You are on page 1of 11

Simplied Damage Plasticity Model for Concrete

Milad Hafezolghorani, PhD Candidate; Farzad Hejazi, Senior Lecturer; Ramin Vaghei, PhD; Mohd Saleh Bin Jaafar, Prof.;
Keyhan Karimzade, Msc, Department of Civil Engineering, Faculty of Engineering, Universiti Putra Malaysia, Selangor, Malaysia.
Contact: farzad@upm.edu.my
DOI: 10.2749/101686616X1081

Abstract Ananiev and Ozbolt,12 and Imran and


Pantazopoulou13 also considered
The past several years have witnessed an increase in research on the nonlinear models that involved the formulation
analysis of the structures made from reinforced concrete. Several mathematical of plasticity in the nominal stress
models were created to analyze the behavior of concrete and the reinforce- space.
ments. Factors including inelasticity, time dependence, cracking and the inter-
active effects between reinforcement and concrete were considered. The Grassel and Jirasek14 introduced dif-
crushing of the concrete in compression and the cracking of the concrete in ten- ferent combinations of plasticity and
sion are the two common failure modes of concrete. Material models were damage applied to the concrete failure
introduced for analyzing the behavior of unconned concrete, and a possible models. They analyzed the local
constitutive model was the concrete damage plasticity (CDP) model. Due to uniqueness conditions of two combi-
the complexity of the CDP theory, the procedure was simplied and a simpli- nations of strain scalar damage and
ed concrete damage plasticity (SCDP) model was developed in this paper. stress based plasticity types. Further-
more, the triaxial damage plastic
The SCDP model was further characterized in tabular forms to simulate the
model accounted for the failure of
behavior of unconned concrete. The parameters of the concrete damage plas-
concrete. A three dimensional inter-
ticity model, including a damage parameter, strain hardening/softening rules,
face model was presented by Grassl
and certain other elements, were presented through the tables shown in the
and Rempling.15 The combination of
paper for concrete grades B20, B30, B40 and B50. All the aspects were dis-
damage mechanics and the theory of
cussed in relation to the effective application of a nite element method in the
plasticity were used as a basis for this
analysis. Finally, a simply supported prestressed beam was analyzed with
model, and it enabled the researchers
respect to four different concrete grades through the nite element program.
to vary the ratio of the permanent
The results showed that the proposed model had good correlation with prior
and total inelastic displacements.
arts and empirical formulations.
Yu et al.16 introduced a modied plas-
Keywords: concrete damage plasticity; concrete failure; unconned concrete;
tic damage model. The theoretical
nite element analysis; crack; crush.
framework of the CDPM was used as
a basis for this model, and the
conned concrete was modeled by
Introduction hardening/softening make the behav- ABAQUS with the conditions of non-
ior of concrete more complicated. uniform connement. The Lubliner
Finite element method based analysis Hence, it becomes difcult to deter- yield criterion was used for the triaxial
of reinforced concrete structures has mine damage in concrete. Constitutive compression stress states by Zhang
signicantly developed since 1970. models are used for this purpose, and et al.17 The Lubliner criterion was
Researchers have tried to analyze the an example of these models is the reworked, improving its limits, and
behavior of concrete and have pub- concrete damage plasticity (CDP) hence many stress states in engineer-
lished several reports and technical model. This model uses the ow the- ing structures could be accounted for
reports on this subject. However, the ory of plasticity and damage mechan- by this revised version of the Lubliner
behavior of concrete is complex, and ics to analyze the concrete criterion. Taddei et al.18 proposed
many parameters must be considered structures.14 three-dimensional nite element mod-
for its analysis. Concrete is composed els for unreinforced and reinforced
of qualitatively and quantitatively dif- Previous Studies walls panels based on concrete dam-
ferent types of materials. These mate- age plasticity constitutive law. Grassel
rials exhibit different properties in Concrete damage plasticity is widely et al.19 utilized a constitutive model
terms of tension and compression. recognized as a precise and practical for concrete structures subjected to
The structural mechanics of concrete constitutive model to simulate con- multiaxial and rate-dependent loading
structures is quite important, and con- crete behavior. Carol et al.5 analyzed by combining an effective stress based
crete identication parameters includ- different combinations of plasticity plasticity model with an isotropic
ing the non-linear stressstrain and damage. Gatuingt and Pijaudier- damage model based on plastic and
relation of the concrete under impos- Cabot,6 and Kratzig and Polling7 used elastic strain measures.
ing stress conditions and strain isotropic damage in their research,
Larsson et al.20 examined the laterally
and elaborated on several types of
loaded lime-cement columns in a
Peer-reviewed by international ex- plasticity combinations.
shear box and used a damage plastic-
perts and accepted for publication
by SEI Editorial Board Models involving the development of ity model to numerically analyze the
stress related plasticity in the effective columns. The model accounted for
Paper received: April 8, 2015 stress space were introduced by some the stiffness degradation of the col-
Paper accepted: December 17, 2015 researchers.810 Lubliner et al.,11 umns. Ming21 analyzed the basic

68 Scientic Paper Structural Engineering International Nr. 1/2017


concepts and properties of the con- nite element software in order to eas- The behavior of the concrete was
crete damage plasticity model using ily understand the mechanical behavior explained by the assumption that con-
inferential reasoning and analysis. of concrete. The SCDP model was then crete damage plasticity
  utilized the
Computational techniques were used veried against the current concrete yield function, f pl, h , , which repre-
to determine the plastic factors for the damage plasticity theory and empirical sented the yield surface in effective
concrete damage in ABAQUS. formulations. stress space to determine the states of
Zhang and Li22 introduced calibration damage or failure.11
techniques for concrete, under uni- In the concrete damage plasticity
form and non-uniform connement,
Damage Plasticity Constitutive model, the ow rule was dened as
based on Lubliner theory through follows:
three dimensional simulation. Grassl Model
et al.23 used a combination of plastic- _ G
The isotropic damaged elasticity and _ pl = : 8
ity and damage mechanics, to create a
constitutive model that was employed the isotropic tensile and compressive
to study the failure of concrete struc- plasticity were used in the concrete In the concrete damage plasticity
tures. The model was designed to ana- damaged plasticity model to study the model, the ow rule was non-associ-
lyze the properties of the failure behavior of concrete in a non-elastic ated. This
 meant  that the yield func-
process in concrete structures experi- manner. The total strain tensor was tion f pl, h , and the plastic
encing multi-axial loading. The plas- comprised of the elastic part el and potential gp did not coincide, and,
ticity model based on effective stress the plastic part pl. therefore, the direction of the plastic
and the plastic and elastic strain mea- ow G was not normal to the yield
sures based damage models were used = el + pl 1
  surface. The plastic potential was
for this purpose. Vaghei et al.24 pro- = D el : pl 2 dened in effective stress space.
posed the nite element model to   However, in this study, due to the
develop a three dimensional version = Del0 : pl 3 complex degradation mechanism of
addressing precast walls and the the uniaxial cyclic behavior of con-
connection. D el = 1 dDel0 4
crete (opening and closing of formed
Shang et al.25 studied the torque of the The nominal stress with the degraded micro-cracks), the uniaxial response
RC girder. They applied the nite ele- elastic tensor from (4) could be of concrete was investigated.
ment analysis software ABAQUS to rewritten as follows: Figure 1 indicated that the uniaxial
concrete damaged plasticity model as a   compressive and tensile response of
constitutive model for the analysis of = 1 dDel0 : pl 5
concrete was assumed to be inu-
concrete material. The outcomes from enced by damaged plasticity, and this
other experiments were used to vali- The damage plasticity constitutive
model was based on the following assumption formed the basis of the
date the results of the nite element model. The uniaxial compressive and
analysis. Tao et al.26 utilized a single, stressstrain relationship:
tensile responses of concrete with
robust nite element (FE) model to = 1 d: ! = 1 dt t + 1 dc c respect to the concrete damage plas-
study the bond behavior subjected to ticity model subjected to compression
shear. Tiwari et al.27 investigated 6
and tension load were given by:
underground tunnels with curved align-
where dt and dc were two scalar dam-  
ment in the longitudinal direction sub- pl, h
age variables, ranging from 0 (undam- t = 1 dt E0 t t 9
jected to blast loading. He observed
aged) to 1 (fully damaged).23 The 
that stress, deformation and damage
damage model used for concrete was c = 1 dc E0 c pl , h 10
responses of tunnel lining through c
based on plasticity and considered the
three dimensional nite element simu-
failure process of tensile cracking and
lations using concrete damage plasticity Given the nominal uniaxial stress, the
compressive crushing.
theory. effective uniaxial stress t and c were
Isotropic hardening variables were derived as follows:
Extant research indicates that it is quite
expressed by inelastic compression  
complicated to represent reinforced , h and cracking strain ck, h , t
strain in pl, h
concrete behavior using concrete c t t = = E 0 t t 11
which include the plastic hardening 1 dt
damaged plasticity models as a consti-
tutive law. This also may not be fully strain pl,h plus the residual strain due c  , h
c = = E0 c pl 12
understood by prospective researchers. to damages. 1 dc c

In the present study, the procedure of  pl, h 


  where compressive strain c equalled
concrete damage plasticity theory was pl, h = tpl, h ; pl = h pl, h , : _ pl , , h + el , and tensile strain equal-
simplied and characterized in tabu- c pl
c c t
led pl , h + el .
lated forms. The ndings were formu- _ = _ el + _ pl 7 t t
lated and presented in a tabular format
for four common concrete grades, Hardening variables were used to
while concrete parameters could be control the development of the yield Simplied Concrete Damage
extended to other concrete grades. or failure of the surface. These vari- Plasticity
Numerical analysis was then conducted ables were connected to the pro-
to investigate the SCDP model, which cesses of tension and compression The values of the hardening and soft-
was implemented in the ABAQUS loading. ening variables were used for the

Structural Engineering International Nr. 1/2017 Scientic Paper 69


(a) (b) c
c t
cu B

cu t0
(1 dt) E0
c0 0.5cu

(1 dc) E0 0.2cu C
D
E0 E0
"
A 0.002 "0 .5 "0.8
E0 E0
F i g. 2: Kent and Park model for conned
"c "t
in,h ck,h and unconned concrete..29 (Units: [])
"c "t
"
pl,h " el
c "
pl,h "tel
c t
the unconned cylinder specimen,
F ig . 1: Response of concrete to a uniaxial loading condition: (a) Compression, (b) respectively.
Tension.11 (Units: []) Park30 reported that c equaled 0.002,
and this value was also assumed in
determination of the cracking and where c and c were nominal this study. Figure 2 indicated a para-
crushing trends, respectively. They compressive stress and strain, respec- bolic increasing trend (AB) for the
were responsible for the loss of the tively, and cu and c were ultimate hardening stage, while a linear behav-
elastic stiffness and the development of compressive strength and the strain of ior (BC) was observed for the
the yield surface. The damage states in
compression and tension were charac-
Materials Plasticity parameters
terized independently by two harden-
ing variables. These were indicated by parameters B20 Dilation angle 31
cpl, h and tpl, h , which referred to Concrete elasticity Eccentricity 0.1
equivalent plastic strains in tension E (GPa) 21.2 fb0/fc0 1.16
and compression, respectively. 0.2 K 0.67
Viscosity parameter 0
Uniaxial Compressive Behavior
Concrete compressive behavior Concrete compression damage
In concrete damage plasticity models,
Yield stress (MPa) Inelastic strain Damage parameter Inelastic strain
the plastic hardening strain in com- C
pression cpl, h played a key role in
nding the relation between the dam- 10.2 0 0 0
age parameters and the compressive 12.8 7.73585E-05 0 7.73585E-05
strength of concrete (see Fig. 1a) as 15 0.000173585 0 0.000173585
follows:
16.8 0.000288679 0 0.000288679
 
c = 1 dc E0 c cpl, h 13 18.2 0.000422642 0 0.000422642
19.2 0.000575472 0 0.000575472
8 c 19.8 0.00074717 0 0.00074717
> in, h
< c = c E
0 20 0.000937736 0 0.000937736
c 1 14
> ,
: c = c
pl h 19.8 0.00114717 0.01 0.00114717
E0 1 dc
19.2 0.001375472 0.04 0.001375472
dc c 18.2 0.001622642 0.09 0.001622642
cpl, h = cin, h 15
1 dc E0 16.8 0.001888679 0.16 0.001888679
Generally, uniaxial compressive behav- 15 0.002173585 0.25 0.002173585
ior could be characterized by either 12.8 0.002477358 0.36 0.002477358
experimental tests or existing constitu- 10.2 0.0028 0.49 0.0028
tive models, such as those proposed by
Hognestad28 and Kent et al.29 for 7.2 0.003141509 0.64 0.003141509
unconned concrete. However, the 3.8 0.003501887 0.81 0.003501887
present study employed the Kent and Concrete tensile behavior Concrete tension damage
Park parabolic constitutive model for
Yield stress (MPa) Cracking strain Damage parameter Cracking strain
unconned concrete, which was
T
expressed by the following equation:
"   # 2 0 0 0
c c 2 0.02 0.000943396 0.99 0.000943396
c = cu 2 16
c c
Table 1: Material properties for concrete with SCDP model in class B20

70 Scientic Paper Structural Engineering International Nr. 1/2017


Plasticity parameters (modulus of elasticity, E0) due to the
Materials damage when plastic strains increased
parameters B30 Dilation angle 31 in brittle materials (such as concrete
Concrete Elasticity Eccentricity 0.1 and concrete-like material) as shown in
E (GPa) 26.6 fb0/fc0 1.16 Fig. 1a. The damage parameter (dc)
0.2 was 0 at the maximum compressive
K 0.67
stress, and thereafter, it began to
Viscosity parameter 0 decrease and continued decreasing until
Concrete compressive behavior Concrete compression damage 0.8 was reached with respect to 20%
remaining strength in large strains.
Yield stress (MPa) Inelastic strain Damage parameter Inelastic strain
C
15.3 0 0 0 Uniaxial Tensile Behavior
19.2 4.8249E-05 0 4.8249E-05 In concrete damage plasticity models,
22.5 0.000119844 0 0.000119844 the plastic hardening strain in tension
25.2 0.000214786 0 0.000214786 tpl, h was derived (see Fig. 1b as
follows:
27.3 0.000333074 0 0.000333074
 
pl, h
28.8 0.000474708 0 0.000474708 t = 1 dt E0 t t 19
29.7 0.000639689 0 0.000639689 8 ck, h t
30 0.000828016 0 0.000828016
>
< t = t
E0
t 1 20
29.7 0.001039689 0.01 0.001039689 > pl, h
: t = t
28.8 0.001274708 0.04 0.001274708 E0 1 dt
27.3 0.001533074 0.09 0.001533074 pl, h ck, h dt t
t = t 21
25.2 0.001814786 0.16 0.001814786 1 dt E0
22.5 0.002119844 0.25 0.002119844 Although concrete had many constitu-
19.2 0.002448249 0.36 0.002448249 tive models in the tension phase, there
were no signicant differences in their
15.3 0.0028 0.49 0.0028
results due to the brittle behavior of
10.8 0.003175097 0.64 0.003175097 concrete. Engineers seldom dene
5.7 0.003573541 0.81 0.003573541 the tension behavior of concrete in
Concrete tensile behavior Concrete tension damage numerical modeling, mainly because
the interaction between reinforcement
Yield stress (MPa) Cracking strain Damage parameter Cracking strain and concrete is simplied and because
T certain changes in the stressstrain
3 0 0 0 relation of concrete in the tension
0.03 0.001167315 0.99 0.001167315 phase must be made to consider bar
slips in concrete. Among the constitu-
Table 2: Material properties for concrete with SCDP model in class B30 tive models for the tension phase con-
sidering tensile strength, 7% to 10%
of maximum compressive strength cu
softening stages of conned and inelastic hardening strain in compres-
was chosen as a tensile strength, t0,
unconned concretes. The softening sion, cin, h was derived as follows:
and in this study, the maximum value
phase continued until 20% of the
c was taken (i.e. t0 = 0.1 cu). In this
unconned cylinder compressive cin, h = c 17 paper, 1% of the tensile strength was
strength (Point C) was reached; that E0
considered during the analysis regard-
is, the stress value was not allowed to
Cyclic behavior contributed to con- less of the realistic condition to pre-
continue to decrease, and perfect plas-
crete behavior, which was dened by vent numerical instability. In contrast,
tic behavior was assumed following
effective parameters, including dam- correspondence strain value, where
the softening trend (CD). For sim-
age in compression and damage in stress is 1% of the ultimate tensile
plicity, the entire constitutive model
tension. Compression damage (dc) strength, was taken as 10 times the
was assumed to be a parabolic curve.
was based on inelastic hardening percentage of the strain, in which
Equation (16) assumed a nonlinear
behavior for concrete from the begin- strain in compression cin, h that con- stress was equal to ultimate tensile
trolled the unloading curve slope. strength.
ning to the end. However, dening
the behavior of concrete up to 40% of Given that dc increased with respect Figure 1b showed that with a further
its strength in the elastic phase was to an increase in cin, h , it could be increase in the hardening cracking
important in determining the effective expressed as follows: ck, h
strain, t the tension damage con-
elastic modulus. In other words, the c tinued to increase, and this could be
constitutive model came into effect dc = 1 18
cu expressed as follows:
when the compressive strength was
60% of the concrete compressive The tangent of the curve decreased t
dt = 1 22
strength. According to Figs. 1a and 2, with respect to the initial tangent t0

Structural Engineering International Nr. 1/2017 Scientic Paper 71


Application of SCDP Parameters grades (20, 30, 40, and 50) were Figure 3 shows a partially prestressed
in FE Programs implemented within the framework in concrete beam that measured 16 m
tabular format, that is, Tables 1, 2, 3, 1 m 0.6 m, and was modeled in
Poissons ratio range for concrete was
and 4, respectively. Accordingly, the ABAQUS. Twelve pieces of steel
between 0.1 and 0.2. Poissons ratio,
hardening and softening rule as well reinforcements sized D = 20 mm with
elasticity modulus, and stressstrain
as the evolution of the scalar damage Grade 460 Mpa and E = 200 GPa
curve of concrete in compression and
variable for compression and tension were placed in two rows with 100 mm
tension were related to loading his-
were presented for concrete grades spacing, horizontally and vertically.
tory, which resulted in strain differ-
20, 30, 40, and 50. The general frame- A prestressed tendon of grade 270
ences. The dilation angle was equal to work of the damage plasticity formu- with seven wire strands (270 ksi =
volume strain over shear strain. The lation was clearly stated and could be 1861 MPa; E = 196.5 GPa; and area
dilation angle for concrete was usually extended to other concrete grades of 1600 mm2) was placed at 200 mm
20 to 40 , which affected material between B20 and B50. eccentricity from the center.
ductility. Consequently, the dilation
angle had considerable effects on the In this study, effective prestressing
entire model. An increase in the dila- Numerical Computation force (Pe = 2667.2 kN), self-weight, and
tion angle increased the system exi- superimposed dead load were applied
bility. From a practical viewpoint, the The performance of the proposed to the simply supported partially pre-
internal dilation angle depended on constitutive model was further evalu- stressed beam. A concrete block and
certain parameters, including plastic ated through the structural analysis of tendon were modeled by a three-
strain and conned pressure. An a simply supported partially pre- dimensional cube element (C3D8), and
increase in plastic strain and conned stressed beam, which was subject to the steel reinforcement elements were
pressure decreased the internal dila- self-weight and superimposed loading. 3D truss (T3D2).
tion angle. The material has a con-
stant dilation angle for a large range Plasticity parameters
Materials
of pressure stresses used for the con-
parameters B40 Dilation angle 31
nement of the material. The default
ow potential eccentricity was = 0.1, Concrete elasticity Eccentricity 0.1
and by raising the value of , the curva- E (GPa) 30 fb0/fc0 1.16
ture of the ow potential increased. If 0.2 K 0.67
the default ow of potential eccentric-
Viscosity parameter 0
ity had a value much lower than the
default value, there could be conver- Concrete compressive behavior Concrete compression damage
gence problems when the conning Yield stress (MPa) Inelastic strain Damage parameter Inelastic strain
pressure is not high. The ratio of ini- C
tial equibiaxial compressive yield 20.4 0 0 0
stress to initial uniaxial compressive
25.6 2.66667E-05 0 2.66667E-05
yield stress was given by fb0/fc0, and
its default value was 1.16 [34]. The 30 0.00008 0 0.00008
ABAQUS software31 used a null 33.6 0.00016 0 0.00016
default viscosity parameter so that the 36.4 0.000266667 0 0.000266667
viscoplastic regularization did not
38.4 0.0004 0 0.0004
occur. This parameter enhanced the
convergence rate of the model when 39.6 0.00056 0 0.00056
the softening process occurred, and it 40 0.000746667 0 0.000746667
gave good results. In the FE program, 39.6 0.00096 0.01 0.00096
the tension recovery parameter equal
to zero denoted that this parameter 38.4 0.0012 0.04 0.0012
contributed to the cyclic behavior and 36.4 0.001466667 0.09 0.001466667
monitored the modulus of elasticity 33.6 0.00176 0.16 0.00176
when compression behavior changed
30 0.00208 0.25 0.00208
to tension behavior and vice versa. A
value of zero for tension recovery 25.6 0.002426667 0.36 0.002426667
implied that material tangent in the 20.4 0.0028 0.49 0.0028
tension phase was completely affected
14.4 0.0032 0.64 0.0032
by compression damages. However,
compression recovery was taken as 7.6 0.003626667 0.81 0.003626667
1, which meant that tension damage Concrete tensile behavior Concrete tension damage
did not affect the material tangent. Yield stress (MPa) Cracking strain Damage parameter Cracking strain
All these assumptions were also T
adopted in reality.
4 0 0 0
According to the aforementioned
0.04 0.001333333 0.99 0.001333333
damage plasticity formulation derived
in this section, four different concrete Table 3: Material properties for concrete with SCDP model in class B40

72 Scientic Paper Structural Engineering International Nr. 1/2017


Plasticity parameters beam failed when the damage in ten-
Materials sion reached 0.99. Accordingly, the
parameters B50 Dilation angle 31 maximum compression stress (S,
Concrete elasticity Eccentricity 0.1 Mises) in concrete and the mid-span
E (GPa) 33.4 fb0/fc0 1.16 displacement of the beam subjected
0.2 to self-weight, superimposed dead
K 0.67
load, and prestressing load were
Viscosity parameter 0 determined.
Concrete compressive behavior Concrete compression damage Figure 8 shows the value of the con-
Yield stress (MPa) Inelastic strain Damage parameter Inelastic strain crete stress and mid-span displace-
C ment of the simply supported partially
25.5 0 0 0 prestressed beam. The plots demon-
strate that raising the concrete grade
32 5.73819E-06 0 5.73819E-06
increases the maximum compressive
37.5 4.13628E-05 0 4.13628E-05 stress in concrete as well as the mid-
42 0.000106874 0 0.000106874 span displacement. The compressive
stress of concrete for different classes,
45.5 0.000202271 0 0.000202271
B20, B30, B40, and B50, reached
48 0.000327555 0 0.000327555 92.35%, 84.53%, 81.475%, and
49.5 0.000482726 0 0.000482726 77.94%, respectively, of their corre-
50 0.000667782 0 0.000667782
sponding ultimate concrete strength.
Furthermore, the mid-span displace-
49.5 0.000882726 0.01 0.000882726 ment shows a gradual increase from
48 0.001127555 0.04 0.001127555 53.8 mm in B20 to 74.64 mm in B50.
45.5 0.001402271 0.09 0.001402271 The force versus displacement graphs
42 0.001706874 0.16 0.001706874 obtained in the service stage due to
self-weight and superimposed dead
37.5 0.002041363 0.25 0.002041363
loads are shown in Fig. 9. The capac-
32 0.002405738 0.36 0.002405738 ity of the partially prestressed con-
25.5 0.0028 0.49 0.0028 crete beam in B20 is 28.20452,
18 0.003224148 0.64 0.003224148 whereas the capacities of B30, B40
and B50 indicated increases of 48.2%,
9.5 0.003678183 0.81 0.003678183 119% and 172.7%, respectively.
Concrete tensile behavior Concrete tension damage
Yield stress (MPa) Cracking strain Damage parameter Cracking strain Validation of Simplied and
T
Tabulated Concrete Damage
5 0 0 0 Plasticity Model
0.05 0.001494322 0.99 0.001494322
Validation of the simplied and tabu-
Table 4: Material properties for concrete with SCDP model in class B50 lated concrete damage plasticity
model was achieved by comparing the
(a) (b) 600 m m
empirical formulations given in previ-
ous research with that presented in
this paper. Table 5 indicated the
results of the empirical formulation as
1m

1000 well as the SCDP model for a simply


mm 200 mm
supported beam in the service stage
16 m
subjected to self weight and super
imposed dead load. The results
0.6 m 6 20 @ 100 mm
showed a good correlation between
Fi g. 3: Schematic view of the partially prestressed concrete beam: (a) Isometric view, (b) these two approaches. The discrep-
Cross section ancy was observed in the analysis
techniques as the NewtonRaphson
method was incorporated in the nite
The effect of the SCDP model on the plotted to interpret the effect of the
element modeling, while the linear
concrete behavior was numerically SCDP method on the aforementioned
technique was employed in the empir-
investigated for four different pre- partially prestressed concrete classes.
ical formulations.
stressed concrete grades, namely, B20, The damage in tension, stress distribu-
B30, B40, and B50. Three key fea- tion, and mid-span deection of the where C, I and W were the distance
tures, including damage in tension, partially prestressed concrete beams of bottom ber to the neutral axis,
Von Mises stress, and maximum dis- are shown in Figs. 47 for the four moment of inertia and service load at
placement, were recorded and com- concrete classes B20, B30, B40, and failure, respectively. Deection due to
pared to determine the effect of the B50, respectively. As demonstrated in prestress load and service load were
SCDP method. The results were these gures, the prestressed concrete sequentially depicted by P, SL.

Structural Engineering International Nr. 1/2017 Scientic Paper 73


(a) DAMAGET
(Avg: 75%)
+9.900e01
+8.910e01
+7.920e01
+6.930e01
+5.940e01
+4.950e01
+3.960e01
+2.970e01
+1.980e01
+9.900e02
+0.000e+00

(b) S, Mises
(Avg: 75%)
+1.847e+01
+1.663e+01
+1.479e+01
+1.295e+01
+1.111e+01
+9.277e+00
+7.439e+00
+5.601e+00
+3.763e+00
+1.925e+00
+8.737e02

(c) U, U2
+0.000e+00
5.385e+00
1.077e+01
1.616e+01
2.154e+01
2.693e+01
3.231e+01
3.770e+01
4.308e+01
4.847e+01
5.385e+01

F ig . 4: The results of the analysis on concrete B20: (a) damage in tension, (b) stress distribution (MPa), and (c) displacement
distribution (mm)

(a) DAMAGET
(Avg: 75%)
+9.900e01
+8.910e01
+7.920e01
+6.930e01
+5.940e01
+4.950e01
+3.960e01
+2.970e01
+1.980e01
+9.900e02
+0.000e+00

(b) S, Mises
(Avg: 75%)
+2.536e+01
+2.283e+01
+2.031e+01
+1.778e+01
+1.526e+01
+1.274e+01
+1.021e+01
+7.687e+00
+5.163e+00
+2.639e+00
+1.147e01

(c) U, U2
+0.000e+00
5.986e+00
1.197e+01
1.796e+01
2.394e+01
2.993e+01
3.591e+01
4.190e+01
4.789e+01
5.387e+01
5.986e+01

F ig . 5: The results of the analysis on concrete B30: (a) damage in tension, (b) stress distribution (MPa), and (c) displacement
distribution (mm)

(a) DAMAGET
(Avg: 75%)
+9.900e01
+8.910e01
+7.920e01
+6.930e01
+5.940e01
+4.950e01
+3.960e01
+2.970e01
+1.980e01
+9.900e02
+0.000e+00

(b) S, Mises
(Avg: 75%)
+3.259e+01
+2.935e+01
+2.610e+01
+2.286e+01
+1.961e+01
+1.637e+01
+1.312e+01
+9.877e+00
+6.632e+00
+3.387e+00
+1.420e01

(c) U, U2
+0.000e+00
6.983e+00
1.397e+01
2.095e+01
2.793e+01
3.492e+01
4.190e+01
4.888e+01
5.587e+01
6.285e+01
6.983e+01

F i g. 6: The results of the analysis on concrete B40: (a) damage in tension, (b) stress distribution (MPa), and (c) displacement
distribution (mm)

74 Scientic Paper Structural Engineering International Nr. 1/2017


(a) DAMAGET
(Avg: 75%)
+9.900e01
+8.910e01
+7.920e01
+6.930e01
+5.940e01
+4.950e01
+3.960e01
+2.970e01
+1.980e01
+9.900e02
+0.000e+00

(b) S, Mises
(Avg: 75%)
+3.897e+01
+3.509e+01
+3.121e+01
+2.733e+01
+2.344e+01
+1.956e+01
+1.568e+01
+1.180e+01
+7.924e+00
+4.044e+00
+1.638e01

(c) U, U2
+0.000e+00
7.464e+00
1.493e+01
2.239e+01
2.986e+01
3.732e+01
4.478e+01
5.225e+01
5.971e+01
6.718e+01
7.464e+01

Fi g. 7: The results of the analysis on concrete B50: (a) damage in tension, (b) stress distribution (MPa), and (c) displacement
distribution (mm)
69.83 74.64
Subscripts H and FE denoted the
59.7774
53.8059
contributions made by the empirical
formulation and nite element model-
ing, respectively.
32.59 38.97
18.47 25.36
Furthermore, SCDP in tabular form
was veried through comparison with
Mid Span Displacement (mm)
nite element models of the available
B20
CDP models given in extant litera-
B30
B40
Maximum Principal Stress (Mpa)
ture. Two comparisons were carried
B50 out for concrete strength classes of
Fi g . 8: Concrete stress and mid-span displacement of the simply supported prestressed 30 MPa and 50 MPa. Von Mises
concrete beam with different concrete grades stress, damage in tension and mid-
span displacement were obtained by
2000 B20 B30 B40 B50 utilizing nonlinear nite element tech-
1800
niques of a partially prestressed con-
1600 crete beam as presented in Figs. 10
Reaction force (kN)

1400 and 10. The FE analysis incorporated


1200 the concrete damage plasticity model
1000 proposed by Tiwari27 and Janko-
800 wiak32 for concrete strength class of
600 30 MPa and 50 MPa. It should be
400 noted that the model proposed by
200 Tiwari assumed that the simple sup-
0 ported beam failed either the damage
0 10 20 30 40 50 60 70 80 in tension reaching 0.82 or the con-
Mid span displacement (mm) crete exceeded the maximum com-
Fi g . 9: Pushover curve of different concrete strength classes pressive stress.

C I W P* SL** H*** FE,SCDP H**** FE,


Grade (mm) (mm4) (N/mm) (mm) (mm) (mm) (mm) (Mpa) SCDP (Mpa)
B20 488.1 5.032 1010 78 16.0 62.4 46.4 53.8059 24.3 18.47
B30 490.5 5.029 1010 101.5 12.8 64.7 51.9 59.7774 31.9 25.36
B40 491.5 5.028 1010 128.8 11.3 72.9 61.6 69.83 40.7 32.59
B50 492.4 5.027 10 10
154.2 10.2 78.3 68.1 74.64 48.9 38.97
2
*P = PeL
8EI
5WL4
**SL = 384EI
***H = P + SL
****
H = A Zt + MZSLt
P Pe

Table 5: concrete stress and mid-span displacement in the SCDP and empirical formulation

Structural Engineering International Nr. 1/2017 Scientic Paper 75


DAMAGET

(a) (Avg: 75%)


+8.254e-01
+7.428e-01
+6.603e-01
+5.778e-01
+4.952e-01
+4.127e-01
+3.302e-01
+2.476e-01
+1.651e-01
+8.254e-02
+0.000e+00

S, Mises

(b) (Avg: 75%)


+2.640e+01
+2.377e+01
+2.114e+01
+1.852e+01
+1.589e+01
+1.326e+01
+1.063e+01
+7.997e+00
+5.367e+00
+2.738e+00
+1.080e-01

U, U2
(c) +0.000e+00
6.526e+00
1.305e+01
1.958e+01
2.610e+01
3.263e+01
3.915e+01
4.568e+01
5.221e+01
5.873e+01
6.526e+01

F ig . 10 : The results of the analysis on concrete B3027 : (a) damage in tension, (b) stress distribution (MPa), and (c) displacement
distribution (mm)

(a) DAMAGET
(Avg: 75%)
+9.900e01
+8.910e01
+7.920e01
+6.930e01
+5.940e01
+4.950e01
+3.960e01
+2.970e01
+1.980e01
+9.900e02
+0.000e+00

(b) S, Mises
(Avg: 75%)
+3.231e+01
+2.910e+01
+2.588e+01
+2.266e+01
+1.944e+01
+1.623e+01
+1.301e+01
+9.791e+00
+6.573e+00
+3.355e+00
+1.378e01

(c) U, U2
+0.000e+00
6.807e+00
1.361e+01
2.042e+01
2.723e+01
3.404e+01
4.084e+01
4.765e+01
5.446e+01
6.126e+01
6.807e+01

F ig . 11 : The results of the analysis on concrete B5032: (a) damage in tension, (b) stress distribution (MPa), and (c) displacement
distribution (mm)

With respect to the results obtained SCDP model and the studied FE Conclusions
and summarized in Table 6, it was models. The maximum reaction force
observed that the compressive stress of the beam with the Tiwari and This paper developed the present
and mid-span displacement values of SCDP models in concrete class B30 damage plasticity model (SCDP). This
the available concrete damage plastic- showed a difference of approximately model simplied the procedure of
ity models and SCDP model had good 4.6%. However, the discrepancy in existing damage plasticity model called
agreement. the concrete class of B50 between CDP. It combined a stress-based plas-
Jankowiak model and SCDP model ticity part with a strain-based damage
Figure 11 clearly demonstrated a rea-
was approximately 16%. mechanics model for the unconned
sonable correlation between the
prestressed concrete beam based on a
tabular format for four different con-
Compressive Compressive Mid span Mid span crete grades (B20, B30, B40, and B50).
stress stress displacement displacement Accordingly, the ndings in this paper
CDP models (MPa)B30 (MPa)B50 (mm) B30 (mm) B50 indicated the following conclusions
from the simplication provided by
Tiwari 26.4 65.26
the proposed tabulated concrete dam-
Jankowiak 32.3 68.07 age plasticity model:
SCDP 25.36 38.97 59.86 74.64 Due to its simplicity, the SCDP
Table 6: Comparison of results between concrete stress and mid-span displacement model was a suitable solution to

76 Scientic Paper Structural Engineering International Nr. 1/2017


B30 B30 (Tiwari 2015) B50 B50 (JANKOWIAK 2005) [2] Kang HD. Triaxial Constitutive Model for
2000 Plain and Reinforced Concrete Behavior Uni-
1800 versity of Colorado: Boulder, 1997.
1600
Reaction force (kN)

[3] Grassl P, Lundgren K, & Gylltoft K. Con-


1400
crete in compression: a plasticity theory with a
1200
novel hardening law. Int. J. Solids Struct. 2002;
1000
39(20): 52055223.
800
600 [4] Chen AC, & Chen WF. Constitutive rela-
400 tions for concrete. J. Eng. Mech. 1975; 101(4):
200 465481.
0 [5] Carol I, Rizzi E, & Willam K. On the for-
0 10 20 30 40 50 60 70 80
mulation of anisotropic elastic degradation.
Mid span displacement (mm)
I. Theory based on a pseudo-logarithmic dam-
Fi g. 12: Validation of force versus displacement curves of the SCDP and available CDP age tensor rate. Int. J. Solids Struct. 2001; 38(4):
models.27,32 491518.
[6] Gatuingt F, & Pijaudier-Cabot G. Coupled
model the cracking and the crush- pl plastic strain tensor damage and plasticity modelling in transient
ing of concrete. Thus, a 3D nonlin- Del degraded elastic tensor dynamic analysis of concrete. Int. J. Numer.
ear FE model was developed to Anal. Methods Geomech. 2002; 26(1): 124.
Del0 initial elastic tensor
assess the performance of concrete nominal stress [7] Krtzig WB, & Plling R. An elasto-plastic
with the SCDP model. All possible damage model for reinforced concrete with
effective stress
nonlinearities, namely, material and minimum number of material parameters. Com-
d scalar damage variable put. Struct. 2004; 82(15): 12011215.
geometric, were considered in the dt scalar tension damage
developed FE model. variable [8] Ju J. On energy-based coupled elastoplastic
The presented 3D nonlinear FE dc scalar compression damage
damage theories: constitutive modeling and
computational aspects. Int. J. Solids Struct.
model was then used to model variable 1989; 25(7): 803833.
different concrete classes. The cu ultimate compressive
developed FE model successfully [9] Lee J, & Fenves GL. Plastic-damage model
strength of unconned for cyclic loading of concrete structures. J. Eng.
predicted the concrete damage cylinder Mech. 1998; 124(8): 892900.
caused by tension, compressive c strain of unconned cylinder
stress in concrete, and mid-span dis- [10] Jason L, Huerta A, Pijaudier-Cabot G, &
at ultimate compressive
placement in the prestressed con- Ghavamian S. An elastic plastic damage formu-
strength lation for concrete: application to elementary
crete beam. pl compressive plastic strain tests and comparison with an isotropic damage
The model realistically described c
tensor model. Comput. Methods Appl. Mech. Eng.
the transition from tensile to pl plastic strain tensor 2006; 195(52): 70777092.
compressive failure. This nding
pl
t
tensile plastic strain tensor [11] Lubliner J, Oliver J, Oller S, & Onate E.
was achieved through the intro- ,h
duction of two separate isotropic in
c
inelastic compression strain A plastic-damage model for concrete. Int.
J. Solids Struct. 1989; 25(3): 299326.
damage variables for tension and pl,h
plastic hardening strain
ck, h cracking strain
compression. t [12] Ananiev S, Ozbolt J. Plastic-damage model
,h for concrete in principal directions. arXiv pre-
The proposed three-dimensional pl
c
plastic hardening strain in
print 2007;0704.2662.
FE model provided the framework compression
pl, h plastic hardening strain in [13] Imran I, & Pantazopoulou S. Plasticity
to develop a realistic model to t model for concrete under triaxial compression.
determine the behavior of the sim- tension
J. Eng. Mech. 2001; 127(3): 281290.
ply supported partially prestressed :: total strain rate
beam. Then, the comparison of the _ el total strain rate in elastic part [14] Grassl P, & Jirsek M. Plastic model with
non-local damage applied to concrete. Int. J.
results between the SCDP model _ pl total strain rate in plastic part
Numer. Anal. Methods Geomech. 2006; 30(1):
and available CDP models includ- :: rate of plastic multiplier 7190.
ing, the two different concrete gp plastic potential
[15] Grassl P, & Rempling R. A damage-
damage plasticity models, namely, FE mid span deection due to plasticity interface approach to the meso-scale
the Tiwari27 and the Jankowiak FE model modelling of concrete subjected to cyclic com-
and Lodygowski32 models, showed FE max concrete compressive pressive loading. Eng. Fract. Mech. 2008;
that the present CDP model had a stress due to FE model 75(16): 48044818.
good correlation with these CDP [16] Yu T, Teng J, Wong Y, & Dong S. Finite
models. Acknowledgements element modeling of conned concrete-II:
plastic-damage model. Eng. Struct. 2010; 32(3):
This work was supported by the University
Nomenclature Putra Malaysia under Putra grant
680691.
No. 9438709. This support is gratefully [17] Zhang J, Zhang Z, & Chen C. Yield crite-
E youngs modulus acknowledged. rion in plastic-damage models for concrete.
Ec concrete youngs modulus Acta Mech. Solida Sin. 2010; 23(3): 220230.
Es steel youngs modulus
[18] Taddei F, Reindl L, Park J,
Ep Prestressing steel youngs References Butenweg C, & Karadogan F. Numerical inves-
modulus tigation of AAC wall panels based on the dam-
[1] Pekau O, & Zhang Z. Strain-space cracking
poissons ratio model for concrete and its application. Comput. age plasticity constitutive law. Cement Wapno
el elastic strain tensor Struct. 1994; 51(2): 151162. Beton 2011; 7: 8691.

Structural Engineering International Nr. 1/2017 Scientic Paper 77


[19] Grassl P, Nystrom U, Rempling R, damage-plasticity approach to modelling the internal blast loading. Adv. Struct. Eng. 2015; 1:
Gylltoft K. A damage-plasticity model for the failure of concrete. Int. J. Solids Struct. 2013; 405415.
dynamic failure of concrete. arXiv preprint 50(24): 38053816.
[28] Hognestad E. Study of combined bending
2011; arXiv:11031288.
[24] Vaghei R, Hejazi F, Taheri H, and axial load in reinforced concrete members
[20] Larsson S, Malm R, Charbit B, & Jaafar MS, & Ali AAA. Evaluate performance University of Illinois: Urbana, 1951.
Ansell A. Finite element modelling of laterally of precast concrete wall to wall connection.
[29] Kent DC, & Park R. Flexural members
loaded limecement columns using a damage APCBEE Procedia 2014; 9: 285290.
with conned concrete. J. Struct. Div. 1971;
plasticity model. Comput. Geotech. 2012; 44:
[25] Shang K, Guo Q, He Z, & Ye Y. A study 97(7): 19691990.
4857. of the torsional behavior of reinforced concrete
[30] Park R. Reinforced concrete structures
[21] Ming C. Research on damage plastic calcu- trough girder based on ABAQUS. CICTP
John Wiley & Sons: New York, 1975.
lation method of ABAQUS concrete damaged 2012; 10: 30293035.
plasticity model. Transport. Stand. 2012; 2: 5154. [31] Simulia D. ABAQUS 6.11 analysis users
[26] Tao Y, & Chen J. Concrete damage plas-
manual. ABAQUS 611 Documentation, 2011.
[22] Zhang J, & Li J. Investigation into Lubli- ticity model for modeling FRP-to-concrete bond
ner yield criterion of concrete for 3D simula- behavior. J. Compos. Construct. 2014; 19(1): [32] Jankowiak T, & Lodygowski T. Identica-
tion. Eng. Struct. 2012; 44: 122127. 04014026. tion of parameters of concrete damage plasticity
constitutive model. Found. Civil Environ. Eng.
[23] Grassl P, Xenos D, Nystrm U, [27] Tiwari R, Chakraborty T, & Matsagar V.
2005; 6: 5369.
Rempling R, & Gylltoft K. CDPM2: a Dynamic analysis of curved tunnels subjected to

78 Scientic Paper Structural Engineering International Nr. 1/2017

You might also like