Analysis of a Semi-rigid Connection for Precast Concrete
-
Upload
pavan2deepuaki -
Category
Documents
-
view
149 -
download
1
Transcript of Analysis of a Semi-rigid Connection for Precast Concrete
Proceedings of the Institution ofCivil EngineersStructures and Buildings 163February 2010 Issue SB1Pages 41–51doi: 10.1680/stbu.2009.163.1.41
Paper 800067Accepted 09/09/2008Received 27/08/2009
Keywords: concrete structures/design methods & aids/buildings,structure & design
Mounir K. El DebsProfessor, Department ofStructural Engineering,University of Sao Paulo,Brazil
Anamaria M. MiottoFormer PhD student,Department of StructuralEngineering, University of SaoPaulo, Brazil
Ana Lucia H. C. El DebsAssociate Professor,Department of StructuralEngineering, University of SaoPaulo, Brazil
Analysis of a semi-rigid connection for precast concrete
M. K. El Debs BEng, MSc, PhD, A. M. Miotto BEng, MSc, PhD and A. L. H. C. El Debs BEng, MSc, PhD
This paper presents a study of a specific type of beam-to-
column connection for precast concrete structures.
Furthermore, an analytical model to determine the
strength and the stiffness of the connection, based on
test results of two prototypes, is proposed. To evaluate
the influence of the strength and stiffness of the
connection on the behaviour of the structure, the results
of numerical simulations of a typical multi-storey
building with semi-rigid connections are also presented
and compared with the results using pinned and rigid
connections. The main conclusions are: (a) the proposed
design model can reasonably evaluate the studied
connection strength; (b) the evaluation of strength is
more accurate than that of stiffness; (c) for a typical
structure, it is possible to increase the number of storeys
of the structure from two to four with lower horizontal
displacement at the top, and only a small increase of the
column base bending moment by replacing the pinned
connections with semi-rigid ones; and (d) although there
is significant uncertainty in the connection stiffness, the
results show that the displacements at the top of the
structure, and the column base moments present low
susceptibility deviations to this parameter.
1. INTRODUCTION
Precast concrete components are characterised by their ability
to be easily produced in factory plants. However, it is necessary
to connect these components to construct a building structure.
These connections are one of the main problems designers have
to face when using precast concrete structures.
The connections affect all stages of structures production, from
the adjacent parts of components manufacture to the assembly
of the structure. This in turn determines the behaviour of the
final structure. Consequently, they can be considered the most
important part of the design of precast concrete structures.
The simplest connections are pinned connections, which
produce structures subject to large bending moments. However,
connections that reproduce the behaviour of cast-in-place
reinforced concrete structures through the continuity of beams
and columns require more labour, reducing the inherent
advantages of precast concrete. The main problems of the
moment-resisting connections are the need for connecting
materials, steel and concrete, and tolerance adjustment, in
addition to the consequences of fragile concrete behaviour.
Thus, the beam-to-column connections of low-rise buildings
are generally designed to be pinned connections, owing to the
ease of production, instead of the better structural behaviour
resulting when the column bending moments are reduced. In
this case, the columns work as cantilevers and are fully
responsible for the structure stability. Consequently, when the
building height rises, the bending moments of the columns
rapidly increase, and the system becomes economically
impracticable. The alternatives to minimise this effect are the
use of special bracing elements or moment-resistant beam-to-
column connections. The adoption of a more difficult
connection is justified by the reduction of the column
bending moments when compared with pinned connection
systems.
The use of moment-resisting connections usually results in
rigid connections. The main objective is to obtain a final
structure that behaves as similarly as possible to the cast-in-
place concrete structures, whose structural analysis processes
are well known.
Between the pinned and the rigid connections, there are
intermediate behaviours, resulting in semi-rigid connections.
In steel structures and in concrete–steel composite structures,
the semi-rigid connection has been thoroughly studied. For
precast concrete structures, this subject was given some
attention in the 1980s (Bljuger, 1988). However, in the 1990s,
there was a massive mobilisation of researchers on behalf of
the COST C1 programme (control of the semi-rigid behaviour
of civil engineering structural connections). This programme
had seven working groups, one of which focused on the study
of connections in prestressed and reinforced concrete. Most of
the results obtained were related to precast concrete
connections (Chefdebien and Daldare, 1994; Elliott et al., 1998;
Lindberg and Keronen, 1992).
However, the use of semi-rigid connections, which has so far
been little explored in precast concrete structures, extends the
design possibilities and can bring significant benefits to the
design of multi-storey structures of small and medium-height
buildings.
The present paper proposes a study of the strength and stiffness
of a specific type of beam-to-column connection for precast
concrete structures by considering tests in two prototypes. The
Structures and Buildings 163 Issue SB1 Analysis of a semi-rigid connection for precast concrete El Debs et al. 41
complete results of the experimental programme can be found
in Miotto (2002).
This paper presents an analytical model to determine both the
strength and stiffness of the connection. It also presents the
influence of these parameters on the structural behaviour by
comparing the numerical simulation results of a typical multi-
storey building with semi-rigid connections and the usual
pinned and rigid connection solutions.
2. STUDIED CONNECTION
The studied connection is a modified beam-to-column
connection currently used in Brazil for precast concrete
structures with heights up to 10 m. Details of the considered
connection are given in Figure 1. Compared with the current
connection, which is generally assumed to be a pinned
connection, it presents the following differences: (a) the
elastomer cushion is replaced by a special mortar cushion and
(b) the space between the beam end recess and the column was
filled with grout.
The cushion used was made of a mortar of cement and sand
with additions of soft aggregate, latex and short fibres to
produce a cement-based material with high deformability and
tenacity (El Debs et al., 2003). El Debs et al. (2006a) present
deformability values, obtained from stress–deformation curves.
This cushion was developed to avoid stress concentration and
also to allow small rotations of the beam.
The modifications do not alter the connection appearance and
the involved tolerances. However, concerning the manufacture,
there is some additional work to fill the upper part of the joint
with grout. Regarding the behaviour of the connection, a
significant transmission of negative bending moment is
expected, as well as a positive bending moment, but at a lower
level.
For the beam design, the negative moments introduced will
cause a reduction in the positive moments in the middle of the
span for the vertical load applied after the connection has been
accomplished.
For the column design, the bending moment transmission
produces a partial, but significant, reduction in the column
bending moments as well as the horizontal displacements at
the top level of the structure. As a consequence, the second-
order effects also tend to be reduced, in comparison to the
pinned connection structures. Further, there are lower bending
moments in the column base. The use of a structural system
with semi-rigid connections could therefore make it possible to
reduce the column cross-section or to increase the total height
of the buildings.
3. EXPERIMENTAL PROGRAMME
3.1. Characteristics of the prototypes
The experimental programme includes two prototypes:
prototype No. 1 represents an interior connection, and
prototype No. 2 represents an exterior connection. The main
geometric characteristics of the specimens are presented in
Figure 2. Owing to the loading procedure, the slab was
separated in such a way that prototype No. 2 would represent
two columns with reduced cross-sections. This reduction in the
column cross-section of the prototype represents a lightly
unfavourable situation. The reason for this option was the ease
of both the prototype manufacture and testing procedures.
The column and the corbel reinforcement, as well as the beam
reinforcement, were designed to represent a span modulation of
6 m 3 6 m. The continuity reinforcement was estimated for
these dimensions by considering the acting load after the
connection was effective. For the flexural continuity
reinforcement distribution, half of the required steel area was
adopted as bars passing through the column, while the rest of
the area was distributed bars in the structural concrete topping.
As a consequence of the available commercial bars, the
effective reinforcement consisted of 2˘16 mm (200 mm2)
passing through the column, and 6˘10 mm (470 mm2) on the
sides of the column. In prototype No. 2, the bars in the concrete
cover (6˘10 mm) were anchored on the edge of the slab and
the beam web bars (2˘16 mm) passed through the column, as
in prototype No. 1, to represent, in the actual situation, a
suitable anchorage in the column. More details of the prototype
construction can be seen in El
Debs et al. (2006b).
3.2. Test and loading
set-up
The test set-up and some
additional details are
presented in Figure 3. A
servo-controlled actuator of
500 kN capacity was used to
apply the loads. The loading
of prototype No. 1 was cyclic
and reversed, with two cycles
corresponding to 20% of the
predicted load, for each side,
two more cycles increasing
the load to 40% of the
predicted load, followed by
two more cycles of 60% of
the predicted load for the
negative moment and then
Fillingwith grout
Cushionof mortar
Bolt
Column
Beam
Hollow coreslab
Holes to passreinforcement rods
Cast-in-placeconcrete
Continuityreinforcement
Figure 1. The proposed connection with details of the alterations from current connection
42 Structures and Buildings 163 Issue SB1 Analysis of a semi-rigid connection for precast concrete El Debs et al.
raising the load monotonically until failure. In prototype No. 2,
the load was also cyclic and reversed, following the same
procedures, but aiming to obtain a failure provoked by the
positive moment.
3.3. Materials
The mechanical properties of
the concrete and the grout
are presented in Table 1. The
properties were determined
using tests on cylindrical
samples of 150 mm diameter
and 300 mm height for the
concrete and 50 mm diameter
and 100 mm height for the
grout on the same day of the
prototype testing. The
compressive strength (fc) and
tensile strength (f t) were
calculated based on an
average of six specimens and
the modulus of elasticity (E)
was calculated as an average
of three samples.
The steel used to produce the
dowels for the prototypes
was SAE 1020 with a
25.4 mm diameter and
nominal yielding stress (fy)
of 250 MPa. The reinforcement steel was Brazilian standard
CA-50 with a nominal yielding stress of 500 MPa. The values
for the yielding stress (fy) and ultimate strength (fu) of the
steel are given in Table 2 and are the average of four tested
specimens. The hollow core slab with an average compressive
strength of 45 MPa was furnished by a precast concrete
company.
3.4. Bending moments–rotation curves for connections
The main results concerning semi-rigid connections are the
bending moment–rotation curves. The bending moment was
calculated considering the reaction to half of the load
introduced by the actuator at a distance of 1.40 m (on the
column face). The rotation of the connection was obtained by
averaging the relative displacements given by the transducers
placed on the upper part of the prototype and on the upper
face of the column corbel and dividing by 450 mm, which is
the distance between these positions.
Figure 4 shows the bending moment–rotation curves for
prototypes No. 1 and No. 2. These curves correspond to the
envelope curves in the cyclic loading stages. As can be
observed, the failure load for prototype No. 1 occurred in the
direction of the negative moment, while the opposite occurred
for prototype No. 2.
190
400
200
50
150
300
250 70
600
200
1050
800
Column Column
20
Bearingcushion
Hollowcore slabCast-in-place
concrete
Bolt (Ø 25 mm)�
Continuityreinforcement
600
600
400 200 1290
100
200
20
(a) (b)
Upper view
Figure 2. Geometry and details of the tested prototypes: (a) prototype No. 1; (b) prototypeNo. 2 (dimensions in mm)
1400 mm
1600 mm
20 mm
Transducers
450 mm
Transducers
Support
Support
Reaction structure
Actuator
Device toapply the load
Figure 3. Test set-up and some details of the prototype tests(dimensions in mm)
Beam and column concrete Concrete topping Grout
Prot. No. 1 Prot. No. 2 Prot. No. 1 Prot. No. 2 Prot. No. 1 Prot. No. 2
fc : MPa 49.0 49.5 33.2 28.5 52.6 59.4f t: MPa 3.2 3.9 2.9 2.5 4.7 4.5E: GPa 32.8 36.3 31.6 24.8 13.4 14.9
Table 1. Mechanical properties of concrete and grout
Structures and Buildings 163 Issue SB1 Analysis of a semi-rigid connection for precast concrete El Debs et al. 43
4. PROPOSED DESIGN MODEL
4.1. Ideal bending moment–rotation curves for
connection
For negative moments, three stages can be observed from the
results
(a) a very short and stiff initial stage
(b) a second stage, with less stiffness
(c) a final stage of yielding.
For positive moments, it is also possible to observe an initial
stage of high stiffness and then a very long stage where the
moment increases slowly, indicating that the material is
yielding.
For negative and positive moments, the initial stiffer stages of
the connections tend to lose stiffness with repeated loading;
therefore, the use of a bi-linear diagram is proposed for both
cases, as shown in Figure 5. In this way, the connection
behaviour can be described by only two parameters, for each
bending moment direction, the yielding moment My (Myp for
positive moment and Myn for negative moment) and the
stiffness k� (k�p for positive moment and k�n for negative
moment).
4.2. Negative moments
The yielding moment can be determined based on Figure 6,
where a rectangular distribution of stresses is assumed for the
grout and the cushion, and the dowel contribution is neglected.
Imposing force equilibrium in the vertical and horizontal
directions, the moment equilibrium related to point C, which is
aligned with Rcu, the solution can be determined. Suppose
Rs ¼ As fy and Rg ¼ ycn fcgbw, the moment at C is
Myn ¼ As fyzn1
with
zn ¼ he � d9e �ycn2
2
and
ycn ¼ As fyfcgbw
3
where As is the cross-section area for the continuity
reinforcement, fy is the yielding stress for the continuity
reinforcement, he is the height of the beam end recess, d9e is the
distance from the reinforcement to the beam face, ycn is the
height of compression stress block at beam end recess, fcg is
Diameter: mm fy:MPa fu: MPa
10 576 68312.5 611 70516 589 764
Table 2. Steel strength
�250
�200
�150
�100
�50
0
50
�4 �2 0 2 4 6 8 10 12
Rotation 10� 3
Ben
ding
mom
ent:
kN m
Prototype No. 1
Prototype No. 2
Figure 4. Envelope of bending moment–rotation curve forconnection
M
Myp
kφn
Myn
φ
Figure 5. Proposed bending moment–rotation relationship
d�e
zn
le
ycn
Rcu
xcu
xcu/2
Vhe
Rq
Rs
M
C
Col
umn
face
Figure 6. Stress diagram in the connection for negativemoments
44 Structures and Buildings 163 Issue SB1 Analysis of a semi-rigid connection for precast concrete El Debs et al.
the compressive strength of the grout, bw is the beam end
recess width and xcu is the length of the compression stress
block under the cushion.
The value of xcu varies with the cushion stiffness. If the
cushion is very deformable, its value tends towards the length
of the beam end recess. If a triangular stress diagram at the
beam end recess is adopted, the value of xcu will be 2‘e=3.
Based on the test results, the beam end recess is supposed to be
fully rigid. The deformability of the connections is therefore
due to: (a) a strain concentration of a main crack in the face of
the column on the tensioned side; (b) the joint deformation on
the compressed side; and (c) cushion deformation. Figure 7
shows the deformed position of the beam end recess. It can be
seen that the following components with their respective
stiffnesses were considered
(a) reinforcement (ks)
(b) grout of the joint (kg) and cushion (kcu).
The position of the centre of rotation CR, which coincides with
point C used to obtain the yielding moment, can be calculated
from
ycr ¼ks(he � d9e)þ kg ycn=2ð Þ
ks þ kg4
The stiffness of the continuity reinforcement is
ks ¼�sAs
wy5
where wy is the opening of the pronounced crack, close to the
column face, and its value can be estimated through the
following expression used by Miotto (2002), where the first
part is given in Federation Internationale du Beton (FIB, 1999)
and the second part is given in Engstrom (1992a)
wy ¼ 2(1þ Æw)s1
Æw�
8(1þ Æers,ef )�s
2
�maxEs
" #1=1þÆw
þ �s
Es4�6
with
Æe ¼Es
Ec,top7
�max ¼ 2:5ffiffiffiffiffiffiffiffiffiffifc,top
q8
rs,ef ¼As
Ac,ef9
where As is the area of the continuity reinforcement, Ac,ef is
the involved area of the continuity reinforcement, �s is the
stress in the continuity reinforcement, � is the average
diameter of the continuity reinforcement, Es is the steel
modulus of elasticity, Ec,top is the concrete topping modulus of
elasticity, fc,top is the compressive strength of concrete topping,
Æw ¼ 0:4 and s1 ¼ 1.
In this case, as the reinforcement is placed in the concrete
topping, Ac,ef can be calculated from the product of the
thickness of the concrete topping and the distance of the
influenced area from the tensioned bars.
Since the objective is to obtain the stiffness of the
reinforcement when yielding, the stress in the continuity
reinforcement �s is the yielding stress (fy).
The stiffness corresponding to the joint filled with grout is
kg ¼ycnbwDgj
10
where bw is the beam width and Dgj is the deformability of the
grout joint, in terms of stress.
The deformability values for the mortar joints can be found in
the technical literature (Barboza, 2002; Bljuger, 1988).
Because the moment equilibrium is related to the rotation
centre CR, the connection stiffness can be calculated using the
expression
k�n ¼ ks he � ycr � d9eð Þ2 þ kg ycr �ycn2
� �2" #
11
4.3. Positive moments
The yielding moment can be determined based on Figure 8,
where a rectangular block of stresses was considered in the
concrete topping. The eventual reaction of the cushion close to
the column is neglected.
φ
φ
φ
φks
hekg
kcu
ycr
xcu/2
ycn/2
VMCR
Figure 7. Deformed position of the beam and deformationcomponents for the negative moment
Structures and Buildings 163 Issue SB1 Analysis of a semi-rigid connection for precast concrete El Debs et al. 45
Imposing the force equilibrium in the vertical and horizontal
directions and the moment equilibrium referring to point C,
which corresponds to the dowel position, the bending moment
in C can be obtained from
Myp ¼ Fsdzp12
with
zp ¼ he �ycp2
13
and
Fsd ¼ c�d2
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffifyd fcc,max
q14
where ycp is the height of the compression stress block at the
top of the beam end recess and is calculated by
ycp ¼ Fsd= fcc,maxbf (bf is composite beam width); c is a
coefficient which can be found in the technical literature
(CEB–FIP, 1991); �d is the dowel diameter; fyb is the yielding
stress of dowel steel; and fcc,max is the higher value between
the compressive strength concrete and compressive strength
grout in contact with the dowel.
The deformed position of the beam end recess, which is
assumed to be rigid, can be obtained from Figure 9. It can be
observed that the following deformation components with the
associated stiffnesses are considered
(a) compressed concrete (kc)
(b) tension in the dowel (ktd)
(c) shear in the dowel (ksd).
As kc approaches infinity, the rotation centre CR is aligned,
which automatically defines its position. In this case, CR also
coincides with the point C used to calculate the yielding
moment.
The stiffness of the connection can be calculated by
considering the equilibrium of moments referred to the centre
of rotation CR, which produces the following expression
k�p ¼ ksd hc �ycp2
� �2
15
The stiffness associated with the shear deformation of the
dowel is
ksd ¼Fsd
avy16
where avy is the transverse displacement of the dowel when the
maximum force is attained.
More information for the calculation of the displacement can
be found in the technical literature (Engstrom, 1992b). This
displacement can also be obtained from CEB model code 1990
(CEB–FIP, 1991), which indicates 0:10�d.
4.4. Comparison with experimental results
Based on the presented equations, the characteristic values of
the moment–rotation curves for prototypes No. 1 and No. 2
can be obtained. The geometry values of the prototypes are
given in Figures 2 and 3. Most of the material mechanical
properties were previously presented. The compressive strength
of the mortar cushion, determined by cylinder samples of
50 mm diameter and 100 mm height, is 25 MPa. Table 3 shows
other values used to obtain the theoretical curves.
The following parameters were also considered
(a) joint deformability of 0.1 3 10�4 m/MPa (Bljuger, 1988)
(b) c coefficient of 1.2 in Equation 14
(c) displacement avy of 0:10�d for Equation 16.
Col
umn
face
le/2
le
Ftd
Fsd
zp
Rc C
ycp V
M
Figure 8. Stress diagram in the connection for positivemoments
le/2
ktdksd
φ
φ
φ
φ
kC � 00
CRV
M
Figure 9. Deformed position of the beam and deformationcomponents for the positive moment
46 Structures and Buildings 163 Issue SB1 Analysis of a semi-rigid connection for precast concrete El Debs et al.
Table 4 presents the connection stiffness and the yielding
moments calculated with the proposed procedures, regarding
points CR ¼ C. The difference between the interior and external
columns is attributable to the material properties.
Figure 10 shows the experimental moment–rotation curves
compared with the theoretical values given in Table 4. For this
comparison, the experimental moments were calculated with
respect to points CR ¼ C, which means lever-arms of 1.246 m
for prototype No. 1 and 1.280 m for prototype No. 2. For
prototype No. 2, which was intended to fail with the positive
moment, the strength and stiffness were well predicted.
Regarding prototype No. 1, the theoretical model is in good
agreement with the connection strength. However, for the
stiffness estimation, the design model furnished values which
are not very close to the experimental ones for both cases. On
the other hand, if the displacement analysis is considered, the
model will also furnish safer values.
With these results, it can be observed that the strength is
relatively well evaluated, but the stiffness evaluation is much
less precise. This is a consequence of the uncertainties present
in the calculation: the crack opening, the deformability of the
joint filled by grout and the displacement of the dowel until
the maximum force is reached.
5. NUMERICAL MODELLING OF A TYPICAL
STRUCTURE
In order to evaluate the influence of the connection stiffness on
the structure behaviour, numerical simulations of a typical
multi-storey building with connections of different grades of
stiffness were performed.
The analysis was restricted to the global stability and the
increase of bending moments in the columns owing to the
structural deformation. The purpose was to show that some
Prototype No. 1Interior column
Prototype No. 2Exterior column
Distance from the reinforcement to the beam top (d9e) 25 mmContinuity reinforcement 870 mm2
2˘16 + 6˘10400 mm2
2˘16Average diameter of the continuity reinforcement 12.5 mm 16 mm
Note: the continuity reinforcement outside the beam web of prototype No. 2 was not takeninto account as it is not anchored appropriately on the slab border vicinity.
Table 3. Values used to obtain the theoretical curves
Interior column Exterior column
Negative moments Positive moments Negative moments Positive moments
Stiffness: MN/rad 60.9 6.4 25.1 6.3Yielding moment: kN m 202.0 36.4 96.3 36.3
Table 4. Theoretical values of connection stiffness and yielding moments
�250
�200
�150
�100
�50
0
50
�10 �8 �6 �4 �2 0 2 4 6 8 10
Rotation 10� 3
Ben
ding
mom
ent:
kN m
Prototype No. 1
Theoretical for internal connection
Prototype No. 2
Theoretical for external connection
Figure 10. Comparison of experimental and theoretical bending moment–rotation curves forconnection
Structures and Buildings 163 Issue SB1 Analysis of a semi-rigid connection for precast concrete El Debs et al. 47
advantages can be obtained when the column-to-beam
connections are considered to be semi-rigid.
Figure 11 represents a typical two-storey frame structure of
three spans. The loads indicated in Figure 11 are already
magnified by the safety factors. Figure 12 shows the structural
models: (a) pinned before the connections between beams and
columns were effective and (b) semi-rigid after the connections
between beams and columns were effective. The dead-load (g),
which corresponds to the weight of the structure itself, is
working in the pinned connections model, while the live-load
(q) and the wind load (W) are working in the semi-rigid
connections model.
The ªz coefficient method proposed by Franco and Vasconcelos
(1991) was chosen for the performed analysis. The analysis also
verifies the need to consider second-order effects by
demonstrating a simplified estimate of these effects. Concisely,
it consists in calculating the ªz coefficient, which evaluates the
deformability of the structure and, multiplied by the horizontal
loads, can also take into account non-linear effects. The ªz
coefficient is given by
ªz ¼1
1� ˜Md=M1dð Þ17
where M1d is the first-order moment at the bottom of the
structure owing to the lateral loads and ˜Md is the first
evaluation of the second-order moments, calculated from the
structure deformations due to the first-order moments.
For design purposes, if ªz is less than 1.1, then there is no need
to consider the overall second-order effects, and if ªz value is
less than 1.2 and greater than 1.1, the moments obtained in the
first-order analysis must be multiplied by ªz . In this study, the
ªz coefficient is applied even if it is lower than 1.1.
The displacements of the structure can be obtained using the
reduced values of flexural stiffness in order to consider the
non-linear behaviour of the materials. The usual values are
(EI )red ¼ 0.4EI for beams and (EI )red ¼ 0.8EI for columns, in a
framed structure, and (EI )red ¼ 0.4EI for fixed-end columns
(cantilever action) and pinned beams (El Debs, 2000). In the
absence of data to establish the stiffness reduction for semi-
rigid connections, the mean value of 0.60 is considered.
The finite-element-based program ANSYS 8.0 was used to
process the calculation. The semi-rigid connections were
simulated by spring elements, COMBIN39, available in the
software library. The COMBIN39 elements are capable of
representing a spring with non-linear behaviour, so the bilinear
moment–rotation curve defined for the connections was used.
In order to analyse the effect of the stiffness of the
connections, the following alternatives were considered
(a) pinned connections
(b) semi-rigid connections with the values present in Table 5
(c) rigid connections (fully restrained connections).
The values in Table 5 were calculated considering the material
properties
(a) precast concrete compressive strength of 35 MPa
(b) cast-in-place concrete compressive strength of 25 MPa
g: kN/m q: kN/m W: kN
Top storey
Intermediatestorey
26
Beam sections: 300 mm 650 mmColumn sections: 300 mm 400 mm
�
�6·0 m 6·0 m 6·0 m
3·75
m3·
75 m
20 10 10
2014
g qt t�
g qi i�
Wt 10 kN�
Wi 10 kN�
Figure 11. Analysed structure and considered loads
(b)
(a)
Figure 12. Structural model before (a) and after (b) theconnections become effective: (a) pinned connections;(b) semi-rigid connections
48 Structures and Buildings 163 Issue SB1 Analysis of a semi-rigid connection for precast concrete El Debs et al.
(c) continuity reinforcement and dowel of same diameters and
strength of the prototype
(d ) elasticity modulus of concrete of 30 GPa, which is
approximately the mean value between the precast and the
cast-in-situ concrete.
To calculate the negative and positive moments, safety factors
for the materials (1.4 for concrete and 1.15 for steel) were
introduced.
Table 6 presents the main results obtained for the analysed
situations. Several observations can be made.
(a) The displacement at the top of the structure for semi-rigid
connection is 13.7% of the value considering pinned
connection.
(b) The ªz coefficient is also significantly reduced.
(c) The bending moment in the column base for semi-rigid
connection is 41.9% of the value considering pinned
connection for the load combination G + Q +W.
(d ) For the analysed parameters, the positive moment at the
connection occurs only for load combination G + G and its
value is low.
Based on the results presented, other similar framed structures
were simulated, with an increasing number of storeys. Vertical
horizontal loads were repeated for the intermediate storeys,
keeping the same load for the top level. Table 7 shows the
results.
Based on the results in Table 7, two primary conclusions can
be drawn. First, it is possible to progress from a two-storey
frame with pinned connections to a four-storey one with semi-
rigid connections. The displacement at the top would be lower,
and the bending moment in the column base would increase
slightly, from 44.65 to 49.29 kNm. Second, even for a five-
storey frame, the positive moments at the connection would be
lower than the yielding moments, which indicate the possibility
of another increase in the height; however, in this case, there
would be a large increase in the column base moments.
Interior column Exterior column
Negative moments Positive moments Negative moments Positive moments
Stiffness: MN/rad 63.5 5.4 26.4 5.4Yielding moment: kN m 147.5 24.2 70.2 24.2
Table 5. Design values of connection stiffness and yielding moments
Loads G + Q + W Loads G + W
Connections a*: mm ªz Mb.ªz†:
kN mMv.ªz
‡:kN m
a*: mm ªz Mb.ªz†:
kN mMv.ªz
‡:kN m
Pinned 29.77 1.19 44.65 0 29.77 1.12 42.02 0Semi-rigid§ 4.07 1.03 18.72 — 4.07 1.02 18.54 3.99Rigid§ 1.99 1.01 15.27 — 1.99 1.01 15.04 15.00
* a – displacement at the structure top level (average of four columns)† Mb – moment at the column bottom (average of four columns)‡ Mv – positive moment at beam–column connection, (—) means that there is only negative moment§ Dead load actuates on pinned connections
Table 6. Main results for the analysed alternatives
Load G + Q + W Load G + W
Connection n* a:mm
ªz Mb.ªz:kN m
Mv.ªz:kN m
a: mm ªz Mb.ªz:kN m
Mv.ªz:kN m
Pinned 2 29.77 1.19 44.65 0 29.77 1.12 42.02 0Semi-rigid 2 4.07 1.03 18.73 — 4.07 1.02 18.54 3.99
3 11.30 1.05 33.94 — 11.30 1.03 33.29 8.52
4 21.81 1.07 49.29 0.06 21.81 1.05 48.37 13.13
5 36.30 1.10 66.26 4.75 36.30 1.06 63.85 17.00
* n – number of storeys
Table 7. Results when the number of storeys is increased for semi-rigid connections
Structures and Buildings 163 Issue SB1 Analysis of a semi-rigid connection for precast concrete El Debs et al. 49
Because of the uncertainty in the calculation of connection
stiffness, a numerical simulation was performed to analyse the
effect of this value on the studied parameters. Therefore, three
different stiffnesses were considered: 0.5, 1.0 and 2.0 times the
calculated values presented in Table 5, while maintaining the
same resistance moments. Table 8 presents the results of this
analysis for the load combination G + Q +W.
The results indicate that the variation of connection stiffness
becomes more important when the structure turns more
deformable as the storey number increases. It is also possible to
observe that the connection stiffness affects the displacements
more than the column base bending moments. For the case of
the four-storey structure, when the connection stiffness
decreases 50%, the displacement at the structure top level
increases 36.8%, while the bending moment in the column base
only increases 14.4%. The results show that the displacements
at the top of the structure and the column base moments
present low susceptibility to deviations of this parameter.
6. CONCLUSIONS
Based on the results obtained, some conclusions can be drawn.
(a) The proposed design models can reasonably evaluate the
studied connection strength for both negative and positive
moments.
(b) The evaluation of strength is more accurate than that of
stiffness.
(c) In the example of a typical structure, it is possible to
increase the number of storeys of the structure from two to
four with lower horizontal displacement at the top and
only a small increase of the column base bending moment
by using semi-rigid connections.
(d ) Although there is significant uncertainty in the connection
stiffness, the results show that the displacements at the top
of the structure and the column base moments present low
susceptibility to deviations of this parameter. In the
example of a typical four-storey structure, the moment in
the column base increases only 14.4% when the connection
stiffness is reduced to half the calculated stiffness.
ACKNOWLEDGEMENTS
The authors would like to acknowledge the Brazilian
government agencies FAPESP and CNPq for the scholarship
and financial support given to this research.
REFERENCES
Barboza ASR (2002) Behaviour of mortar joints subjected to
compression in precast concrete connections. PhD thesis,
University of Sao Paulo, Sao Carlos. (In Portuguese:
Comportamento de juntas de argamassa solicitadas a
compressao na ligacao entre elementos pre-moldados.)
Bljuger FE (1988) Design of precast concrete structures. Wiley,
New York.
CEB–FIP (Comite Euro-International du Beton – Federation
Internationale de la Precontrainte) (1991) CEB–FIP model
code 1990, Bulletin d’Information, No. 203-205. CEB–FIP,
Lausanne.
Chefdebien A and Daldare J (1994) Experimental investigations
on current connections between precast concrete
components. Proceedings of the 2nd Workshop on Semi-rigid
Behaviour of Civil Engineering Structural Connections,
Prague, 21–30.
El Debs MK (2000) Precast concrete: fundaments and
application. EESC/USP, Sao Carlos. (In Portuguese: Concreto
pre-moldado: fundamentos e aplicacoes.)
El Debs MK, Barboza ASR and Miotto AMM (2003)
Development of material to be used as bearing pad in precast
concrete connections. Structural Concrete 4(4): 185–193.
El Debs MK, El Debs ALHC and Miotto AM (2006a)
Experimental analysis of beam-to-column connection with
semi-rigid behaviour of precast concrete structures.
Proceedings of the 2nd International Congress of Federation
Internationale du Beton, Naples, 1–11.
El Debs MK, Montedor LG and Hanai JB (2006b) Compression
tests of cement-composite bearing pads for precast concrete
connections. Cement and Concrete Composites 28(7): 621–
629.
Elliott KS, Davies G, Mahdi A, Gorgun H, Virdi K and
Ragupathy P (1998) Precast concrete semi-rigid beam-to-
column connections in skeletal frames. Proceedings of the
International Conference, Control of Semi-rigid Behaviour of
Civil Engineering Structural Connections, Liege, pp. 45–54.
Engstrom B (1992a) Anchorage of ribbed bars in the post yield
stage. Proceedings of Workshop on Semi-rigid Behavior of
Civil Engineering Structural Connections, Strasbourg,
pp. 65–76.
Engstrom B (1992b) Combined effects of dowel action and
friction in bolted connections. Proceedings of the 1st
Workshop Semi-rigid Behaviour of Civil Engineering
Structural Connections, Strasbourg, 77–98.
FIB (Federation Internationale du Beton) (1999) Structural
a: mm ªz Mb.ªz: kN m
n* 0.5 1.0 2.0 0.5 1.0 2.0 0.5 1.0 2.0
2 5.20 4.07 3.33 1.03 1.03 1.02 20.40 17.44(27.6%) (�18.2%) (9.3%) 18.65 (�6.5%)
3 15.21 11.30 8.75 1.06 1.05 1.04 38.18 30.76(34.6%) (�22.6%) (12.8%) 33.86 (�9.2%)
4 29.82 21.81 16.89 1.10 1.07 1.06 56.42 49.34 44.72(36.8%) (�22.6%) (14.4%) (�9.4%)
5 50.44 36.30 27.78 1.14 1.10 1.07 77.08 66.11 59.22(39.0%) (�23.5%) (16.6%) (�10.4%)
* n – number of storeys
Table 8. Analysis for G + Q + W loads using half and twice times the stiffness
50 Structures and Buildings 163 Issue SB1 Analysis of a semi-rigid connection for precast concrete El Debs et al.
concrete. The textbook on behaviour design and
performance. Fib Bulletin, vol. 1-3, fib, Lausanne.
Franco M and Vasconcelos AC (1991) Practical assessment of
second order effects in tall buildings. Proceedings of the
Colloquium on the CEB–FIP MC90, Rio de Janeiro, pp. 307–
323.
Lindberg R and Keronen A (1992) Semi-rigid behaviour of a RC
portal frame. Proceedings of the 1st Workshop on Semi-rigid
Behaviour of Civil Engineering Structural Connections,
Strasburg, 53–63.
Miotto AM (2002) Beam-to-column connections of precast
concrete structures: analysis of deformability regarding
bending moment. PhD thesis, University of Sao Paulo, Sao
Carlos. (In Portuguese: Ligacoes viga-pilar de estruturas de
concreto pre-moldado: analise com enfase na
deformabilidade ao momento fletor.)
What do you think?To discuss this paper, please email up to 500 words to the editor at [email protected]. Your contribution will be forwarded to theauthor(s) for a reply and, if considered appropriate by the editorial panel, will be published as discussion in a future issue of thejournal.
Proceedings journals rely entirely on contributions sent in by civil engineering professionals, academics and students. Papers should be2000–5000 words long (briefing papers should be 1000–2000 words long), with adequate illustrations and references. You cansubmit your paper online via www.icevirtuallibrary.com/content/journals, where you will also find detailed author guidelines.
Structures and Buildings 163 Issue SB1 Analysis of a semi-rigid connection for precast concrete El Debs et al. 51