Soil Dynamics and Earthquake Engineering 113 (2018) 136–147

Contents lists available at ScienceDirect

Soil Dynamics and Earthquake Engineering

journal homepage: www.elsevier.com/locate/soildyn

The influence of moment resisting beam-to-column connections on seismic T

behavior of chevron concentrically braced frames
⁎
Silvia Costanzo, Mario D’Aniello , Raffaele Landolfo
Department of Structures for Engineering and Architecture, University of Naples “Federico II”, Via Forno Vecchio 36, 80134 Naples, Italy

A R T I C LE I N FO A B S T R A C T

Keywords: Chevron concentrically braced frames (C-CBFs) are very popular lateral-resisting systems, often opted for steel
Chevron concentrically braced frames buildings due to their improved architectural functionality and reduced cost of fabrication and erection as X
Eurocode 8 bracings. According to EN 1998-1 these systems are expected to provide limited ductility, thus penalizing the
Bracings design of C-CBFs with smaller values of the behavior factor (namely q = 2 and q = 2.5 for medium “DCM” and
Ductility
high ductility class “DCH”, respectively). To improve the ductility and the energy dissipation capacity of chevron
Seismic design
Beam-to-column joints
bracings, the influence of the secondary frame effect provided by moment resisting beam-to-column connections
belonging to the braced bays is investigated in this paper. To this aim, parametric non-linear analyses are carried
out on a set of structural archetypes representative of low, medium and high-rise multi-story buildings. The
results show that fully restrained joints can have beneficial effects providing an additional reserve of strength,
stiffness and ductility.

1. Introduction
Chevron concentrically braced frames (C-CBFs), also known as in-
verted-V bracings, are very popular lateral-resisting systems, widely
opted for steel buildings in seismic areas. Inverted V bracings are often
preferred to X-CBFs since the location of openings (doors, windows) is
easier and it requires fewer connections, as well as less material for the
braces [1–3]. Moreover, for the values of interstory height and span
length commonly used in low and medium-rise steel buildings, the
chevron configuration guarantees appropriate slopes of the diagonal
members (i.e. in the range 30° ÷ 60°, as shown by [4]) that are suitable
for the proper design of gusset plate connections.
Chevron bracings provide large lateral stiffness, which allows easily
meeting both the drift limitations and the stability requirements. On the
other hand, as widely observed in previous studies [5–15] such struc-
tural systems may exhibit relatively poor inelastic response under se-
vere seismic ground motions.
The seismic performance demand of C-CBFs is strongly influenced
by the type of developed plastic mechanism, which strictly depends on
the flexural behavior of the brace-intersected beam. Indeed, if that
beam experiences flexural yielding following the brace buckling, the
structure can undergo significant loss of strength and stiffness and very
poor energy dissipation capacity [5–7]. To prevent such detrimental
behavior, current seismic codes [16–18] provide capacity design rules
to assure “strong beam mechanism”, which aims at enforcing the
energy dissipation capacity in the braces under tension, while beams,
columns and connections are kept in elastic range. Moreover, as high-
lighted by [5–8], even the beam flexural stiffness, beside its strength,
should be controlled to guarantee an effective seismic response of
chevron bracings. Indeed, large vertical deflection of the beam can
prevent the yielding of brace under tension and impose severe ductility
demand to the compression diagonal, thus leading to a very poor
overall performance due to the brace deterioration.
Furthermore, both numerical and experimental evidences show that
C-CBFs are prone to soft-story mechanisms in the most of cases. Several
Authors [9–11,19–22] observed that the design rules currently codified
in Eurocode 8 are less effective in assuring uniform distribution of
plastic demand along the building height.
In the framework of EN 1998-1, chevron concentrically braced
frames are expected to provide limited ductility and smaller behavior
factors are recommended (i.e. q = 2 and q = 2.5 for medium “DCM”
and high ductility class “DCH”, respectively) as respect to X-CBFs (i.e. q
= 4 for both DCM and DCH). On the contrary North-American seismic
codes do not adopt such distinction [17,18], since these codes stipulate
that the ductility class solely depends on the design requirements, and
the behavior factor is assumed the same for both configurations.
Several researchers [7,8,21,23–28] proposed and investigated al-
ternative design criteria to improve the seismic response of chevron
bracings under severe ground motions. The outcomes of a recent
comprehensive numerical parametric study [11] showed that the

⁎
Corresponding author.
E-mail addresses: silvia.costanzo@unina.it (S. Costanzo), mdaniel@unina.it (M. D’Aniello), landolfo@unina.it (R. Landolfo).

https://doi.org/10.1016/j.soildyn.2018.06.001
Received 6 April 2018; Received in revised form 19 May 2018; Accepted 2 June 2018
Available online 14 June 2018
0267-7261/ © 2018 Elsevier Ltd. All rights reserved.
S. Costanzo et al. Soil Dynamics and Earthquake Engineering 113 (2018) 136–147

design rules given by North American codes [17,18] are more effective
than those recommended by the current version of EC8, which leads to
the poorest energy dissipation capacity. In the light of such results,
Costanzo et al. [23] proposed alternative design criteria to improve the
seismic performance of medium ductility designed chevron bracings.
Further recent research findings concern the influence of the con-
tribution to lateral strength and stiffness of the secondary action frame
provided by the gusset plates and beam-to-column connections of the
braced bays for both high [29–31] and moderate [32–35] seismicity.
In the US practice, moment-resisting (either full or semi rigid)
beam-to-column joints into the braced bays are adopted for Special
Concentrically braced frames SCBFs (namely for high ductility class).
Uriz and Main [36] investigated by full scale test the contribution of
beam-to-column gusset plate connections to the lateral load response of
2-story chevron special concentrically braced frame designed according
to AISC341-10 [37]. Tests results showed that significant contribution
was provided by the beam-to-column connections which resisted about Fig. 1. Plan layout of the examined structural archetypes.
30% of the peak lateral load after the buckling of braces.
At current stage, Eurocode 8 does not specifically account for the
bracing members at the i-th and roof level; NEd,br,E,i and NEd,br,E,rf are
secondary frame contribution and does not provide any requirement for
the axial force at the i-th and roof level due to the seismic action;
beam-to-column joints belonging to the braced bays. On the contrary,
NEd,br,G,i and NEd,br,G,rf are the axial force at the i-th and roof story
Japanese Building Code [38,39] assigns the behavior factor even de-
due to the non-seismic actions included in the combination of ac-
pending on the contribution of the secondary frame action: the larger is
tions for the seismic design situation. The condition expressed by Eq.
such contribution and the larger is the expected dissipation capacity.
(2) aims at keeping the bracings at the roof story in the elastic range
In the light of these considerations, the research presented in this
to limit the damage concentration at the upper stories. Thus, the top
paper is addressed to investigate the possibility to furtherly improve the
story behaves similarly to an outrigger beam that modifies the shape
Eurocode 8 rules for chevron concentrically braced frames, by ac-
of lateral displacements from cantilever-like to shear-type. The re-
counting for the extra strength and stiffness, as well as the ductility
quirement expressed by Eq. (2) is disregarded in low rise frames,
reserve, given by the secondary frame action, provided that beam-to-
since stocky and short structures do not generally exhibit cantilever-
column joints belonging to the braced bays are designed to be fully
type behavior. The EN 1998-1 limitation on the non-dimensional
moment restrained.
slenderness of bracing members (which should be less than or equal
With this aim, a set of low, medium and high-rise frames equipped
to 2.0) is retained.
with chevron bracings were alternatively designed assuming either
– To enforce a uniform sequence of buckling of braces and the cor-
pinned or moment-resisting beam-to-column joints into the braced
responding variation of the post-buckling stiffness along the
bays. Both nonlinear static and dynamic analyses were performed to
building height, the following condition is imposed:
evaluate and compare the seismic performance of examined cases.
The paper is organized in two main parts: the first part briefly [(Ωb, i − Ωb)/ Ωb] ≤ 0.25 (3)
summarizes the assumed design criteria, while the results of non-linear
analyses are presented and discussed in the second part. Where Ωb = min(Ωb, i ) = min ( χ ⋅ Npl, br , Rd, i
NEd, br , i ) with i ∈ [1, (n − 1)] and
Ωb,i is capacity-to-demand ratio at the i-th story.
2. Design criteria
The requirement given by Eq. (3) differs from the EC8-compliant
The investigated structures have been designed according to the condition for the story-to-story variation of brace overstrength. Indeed,
criteria recently proposed by Costanzo et al. [23], formerly developed Eq. (3) aims at controlling the sequence of the buckling of the diagonal
with the aim to revise the capacity design rules of the current EN1998-1 members, while according to EN1998-1 the designer should check the
[16]. Furthermore, additional requirements for beam-to-column joints variation of the brace yield strength. However, the brace buckling
belonging to the braced bays are introduced in this study. For the sake under compression is the first nonlinear event. Once occurred, the story
of clarity, the adopted design assumptions are briefly summarized as lateral stiffness drastically decreases, and the seismic demand tends
follows: concentrating in those stories. Hence, as demonstrated by [23], the
compression-based approach to define the capacity-to-demand ratio
– The seismic-induced effects on bracing members are calculated by allows obtaining almost uniform sequence of the buckling of braces and
performing a linear modal analysis considering both diagonals ac- satisfactory distribution of lateral displacements along the building
tive in tension and compression. height, also reducing the tendency to soft-story mechanisms. Moreover,
– Cross-sectional Class 1 according to EN1993:1-1 [40] are selected it is also worth noting that the requirement given by Eq. (3) simplifies
for braces. the design process. Indeed, since the iterations to select the cross sec-
– The design resistances of the braces are assumed to satisfy the fol- tions of diagonal members decreases, being easier to satisfy both
lowing conditions: strength checks and the slenderness requirements of the braces.
Nb, br , Rd, i ≥ NEd, br , i = NEd, br , E , i + NEd, br , G, i at the i−th story (1)
– The non-dissipative members (beams, columns and connections) are
designed to withstand the most unfavorable condition between
Nb, br , Rd, rf ≥ NEd, br , rf = NEd, br , E , rf ⋅q + NEd, br , G, rf at the roof (1) the seismic-induced effects evaluated by means the former
(2)
elastic analysis and magnified by tension overstrength factor

– Where Nb,br,Rd,i and Nb,br,Rd,rf are the factored buckling capacity of

Ωy = min ( Npl, br , Rd, i
NEd, br , i ), without any requirements for its story-to-
story variation;
the braces at the i-th and roof level, that are evaluated according to
(2) the internal forces calculated performing a plastic mechanism
EN 1993:1-1 [40]; NEd,br and NEd,br,rf are the axial force acting in the

137
S. Costanzo et al. Soil Dynamics and Earthquake Engineering 113 (2018) 136–147

Fig. 2. Vertical layout of the examined structural archetypes.

Table 1
Cross-section properties of structural members of three-story cases.
Columns Beams Braces (d × t) Gravity members

Storey q = 4 (pinned) q = 4 (MRJ) q = 6 (MRJ) q = 4 (pinned) q = 4 (MRJ) q = 6 (MRJ) q = 4 (pinned) q = 4 (MRJ) q = 6 (MRJ) Columns Beams
S355 S355 S355 S355 S355 S355 S235 S235 S235 S355 S355

3 HE 220 B HE 220 B HE 220 B HE 550 A HE 340 A IPE 400 168.3 × 4 168.3 × 4 139.7 × 4 HE 200 B IPE 330
2 HE 260 B HE 260 B HE 240 B HE 550 B HE 360 A IPE 400 168.3 × 6 168.3 × 6 168.3 × 4 HE 240 B IPE 330
1 HE 260 B HE 260 B HE 240 B HE 550 M HE 400 B IPE 450 177.8 × 8 177.8 × 8 177.8 × 5 HE 240 B IPE 330

Table 2
Cross-section properties of structural members of six-story cases.
Columns Beams Braces (d × t) Gravity members

Storey q = 4 (pinned) q = 4 (MRJ) q = 6 (MRJ) q = 4 (pinned) q = 4 (MRJ) q = 6 (MRJ) q = 4 (pinned) q = 4 (MRJ) q = 6 (MRJ) Columns Beams
S355 S355 S355 S355 S355 S355 S355 S355 S355 S355 S355

6 HE 300 B HE 300 B HE 260 A HE 500 M HE 450 B IPE 600 219.1 × 8 219.1 × 8 219.1 × 8 HE 260 A IPE 330
5 HE 300 B HE 300 B HE 260 A HE 450 M HE 400 A IPE 550 177.8 × 6 177.8 × 6 139.7 × 8 HE 260 A IPE 330
4 HE 340 B HE 340 B HE 280 B HE 550 M HE 450 A IPE 550 177.8 × 8 177.8 × 8 168.3 × 6 HE 280 B IPE 330
3 HE 340 B HE 340 B HE 280 B HE 550 M HE 500 A IPE 600 177.8 × 10 177.8 × 10 168.3 × 8 HE 280 B IPE 330
2 HE 340 M HE 340 M HE 280 M HE 600 M HE 500 B IPE 600 177.8 × 12 177.8 × 12 168.3 × 8 HE 280M IPE 330
1 HE 340 M HE 340 M HE 280 M HE 600 M HE 500 B IPE 600 219.1 × 8 219.1 × 8 193.7 × 8 HE 280M IPE 330

Table 3
Cross-section properties of structural members of twelve-story two braced bays cases.
Columns Beams* Braces (d × t)

Storey q = 4 (pinned) q = 4 (MRJ) q = 6 (MRJ) q = 4 (pinned) q = 4 (MRJ) q = 6 (MRJ) q = 4 (pinned) q = 4 (MRJ) q = 6 (MRJ)
S355 S355 S355 S355 S355 S355 S355 S355 S355

12 HE 300 B HE 300 B HE 240 B HE 600 B HE 450 A HE 400 A 168.3 × 8 168.3 × 8 168.3 × 6

11 HE 300 B HE 300 B HE 240 B HE 500 B HE 360 A HE 340 A 139.7 × 6 139.7 × 6 114.3 × 6.3
10 HE 300 B HE 300 B HE 240 B HE 550 B HE 400 A HE 340 A 168.3 × 6 168.3 × 6 139.7 × 5
9 HE 400 B HE 400 B HE 340 B HE 550 B HE 400 A HE 360 A 168.3 × 6.3 168.3 × 6.3 139.7 × 6
8 HE 400 B HE 400 B HE 340 B HE 600 B HE 450 A HE 400 A 168.3 × 8 168.3 × 8 139.7 × 8
7 HE 400 B HE 400 B HE 340 B HE 600 B HE 450 A HE 400 A 177.8 × 8 177.8 × 8 139.7 × 8
6 HE 400 M HE 400 M HE 360 M HE 600 B HE 500 A HE 400 A 193.7 × 8 193.7 × 8 168.3 × 6
5 HE 400 M HE 400 M HE 360 M HE 600 B HE 500 A HE 400 A 193.7 × 8 193.7 × 8 168.3 × 6
4 HE 400 M HE 400 M HE 360 M HE 600 B HE 500 A HE 400 A 193.7 × 8 193.7 × 8 168.3 × 6
3 HD400 × 347 HD400 × 347 HE 450 M HE 600 B HE 500 A HE 400 A 193.7 × 8 193.7 × 8 168.3 × 6
2 HD400 × 347 HD400 × 347 HE 450 M HE 600 M HE 500 B HE 400 A 193.7 × 10 193.7 × 10 168.3 × 6.3
1 HD400 × 347 HD400 × 347 HE 450 M HE 600 M HE 500 B HE 500 A 219.1 × 8 219.1 × 8 168.3 × 8

* All gravity resistant beams are IPE 330.

analysis, namely considering a free-body distribution of plastic detail the brace under tension is assumed attaining its full plastic
forces transmitted by the braces yielded under tension and those strength (namely given by γovNpl,br,Rd), while the brace under
under compression behaving in the post-buckling range. In compression attains its average post-buckling strength

138
S. Costanzo et al. Soil Dynamics and Earthquake Engineering 113 (2018) 136–147

Table 4
Cross-section properties of structural members of twelve-story one braced bay cases.
Columns Beams* Braces (d × t)

Storey Gravity q = 4 (pinned) q = 4 (MRJ) q = 6 (MRJ) q = 4 (pinned) q = 4 (MRJ) q = 6 (MRJ) q = 4 (pinned) q = 4 (MRJ) q = 6 (MRJ)
S355 S355 S355 S355 S355 S355 S355 S355 S355 S355

12 HE 200 B HE 400 B HE 400 B HE 320 A HE 600 M HE 450 B HE 500 A 177.8 × 10 177.8 × 10 177.8 × 10
11 HE 200 B HE 400 B HE 400 B HE 320 A HE 550 B HE 400 A HE 360 A 168.3 × 6 168.3 × 6 139.7 × 6.3
10 HE 200 B HE 400 B HE 400 B HE 320 A HE 500 M HE 450 A HE 400 A 168.3 × 8 168.3 × 8 168.3 × 6
9 HE 220 B HE 400 M HE 400 M HE 360 B HE 550 M HE 500 A HE 450 A 193.7 × 8 193.7 × 8 168.3 × 8
8 HE 220 B HE 400 M HE 400 M HE 360 B HE 550 M HE 500 A HE 450 A 193.7 × 8 193.7 × 8 168.3 × 8
7 HE 220 B HE 400 M HE 400 M HE 360 B HE 550 M HE 500 A HE 450 A 219.1 × 8 219.1 × 8 177.8 × 8
6 HE 260 B HD400 × 347 HD400 × 347 HE 360 M HE 550 M HE 500 A HE 500 A 219.1 × 8 219.1 × 8 177.8 × 10
5 HE 260 B HD400 × 347 HD400 × 347 HE 360 M HE 600 M HE 500 B HE 500 A 244.5 × 8 244.5 × 8 177.8 × 10
4 HE 260 B HD400 × 347 HD400 × 347 HE 360 M HE 600 M HE 500 B HE 550 A 244.5 × 8 244.5 × 8 177.8 × 12
3 HE 300 B HD400 × 421 HD400 × 421 HD400 × 347 HE 600 M HE 500 B HE 550 A 244.5 × 8 244.5 × 8 177.8 × 12
2 HE 300 B HD400 × 421 HD400 × 421 HD400 × 347 HE 600 M HE 500 B HE 550 A 244.5 × 8 244.5 × 8 177.8 × 12
1 HE 300 B HD400 × 421 HD400 × 421 HD400 × 347 HE 700 M HE 550 B HE 550 A 244.5 × 10 244.5 × 10 219.1 × 8

* All gravity resistant beams are IPE 330.

where kb is the flexural stiffness of the beam accounting for the

deformability of supports and kbr is the vertical rigidity of the bra-
cing-system (more details can be found in [7,23]).
– To enhance the redundancy and the ductility of the system, addi-
tional requirement is imposed for the beam-to-column connections
belonging to the braced bays, which should be fully restrained
moment-resisting type (namely full strength and full rigid).

3. Parametric study

A comprehensive set of nonlinear static and dynamic analyses was

performed on low, medium and high rise residential buildings equipped
with bracings in chevron configuration.
The examined cases consist of three, six and twelve-story planar
Fig. 3. Comparison between selected ground motion and EC8-compliant spec- frames, which were extracted from the perimeter in Y direction of a
trum. reference residential building, whose structural layout is shown in plan
and vertical configuration in Figs. 1 and 2, respectively. As shown in
Npb = γov⋅0.3⋅χ ⋅Npl, br , Rd ; the steel randomness coefficient γov is Fig. 2, two different configurations were considered for the 12-story
assumed equal to 1.25; cases, namely one equipped with single braced bay and another with
two braced bays per story. Indeed, differently from low and medium-
– The flexural stiffness of the brace-intersected beam complies with rise buildings, the seismic performance of high-rise frames is more in-
the following requirement fluenced by the higher modes of vibration. Thereby, the double braced
bay configuration was considered to investigate the influence of the
increase of lateral stiffness and redundancy on the response of 12-story
kb frames.
KF = ≥ 0.2
kbr (4) The span length is equal to 8 m, while the interstory height is equal

Table 5
Data of the selected ground motions.
Earthquake name Date Station Name Station Country Magnitude Mw Fault mechanism

Alkion 24.02.1981 Xylokastro-O.T.E. Greece 6.6 Normal

Montenegro 24.05.1979 Bar-Skupstina Montenegro 6.2 Reverse
Opstine
Izmit 13.09.1999 Yarimca (Eri) Turkey 5.8 Strike-Slip
Izmit 13.09.1999 Usgs Golden Turkey 5.8 Strike-Slip
Station Kor
Faial 09.07.1998 Horta Portugal 6.1 Strike-Slip
L'Aquila 06.04.2009 L'Aquila - V. Italy 6.3 Normal
Aterno - Aquila
Park In
Aigion 15.06.1995 Aigio-OTE Greece 6.5 Normal
Alkion 24.02.1981 Korinthos-OTE Building Greece 6.6 Normal
Umbria-Marche 26.09.1997 Castelnuovo-Assisi Italy 6.0 Normal
Izmit 17.08.1999 Heybeliada-Senatoryum Turkey 7.4 Strike-Slip
Izmit 17.08.1999 Istanbul-Zeytinburnu Turkey 7.4 Strike-Slip
Ishakli 03.02.2002 Afyon-Bayindirlik ve Iskan Turkey 5.8 Normal
Olfus 29.05.2008 Ljosafoss-Hydroelectric Power Iceland 6.3 Strike-Slip
Olfus 29.05.2008 Selfoss-City Hall Iceland 6.3 Strike-Slip

139
S. Costanzo et al.
Fig. 4. Pushover response curves for the examined cases.
Fig. 5. Plastic redistribution parameters obtained from pushover analyses.
to 3.5 m, except for the first floor that is 4.0 m high. At each floor, the
rigid diaphragm transmitting the horizontal actions is obtained by
means of composite slabs with profiled steel sheetings supported by the
hot rolled steel beams (primary and secondary), which are restrained to
avoid flexural-torsional buckling.
The structural design against gravity loads was carried out ac-
cording to the non-seismic Eurocodes [40–43], without reference to a
specific National Annex. Hence, the values of the safety factors re-
commended by Eurocode were used. Permanent loads (Gk) and live
loads (Qk) were assumed equal to 5.20 kN/m2 and 2.00 kN/m2, at each
story. The inertial effects in the seismic design situation was evaluated
according to EN 1998-1 [16].
A reference peak ground acceleration equal to agR = 0.35g (being g
the gravity acceleration), a type C soil, a type 1 spectral shape were
considered. The importance factor γI was assumed equal to 1.0.
In addition, the following parameters were varied:
140
Soil Dynamics and Earthquake Engineering 113 (2018) 136–147
– The degree of restraint of the beam-to-column joints belonging to
the braced bays: ideal full pinned connections with rotational stiff-
ness Sj = 0 and full fixed moment-resisting joints (MRJs) were al-
ternatively considered.
– Two different values of the behavior factor were considered, namely
q = 4 (consistently to the design criteria developed and validated in
[23]) and q = 6.
Tables 1–4 report the geometrical and mechanical features of the
structural members obtained for the designed structures. Cold formed
circular hollow profiles were used for the diagonals members; IPE and
HE profiles were used for the beams, while HE and HD profiles for the
columns. All steel members have cross section meeting Class 1 re-
quirements according to EN 1993:1-1 [40]. It is trivial to observe that
using moment-resisting joints, allows selecting smaller beam profiles
even at the same behavior factor. This feature can be explained con-
sidering that the design of brace-intersected beams is mainly ruled by
the requirement on flexural stiffness expressed by Eq. (4). Indeed, full
fixed restrains at both ends of the brace-intersected beam significantly
increase the beam flexural stiffness, i.e. four times larger respect the
pinned-case at the same cross section.
4. Seismic performance evaluation
4.1. Modelling assumptions
The nonlinear behavior of examined planar frames was simulated
by means of 2D models developed in Seismostruct [44].
Masses were considered as lumped at each floor, due to the in-plane
rigidity of the floor diaphragms.
The P–Δ effects were accounted for by means of a leaning-column,
namely by assigning the gravity loads on the interior frames to a fic-
titious column whit no-lateral stiffness, connected to the main frame
using pinned rigid links. Beam-to-column joints were assumed
S. Costanzo et al.
Fig. 6. Interstory drift ratios at (a) DL, (b) SD, and (c) NC limit state.
consistently to the relevant design assumption, while the bracing
members were modelled as fixed in-plane. The gusset plates were also
modelled by using rigid segments to account for their geometrical di-
mensions and a rotational elastic-perfectly plastic spring to simulate the
out-of-plane stiffness and the strength. The springs were inserted at the
theoretical position of the yield line of the gusset plates. The mechan-
ical features of these non-linear springs were evaluated considering the
equivalent cantilever determined according to Whitmore assumption
(i.e. distribution of the inner forces of the gusset with two lines ra-
diating outward at 30° angles from the brace-to-gusset connection). The
columns were considered continuous through each floor beam.
The nonlinear behavior of structural members was reproduced by
using force-based (FB) distributed inelasticity elements [45]. The cross-
section response was simulated by means of the fiber approach, by
assigning a uniaxial stress-strain relationship at each fiber. The steel
behavior was modelled by using Menegotto-Pinto [46] hysteretic stress-
strain relationship. The average value of yield steel stress was assumed,
and it was obtained by multiplying the characteristic yield stress of the
material by the randomness coefficient γov, set as recommended by
EN1998-1 [16].
The behavior of diagonal members is reproduced by mean of the
physical-theory model (PTM) developed by D’Aniello et al. in [47-49].
The accuracy of the numerical model was validated against experi-
mental data as described in former studies carried out by the Authors
[8,10,23,47-49].
Nonlinear-static analyses were performed according to EN 1998-1
[16]; lateral force pattern was considered as proportional to the shape
141
Soil Dynamics and Earthquake Engineering 113 (2018) 136–147
of the first mode of vibration.
A suite of 14 natural accelerograms was used to perform the non-
linear dynamic analyses. The records were obtained from the RESORCE
ground motion database [50] and selected according to procedure de-
veloped by [51] to match the elastic acceleration spectrum provided for
by EN 1998-1 [16] for the Severe Damage limit state (i.e. 10% prob-
ability of exceedance in 50 years).
Furthermore, to calculate the residual interstory drift ratios from the
dynamic time history analyses, each record was fictitiously extended by
10 s at zero acceleration. The Rayleigh tangent stiffness damping was
used and set equal to 2% for both first and second modes of vibration.
Fig. 3 compare the spectra of the selected record with the spectrum
of EN 1998-1 [16], while the data of the records are summarized in
Table 5.
4.2. Nonlinear static analyses
Fig. 4 depicts the monotonic response curves obtained for all ex-
amined structures in terms of base shear-to-design base shear ratios (V/
Vd) versus roof drift ratio (RDR), i.e. top displacement normalized to
the building height. As it can be observed, the contribution of the
secondary frame (evaluated at each analysis step as the sum of shear
forces developed within the base columns) to the lateral capacity is also
reported (see dashed lines) for the cases with beam-to-column moment-
resisting joints.
The comparison of the response curves shows slight beneficial ef-
fects of using beam-to-column moment resisting joints in the braced
S. Costanzo et al.
Fig. 7. Residual interstory drift ratios at (a) DL, (b) SD, and (c) NC limit state.
bays. As a general remark, the cases equipped with MRJs show larger
resistance. In addition, this effect is more evident for 12-story frames
because in these cases the lateral stiffness of the secondary frame (i.e.
beams and columns) is larger than the stiffness of the bracings.
It is interesting to note that the frames designed assuming q = 6
show a slightly larger normalized lateral capacity, which depends on
the plastic engagement of the braces under tension. Indeed, for q = 6
the braces are weaker with lower post-buckling resistance, thus the
diagonals in tension can experience the larger strain hardening.
Fig. 5 reports the values of plastic redistribution parameter αu (being
α1
α1 the multiplier of the horizontal seismic design action to first reach
nonlinear event in any member of the structure and αu the multiplier of
the horizontal seismic design action necessary to form a global me-
chanism) obtained from the nonlinear static pushover analyses. As ex-
pected, the ratio αu is larger for the cases with MRJs, ranging between
α1
[1.04–1.31] and [1.05–1.27] for frames designed assuming behavior
factor equal to 6 and 4, respectively. The plastic redistribution para-
meter is smaller than unit for frames with pinned connections.
4.3. Nonlinear dynamic analyses
The seismic performance of the examined systems was evaluated for
142
Soil Dynamics and Earthquake Engineering 113 (2018) 136–147
the three limit states defined by Eurocode 8, namely: damage limitation
(DL), severe damage (SD) and near collapse (NC).
According to EC8, the seismic hazard is expressed as agd = γI agR for
ground of type A (i.e. rocks), where agR is the reference peak ground
acceleration on bedrock that corresponds to the reference probability of
exceedance in 50 years of the seismic action for the no collapse re-
−1/3
quirement. The hazard curve was assumed as γI = TLR
L ( )
T
, being TL
the return period and TLR the reference return period, which corre-
sponds to the following values associated per limit state: 0.59 at DL, 1 at
SD and 1.73 at NC.
Both global and local response indicators were measured to monitor
the seismic behavior at each limit state. The average demand obtained
for the 14 considered ground motions for each performance parameter
is reported hereinafter. The monitored response indicator are the fol-
lowing:
– transient interstory drift ratio IDR (given by the horizontal relative
displacement at each story divided by the interstory height);
– residual interstory drift ratios IDRRES, which are defined as the
average value of relative horizontal displacements at each story
experienced during the last 10 s at zero acceleration fictitiously
added to each record, divided by the interstory height.
S. Costanzo et al. Soil Dynamics and Earthquake Engineering 113 (2018) 136–147

Fig. 8. Brace ductility demand at (a) DL, (b) SD, and (c) NC limit state.

– braces ductility demand (μ) that is given by the following ratio: uniform distribution of the demand at intermediate stories, although
the frames designed with q = 6 show larger IDR with a less uniform
d
μ= distribution, especially for 12 story-1 braced bay frame where the
dy (5) higher modes of vibration have a larger influence on the overall re-
being d the axial displacement and dy the displacement corresponding sponse due to the larger lateral flexibility. All examined structures sa-
to the brace yielding. tisfy the requirements for non-structural damage at DL limit state (i.e.
1%, considering buildings having non-structural elements fixed in a
– fracture of braces is controlled a-posteriori by monitoring the ratio way so as not to interfere with structural deformations) except for the
μc
, where μc is given as: 12 story-1 braced bay frame that exhibits excessive lateral flexibility.
μfr
Results reported in Fig. 6 show that the stiffness of beam-to-column
d joints of the braced bays negligibly influences the seismic response in
μc =
χd y (6) terms of transient interstory drift ratios; conversely, at the same beha-
vior factor (i.e. q = 4), the frames with MRJs exhibit significantly re-
Being the χ buckling reduction factor defined according to EN duced residual deformation (see IDRRES in Fig. 7) showing larger lateral
1993:1-1 [40]; stiffness, degree of redundancy and more uniform distribution of plastic
μfr is conservatively defined as half time (i.e. assuming symmetric deformations along the building height (see also Fig. 8).
imposed cyclic loadings) the maximum ductility at last cycle prior Fig. 8 shows the ductility demand (Eq. (5)) of both diagonals under
fracture. The latter is given as function of brace slenderness ratio, ac- tension and compression for the analyzed structures. The μ profiles are
cording to the empirical formulation provided by [52]. consistent with the relevant displacement shape profiles. The frames
Fig. 6 depicts the average IDR profiles experienced by three, six and designed assuming q = 6 exhibit the larger energy dissipation capacity,
twelve-story buildings at each limit state. All the frames exhibit sa- with the larger plastic engagement of braces under tension at both SD
tisfactorily performance with adequate lateral stiffness with almost

143
S. Costanzo et al.
Fig. 9. Compression displacement demand of the braces at (a) DL, (b) SD, and (c) NC limit state.
and NC limit state. Small damage concentration is recognized for braces
under compression at intermediate stories in case designed using q = 6
with twelve story and one braced bay, as also confirmed by the corre-
sponding residual interstory drift ratio (see IDRRES in Fig. 7).
The braces ductility demand under compression at the last cycle
prior fracture is shown in Fig. 9. The diagonals of lightest frames
(namely designed with q = 6) experience slightly larger deterioration
under compression, but no fracture occurs for all the examined cases
even at NC limit state. Results reported in Figs. 8 and 9 show that at the
same behavior factor, the brace ductility demand is slightly reduced by
using MRJs as respect to the corresponding cases with pinned beam-to-
column joints.
4.4. Behavior factors
Incremental dynamic analyses (IDAs) were performed to define the
values of the behavior factor for the examined cases. The q factors were
evaluated according to [53,54] as the ratio between the peak ground
acceleration leading to the accepted failure for the selected perfor-
mance level (Au) and the peak ground acceleration corresponding to the
yielding of the frame (Ay). The values of ground accelerations Ay and
144
Soil Dynamics and Earthquake Engineering 113 (2018) 136–147
Au. were derived by the IDAs. In detail, the acceleration corresponding
to the failure was determined as follows:
Au = min(Ac , Abr , Aθ , Ab ) (7)
where Ac is the ground acceleration corresponding to column buckling;
Abr corresponds to bracing member attaining its maximum axial de-
formation; Aθ corresponds to the maximum permitted interstory drift
ratio (i.e. 1.5% at SD and 2% at NC as recommended by [55]); Ab is the
acceleration corresponding to the flexural yielding of the brace-inter-
sected beam.
Fig. 10 depicts the peak displacements (dmax) obtained from IDAs on
the abscissas and the normalized spectral acceleration Sa (T1) on the
Sa, d (T1)
ordinates, where Sa(T1) is the spectral acceleration (per record) corre-
sponding to the period of the first mode of vibration T1, and Sa,d(T1) is
the design acceleration from the EC8 design spectrum. The average
curve is also given in bold black line.
Fig. 11 reports the values of behavior factors calculated from IDAs
for all the examined cases, while in Fig. 12 the obtained q factors are
separately shown for three, six and twelve-story frames. The relevant
average and average ± standard deviation (STdev) values are also
highlighted in both figures. It is interesting to note that the average
S. Costanzo et al.
Fig. 10. Peak roof displacements (dmax) obtained from IDAs for (a) 3-story frames, (b) 6-story frames, (c) 12-story frames with 2 braced bays, and (d) 12-story frames
with 1 braced bay.
values of the behavior factor are larger than those assumed at design
stage, namely the evaluated q factors are equal to 5.78, 6.17, 7.26 for
the case designed with q = 4 and pinned connections, q = 4 and MRJs,
and q = 6 and MRJs, respectively (see Fig. 11).
The comparison between the average minus standard deviation q
factors and the values adopted at design stage shows that q = 4 can be
conservatively assumed. Indeed, qAv-StDev values are equal to 4.20 and
4.48 for all cases designed with pinned and moment resisting connec-
tions, respectively. On the contrary, qAv-StDev is equal to 4.89 for the set
of frames design with q = 6, showing also the larger standard deviation
(i.e. equal to 2.37, see Fig. 11). The large dispersion measured for this
set of structures can be explained by comparing the disaggregate data in
Fig. 12, where the 12-story frames show the poorer response.
This result is mainly due by the influence of the higher modes of
vibration, whose contribution on the global performance increases in
145
Soil Dynamics and Earthquake Engineering 113 (2018) 136–147
these cases due to the lower lateral stiffness of the frames as respect to
the structures designed with q = 4. Hence, using q = 6 can be un-
conservative for the taller buildings (e.g. 12 stories). With this regard, it
is worth noting that the examined structures are regular in elevation
according to the current EN1998-1 [16], thus it would be not necessary
to decrease by 0.8 the value of the reference behavior factor (see
EN1998-1 clause 4.2.3.1 (7)). This finding highlights another criticism
of the current EN1998-1 [16], which does not differentiate the re-
ference behavior factor with the height of the building, as on the con-
trary specified by North American codes (e.g. ASCE-07 [56]). Indeed,
the EC8-compliant rule of vertical regularity does not effectively allow
accounting for the influence of higher modes of vibration on the global
response of multi-story frames, which affect the global response espe-
cially in nonlinear range of behavior of braced structures that are
characterized by significant loss of interstory lateral stiffness following
S. Costanzo et al. Soil Dynamics and Earthquake Engineering 113 (2018) 136–147

Fig. 11. Behavior factors for all examined structures.

the brace buckling [57,58].

5. Conclusive remarks

The research presented in this paper is addressed to investigate the

influence of moment resisting beam-to-column joints (MRJs) into the
braced bays on the seismic response of chevron concentrically braced
frames (C-CBFs). With this aim, a set of low, medium and high-rise 2D
frames extracted from a reference residential building where designed
(alternatively considering either pinned or MRJs) and numerically
analyzed.
Results from nonlinear static and dynamic analyses inferred the
following remarks:

– Using MR beam-to-column joints in the braced bays allows slightly

enhancing the lateral strength of C-CBFs; the beneficial effects of the
secondary action frame is more evident for the 12-story cases, which
exhibit larger lateral extra-strength and stiffness as respect to the
corresponding frames with pinned connections.
– The overstrength due to the redundancy of the system can be also
improved, with values of the plastic redistribution parameters αu
α1
significantly larger than those provided by the corresponding cases
with pinned connections at the ends of the brace-intersected beam.
– The most of examined frames exhibited satisfactory performance
with adequate lateral stiffness, displacement profile and distribution
of plastic demand along the building height. The poorer perfor-
mance was observed for the taller buildings designed assuming the
q = 6, due to the detrimental influence of higher modes of vibration.
– The use of MRJs does not appreciably affect the seismic response in
terms of transient interstory drift ratio, while the residual dis-
placements (see Fig. 7) are reduced due to the improved redundancy
of the system. The ductility demand on bracing members (see Figs. 8
and 9) is also slightly smaller in the frames equipped with MRJs.
– The cases designed assuming q = 6 experienced the larger plastic
engagement of diagonals under tension; on the other hand, the
average demand of the compression deformation in the braces
shows that the adopted design criteria can sufficiently keep the
braces far from fracture on the average for all the examined frames.
– The behavior factors calculated from IDAs show that using MRJs in
the braced bays slightly improves the reserve of ductility and the
energy dissipation capacity of chevron concentrically braced frames.
However, the analyses show that assuming the behavior factor equal
to 6 is unconservative, while the value equal to 4 is appropriate and
Fig. 12. Behavior factors for each examined structure.
suitable for both the cases with and without MRJs.
– The results of IDAs also highlight the need to update the current
version EN1998-1 [16], which does not provide any limitation on
the values of behavior factors with the height of the building.
However, this aspect deserves and needs further research.
References
[1] Tremblay R, Robert N. Seismic design of low- and medium-rise chevron braced steel
frames. Can J Civ Eng 2000;27:1192–206.
[2] Allison H. Steel design – special considerations. In: White R, CG, editors. Building
structural design handbook; 1987.
[3] Stafford Smith B, Coull A. Tall building structures: analysis and design. NY: John
Wiley & Sons; 1991.
[4] Astaneh-Asl A, Cochran ML, Sabelli R. Seismic detailing of gusset plates for special
concentrically braced frames. Struct Steel Educ Counc – Steel Tips 2006.
[5] Shen J, Wen R, Akbas B, Doran B, Uckan E. Seismic demand on brace-intersected
beams in two-story X-braced frames. Eng Struct 2014;76:295–312.
[6] Shen J, Wen R, Akbas B. Mechanisms in two-story X-braced frames. J Constr Steel
Res 2015;106:258–77.
[7] D’Aniello M, Costanzo S, Landolfo R. The influence of beam stiffness on seismic
response of chevron concentric bracings. J Constr Steel Res 2015;112:305–24.

146
S. Costanzo et al.
[8] Costanzo S, D’Aniello M, Landolfo R. Critical review of seismic design criteria for
chevron concentrically braced frames: the role of the brace-intercepted beam. Ing
Sismica: Int J Earthq Eng 2016;33(1–2):72–89.
[9] Costanzo S, Landolfo R. Concentrically braced frames: European vs North American
seismic design provisions. Open Civ Eng J 2017;11(Suppl 1: M11):S453–63.
[10] Costanzo S, D'Aniello M, Landolfo R, De Martino A. Remarks on seismic design rules
of EC8 for inverted-V CBFs. Key Eng Mater 2018;763:1147–54.
[11] Costanzo S, D’Aniello M, Landolfo R. Seismic design criteria for chevron CBFs:
European vs North American codes (part-1). J Constr Steel Res 2017;135:83–96.
http://dx.doi.org/10.1016/j.jcsr.2017.04.018.
[12] Hwang S-H, Lignos DG. Effect of modeling assumptions on the earthquake-induced
losses and collapse risk of steel-frame buildings with special concentrically braced
frames. J Struct Eng (U S) 2017;143(9).
[13] Karamanci E, Lignos DG. Computational approach for collapse assessment of con-
centrically braced frames in seismic regions. J Struct Eng (U S)
2014;140(8):A4014019.
[14] Lignos DG, Karamanci E. Drift-based and dual-parameter fragility curves for con-
centrically braced frames in seismic regions. J Constr Steel Res 2013;90:209–20.
[15] Okazaki T, Lignos DG, Hikino T, Kajiwara K. Dynamic response of a chevron con-
centrically braced frame. J Struct Eng (U S) 2013;139(4):515–25.
[16] EN 1998-1-1. Eurocode 8: design of structures for earthquake resistance – part 1:
general rules, seismic actions and rules for buildings. CEN; 2005.
[17] American Institute of Steel Construction, Inc. (AISC). Seismic provisions for struc-
tural steel buildings. ANSI/AISC Standard 341-16. AISC, Chicago, Illinois; 2016.
[18] CSA. Design of steel structures. CSA-S16-14, Canadian Standards Association,
Toronto, ON; 2014.
[19] Elghazouli AY. Assessment of European seismic design procedures for steel framed
structures”. Bull Earthq Eng 2010;8:65–89.
[20] Broderick BM, Elghazouli AY, Goggins JM. Earthquake testing and response ana-
lysis of concentrically-braced-sub-frames. J Constr Steel Res 2008;64(9):997–1007.
[21] Brandonisio G, Toreno M, Grande E, Mele, De Luca A. Seismic design of concentric
braced frames. J Constr Steel Res 2012;78:22–37.
[22] Bosco M, Brandonisio G, Marino EM, Mele E, De Luca A. Ω* method: an alternative
to Eurocode 8 procedure for seismic design of X-CBFs. J Constr Steel Res
2017;134:135–47.
[23] Costanzo S, D’Aniello M, Landolfo R. Seismic design criteria for chevron CBFs:
proposals for the next EC8 (part-2). J Constr Steel Res 2017;138:17–37. http://dx.
doi.org/10.1016/j.jcsr.2017.06.028.
[24] Longo A, Montuori R, Piluso V. Plastic design of seismic resistant V-braced frames. J
Earthq Eng 2008;12(8):1246–66.
[25] Tenchini A, D'Aniello M, Rebelo C, Landolfo R, da Silva L, Lima L. High strength
steel in chevron concentrically braced frames designed according to Eurocode 8.
Eng Struct 2016;124:167–85.
[26] Marino E. A unified approach for the design of high ductility steel frames with
concentric braces in the framework of Eurocode 8. Earthq Eng Struct Dyn
2014;43:97–118.
[27] Bosco M, Ghersi A, Marino EM, Rossi PP. A capacity design procedure for columns
of steel structures with diagonals braces. Open Constr Build Technol J
2014;8:196–207.
[28] Zhang L, Kurata M, Marino EM, Takeda T. Development of a minimal-disturbance
rehabilitation system for sustaining bidirectional loading. J Struct Eng (U S)
2018;144(6):04018054.
[29] Malaga-Chuquitaype C, Elghazouli AY, Enache R. Contribution of secondary frames
to the mitigation of collapse in steel buildings subjected to extreme loads. Struct
Infrastruct Eng 2014;2479:1–16. http://dx.doi.org/10.1080/15732479.2014.
994534.
[30] Hsiao PC, Lehman DE, Roeder CW. Evaluation of the response modification coef-
ficient and collapse potential of special concentrically braced frames. Earthq Eng
Struct Dyn 2013;42:1547–64. http://dx.doi.org/10.1002/ eqe.2286.
[31] Vargas RE, Bruneau M. Experimental validation of the structural fuse concept. In:
Proceedings of the 14th world conference on earthquake engineering: innovation
practice safety; 2008.
[32] Kanyilmaz A. Secondary frame action in concentrically braced frames designed for
moderate seismicity: a full scale experimental study. Bull Earthq Eng
2017;15:2101–27.
147
Soil Dynamics and Earthquake Engineering 113 (2018) 136–147
[33] Yoo JH, Lehman DE, Roeder CW. Influence of connection design parameters on the
seismic performance of braced frames. J Constr Steel Res 2008;64:607–23. http://
dx.doi.org/10.1016/j.jcsr.2007.11.005.
[34] Yoo JH, Roeder CW, Lehman DE. Simulated behavior of multi-story X-braced
frames. Eng Struct 2009;31:182–97. http://dx.doi.org/10.1016/j.engstruct.2008.
07.019.
[35] Stoakes CD. Beam-column connection flexural behavior and seismic collapse per-
formance of concentrically braced frames (Ph.D. thesis). Urbana-Champaign:
University of Illinois at Urbana-Champaign, ProQuest Dissertations Publishing;
2012. p. 3570612.
[36] Uriz P, Mahin SA. Toward earthquake-resistant design of concentrically braced
steel-frame structures. PEER report 2008/08. Earthquake Engineering Research
Center, University of California, Berkeley.
[37] American Institute of Steel Construction, Inc. (AISC). Seismic provisions for struc-
tural steel buildings. ANSI/AISC Standard 341-10. AISC, Chicago, Illinois; 2010.
[38] BCJ. Structural provisions for building structures. Tokyo; 1997 (in Japanese).
[39] Marino EM, Nakashima M, Mosalam KM. Comparison of European and Japanese
seismic design of steel building structures. Eng Struct 2005;27:827–40. http://dx.
doi.org/10.1016/j.engstruct.2005.01.004.
[40] EN 1993:1-1. Eurocode 3: design of steel structures – part 1-1: general rules and
rules for buildings. CEN; 2005.
[41] EN 1990. Eurocode 0: basis of structural design; 2001.
[42] EN 1991-1-1. Eurocode 1: actions on structures – part 1-1: general actions – den-
sities, self-weight, imposed loads for buildings; 2002.
[43] EN 1994-1-1. Eurocode 4: design of composite steel and concrete structures – part 1.
1: general rules and rules for buildings; 2004.
[44] Seismosoft. SeismoStruct – a computer program for static and dynamic nonlinear
analysis of framed structures. Available from URL: 〈www.seismosoft.com〉; 2011.
[45] Spacone E, Ciampi V, Filippou FC. Mixed formulation of nonlinear beam finite
element. Comput Struct 1996;58(I):71–83.
[46] Menegotto M, Pinto PE. Method of analysis for cyclically loaded R.C. plane frames
including changes in geometry and non-elastic behavior of elements under com-
bined normal force and bending. In: Proceedings of the symposium on the re-
sistance and ultimate deformability of structures acted on by well defined repeated
loads; 1973.
[47] D’Aniello M, Portioli F, Landolfo R. Modelling issues of steel braces under extreme
cyclic actions. In: Proceedings of the COST-C26 final conference. Naples; 16–18
September 2010. p. 335–41.
[48] D’Aniello M, La Manna Ambrosino G, Portioli F, Landolfo R. Modelling aspects of
the seismic response of steel concentric braced frames. Steel Compos Struct Int J
2013;15(5):539–66.
[49] D’Aniello M, La Manna Ambrosino G, Portioli F, Landolfo R. The influence of out-of-
straightness imperfection in physical-theory models of bracing members on seismic
performance assessment of concentric braced structures. Struct Des Tall Spec Build
2015;24(3):176–97.
[50] Akkar S, Sandıkkaya MA, Şenyurt M, Azari Sisi A, Ay BÖ, Traversa P, et al.
Reference database for seismic ground-motion in Europe (RESORCE). Bull Earthq
Eng 2014;12(1):311–39.
[51] Fulop L. Selection of earthquake records for the parametric analysis. Research re-
port VTT-R-03238-10, VTT, Espoo; 2010.
[52] Goggins JM, Broderick BM, Elghazouli AY, Lucas AS. Behaviour of tubular steel
members under cyclic axial loading. J Constr Steel Res 2006;62:121–31.
[53] Salvitti LM, Elnashai S. Evaluation of behaviour factors for RC buildings by non-
linear dynamic analyses. Eng Struct 2000;22:1244–60.
[54] Tenchini A, D’Aniello M, Rebelo C, Landolfo R, da Silva LS, Lima L. Seismic per-
formance of dual-steel moment resisting frames. J Constr Steel Res
2014;101:437–54.
[55] FEMA 356. Prestandard and commentary for the seismic rehabilitation of buildings;
2000.
[56] ASCE/SEI 7-10. Minimum design loads for buildings and other structures.
[57] Longo A, Montuori R, Piluso V. Seismic reliability of V-braced frames: influence of
design methodologies. Earthq Eng Struct Dyn 2009;38:1587–608.
[58] Giugliano MT, Longo A, Montuori R, Piluso V. Seismic reliability of traditional and
innovative concentrically braced frames. Earthq Eng Struct Dyn
2011;40(13):1455–74.