Académique Documents
Professionnel Documents
Culture Documents
Engineering Structures
journal homepage: www.elsevier.com/locate/engstruct
a r t i c l e i n f o a b s t r a c t
Article history: Steel frames with reinforced concrete infill walls (SRCW) are an interesting seismic-resistant structural
Received 27 April 2016 solution. However, an effective seismic design of SRCWs is not easy due to the current lack of specific
Revised 9 March 2017 capacity design rules that allow controlling the formation of a proper energy dissipating mechanism.
Accepted 13 March 2017
In order to overcome such an issue, a ductile design procedure is presented in this paper. The proposed
procedure leads to innovative SRCW systems where energy dissipation is expected to take place only in
the vertical elements of the steel frame, which are subjected mainly to axial forces. The non-ductile com-
Keywords:
ponents, i.e. reinforced concrete wall and steel-to-concrete connections, are expected to suffer negligible
Ductile design
Earthquake resistant systems
damage. Accordingly, the system is designed to control the formation of diagonal struts in the infill walls
Experimental tests and behaves as a lattice brace instead of a shear wall. Experimental test results and nonlinear finite ele-
Hybrid systems ment analyses are illustrated to support the developed ductile design approach and highlight the advan-
Reinforced concrete infill walls tages of SRCWs.
Ó 2017 Elsevier Ltd. All rights reserved.
1. Introduction interactions between the steel frame and the compression strut
in the RC infill walls; (3) interactions between steel frames and
During the two past decades increasing interest has been given the RC infill walls through friction and shear connectors. The con-
to hybrid steel and concrete structures, obtained through the com- tribution of each resisting mechanism to the overall horizontal
bination of structural elements made of reinforced concrete (RC), resistance depends on the mechanical properties and geometrical
steel, and composite steel-concrete, arranged to create more effi- configuration of the SRCW. Accordingly, different typologies of
cient and economical seismic resistant solutions, e.g. [1–9]. Atten- SRCWs can be identified depending on the beam-to-column joint
tion is here focused on hybrid steel and concrete systems made by typology and to the distribution of shear studs along the interface
steel frames with reinforced concrete infill walls (SRCWs) owing to between the frame and the infill wall. Systems in which the beam-
their many advantages as seismic-resistant systems [1,7], i.e. high to-column joint is able to transfer significant portion of the flexural
initial stiffness beneficial in reducing building damage under low- moment without important relative rotations are usually referred
intensity earthquakes, effective damping characteristics, and to as Fully-restrained SRCWs (FSRCWs); otherwise they are
potentially easy repairs after moderate damage through the use referred to as Partially-restrained SRCWs (PSRCWs) [12,13]. Inte-
of epoxy resins on the cracked wall. In this way, SRCWs provide gral infilled frames are characterized by connectors distributed
a promising structural solution potentially able to reduce the seis- along the interfaces between the frames and the infill walls, while
mic risk in all its aspects, i.e., economic loss and fatalities, as for non-integral infilled frames are not provided with connectors.
example contemplated in recent studies [10,11] following the Intermediate configurations can be classified as semi-integral
2012 Emilia earthquakes in Italy and highlighting the necessity infilled frames.
of an adequate structural behaviour also for low-to-moderate seis- The seismic behaviour of SRCWs has been the object of many
mic events. theoretical and experimental studies over more than four decades.
In SRCWs three different lateral resisting mechanisms can be Test results of Mallick and Severn [14] showed that the presence of
identified (Fig. 1): (1) contribution of the steel frame; (2) direct shear connectors in the corner of the steel frame prevented the
rotation of the infill walls, increased the overall stiffness, and did
⇑ Corresponding author at: University of Camerino, Viale della Rimembranza, not affect the lateral strength. Furthermore, integral infilled frames
Ascoli Piceno 63100, Italy. exhibited shear failure of the infill walls, while non-integral infilled
E-mail address: alessandro.zona@unicam.it (A. Zona).
http://dx.doi.org/10.1016/j.engstruct.2017.03.019
0141-0296/Ó 2017 Elsevier Ltd. All rights reserved.
A. Dall’Asta et al. / Engineering Structures 141 (2017) 144–158 145
in which vertical steel elements contribute to resist the overturn- 2.2. Construction issues
ing moments, acting as reinforcements, whereas the RC infill walls
have to resist the storey shear, might be far from reality especially The construction of the proposed innovative SRCW does not
for medium and high rise buildings and when not encased ele- involve particular difficulties in the shop prefabrication nor in
ments are adopted. In such cases, the distribution of shear connec- the construction site where the assembling of steel elements and
tors along the wall edge strongly affect the behaviour of the system the casting of concrete wall can be carried out by means of usual
and the provision (usually required by design codes) that they have equipment and workmanship. The plan configuration of the
to resist the storey shear to be transmitted to the RC infill wall is required SRCWs can be arranged to accommodate functional and
not sufficient to ensure the formation of an inclined uniform com- architectural needs, and can be effectively coupled with gravita-
pression field. Slip between the steel frame and the RC infill wall, in tional structures provided that rigid floors are suitably connected
fact, produces a concentration of the forces at the wall corners and to them as usual in steel construction.
failure of shear connectors or concrete crushing typically before The steel side elements can be made up of hot rolled or hollow
the yielding of the vertical elements. The basic idea of this paper profiles. As these elements act as fuses for the system, they have to
is to design a system in which the objective resisting mechanism be designed in order to strictly fulfil strength verification limiting
at the ultimate limit state is fostered by a suitable configuration their over-strength as well as to ensure that energy dissipation is
of the elements, allowing the introduction of a clear capacity uniformly activated among the adopted SRCWs and along their ele-
design procedure that should lead to reliable ductile structural vation. For this purpose, they could be realised with welded pro-
solutions. files using low-grade steels. As already mentioned, these
The novel SRCW depicted in Fig. 3a is proposed. Because the for- elements are not directly connected to the RC wall, the local buck-
mation of a main diagonal strut within the RC infill wall cannot be ling is limited by the formation of RC vertical struts, and, even
avoided when yielding of the side steel element is enforced, the though in principle no specific provisions are necessary, it is better
structural system should be conceived to properly control the com- to assure their cross section being compact, e.g. at least class 2
pression field in the concrete. For this purpose, joints of the steel according to Eurocode 3 [25].
frame are shaped to support the diagonal strut formation, i.e. an The end connections of the ductile steel elements should be
inclined and appropriately stiffened plate permits force exchanges designed to remain elastic to permit the replacement of the ductile
between the RC infill and the steel frame with an imposed effective elements after seismic damage. End-plates connections should be
contact area (Fig. 3b). Accordingly, the infill wall, without the sharp preferred and the ductile elements should be connected to the
corners, does not need to undergo significant slip for the formation end plate by means of full penetration welding. The adjacent ver-
of the diagonal strut. Shear connectors are not essential for the for- tical elements to which the ductile elements are connected should
mation of such resisting mechanism and their distribution along be over-strengthened; this can be assured by using a higher steel
the steel frame depends on two main issues further discussed in grade or by suitably enlarging the resisting cross section that
the following paragraphs: the restraining of possible out-of-plan should have width equal or greater than the infill wall thickness;
overturning of the infill wall and the formation of a RC vertical for this purpose it can be advantageous using the same profile
strut in compression (Fig. 3b). adopted for the horizontal beams.
The energy dissipation is attained when the dissipative fuses Horizontal beams, including the stiffened joints and the shear
placed within the steel vertical elements yield in tension; the axial connectors, necessary for the stability of the wall during a seismic
forces in compression are contrarily limited by the formation of event, can be entirely prefabricated in the shop by using hot rolled
vertical struts that act in parallel to the compressed side steel ele- standard steel profiles. As it is better that their width is compliant
ments (Fig. 3b). with the wall thickness, HE series are usually preferable. However,
In order to allow the replacement of damaged steel fuses after a when the wall thickness is bigger than 300 mm (the max available
seismic event, they are not connected to the infill wall and suitable width of HE profiles), welded sections may be also adopted instead
end connections are realised. Therefore, the proposed SRCW sys- of coupled elements.
tem behaves as a truss even from the very first lateral sways As for the RC infill walls, detailing rules suggested by Eurocode
and, taking advantage from the stiffness of the concrete in com- 8 [24] are not mandatory as the system is conceived as a lattice lat-
pression and the energy dissipation capacity of the steel elements eral resisting structure and it is not expected that energy dissipa-
in tension, represents a potentially efficient steel and concrete tion takes place in wall critical regions. In addition to a double
hybrid seismic-resistant structure. skin welded mesh placed to limit crack widths and complying with
Fig. 3. (a) Innovative SRCW systems; (b) limit behaviour under seismic action of the innovative SRCW system.
A. Dall’Asta et al. / Engineering Structures 141 (2017) 144–158 147
provisions for RC walls in non-seismic areas, reinforcements must (Eurocode 8, paragraph 7.3). Once the lateral force system is
be provided in order to prevent concrete crushing by means of a defined, forces in the elements of the lattice-like structure due to
suitable confinement action. In particular, attention must be paid the earthquake action can be promptly evaluated. It is worth
to regions in contact with the bearing plates where compression remembering that, being a capacity-based design, the forces eval-
stress diffusion takes place. For this purpose different reinforce- uated in this step acting on the non dissipative elements, will be
ment layouts may be used as suggested in the following section. suitably magnified as described in the following points.
It is important to remark that the adoption of transverse confine- Step 2: design of the cross sections of the steel dissipative elements
ment stirrups is possible only for thick walls due to the minimum in tension
space required by the stirrups themselves. Design of the steel dissipative elements is carried out by super-
Finally, shear connectors must be placed at the plates where the imposing tensile forces evaluated in the previous step with forces
wall struts interact with the steel frame with the aim of providing due to gravity loads (compression forces). Due to the formation of
out-of-plane stability. In the case in which the wall has to bear a the diagonal strut in the RC wall, the element on the opposite side
part of the compression, as described in the following Step 8 of of the wall is subjected to a compression force for which a specific
the design procedure, shear connectors must be welded to the ver- verification is carried out at Step 8 for what concern possible insta-
tical elements, outside the dissipative zones, in order to transmit bility. In any case, at least class 2 sections, defined according to
forces from the frame to the wall. In this case it is important to Eurocode 3 [25], are to be used for these elements.
remark that local failure mechanisms can be characterised by split- Step 3: capacity design of the connection of the ductile elements
ting of the wall. This mechanism can be controlled with suitable and of the adjacent elements
transverse reinforcements designed according to Eurocode 4 Part The design of the connection of the ductile elements and of the
2 (6.6.4 and Annex C) [26] and by adopting stud connectors with adjacent elements is performed with the formula
suitable length.
Rd P 1:1cov Rfy ð1Þ
3. Design of the innovative SRCW systems suggested by Eurocode 8 (paragraph 6.5.5) [24] where cov is the
over-strength coefficient of the element with plastic resistance Rfy
3.1. Details of the design procedure of the connected dissipative member based on the design yield
stress of the material.
As discussed in the previous section, the proposed SRCW should Step 4: calculation of geometric over-strength factors
develop a resisting mechanism in which the steel elements of the The over-strength factor Xi for the i-th dissipative steel side ele-
frame are subjected mainly to axial forces and the concrete walls ment is calculated as usual for steel structures, as suggested by
are characterised by the formation of diagonal struts (Fig. 3b). A Eurocode 8 [24], by the ratio of the plastic resistance of the ductile
statically determinate truss model (Fig. 4a) describes such a limit element, Npl,Rd,i, and its design force, NEd
behaviour of the system. This scheme can be adopted in a force-
Npl;Rd;i
based procedure because the stress resultants in the elements do Xi ¼ ð2Þ
NEd;i
not depend on the dimension of members and because a capacity
design is straightforward once the required strengths of the mem- The maximum over-strength Xi should not differ from the min-
bers are defined. The procedure hereafter described is organized imum value by more than 25% in order to guarantee yielding of the
into 9 steps and follows a capacity design path leading to systems edge steel elements at the different levels. The same condition is
in which overall dissipative mechanisms can be developed. adopted in the design of steel braced frames in order to avoid
Step 1: definition of the static equivalent lateral loads and calcula- the formation of soft-storey [24] and its influence investigated in
tion of the truss actions a number of recent studies, e.g. [27–29]. For the proposed innova-
A suitable distribution of lateral loads can be defined according tive SRCW, such a provision is investigated in the analyses
to Eurocode 8 [24] assuming a linear shape of the main vibration addressed in Section 3.2. The proposed SRCW system is less sensi-
mode. The resulting base shear is derived from the seismic mass tive to the concentration of the deformations in a single storey, i.e.
of the structure by considering the spectral acceleration corre- soft-storey formation, due to the presence of the non-dissipative
sponding to an estimated fundamental period. As usual, the design RC infill walls.
spectrum is reduced by a suitable behaviour factor q in order to Step 5: calculation of axial forces in non-ductile elements by com-
account for the ductility of the system while designing the struc- bining the effects of gravity loads with those of the seismic action suit-
ture in the linear range. To a first approximation, the behaviour ably magnified
factor can be chosen as suggested for Type 1 composite walls The non-ductile elements of the structure are the diagonal
struts developed in the RC infill walls and the horizontal beams.
The forces are calculated by a suitable increment of the seismic
design component, NEd,E, accounting for the material and geometric
over-strength of the ductile elements with the usual formula
Fig. 5. (a) Diagonal struts within the infill walls; (b) compression fields involved to resist the axial force in the case lateral elements failed due to instability.
wall should be equal to the bearing plate width lb at the diagonal Two different reinforcement layouts may be adopted (Fig. 6),
ends whereas the effective width is obtained multiplying lb by a the former is constituted by two sets of orthogonal reinforcements
coefficient a > 1 at mid diagonal. It is observed that in this region whereas the latter is constituted by a set of specific transverse
of the wall the concrete stress field is also characterised by trans- (with respect to the strut direction) reinforcements. In the first
verse tension that reduces the available compression strength. The case, vertical and horizontal reinforcements should fulfil the
design formula conditions
f f f T2 Ahsl L
0:85 ck tw lb ; 0:85 ck m 1 ck ðat w lb Þ ¼ Aslv 2 þ Ah2 ¼ ð6Þ
cc cc 250 2 sl
Avsl H
f yd
P NEd;G þ 1:1cov XNEd;E ð4Þ
where the meaning of symbols is explained in Fig. 6 and fyd is the
is derived from Eurocode 2 (paragraph 6.5) [30], where fck is the
design tensile strength of reinforcing bars. It is worth noting that
characteristic compressive cylinder strength of concrete, cc is the
the first reinforcement layout (orthogonal bars, see Fig. 6a) is sim-
partial factor for concrete, tw is the infill wall thickness, m is a coef-
pler but possibly less stiff than the second (specific transverse rein-
ficient that takes into account the influence of transverse tension.
forcements, see Fig. 6b), which instead requires a third layer of
The second value of the left term of Eq. (4), representing the
reinforcements that can be placed only in the case of sufficiently
compressive strength at mid-strut, takes into consideration the
thick walls.
transverse tension (m = 0.6 may be assumed, as suggested by Euro-
Step 7: design of the beams in tension
code 2 [30]) whereas the first value, representing the compressive
These elements are designed to resist the magnified tensile
strength at strut end, considers a simple compression field. The
axial forces calculated in step 5.
two design parameters lb and a can be determined with an itera-
Step 8: check of side steel elements in compression and possible
tive procedure or by imposing a tentative value for a (e.g. a = 2).
design of the infill walls to assure the formation of a vertical struts
The bearing plate should then be proportioned and suitably stiff-
The ductile elements are checked for instability by using the
ened to avoid stress localization in the concrete.
formula
The wall reinforcing bars should be checked to guarantee the
diffusive mechanism that depends on the choice of the parameter Afy v
P NEd;G þ 1:1cov XNEd;E ð7Þ
a; for this purpose rules for partial discontinuity regions suggested cM1
in Eurocode 2 (paragraph 6.5.2) [30] are considered. In this case,
tensile force, T, to be resisted by reinforcements is evaluated by where fy is the yield strength of steel, A is the area of the element
means of the formula cross section, v is the reduction factor for relevant buckling mode
and cM1 is partial factor for resistance of members to instability as
1 1 defined by Eurocode 3 [25].
T¼ 1 N ð5Þ
4 a Ed The effective length of the element can be selected to be equal
to the distance between the beam-to-column intersection nodes
Fig. 6. (a) Reinforcements constituted by orthogonal rebar layout; (b) reinforcements constituted by additional stirrups.
A. Dall’Asta et al. / Engineering Structures 141 (2017) 144–158 149
reduced by the real dimensions of the joint. In case the verification the wall infills are considered, namely w/h = 0.66, 1, 1.5 and 2 (w is
is not satisfied, it is expected that the adjacent strip of the concrete the wall width and h the storey height respectively); consequently,
wall collaborates to bear the compression force (Fig. 5b); in such a the overall aspect ratios of the SRCWs are 0.165, 0.25, 0.375 and
case, the design of the shear connection between the wall and the 0.5, for 4-storey systems, and 0.083, 0.125, 0.188 and 0.25, for 8-
frame as well as the check of the vertical strut developing in the storey systems, respectively. The systems are designed for three
wall have to be carried out. different base shears in order to cover low, medium, and high seis-
The shear connection is designed to involve an adjacent strip of mic intensities. The design procedure is aimed at obtaining a global
the infill wall in resisting part of the compression force. In partic- dissipative mechanism in which the lateral elements of the wall
ular, this should be able to transmit the force in excess with respect (ductile elements) may undergo yielding at each storey; in order
to the bearing capacity of the ductile element given by to check the effectiveness of limits imposed to the over-strength
ratios for the ductile elements (see Step 4 of the design procedure),
Afy v
tRd P NEd;G þ 1:1cov XN Ed;E ð8Þ four different design scenarios satisfying the condition Xmax/Xmin -
cM1 1.25 are considered namely: (1) all the ductile elements are
In this case, the shear connection has to be placed at the vertical designed to fulfil the optimal condition Xi = 1; (2) half of the ele-
elements outside the dissipative zone. It has to be designed by tak- ments are optimal and the others are such that Xi = 1.25; (3)
ing into account that possible splitting failure mechanisms of the one-fourth of the elements are optimal and the others are such that
wall might occur instead of the usual failures due to concrete Xi = 1.25; and (4) one ductile element is optimal and the other
crushing and stud yielding. For this purposes, and in order to avoid ones are such that Xi = 1.25. Steel S235 (fyk8235 MPa) is consid-
the splitting failure mechanism, rules suggested by Eurocode 4 Part ered for the ductile elements whereas higher grades are considered
2 (paragraph 6.6.4 and Annex C) [26] could be considered. for the non-ductile elements. Concrete C35/45 (fck = 35 MPa) and
In the verification of the vertical strut developing in the wall, steel B450 (fyk = 450 MPa) were used for RC infill, furthermore 8-
this element withstands the same force calculated from Eq. (8) storey systems were designed also for concretes C40/50 (fck = 40 -
and has to be suitably reinforced with confinement stirrups. The MPa) and C45/54 (fck = 45 MPa). 180 cases were completely
same detailing rules suggested in Eurocode 8 (paragraph designed but, due to space limitation, hereafter only a summary
5.4.3.4.2) [24] for RC walls might be adopted. of the outcomes is reported.
Step 9: calculation of the length of the dissipative elements, in order Table 1 shows the results for the 4-storey systems. The design
to ensure the compliance between local and global ductility procedure behaves well for the selected cases with only some
This is the second crucial point in the design as it permits a first problems in the case of walls with aspect ratio 0.66 designed for
control of the behaviour factor used in Step 1. The length of the dis- the highest seismic intensity since the use of custom-fabricated
sipative elements can be evaluated assuming that: (i) the system profiles would be needed for the non-ductile elements and the
achieves the plastic range with all the fuses simultaneously shear connection is not feasible for the impossibility of fulfilling
yielded, (ii) the steel has a perfectly plastic behaviour with a fixed the splitting verification. It is noted that the four scenarios
ductility, and (iii) the equal displacement assumption is fulfilled. assumed for the regularity of the dissipative elements do not lead
For this purpose, the formulas derived considering this simplified to different results.
mechanisms (Fig. 4) can be adopted. In the case of 8-storey systems (Table 2) a very different picture
Under the hypothesis that all the fuses have the same length H0 , can be observed. As expected, wall infills with aspect ratio 0.66 are
and in order to assure that the strain demand on the dissipative not suitable for all the three selected levels of seismicity due to the
elements is lower than the material ductility lm, the following for- very low overall aspect ratio 0.083. Custom-fabricated profiles
mula is obtained: should be used at the first three storeys and the wall is involved
to resist compression axial forces in the vertical elements and
ðls 1Þ del L the shear connection system cannot be properly designed. Further-
H0 ¼ PN ð9Þ
ðlm 1Þ ey i¼1 ðzN zi1 Þ more, due to the concrete strength, it is not possible to proportion
the bearing plates needed to avoid the local concrete crushing and,
where ls is the ductility of the structure; N is the number of storeys, at the same time, to keep dimensions that may comply with those
del is the structure elastic displacement evaluated for the static of the connected steel elements. In the case of wall infills with
equivalent loading inducing the first yielding; dpl is the structure higher aspect ratio, the system can be designed only for low and
plastic displacement. medium seismicity levels whereas for the high seismicity cases
It is worth noting that the statically determinate model custom-fabricated profiles would be needed. In an attempt to
depicted in Fig. 4 can be profitably used also by considering more understand if the problem can be overcome by considering more
sophisticated assumptions involving the behaviour of the compo- resisting concretes, SRCWs are redesigned by using C40/50 and
nents. In particular, by introducing the hardening in the behaviour C45/55 concrete classes. In these cases the walls with the lowest
of the steel, it is possible to plot a simplified capacity curve of the aspect ratio cannot be properly designed.
system in which the development of the lateral-resisting mecha- The previous results are of course not exhaustive as the design
nism can be controlled. of such systems could be refined in real cases adopting solutions
that should be validated case-by-case. Nevertheless, they give the
3.2. Applicability of the procedure overall information that a feasible design is pursuable for the wall
with overall aspect ratios higher than 0.15.
The applicability of the proposed procedure for the design of
feasible and safe SRCWs is assessed by considering a large set of
cases defined by changing the wall aspect ratio, the number of 4. Validation of the design procedure
storeys, and the base shear. 4-storey and 8-storey SRCWs,
extracted from a regular building with interstorey height The innovative SRCW system is designed by considering an
h = 3.40 m, are considered. The mass associated to each system is equivalent lattice-like lateral resisting structure. Nevertheless,
200 kN s2/m for each floor. In order to investigate whether the sys- the real system is characterised by the presence of the reinforced
tem can be properly designed and at what extent the wall slender- concrete infill walls that may behave differently from simple diag-
ness may compromise its efficiency, four different aspect ratios of onal struts since biaxial stress fields are expected to develop. Fur-
150 A. Dall’Asta et al. / Engineering Structures 141 (2017) 144–158
Table 1
Qualitative outcome of the design of 4-storey systems.
Table 2
Qualitative outcome of the design of 8-storey systems.
thermore, joints of the steel frame are not hinged but are rather forcements of the walls are B450C and are made up of a double
characterised by the presence of the diagonal welded plates mesh of /10 reinforcements spaced to comply with Eurocode 2
(placed to control the formation of the diagonal struts within the [30]; furthermore, side strips are confined with stirrups /8 suitably
walls) that make the joint very stiff. Models capable of reproducing spaced [24]. Shear connectors are welded to the diagonal plates
such mechanisms and experimental tests are necessary in order to only at the steel frame joints.
assess the actual behaviour of the proposed SRCWs. Some cases Main data of the analysed cases, profiles adopted for the ductile
selected from the set previously considered are analysed for this and non-ductile elements, thickness of the walls (tw), width of the
purpose and the results obtained for one 4- and one 8-storey sys- diagonal bearing plates (lb), as well as the over-strength factors
tem are hereafter discussed. Moreover, experimental tests on two (coefficients X) are summarised in Tables 3 and 4. Detailing draw-
downscaled specimens design adopting the proposed procedure ings, not reported here due to space limitations, are available in [7].
are carried out and results detailed analyzed.
4.1.2. Advanced model
4.1. Numerical analysis Numerical models are developed in the software for finite ele-
ment analysis ABAQUS [31]. The geometry of the system is closely
4.1.1. Cases analysed reproduced by using shell finite elements both for the steel frame
Cases analysed are marked with asterisks in Tables 1 and 2; in and for the RC infill walls; in particular, 4-node linear shell ele-
particular the 4-storey system with the wall width w = 3.40 m ments with 5 degrees of freedom per node (reduced integration,
under medium seismicity is considered; as for the 8-storey system, small membrane strains) are used. The walls are assumed to be
w = 5.10 m and low seismicity are considered. Steel classes S235 connected to the frame at the inclined bearing plates where stud
and S355 are used for the ductile and non-ductile elements, connectors are placed. Wall reinforcements are considered by
respectively. As for the walls, concretes C35/45 and C40/50 are introducing two layers of reinforcements according to the design.
used for the 4-storey and the 8-storey buildings, respectively. Rein- A coarse mesh (mean size of 0.5 m) is adopted for the RC walls
A. Dall’Asta et al. / Engineering Structures 141 (2017) 144–158 151
Table 3
Data of the 4-storey system discussed.
Table 4
Data of the 8-storey system discussed.
to avoid numerical convergence problems, whereas a more refined the Mises equivalent deviatoric stress. When a crack is detected, its
mesh (average size 0.1 m) is adopted for the steel members. Repre- orientation is stored for successive calculations. Subsequent crack-
sentative perspective views of one of the models are presented in ing at the same point is restricted to being orthogonal to this direc-
Fig. 7a. tion because stress components associated with an open crack are
The behaviour of the concrete is described adopting a smeared not included in the definition of the failure surface used for detect-
cracking model with full shear retention. The concrete behaviour is ing the additional cracks. As already mentioned, the behaviour of
assumed to follow Mander’s law in compression and a linear elastic cracked concrete is modelled with a strain-softening branch that
law in tension; a softening branch is adopted after cracking to sim- allows to simulate the interaction between reinforcements and
ulate the tension-stiffening effect. In particular, cracking is concrete. In particular, it is assumed a linear strain softening with
assumed to occur when the stress reaches a failure surface defined zero stress at a total strain of about 10 times the strain at tensile
by a linear relationship between the equivalent pressure stress and failure. As for the steel, elastoplastic with hardening models are
40 400
30 200
20 0
10 -200
0 -400
-10 -600
-0.001 0 0.001 0.002 0.003 0.004 -0.25 -0.15 -0.05 0.05 0.15 0.25
400 400
200 200
0 0
-200 -200
-400 -400
-600 -600
-0.25 -0.15 -0.05 0.05 0.15 0.25 -0.25 -0.15 -0.05 0.05 0.15 0.25
(a) (b)
Fig. 7. Finite element models: (a) geometry and finite element mesh; (b) material constitutive laws.
152 A. Dall’Asta et al. / Engineering Structures 141 (2017) 144–158
considered for the reinforcements and the steel frame. The yielding 4.1.3. Results of the analyses
point, the material plastic hardening, and the ultimate stress for Fig. 8 shows results obtained for the 4-storey system; the yield-
each steel grade considered are defined according to the mechan- ing pattern of the steel frame for three different displacements
ical characteristics of materials adopted in the design. The nonlin- measured at the top level is depicted in Fig. 8a. It is worth noting
ear behaviour for all the materials considered is presented in that plastic strains are attained only at the ductile elements
Fig. 7b. whereas the other elements remain in their linear elastic phase.
It is worth noting that, the model adopted is aimed at investi- Such behaviour fully agrees with the dissipating mechanism taken
gating the overall behavior of the system, with some insight in as objective of the proposed design procedure. It is worth noting
the behaviour of steel elements and the infill wall, but is not suit- that yielding does not occur at the last storey despite the over-
able for capturing interface slip as well as the local behaviour of the strength coefficient X is equal to 1.02; nevertheless, the develop-
studs. In fact, as finite elements are simply connected in regions in ment of the plastic mechanism at the first three storeys can be
which shear studs are placed, the model is not able to describe the deemed very satisfactory and follows the expected sequence in
deformability of the connectors but can roughly capture effects of consideration of the coefficients X relevant to the first three
perfect contact in compression and damage of the concrete in storeys.
tension. The resisting mechanism adopted in the design is also corrobo-
Pushover analyses are performed by considering triangular dis- rated by the stress field in the wall (Fig. 8b) that clearly depicts the
tributions of lateral forces. The loads concentrated at the reference formation of the diagonal struts. It can also be observed that the
points located at the two ends of each beam at the floor levels and values of the principal stresses do not increase with the overall dis-
are ramped linearly in time. To overcome numerical convergence placement demonstrating that the wall is protected against crush-
problems, the nonlinear solution algorithm is forced to allow more ing due to the yielding of the side ductile elements. The same
iterations per load increment until convergence is achieved. remarks hold for the 8-storey system.
(a)
(b)
Fig. 8. Results for the 4-storey system (u is the horizontal displacement of the top floor): (a) yielding pattern in the steel frame, the red circles indicates the first yielding of
the elements; (b) stress field in the wall.
A. Dall’Asta et al. / Engineering Structures 141 (2017) 144–158 153
A comparison between the pushover curves obtained using the the wall, supplemental confining reinforcement in the two vertical
proposed design method for SRCW systems and the finite element portions of the wall close to the dissipative elements, and open
analysis is presented in Fig. 9. Not only the global behaviour but stirrups all along the upper and lower edges of the steel frame.
also the sequence of yielding of ductile elements are well predicted The yield strength, fym, of the dissipative elements, obtained
in the design phase. It is worth noting that the numerical model through tensile tests on a coupon, is 243 N/mm2. The mean value
exhibits an initial stiffness higher than the one predicted by the of the concrete compressive strength, obtained by two compres-
simplified model, mainly due to the contribution of the concrete sive tests on cubic specimens (one for each wall), fcm, is 33.2 N/
tensile resistance. However, after the concrete cracking, the SRCW mm2.
stiffness evaluated through the refined model is very close to the
one estimated using the simplified model. Furthermore, the hard- 4.2.1. Test setup
ening behaviour subsequent to the yielding of the ductile elements The test setup is shown in Fig. 12. The SRCW specimen is bolted
in tension is well reproduced up to the first yielding in compres- to a steel base firmly connected to the strong floor. The external
sion of one of the ductile element. Finally, the presence of the rein- force is applied by an hydraulic jack system to the upper beam of
forcement mesh does not introduce any contribution in terms of the steel frame. The tests are carried out in displacement control
stiffness and strength because the stress field in the wall is mainly and the displacement history imposed to the jacks end is reported,
characterized by the diagonal compression field. for both tests, in Fig. 13. An initial maximum displacement equal to
For 8-storey systems the comparisons are still good; also in this 20 mm is imposed cyclically in order to assess a displacement duc-
case the plastic mechanism develops almost completely and only tility equal at least to 3 (during the test a yield displacement equal
the elements at the last two levels remain in the elastic field (see to about 6 mm is observed). The imposed maximum displacement
Fig. 10). is then raised to about 30 mm.
The numerical analysis presented are surely affected by some
limitations, as the simplified model of the shear connectors, the 4.2.2. Experimental results
neglect of contact issues between steel and concrete, the adoption In Fig. 14 the experimental cyclic behaviour of specimen 1 is
of shell elements with coarse mesh, etc. However, the good agree- shown. The first loading cycle highlights a relatively ‘‘fat” hys-
ment between the design procedure objectives and the results of teretic behaviour, while pinching phenomena, with the maximum
the refined and simplified models demonstrates that the design resistance that remains practically constant, is exhibited during the
procedure proposed may be suitable for dimensioning the system subsequent cycles. The first semi-cycle (Fig. 14b) shows that the
elements and that the simplified lattice-like system can also pre- system is characterized by a behaviour very close to an ideal
dict with adequate approximation the behaviour of the structure. elastoplastic one with a displacement ductility equal, at least, to
3. At the end of the first unloading phase, the concrete wall exhibits
practically no damage, excluding a little detachment from the lat-
4.2. Experimental assessment eral steel boundary elements. During the cyclic test, specimen 1
shows the tendency to maintain some plastic deformation in the
Results of an experimental campaign carried out on downscaled dissipative elements and vertical displacements (Fig. 15) are accu-
specimens (scaled down by 2/3 from the full scale) are presented in mulated in the lower interface between the steel frame and the
order to: validate the predicted behaviour of the proposed SRCW, infill wall. No cracks are detected within the concrete wall. It can
evaluate possible problems related to the realization of such sys- be inferred that, mainly due to the low number of shear studs con-
tem, and highlight the influence of different shear stud distribu- necting the RC wall to the steel boundary elements, the wall
tions along the steel frame perimeter. The specimens, behaves as a rigid body within the steel frame, avoiding any dam-
respectively named ‘‘1” and ‘‘2” are shown in Fig. 11 and can be age, except for the corner zones. The first cycles confirm so the lat-
both classified as semi-integral infilled wall. Specimen 1 is charac- tice brace resisting mechanism foreseen during the design of the
terized by the presence of shear studs only in the steel frame cor- specimen adopting the design methodology described within this
ner zones (situation very close to a non-integral infilled wall) while paper. However, due to the continuous accumulation, cycle after
specimen 2 has shear studs distributed all along the steel frame cycle, of the vertical displacements in the lower edge, the force
perimeter, excluding the dissipative zones (situation very close to application point of the compressed concrete diagonal strut moves
a integral infilled wall). from the reinforced and stiffened steel corner, to the non-
In both cases the RC wall is 12 cm thick and the reinforcement dissipative column of the steel frame, as schematically represented
layout is made up of a couple of welded steel meshes in Fig. 16. The change of the force application point causes the
150 mm 150 mm of diameter 8 mm bars, one for each side of unforeseen and unwanted failure of the specimen due to an exces-
sive shear deformation of the non-dissipative vertical steel ele-
ment (Fig. 17a). At the same time, the spalling of the concrete on
2500
the opposite lower corner of the infill wall and the complete
First yielding at 1st floor (compression) detachment of the infill wall from the steel frame occurs (Fig. 17b).
2000 Practically no other damage is visible within the RC wall.
Base Shear [kN]
2500
2000
1000
Fig. 11. SRCW Steel frame specimens: (a) Specimen 1 with shear studs only in the corner; (b) Specimen 2 with shear studs all along the perimeter (excluded the dissipative
zones).
in the dissipative elements (Fig. 20), but the resulting vertical dis- frame. Due to the developed displacement, the vertical and hori-
placement causes the gradual opening of the main cracks instead zontal reinforcing bars crossing the main crack break (Fig. 21),
of the detachment of the wall from the lower edge of the steel causing the loss of some horizontal force carrying capacity. After
A. Dall’Asta et al. / Engineering Structures 141 (2017) 144–158 155
Fig. 13. Loading history of tests on (a) Specimen 1 and (b) Specimen 2.
Fig. 14. Specimen 1: (a) cyclic force-displacement curve; (b) first loading and unloading phase.
Fig. 16. Specimen 1: (a) initial configuration; (b) configuration anticipating failure.
Fig. 17. Lower corners of Specimen 1 after failure: (a) shear failure of the non-dissipative zone and (b) spalling of the concrete and complete detachment by the steel frame.
Fig. 18. Specimen 2: (a) cyclic force-displacement curve; (b) first loading and unloading phase.
of the reinforcing bars and the consequent modification of the by the presence of few shear studs, the vertical loads can likely
resisting mechanism, Specimen 2. In real applications, the presence bring also to the increasing of the ductility capacity, avoiding
of the gravity loads transmitted at the different levels of the struc- unwanted failure mode such as the one evidenced by the experi-
ture by the floors should so bring positive effects on the resisting mental test on Specimen 1. Further studies are so needed to
system, mitigating the formation of horizontal cracks within the numerically and experimentally assess the influence of the vertical
wall, increasing the horizontal load carrying capacity and the re- loads on both the system configurations.
centering one. In system similar to Specimen 1, so characterized
A. Dall’Asta et al. / Engineering Structures 141 (2017) 144–158 157
Fig. 19. Specimen 2: cracking (a) at the end of the first unloading phase and (b) at the end of the first cycle (red line: main crack; blue lines: diffused cracking). (For
interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Fig. 21. Specimen 2 at the end of the test: (a) global view; (b) failure of the steel reinforcements in tension crossing the main crack.
158 A. Dall’Asta et al. / Engineering Structures 141 (2017) 144–158
imen 1 and Specimen 2) and to experimentally assess the influence [14] Mallick DV, Severn RT. Dynamic characteristics of infilled frames. Proc Inst Civ
Eng 1968;39:261–88.
of vertical loads on the system behaviour.
[15] Liauw TC, Lee SW. On the behavior and analysis of multi-story infilled frames
subject to lateral loading. Proc Inst Civ Eng 1977;63:641–56.
Acknowledgments [16] Liauw TC. Test on multistory infilled frames subject to dynamic lateral loading.
ACI J 1979;76:551–64.
[17] Liauw TC, Kwan KH. Static and cyclic behaviors of multistory infilled frames
Support for this research from the European Commission, with different interface conditions. J Sound Vib 1985;99:275–83.
Research Fund for Coal and Steel, Steel Technical Group TGS 8 [18] Makino M, Kawano A, Kurobane Y, Saisho M, Yoshinaga K. An investigation for
(RFSR-CT-2010-00025) and from the Italian Department of Civil the design of framed structures with infill walls. In: Proceedings of the seventh
world conference on earthquake engineering, vol. 4; 1980. p. 369–72.
Protection within the Italian Research Project RELUIS-DPC 2014- [19] Makino M. Design of framed steel structures with infilled reinforced concrete
2018, is gratefully acknowledged. walls. In: Roeder CW, editor. Composite and mixed construction. New York
(NY): ASCE; 1985. p. 279–87.
[20] Tong X, Hajjar JF, Schultz AE, Shield CK. Cyclic behavior of steel frame
References structures with composite reinforced concrete infill walls and partially-
restrained connections. J Constr Steel Res 2005;61:531–52.
[1] Morino S. Recent developments in hybrid structures in Japan: research, design [21] Saari WK, Hajjar JF, Schultz AE, Shield CK. Behavior of shear studs in steel
and construction. Eng Struct 1998;20:336–46. frames with reinforced concrete infill walls. J Constr Steel Res
[2] Hajjar JF. Composite steel and concrete structural systems for seismic 2004;60:1453–80.
engineering. J Constr Steel Res 2002;58:703–23. [22] Sun G, He R, Qiang G, Fang Y. Cyclic behavior of partially-restrained steel frame
[3] Deierlein GG, Noguchi H. Overview of US-Japan research on the seismic design with RC infill walls. J Constr Steel Res 2011;67:1821–34.
of composite reinforced concrete and steel moment frame structures. J Struct [23] Braga F, Morelli F, Salvatore W, A Macroseismic approach for the evaluation of
Eng 2004;130(2):361–7. seismic risk. In: Kruis J, Tsompanakis Y, Topping BHV, editors. Proceedings of
[4] Spacone E, El-Tawil S. Nonlinear analysis of steel-concrete composite the fifteenth international conference on civil, structural and environmental
structures: state of the art. J Struct Eng 2004;130:159–68. engineering computing, Paper 91. Stirlingshire, UK: Civil-Comp Press; 2015.
[5] Zona A, Barbato M, Conte JP. Nonlinear seismic response analysis of steel- http://dx.doi.org/10.4203/ccp.108.91.
concrete composite frames. J Struct Eng 2008;134:986–97. [24] European Committee for Standardization, Eurocode 8: Design of structures for
[6] El-Tawil S, Harries K, Fortney P, Shahrooz B, Kurama Y. Seismic design of earthquake resistance – Part 1: General rules, seismic actions and rules for
hybrid coupled wall systems: state of the art. J Struct Eng 2010;136:755–69. buildings. EN 1998–1, December 2004.
[7] Dall’Asta A, Leoni G, Zona A, Hoffmeister B, Bigelow H, Degée H, Braham C, [25] European Committee for Standardization. Eurocode 3: Design of steel
Bogdan T, Salvatore W, Morelli F, Tsintzos P, Karamanos SA, Varelis GE, Galazzi structures – Part 1–1: General rules and rules for buildings. EN1993-1-1,
A, Medici E, Boni P, Innovative hybrid and composite steel-concrete structural August 2005.
solutions for building in seismic area, Final Report, EUR 26932 EN, European [26] European Committee for Standardization, Eurocode 4 – Design of composite
Commission; 2015. http://dx.doi.org/10.2777/85404. steel and concrete structures – Part 2: General rules and rules for bridges. EN
[8] Zona A, Degeé H, Leoni G, Dall’Asta A. Ductile design of innovative steel and 1994–1, January 2006.
concrete hybrid coupled walls. J Constr Steel Res 2016;117:204–13. [27] Rossi PP, Lombardo A. Influence of the link overstrength factor on the seismic
[9] Morelli F, Manfredi M, Salvatore W. An enhanced component based model of behaviour of eccentrically braced frames. J Constr Steel Res 2007;63:1529–45.
steel connection in a hybrid coupled shear wall structure: development, [28] Elghazouli AY. Assessment of European seismic design procedures for steel
calibration and experimental validation. Comput Struct 2016;176:50–69. framed structures. Bull Earthquake Eng 2010;8:65–89.
http://dx.doi.org/10.1016/j.compstruc.2016.08.002. [29] Zona A, Ragni L, Dall’Asta A. Sensitivity-based study of the influence of brace
[10] Braga F, Gigliotti R, Monti G, Morelli F, Nuti C, Vanzi I, et al. Post-seismic over-strength distributions on the seismic response of steel frames with BRBs.
assessment of existing constructions: evaluation of the shakemaps for Eng Struct 2012;37:179–92.
identifying exclusion zones in Emilia. Earthquakes Struct 2014;8:37–56. [30] European Committee for Standardization, Eurocode 2 – Design of concrete
[11] Braga F, Gigliotti R, Monti G, Morelli F, Nuti C, Vanzi I, et al. Speedup of post structures - Part 1–1: General rules and rules for buildings. EN 1992-1-1,
earthquake community recovery. The case of precast industrial buildings after November 2005.
the Emilia 2012 earthquake. Bull Earthq Eng 2014;12:2405–18. [31] ABAQUS user’s manual, version 6.8EF-2. Providence, RI, USA: Dassault
[12] Liauw TC, Kwan KH. Plastic theory of non-integral infilled frames. Proc Inst Civ Systèmes Simulia Corp.; 2008.
Eng Lond UK Part 1983;2(75):379–96.
[13] Liauw TC, Kwan KH. Plastic theory of infilled frames with finite interface shear
strength. Proc Inst Civ Eng Lond UK Part 1983;2(75):707–23.