Page 157 - Riem-Vol6_nº3

502 IBRACON Structures and Materials Journal • 2013 • vol. 6 • nº 3
Numerical approach of the bond stress behavior of steel bars embedded in self-compacting concrete
and in ordinary concrete using beam models
for each series, according to the concrete compressive strength.
So, for SCC1 and OC1 series (fc = 30 MPa), the failure was
characterized by high slip of the steel bar with low displacement
of the beam and, and for SCC2 and OC2 series (fc = 60 MPa),
the failure was characterized by the yield of the steel bar with
low slip and the beam presenting high vertical displacement.
3. Numerical approach
The failure in steel-concrete interface could be attained by
combining Coulomb’s frictional hypothesis with a bound for the
maximum tensile stress, resulting in two different failure modes
that could be called sliding failure and separation failure [10].
The sliding failure is assumed to occur in a section when the
shear stress exceeds the sliding resistance and should be de-
termined by two parameters: the cohesion (c) and the friction
coefficient (m ).
In previous studies [15-18] the variation of the frictional coef-
ficient and the cohesion seemed not to affect the general re-
sponse of the bond in the contact surface. However, the number
of elements in the contact surface, and parameters like FKN
(normal contact stiffness factor), FKT (tangent contact stiffness
factor) and IT (iteration number), presented in Ansys® software,
affect directly the load vs. slip behavior, according to the adopt-
ed bond model [15-16]. But, it is worth to mention that the differ-
ences between the two materials (SCC and OC) are restricted
to the material’s properties.
3.1 Materials
Compressive strength and elasticity modulus of concrete were ob-
steel bar, near the bonded region and in the middle of the bar,
as shown in Figure 3. According to the technical literature, the
application of strain gages on the steel-concrete interface must
be avoided due to its high influence on the bond stresses; how-
ever, some researches good estimation of the bond stress dis-
tribution with this procedure [13-14]. So, the strain gages where
placed at the beginning and at the end of the embedment length
to measure the strain variation, and an additional strain gage
was placed in the middle of the bar (Figure 3).
The used cement was Ciminas CP-V Ari Plus (initial high
strength cement). The used siliceous sand had density of 2.63
kg/dm3 and absorption of 4.0% and the used crushed gravel
had density of 2.83 kg/dm3 and absorption of 1.71%. The used
superplasticizer was based on carboxylate, with density of 1.1
kg/dm3 and 20% of solid content. Table 1 shows the materials
contents and the results for fresh SCC. Table 2 shows the hard-
ened properties of SCC series and OC series.
Figure 4 shows some of the specimens, a beam cast with OC
with 10 mm steel bar and a beam cast with SCC with 16 mm
steel bar the beams, during tests at the universal test ma-
chine, Instron.
According to the experimental results, the failure was different
Figure 3 – Steel bar instrumentation
10 mm steel bar
16 mm steel bar
Unbonded zone Bonded zone
SG SG SG SG
Unbonded zone Bonded zone
SG SG SG SG
25,0 22,0 25,0 16 22,0 16
13,75 10 16,25 13,75 10 16,25
SG - Strain Gage
Table 1 – Materials content and fresh results for SCC series
Material Tests OC1 OC2 SCC1 SCC1 SCC2 SCC2
Cement (kg)
Sand (kg)
Gravel (kg)
Water (kg)
Superplasticizer (%)
Filler (kg)
Silica fume (kg)
Slump test
Slump flow (cm)
T (s)50
L-Box test
T (s)60
RB
V-Funnel
T (s)v
365.3
883.9
942.3
260.8
–
–
–
488.3
766.6
942.4
227.0
–
–
–
338.8
854.8
919.1
273.6
0.4%
101.6
–
–
67.5
1.0
–
1.0
0.95
–
1.5
365.1
815.3
876.7
146.1
0.75%
146.1
36.5
–
61.0
1.0
–
1.0
0.9
–
2.0
Table 2 – Hardened results for SCC
and OC series
Hardened
properties OC1 SCC1 OC2 SCC2
f (MPa)c
E (MPa)c
f (MPa)ct
32.02
27.24
2.182
30.10
27.87
2.450
50.20
34.30
3.920
53.30
36.68
4.990