Professional Documents
Culture Documents
Nonlinear FEM analysis of RC beam-column joint strengthened by cast in-situ joint expansion
146
Journal of Advanced Concrete Technology Vol. 12, 146-157, May 2014 / Copyright 2014 Japan Concrete Institute
Scientific paper
doi:10.3151/jact.12. 146
Abstract
Part-1 of this study presents the experimental results of beam-column joints in soft-first story buildings, in which the
first-story column depth is twice that of the upper stories. In Part-2, strut-and-tie models (STMs) are developed for the
specimens. The flow of internal forces shown by the developed strut-and-tie models agreed with the observed cracks.
The developed models for the joints are different from those for the usual joint: the main difference is that, in the
specimens large struts were extended into the wall panel. STMs are also developed for the total frame to understand the
overall equilibrium. Based on the STMs, new design equations are proposed, where the joint strength is evaluated as the
sum of the flexural strengths of the beam and the second-story column. In I-type joint, the effect of the wall panel and
beam stirrups is also considered. In O-type joint, the effect of the hoops in the joint is also considered.
1. Introduction
Specimen O-1
Specimen I-1
Analytical Strength
Maximum Strength
400
Closing
436 kN
436 kN
Opening
Specimen I-1 Specimen O-1
-400
-800
-950 kN
-1200
-40
147
S. Halim, S. Takahashi, T. Ichinose, M. Teshigawara, T. Kamiya and H. Fukuyama / Journal of Advanced Concrete Technology Vol. 12, 146-157, 2014
Wstrut =
Cstrut
0.85 f c ' b
(1)
S. Halim, S. Takahashi, T. Ichinose, M. Teshigawara, T. Kamiya and H. Fukuyama / Journal of Advanced Concrete Technology Vol. 12, 146-157, 2014
struct the STM are the same as before and the reinforcement details are similar to those of the test specimens. The biggest difference is that the wall panel and
the second-story column are removed.
The strength of the knee joint is approximately threefourth of that of the test specimen. The capacity of the
knee joint is governed by beam flexural capacity. The
difference of the capacities in the opening load is attributable to the difference of the location of node C. Note
that strut BC in Fig. 2a is less inclined than that in Fig.
4. This leads to a larger compressive force in strut AB
and a larger shear force in the column. Moreover, note
that the inclination of strut CD in Fig. 2a is opposite to
that of strut AB, indicating that the sign of the shear
force in the second-story column is opposite to that of
the first-story column. The cracks observed in the test
also confirm this tendency. The negative shear force is
compensated by strut HE in the wall panel (Fig. 2a).
148
S. Halim, S. Takahashi, T. Ichinose, M. Teshigawara, T. Kamiya and H. Fukuyama / Journal of Advanced Concrete Technology Vol. 12, 146-157, 2014
149
and TAD in Fig. 7d) resist the tension. The beam top bars
(TCH) also resist the tension.
(3') The locations of nodes B, I, F, and H are determined so that the strength is maximized.
S. Halim, S. Takahashi, T. Ichinose, M. Teshigawara, T. Kamiya and H. Fukuyama / Journal of Advanced Concrete Technology Vol. 12, 146-157, 2014
150
first-story column main bars does not resist the compression force because it terminates in the joint panel.
Figure 9b shows the STM with compression zones. The
strut distribution agrees with the crack patterns observed
in the test. The computed forces in the ties also agree
with the observed strains (e.g., the beam top bars as well
as the first-story column main bars yielded as was observed). The strength based on STM is 80% of the observed strength (Fig. 3). The STM as well as the observed cracks show that a large shear force is transferred
to the wall panel in the second story. As concluded in
paper-1 the additional beam top bars provided in I-1t
prevented the compressive failure of the wall panel.
This phenomenon can also be discussed in term of the
strut and tie model (Fig. 9b). Increasing tensile force in
tie TCH would result in larger forces in strut IC and IH,
and smaller compressive force in strut IE. Similarly,
considerable amount of vertical force can be resisted by
the beam stirrups making a truss model along the beam
length which is similar to Fig. 5a but upside down. The
truss is not shown in Fig. 9b because the effect in specimen I-1 is much smaller than those of the other struts.
Figure 10 shows the STM for the closing load of the
exterior joint. The procedure and assumptions are similar to those for Specimen I-1. The strength of the exterior joint (110kN) is approximately 15% of the specimens strength (755 kN). The strength in Fig. 10 is governed by the flexural capacity of the beam. This difference is attributable to the following reasons:
(a) Struts IH and IE in Fig. 9 are much larger than
struts BE and BF in Fig. 10 because they extend the
forces into the wall panel.
(b) The vertical tie BD in Fig. 9 is located on the right
of the strut CD, producing a bending moment opposite to that of the usual joint (Fig. 10).
S. Halim, S. Takahashi, T. Ichinose, M. Teshigawara, T. Kamiya and H. Fukuyama / Journal of Advanced Concrete Technology Vol. 12, 146-157, 2014
8. Simplified equations
On the basis of the observed failure modes and the results of the strut-and-tie analysis, simplified equations
are proposed. The equations are exclusively utilizing
conventional flexural analyses.
The observed failure modes can be classified into two
groups: (1) the failure associated with horizontal cracks
151
(2)
Qc =
M c1
L + d co
(3)
S. Halim, S. Takahashi, T. Ichinose, M. Teshigawara, T. Kamiya and H. Fukuyama / Journal of Advanced Concrete Technology Vol. 12, 146-157, 2014
Deff
152
Dc1
Dc1
dco
Tv
Cv
dco
Cv
Tv
This distance is
almost dco
Qc
Qc
Qc
Tv
Cv
was prominent in Specimen O-1, where the hoop reinforcement ratio in the joint was 0.17%; the inclined
cracks propagated toward the point of anchorage (Fig.
16b in Part-1). In Specimen O-1t, where the hoop reinforcement ratio was 0.80%, the hoop reinforcement in
the joint reduced the opening of such cracks (Fig. 21b
in Part-1). Figure 13a shows the assumed strut created
by the contribution of the hoops in the joint, where
awy represents the horizontal force of the hoops in the
joint and Cout represents the compressive force acting
outside the point of anchorage. This assumption leads to
Cout =
Db
aw y
2(lb Dc 2 )
(4)
(5)
Db
aw y Dc1 (6)
2 f c ' b(lb Dc 2 )
In case Dc1 equals to Dc2, the strut outside the anchorage point would be straight as shown in Fig. 13b and
therefore Deff would be equal to Dc1.
Computing the moment capacity of the first-story column, Mc1, the moment due to eccentricity of the axial
force should also be considered because the center of
the effective section does not coincide with that of the
full section of the first-story column. Therefore,
M c1 = M ef f N i
Dc1 Def f
2
(7)
S. Halim, S. Takahashi, T. Ichinose, M. Teshigawara, T. Kamiya and H. Fukuyama / Journal of Advanced Concrete Technology Vol. 12, 146-157, 2014
in Fig. 14a. Ch and Cv in Fig. 14a represent the horizontal and vertical components of the strut acting on nodes
C, respectively. Th and Tv represent the tensile forces of
the beam and column reinforcement, respectively. Thj
represents the tensile forces of the hoops in the joint
assuming that the crack in the joint penetrates till about
1/3 of the beam depth (height of the joint panel).
Thj = Pj y bc
Db
3
(8)
Qj =
Mj
(9)
L + Db / 2
(10)
Ch
Cv
Cv
CC
Tv
Ch
.Db
Dc2
Ch
Cv
C
Db
Thj
Th
CI
Th
Cv
Tv
I
Ch
.Db
Qj
Qj
Db
2
(11)
Db
) (Th Thj )
2
(12)
M b = Db + d b
2
(Th + Thj ) Db Q j
(13)
Th Db Q j
(14)
b
B L+ 2
Dc1 Dc 2
2
M b = Db + d b
2
Thj
C
M c2 = M c 2 N i
153
O-Type
I-Type
Opening
0.4
Closing
0.2
0.3
154
S. Halim, S. Takahashi, T. Ichinose, M. Teshigawara, T. Kamiya and H. Fukuyama / Journal of Advanced Concrete Technology Vol. 12, 146-157, 2014
Dc2
Cv
.Db
Cv
Tv C v
Ch Tw
Tst1
Ch
Th
45
Db d
C
Tst2
Db
2
L+
Db
2
Ch
.Db
Qj
Tv
Th
0.4Db
Qj
(15)
Dc1 + Dc 2
+ Tst1 lst1 + Tw lw
2
(16)
panel concrete were prominent (Fig. 8 of Part-1). Therefore, the critical section is assumed as shown in Fig.
15b. Ch and Cv in Fig. 15b represent the horizontal and
vertical components of the strut acting on node I, respectively. For simplicity, Ch is assumed to act at the
beam section and Cv is assumed to act at the T-section
composed of the wall panel and the second-story column. The length of the wall panel L is assumed as the
sum of the length above the joint (Dc1-Dc2) and 80% of
the beam depth.
L = ( Dc1 Dc 2 ) + 0.8Db
(17)
lst1=0.5ld dc+0.9Dc1
Tw
Cv
Tst1
ld
45
dc
ld 2dcr
dc
S. Halim, S. Takahashi, T. Ichinose, M. Teshigawara, T. Kamiya and H. Fukuyama / Journal of Advanced Concrete Technology Vol. 12, 146-157, 2014
155
Shear Force
Moment
Qc
(Eq. 3)
M c1
Qj
(Eqs. 9-10)
M c2
Mb
Joint Type
O-Type
I-Type
O-Type
I-Type
O-Type
I-Type
Loading Cases
Opening
Closing
Eqs. 4-7 and Flexural Analysis
with Deff
Flexural Analysis with Full Section
Eq. 11 and Flexural Analysis of Rectangular Section
Eqs. 15-16 and Flexural Analysis
Eqs. 17-19 and Flexural Analysis
of Rectangular Section
of T-Section
Eq. 13 + Table 1
Eq. 14 + Table 1
(18)
Dc 2 b ( Dc1 Dc 2 ) + L2 t
2( Dc 2 b + L t )
(19)
moment capacity of the rectangular section of the second-story column (Fig. 14). The second story moment
(Mc2') of I-type joints for the opening load should be
calculated from Eqs. 15-16. Mc2 in Eq. 16 is the moment
capacity of the second-story column with rectangular
section (Fig. 15a). For the closing load of I-type joints,
Eqs. 17-19 should be used to find Mc2'. Mc2 in Eq. 18
should be computed by flexural analysis of the T-section
(Fig. 15b).
The beam bending moment Mb is defined by Eq. 13.
The values are given in Table 1 for different joint
types and load cases. Because Eq. 13 includes the joint
strength (Qj), Qj is computed with an iteration method.
Figure 17 shows the comparison between the observed and calculated strengths. Qexp in the vertical axis
is the observed strength, while Qj in the horizontal axis
is the computed joint strength. The denominator Qc is
the computed column strength. Assume that the proposed equations are perfect (Qu= Qexp). Then Eq. 2 leads
to the following:
If Qj<Qc, Qexp= Qj, which is represented by the inclined red line connecting (0, 0) and (1, 1) in Fig. 17.
If Qj>Qc, Qexp= Qc, which is represented by the horizontal black line connecting (1, 1) and (2, 1) in Fig. 17.
These lines constitute the border between the green
and yellow zones. A plot in the green zone indicates that
the observed strength is larger than the expected value
and vice versa. Because most plots are located around
the border, we conclude that the proposed design equations appropriately evaluate the strengths of the speci-
1.5
1.0
I-1
0.5
Hanai-1
Hanai-2
Hanai-3
Joint Column
Load Opening
Case Closing
156
S. Halim, S. Takahashi, T. Ichinose, M. Teshigawara, T. Kamiya and H. Fukuyama / Journal of Advanced Concrete Technology Vol. 12, 146-157, 2014
Qtest
( N / mm2 )
Ac1
-10%
O-1
I-1
I-1t
O-1t
9. Conclusions
(1) The STM for a prototype frame validates the
specimens constructed in this study. In other words,
the specimens (part of a structural frame) represent
the structural behavior of beam-column joint and
connecting elements appropriately. The model also
implies that the vertical wall reinforcement plays an
important role in a wall with an open frame.
(2) The comparison between STMs for test specimens
and those for usual exterior and knee joints shows
that the flow of the forces of the specimens is dif-
I-1t
O-1 I-1O-1t
Hanai-2 Hanai-1
Hanai-3
Hinge
Qcomputed
Ac1
dco
Column
end speing
Elastic
( N / mm2 )
Joint
rotational
spring
Rigid
b) Joint details
Mj
Mc1
Mc1
O
y
(rad/m)
O
cy
(rad)
(rad)
S. Halim, S. Takahashi, T. Ichinose, M. Teshigawara, T. Kamiya and H. Fukuyama / Journal of Advanced Concrete Technology Vol. 12, 146-157, 2014
ferent from that of the usual joints. All the specimens were stronger than the usual joints and a huge
amount of shear force was transferred to the wall
panel.
(3) In the case of the opening load of O-type joint, the
effective depth of the first-story column is defined
by the embedment length of the beam bottom bars
and shear reinforcement in the joint panel.
(4) The strength related to the beam failure can be approximated as the sum of the beam and secondstory column strengths including the effects of the
wall panel, hoops in the joint, and stirrups in the
beam.
(5) The strength and failure modes predicted by the
proposed procedure (Table 2) are in good agreement with the test results.
References
Giberson, M. F., (1974). Two nonlinear beams with
definition of ductility. Journal of the Structural
Division, ASCE, 95(7), 137-157.
Hanai, N., Goto, Y. and Ichinose, T., (2009).
Anchorage of column re-bars where column depth
greatly reduces and structural wall appears. AIJ J.
Technol. Des., 15(29), 143-146. (in Japanese)
Ichinose, T., Kawai, T., Takahashi, S. and Teshigawara,
M., (2013). Second stiffness of beam-column joint
in soft-first story RC building. Proceedings of the
Japan Concrete Institute, 35(2), 265-270. (in
Japanese)
Kabeyasawa, T., Shiohara, H., Otani, S. and Aoyama, H.,
(1983). Analysis of the full-scale seven-story
reinforced concrete test structure. Journal of the
Faculty of Engineering, University of Tokyo, 36(2),
432-478.
Liang, Q., (2005). Performance-based optimization of
157