Y. Kitsutaka, Fracture parameters by polylinear tension-softening analysis, J. compressive strength relationships at early ages, Mater. J. 88 (2) (1991) 115
Eng. Mech. 123 (5) (1997) 444450. 121.
[21] W.F. Chen, A.F. Saleeb, Constitutive Equations for Engineering Materials, [39] R. Hill, A general theory of uniqueness and stability in elastic-plastic solids, J.
Plasticity and Modeling, vol. 2, Elsevier, Amsterdam, 1994. Mech. Phys. Solids 6 (3) (1958) 236249.
[22] M.H. Yu, Advances in strength theories for materials under complex stress [40] S. Pietruszczak, J. Jiang, F.A. Mirza, An elastoplastic constitutive model for
state in the 20th Century, Appl. Mech. Rev. 55 (3) (2002) 169218. concrete, Int. J. Solids Struct. 24 (7) (1988) 705722.
[23] I. de-Pouplana, E. Oate, Combination of a non-local damage model for quasi- [41] Y. Zhang, Modelisation Du Comportement Elastoplastique Dune Pate De
brittle materials with a mesh-adaptive finite element technique, Finite Elem. Ciment Soumise a La Degradation Chimique, University of Science and
Anal. Des. 112 (2016) 2639. Technology of Lille, Lille, 2008.