Journal of Building Engineering 40 (2021) 102334

Contents lists available at ScienceDirect

Journal of Building Engineering

journal homepage: http://www.elsevier.com/locate/jobe

Earthquake-resistant CLT buildings stiffened with vertical steel ties

Stefano Pacchioli a, Luca Pozza a, Davide Trutalli b, *, Andrea Polastri c
a
Department of Civil, Chemical, Environmental and Materials Engineering, University of Bologna, viale Risorgimento 2, Bologna, 40136, Italy
b
Department of Civil, Environmental and Architectural Engineering, University of Padova, via Marzolo 9, Padova, 35131, Italy
c
Institute of Bioeconomy of the Italian Research Council (CNR IBE), via Biasi 75, San Michele all’Adige, TN, 38010, Italy

A R T I C L E I N F O A B S T R A C T

Keywords: The interest in multi-storey CLT buildings in seismic areas is leading to the development of new strategies to
Cross-laminated timber (CLT) increase the lateral stiffness of shear walls and to resist high tensile forces due to rocking. Both these purposes
Seismic design can be achieved with vertical steel ties placed at each shear-wall end, to directly transfer tensile forces from each
Steel ties
storey to the foundation. Three technologies are proposed for transferring forces from CLT panels to the ties: the
Multi-storey timber buildings
Timber structures
use of nailed plates, of screwed connectors, or directly by contact with a thick plate at the top of each storey wall.
The dynamic behaviour of CLT shear walls, representing the bracing system of a building and anchored with the
aforementioned technologies, has been investigated by means of dynamic analyses and a comparison with the
use of common nailed plates or screwed connections without ties. Results, varying the number of storeys and the
seismic mass, show that the proposed technology is an effective strategy to increase the feasibility of multi-storey
CLT buildings. Complementary non-linear static analyses have been performed to evaluate the actual displace­
ment capacity and ductility of the systems.

1. Introduction height and slenderness of the shear walls. Finally, the increasing
architectural needs for internal free spaces are leading to optimize the
Wood-based engineered products like Cross-Laminated Timber (CLT) number of shear walls in the building, increasing further the strength
are characterized by high in-plane strength and stiffness of the timber demand to connections.
panel conferred by the cross-lamination of solid-wood boards [1], which Various strategies have therefore been developed for slender multi-
give the multi-storey structure good resistance to vertical and lateral storey CLT buildings in high-seismicity areas [6,7]:
loads [2].
A key role in the mechanical behaviour of CLT structures subjected to 1. Hybrid timber-concrete [8] or timber-steel systems [9,10];
lateral loads (wind or earthquake) is assumed by connections, generally 2. Use of post-tensioning bars, to reduce tensile forces at connections
hold-downs and angle brackets, which are assigned the function of and assuring recentering of shear walls [11–16];
connecting the panels together and to concrete foundation. These nailed 3. Rocking coupled shear walls with vertical joints [17,18];
brackets were originally developed for the use in light-frame systems 4. Reduction of inertial forces using connections with high ductility and
and were not originally conceived to be particularly strong and ductile dissipative capacity, employing the post-elastic hysteretic behaviour
[3]. The adoption of these connections in CLT structures has required to of steel or the friction to dissipate energy during a strong earthquake,
increase thickness, steel grade, number of nails and to add new stiffeners preserving the structure from damages [19–25].
to reduce local deformations. Nevertheless, it has been demonstrated
that stiffness and load-bearing capacity of such connections may not be All these strategies require the adoption of special technologies,
sufficient to realize CLT buildings taller than five or six storeys in innovative connections or the coupling of different materials, and the
high-seismicity areas, due to the very high tensile forces at hold-downs consequent development of new design methodologies, normally not
and the excessive lateral flexibility of the structure [4,5]. The rocking implemented by regulations.
contribution takes on great importance with increasing seismic action, In this work, an original earthquake-resistant system based on the

* Corresponding author.
E-mail addresses: stefano.pacchioli2@unibo.it (S. Pacchioli), luca.pozza2@unibo.it (L. Pozza), davide.trutalli@dicea.unipd.it (D. Trutalli), andrea.polastri@ibe.
cnr.it (A. Polastri).

https://doi.org/10.1016/j.jobe.2021.102334
Received 28 May 2020; Received in revised form 10 February 2021; Accepted 22 February 2021
Available online 6 March 2021
2352-7102/© 2021 Elsevier Ltd. All rights reserved.
S. Pacchioli et al. Journal of Building Engineering 40 (2021) 102334

use of vertical steel ties as alternative or in addition to traditional nailed

plates or screwed connections is analysed. According to this technology,
high-strength vertical CLT cantilevers can be realized to brace the
structure, optimizing the amount of connections in the building, without
the use of prestress, special connections or hybrid structures.
A technology based on the use of strong and stiff steel ties may
reduce the dissipative capacity of the structure, compared to the use of
nailed plates or the dissipative connections mentioned above. For this
reason, a non-dissipative seismic design of the wall systems has been
followed in the calculations presented in the following Sections,
adopting a behaviour factor q equal to 1.5 according to Eurocode 8 [26].
It is worth noting anyway that this choice does not penalize particularly
the design of the structure, as current regulations limit significantly the
behaviour factor associated to CLT buildings. For example, current Eu­
ropean Seismic Code gives a maximum value equal to 2 [26], and cur­
rent Italian Code for Constructions equal to 2.5 [27]. The design with a
behaviour factor equal to 1.5, i.e., in accordance with low-dissipative
Fig. 1. Conceptual representation of the steel ties coupled with nailed plates.
structural behaviour concept [26], leads to a significant increase of
3D view and detail (only the front face is shown).
the lateral stiffness of the structure: damages after an earthquake are
limited and the structure is less flexible also for wind action, with a clear
benefit also in terms of reduced discomfort for wind-induced vibrations withstand only the tensile force deriving from rocking at that storey,
in high-rise buildings [28]. whereas the tensile forces derived from the storeys above are directly
transferred to the foundation through the ties. This results in reducing
2. Description of the technology the tensile forces at the nailed connections, permitting of overcoming
the typical limitations in the realization of multi-storey buildings.
In low- and mid-rise CLT structures, tensile and shear forces, which To avoid eccentricities, the nailed plates and the steel ties should be
are generated at each storey level from lateral loads, are generally placed at both sides of the wall. Since the steel tie is continuous from the
resisted respectively by hold-downs and angle brackets. Alternative foundation to the top storey, it is necessary to realize a slot through the
connections can be used, suitably designed to optimize one or more timber floors. Furthermore, if required, it is possible to realize the tie
targets: strength, ductility, dissipative capacity. Independently from the with various steel elements and then to restore its continuity by welding
type of connection, they are attributed both to resist the above­ or bolting.
mentioned forces and to limit the lateral displacements of the structure.
As the number of storeys of the building increases, there is an increase in 2.2. Steel ties coupled with the screwed connection X-RAD
tensile forces at connections especially for slender shear walls. As
already mentioned, the rocking contribution to lateral displacements The system is based on the combination of the screwed connection X-
takes on increasingly importance as the height of the building grows, RAD [29] and the steel ties, which work together, Fig. 2.
with a consequent increase in the deformability of the structure. In detail, the steel ties are bolted to the X-RAD at each floor level. As
To limit the lateral displacements and to withstand the high tensile the strategy described in Section 2.1, the ties are continuous along the
forces in multi-storey CLT buildings, connections with high stiffness and height of the shear wall and can be composed by various bolted or
strength are needed. Traditional hold-downs, composed of a steel plate welded elements. At the corners of the building, the steel ties can be
and many nails or screws, may not be enough. For this reason, steel ties, opportunely realized with steel L profiles, whereas elsewhere steel plates
which may consist of steel plates, bars or profiles have been analysed, to can be placed at the narrow sides (thickness) of the wall, Fig. 2. Ac­
work in addition to or in place of traditional hold-downs or other con­ cording to this arrangement of the ties, eccentricities are avoided. As for
nections with the same purpose. the ties combined with nailed plates, the X-RAD for each storey must
In detail, three original connection systems are proposed to resist the withstand only the tensile force deriving from rocking at that storey,
tensile forces due to rocking: whereas the tensile forces derived from the storeys above are directly
transferred to the foundation through the tie.
a) steel ties coupled with nailed plates;
b) steel ties coupled with the screwed connection X-RAD [29];
c) steel ties only.

The proposed systems are conceived to withstand high seismic ten­

sile forces that arise in multi-storey CLT buildings braced with a limited
number of shear walls, limiting rocking, increasing the lateral strength
and stiffness of the structure and therefore reducing inter-storey drifts
and damages to structural and non-structural components, permitting to
increase the height of the building, also in high-seismicity areas. The
following subsections present the concept of these three technologies.

2.1. Steel ties coupled with nailed plates

The system is composed by steel ties welded to steel plates, which are
nailed to the panels at each storey level, Fig. 1. In this way, the steel ties
connect the nailed plates together and transmit the tensile forces derived
from them directly to the concrete foundation. According to this strat­ Fig. 2. Conceptual representation of the steel ties coupled with X-RAD. 3D
egy, the nailed plates work as hold-downs at each storey but have to view and detail (only the front face is shown).

2
S. Pacchioli et al. Journal of Building Engineering 40 (2021) 102334

2.3. Steel ties only Table 1

Seismic mass per storey relative to the modelled shear wall.
This strategy, unlike the previous ones, does not involve the coupling Building A Building B
of ties with other connections. The rocking behaviour at each storey is
Dimensions 17.1 × 15.5 m 23.3 × 15.5 m
directly prevented by thick steel plates, in direct contact with the top Floor mass 14.0 t 19.0 t
narrow side of each wall panel. In this case, the ties can be realized with Roof mass 11.3 t 15.3 t
two steel bars placed at both sides of the wall; the bars are secured to the Wall mass 1.5 t 1.5 t
plates at each floor level, Fig. 3.
Compared to the other two systems (i.e., nailed plates or X-RAD
panels of the shear walls to the foundation and in-between storeys and to
combined with steel ties), this is definitely the stiffest, since the defor­
withstand the seismic forces: traditional hold-downs and angle brackets
mation contribution related to nails or screws is avoided. Moreover, this
[31] or the innovative strong connection X-RAD [29]. Since for the
strategy should be also the most cost-effective, since dowel-type fas­
considered high earthquake intensity, the seismic demand for the
teners and additional brackets are avoided.
five-storey shear walls (in terms of force and/or stiffness) exceeded the
capacity of these two connection systems, steel ties have been used to
3. Numerical model and analyses
strengthen and stiffen the structure, according to the conceptual repre­
sentations in Section 2. The use of steel ties allowed also to extend the
In this Section, the effectiveness of the technology is evaluated,
analyses to the seven-storey shear walls.
presenting parametric Linear Dynamic Analyses (LDA) of multi-storey
Table 2 summarizes all the analysed configurations, with the
CLT shear walls, which represent the bracing system of a building,
following meaning of the labels:
supposed to be realized in an area with high earthquake hazard. Results
are reported in Section 4. Then, in Section 5 results of non-linear static
- The first letter (A or B) identifies the dimension of the building, that
analyses are discussed to assess the actual capacity of the system and to
is, the different storey mass assumed for the single shear wall;
compare the response of the new strategies with traditional systems.
- The number (3, 5 or 7) identifies the number of storeys of the shear
wall;
3.1. Description of the buildings - The last letters identify the connection system used or the combi­
nation of them: “H” is for hold-down, “X” is for X-RAD (i.e., screwed
Two CLT buildings have been considered, having rectangular plan connection), “T” is for steel ties (see Section 2, Fig. 3), “NP + T” and
with dimensions of 17.1 × 15.5 m (Building A) and 23.3 × 15.5 m “X + T” are for a combination of nailed plates or X-RAD with the steel
(Building B), and a variable number of storeys (3, 5 and 7 storeys), with ties (hereafter called also combined configurations or strengthened
an inter-storey height of 3 m. The structure is braced with six shear walls configurations, see Section 2, Figs. 1 and 2).
per direction for a total of 12 shear walls in the building, symmetrically
arranged in plan. The base length of the walls is equal to 6 m. They are
3.3. Numerical models
intended to act only as lateral-load resisting systems, whereas all the
vertical gravitational loads are resisted by a frame, which is supposed to
The numerical models of the buildings were developed and analyses
be perfectly pendular, in order to neglect its effects in the lateral stiffness
were performed using the SAP2000 Finite Element code [32]. The
of the building. The perfect regularity of the buildings and the rigid floor
following finite elements were used: two-dimensional elastic shell ele­
diaphragms allowed to study the behaviour of the whole structure
ments to model the CLT panels; uniaxial elements to simulate the actual
modelling only one shear wall, with the same number of storeys of the
stiffness of the connections; one-dimensional truss elements to simulate
building and one sixth of the mass.
the steel ties. Fig. 4 shows the models of the shear walls, performed with
The mass was computed according to the seismic combination of EN
the characteristics hereafter summarized.
1990 [30], assuming floor and roof dead loads equal to 2.5 kN/m2, floor
In the configurations H and NP + T (Fig. 4 a,c), the connections have
live loads equal to 2 kN/m2 and roof live loads equal to 0.5 kN/m2. The
been modelled with uniaxial elements according to the component
seismic storey masses are listed in Table 1.
approach presented in Polastri et al. (2019) [4]. It can be noted that,
according to the technology presented in Section 2.1, in the model NP +
3.2. Earthquake-resistant connections
T (Fig. 4 c), the tensile forces due to rocking at each storey level are
firstly resisted by the nailed plates and then introduced in the steel tie
Two technologies have been preliminarily considered to connect the
and directly transferred to foundation. The values to be assigned to these
models are: the axial stiffness of the hold-down kH,i or of the nailed plate
kNP,i; the shear stiffness of the element representing all the
angle-brackets in a storey kA,i; the Modulus of Elasticity ET and the
cross-section area AT of the truss elements representing the steel ties. In
the configurations X and X + T (Fig. 4 b,d), the X-RAD system that
connects two panels at the same corner is represented by an assembly of
four uniaxial elements: the two vertical elements simulate the stiffness at
90◦ to the horizontal (kX,v); the other two elements the stiffness at 0◦ to
the horizontal (kX,h). The steel tie is represented also in this case with ET
and AT. Finally, in the configurations T (Fig. 4 e), the steel ties are
directly fixed to the top of each panel, for all the storeys, in order to
transfer directly the force from the wall to the tie, according to the
technology described in Section 2.3.

3.4. Design of the shear walls and stiffness of the connection systems

Fig. 3. Conceptual representation of the steel ties. 3D view and detail (only the The shear walls were designed performing iterative LDA [26], as­
front face is shown). suring, for each iteration, the consistency between strength and stiffness

3
S. Pacchioli et al. Journal of Building Engineering 40 (2021) 102334

Table 2
Analysed configurations of the shear walls and labels.
Labels

Config. Connection system Building A Building B

Rocking Sliding 3 storeys 5 storeys 7 storeys 3 storeys 5 storeys 7 storeys

H Hold-downs Angle brackets A3H A5H - B3H B5H -

NP + T Nailed plates and steel ties Angle brackets - A5NP þ T A7NP þ T - B5NP þ T B7NP þ T
X X-RAD X-RAD A3X A5X - B3X B5X -
X+T X-RAD and steel ties X-RAD - A5X þ T A7X þ T - B5X þ T B7X þ T
T Steel ties Angle brackets - A5T A7T - B5T B7T

Fig. 4. Numerical models of the shear walls. Configurations with: (a) hold-downs “H”; (b) X-RAD “X”; (c) nailed plates and steel ties “NP + T”; (d) X-RAD and steel
ties “X + T”; (e) steel ties “T′′

of connections in the model, up to fulfilment of verifications for Ultimate

Limit State (ULS) and Damage Limitation State (DLS) [33]. As antici­
pated in Section 3.2, H and X configurations for the five-storey buildings
did not fully comply with these verifications and the strengthening with
the ties was required. Nevertheless, the results are anyway shown also
for these configurations for comparative purposes.
The site location of Perugia in the region of Umbria has been chosen
because it is representative of a high seismic hazard zone in Italy. It is
characterised by the seismic frequency spectra reported in Fig. 5, ac­
cording to Italian Regulation for Constructions [27], assuming building
foundations resting on ground type C and behaviour factor q equal to 1.5
or 2.0. A q equal to 2.0 was assumed for all the non-strengthened con­
figurations (H and X), according to a dissipative structural behaviour in
medium ductility class [26]. For the configurations with vertical steel
ties (NP + T, X + T and T), it was assumed equal to 1.5. As mentioned
above, the choice of designing these configurations according to a
low-dissipative structural behaviour [26] was made since a reduced Fig. 5. Design seismic spectra of Perugia in Italy, ground type C, behaviour
dissipative capacity is expected by using steel ties, which represent the factors q equal to 1.5 or 2.
leading earthquake-resistant system in the building.
The following connections have been considered in the models: - Angle brackets TTF200 for floor-wall joints or TCF200 for wall-to-
foundation joints, with 30 nails [35];
- Hold-downs WHT340 with 20 nails or WHT620 with 52 nails [34]; - 4-mm thick nailed plates with 12 nails (NP12) or 26 nails (NP26);
- X-RAD [36].

4
S. Pacchioli et al. Journal of Building Engineering 40 (2021) 102334

Table 3
Strength and stiffness of the earthquake-resistant connections.
Rocking Sliding Rocking and sliding

Metal connector WHT340 WHT620 NP12 nailed plate NP26 nailed plate TTF200 TCF200 X-RAD [29]
hold-down [4] hold-down [4] [37] [37] angle bracket [4] angle bracket [4]
Elastic stiffness (kN/mm) 5.70 13.25 5.80 12.58 8.21 8.48 15.2 (0◦ /90◦ )
17.3 (180◦ /270◦ )
Strength (kN) 46.20 93.70 41.58 90.09 39.10 39.10 Capacity domain

Table 3 lists strength and stiffness values assumed for connections, actually leads to a significant advantage in terms of rocking behaviour.
according to values given in Polastri et al. (2019) [4] for hold-downs and Contrary to the non-strengthened five-storey buildings, both ULS and
angle brackets and in Polastri et al. (2018) [29] for X-RAD. Strength and DLS verifications are fulfilled for all the five- and seven-storey NP + T
stiffness of nailed plates were obtained multiplying the values available and X + T configurations. Finally, as expected, the T configuration
in Izzi et al. (2016) [37] relative to a steel-timber connection with one 4 resulted the most cost-effective strategy in terms of optimization in the
× 60-mm ring-shank nail, by twelve or twenty-six nails for NP12 and number of connections.
NP26 respectively. The steel ties were assumed to be realized with
rectangular steel plates (or alternative shapes such as L or UPN profiles), 4. Results of linear dynamic analyses
realized in class S355, Modulus of Elasticity ET equal to 210,000 MPa
[38]. In this Section, results from LDA are summarized and discussed in
The iterative design procedure resulted in the arrangement of con­ terms of fundamental periods of vibration, tensile forces at connections,
nections listed in Table 4 for the configurations H, NP + T and T. With inter-storey drifts and displacements.
reference to the configurations X and X + T, one X-RAD per panel corner
was assumed for all configurations and storeys, with the exception of the 4.1. Fundamental periods of vibration
7-storey shear walls (A7X + T and B7X + T), for which a couple of X-
RAD was added for all the storeys at the middle of the panel base and Fig. 6 shows the results of the analyses in terms of fundamental
panel top, to increase the shear strength and comply with the seismic period of vibration of the structure, which has been proven to be
demand. For all the configurations with steel ties, an overall cross- strongly dependent on the rocking behaviour of the CLT shear wall and
section area of the ties AT equal to 1500 mm2 was assumed for the 5-sto­ stiffness of connections that resist this mechanism [39]. The periods TB
rey shear wall, and equal to 2250 mm2 for the 7-storey shear wall. and TC of the seismic spectra defined by Eurocode 8 [26] (see Fig. 5) are
It is worth noting that the design of connections and ties depends also represented in Fig. 6 to identify the limit of the constant spectral
strongly on the chosen aspect ratio of the shear wall: slender panels may acceleration branch (plateau). The configurations without steel ties (H
require very strong ties; on the contrary, large panels may require high and X) for both the buildings show similar fundamental periods of vi­
lateral forces to activate rocking, even if vertical loads are supported by brations. The use of steel ties combined or not with the nailed plates and
the frame. A linear design is anyway sufficient to have a correct balance X-RAD (NP + T, X + T, T) increases strongly the global stiffness of the
among strength and stiffness of all the connections used [33]. shear walls, resulting in a substantially lower fundamental period and a
To compare the configurations in terms of resulting amount of con­ higher spectral acceleration. This is evident comparing results for the
nections per panel (Table 4), it is necessary to remember that the five-storey configurations in Fig. 7 (dashed black lines), where the
strengthening with the steel ties leads to an increase of the seismic de­ fundamental periods of vibration have been superimposed to the design
mand due to an increase of stiffness and consequent decrease of periods spectra.
of vibration and due to the lower behaviour factor used, Fig. 5. The
consequent increase of strength demand clearly results in an increase of 4.2. Tensile forces
the number of angle-brackets per storey to resist higher shear forces, as
can be noted by comparing results among the five-storey configurations. Fig. 8 shows the results of the analyses in terms of maximum tensile
However, despite this apparent disadvantage, the addition of steel ties forces due to rocking. The maximum values were recorded always in the

Table 4
Arrangement of the earthquake-resistant connections per panel of the multi-storey shear wall. Configurations H, NP + T and T.
Storey 1 2 3 4 5 6 7

A3H 4 WHT620 4 WHT340 2 WHT340 – – – –

3 TCF200 3 TTF200 2 TTF200
A5H 4 WHT620 4 WHT620 4 WHT340 4 WHT340 2 WHT340 – –
3 TCF200 3 TTF200 2 TTF200 2 TTF200 2 TTF200
A5NP + T 4 NP26 2 NP26 2 NP26 2 NP12 2 NP12 – –
5 TCF200 5 TTF200 4 TTF200 3 TTF200 2 TTF200
A5T 5 TCF200 5 TTF200 4 TTF200 3 TTF200 2 TTF200 – –
A7NP + T 4 NP26 4 NP26 4 NP26 4 NP26 2 NP26 2 NP26 2 NP12
7 TCF200 6 TTF200 6 TTF200 5 TTF200 4 TTF200 3 TTF200 2 TTF200
A7T 7 TCF200 6 TTF200 6 TTF200 5 TTF200 4 TTF200 3 TTF200 2 TTF200
B3H 4 WHT620 4 WHT340 2 WHT340 – – – –
3 TCF200 3 TTF200 2 TTF200
B5H 4 WHT620 4 WHT620 4 WHT340 4 WHT340 2 WHT340 – –
3 TCF200 3 TTF200 2 TTF200 2 TTF200 2 TTF200
B5NP + T 4 NP26 2 NP26 2 NP26 2 NP12 2 NP12 – –
7 TCF200 6 TTF200 5 TTF200 4 TTF200 3 TTF200
B5T 7 TCF200 6 TTF200 5 TTF200 4 TTF200 3 TTF200 – –
B7NP + T 4 NP26 4 NP26 4 NP26 4 NP26 2 NP26 2 NP26 2 NP12
8 TCF200 7 TTF200 7 TTF200 6 TTF200 5 TTF200 4 TTF200 3 TTF200
B7T 8 TCF200 7 TTF200 7 TTF200 6 TTF200 5 TTF200 4 TTF200 3 TTF200

5
S. Pacchioli et al. Journal of Building Engineering 40 (2021) 102334

Fig. 6. Fundamental periods of vibration. Units: s.

the B5H configuration, which is very close to the limit of about 187 kN of
a couple of WHT 620 in the wall end, see Table 3. Despite the higher
spectral acceleration and consequent higher inertial forces due to the
lower period of vibration and to the lower behaviour factor assumed, the
strengthening with the steel ties allows to reduce the forces in the
connections, and to have the possibility of increasing the number of
storeys.
The T configurations behave similarly to the NP + T ones. The
maximum tensile force at the base is almost the same, but in the case of
the T configurations, the force must be resisted entirely by the tie.
Conversely, in the NP + T and X + T configurations, tension in the ties is
lower, due to the contribution of nailed plates and X-RAD at the base of
the building, which are directly anchored to the foundation. The T
configuration is confirmed to be more effective than the NP + T and X +
T configurations, since only the ties are used, avoiding additional nailed
or screwed connections to withstand tensile forces. Nevertheless, the
combined configurations may become further interesting, when used in
combination with dissipative device (e.g. Refs. [19–25]). In this way, the
dissipators permit to reduce at each storey the inertial force transmitted
from the panels to the tie. It is worth recalling also the possibility of
pre-assembling the connections to the panels, making the X + T
Fig. 7. Fundamental periods of vibration and design spectra. (a) Building A; (b)
configuration particularly advantageous in terms of prefabrication and
Building B.
fast installation of the ties to the structure, see Section 2.2.

connections at the base of the shear walls. The tensile force is showed for
the rate resisted by the hold-downs (H), by the X-RAD (X), by the nailed 4.3. Displacements and inter-storey drifts
plates (NP) and by the steel tie (T).
The comparison between the configurations without steel ties (H and Fig. 9 and Fig. 10 show the results of the analyses in terms of
X) for both the buildings shows similar forces. It is worth noting that for maximum inter-storey lateral drift and maximum top displacement of
the 5-storey shear walls, very high tensile forces in the hold-downs and the shear walls.
X-RAD have been recorded, reaching about 180 kN in the hold-downs of The configurations without steel ties (H and X) for both the buildings

Fig. 8. Maximum tensile forces at hold-downs, nailed plates, X-RAD connections and steel ties. Units: kN.

6
S. Pacchioli et al. Journal of Building Engineering 40 (2021) 102334

Fig. 9. Maximum inter-storey drifts and DLS limit. Units: mm.

Fig. 10. Maximum top displacements. Units: mm.

show similar top displacements, with an inter-storey drift slightly higher comply fully with the limit of 0.5% in Eurocode 8 [26]. In these
for the H configurations, mainly due to the greater rocking deformation. strengthened configurations, shear deformations take on greater
In these configurations, the global lateral displacements are mainly due importance and contribute for about 50% to the global lateral
to rocking for about an 80% rate and the contribution of shear de­ displacements.
formations is limited to about 20%.
It can be noted that the use of the steel ties and the design of the 5. Complementary analyses and discussions
structure according to a low-dissipative structural behaviour, result in
limiting the lateral displacements of the building, and therefore in To complete and validate the analyses discussed in Section 4, com­
limiting the damage to structural and non-structural components. plementary analyses have been performed. Section 5.1 presents the re­
The 5-storey shear walls analysed in this work demonstrate that the sults of linear analyses to evaluate the effects of variation of stiffness of
addition of the steel ties almost halves the top displacements and inter- ties per storey; Section 5.2 presents a validation of analyses presented in
storey drifts, and, also for the 7-storey shear walls, displacements and Section 4 with non-linear static analyses, to evaluate the actual capacity
drifts are smaller than the 5-storey configurations without ties and of the analysed techniques.

Table 5
Cross-section areas AT of the steel ties, units mm2. In brackets the working rate of the steel ties.
Storey 1 2 3 4 5 6 7

A5NP + T 1500 (60%) 1500 (60%) 1500 (38%) 1500 (21%) 1500 (8%) – –
A5NP + T* 1500 (54%) 1500 (54%) 950 (53%) 550 (53%) 200 (48%) – –

A5X + T 1500 (53%) 1500 (53%) 1500 (35%) 1500 (21%) 1500 (10%) – –
A5X + T* 1500 (48%) 1500 (48%) 1000 (47%) 600 (49%) 300 (46%) – –

A5T 1500 (82%) 1500 (58%) 1500 (38%) 1500 (21%) 1500 (7%) – –
A5T* 1500 (74%) 1100 (68%) 700 (67%) 400 (68%) 150 (65%) – –

A7NP + T 2250 (58%) 2250 (58%) 2250 (44%) 2250 (33%) 2250 (22%) 2250 (12%) 2250 (6%)
A7NP + T* 2250 (55%) 2250 (55%) 1750 (54%) 1300 (54%) 850 (56%) 500 (50%) 200 (48%)

A7X + T 2250 (60%) 2250 (60%) 2250 (47%) 2250 (35%) 2250 (24%) 2250 (14%) 2250 (7%)
A7X + T* 2250 (56%) 2250 (56%) 1750 (56%) 1300 (55%) 900 (55%) 600 (48%) 250 (49%)

A7T 2250 (75%) 2250 (65%) 2250 (50%) 2250 (36%) 2250 (23%) 2250 (13%) 2250 (5%)
A7T* 2250 (70%) 2000 (69%) 1500 (70%) 1100 (68%) 700 (70%) 400 (67%) 150 (66%)

7
S. Pacchioli et al. Journal of Building Engineering 40 (2021) 102334

Table 6
Results for the optimized configurations and percentage change from the non-optimized counterparts.
Period Maximum Tensile Force Top displacement Inter-storey drift

(s) (kN) (mm) (mm)

A5NP + T* 0.71 9.23% 385.6 − 9.93% 32.8 8.25% 8.6 11.69%

A5X + T* 0.77 10.00% 312.2 − 9.06% 34.0 7.26% 9.2 8.24%
A5T* 0.70 11.11% 387.2 − 11.38% 31.4 8.65% 8.3 12.16%
A7NP + T* 0.90 4.65% 549.7 − 4.48% 43.1 9.67% 8.3 9.21%
A7X + T* 0.96 6.67% 514.2 − 6.48% 43.6 11.22% 8.5 8.97%
A7T* 0.88 6.02% 560.9 − 6.09% 42.8 11.46% 8.1 14.08%

5.1. Effects of optimizing the capacity of steel ties per storey and 7-storey Building A is given in Fig. 11. Three dimensionless pa­
rameters have been introduced, namely: the period index τ equal to T1/
In order to optimize the steel ties, the cross-section can be gradually Tc that shows how far the fundamental period of vibration is from the
reduced from the foundation to the top of the building, keeping almost plateau; the tensile force index φ equal to Fmax/(M⋅PGA); and the drift
constant among the storeys the ratio between the tensile strength de­ index δ equal to the actual drift divided by 0.5%. In these parameters, T1
mand to the capacity of the ties; this ratio is hereafter called “working is the fundamental period of vibration, Tc is equal to 0.476 s, Fmax is the
rate”. maximum tensile force, M is the total seismic mass relative to the shear
To evaluate the influence of such optimization on the fundamental wall according to Table 1, PGA is the peak ground acceleration equal to
period of vibration, tensile forces and displacements, additional LDA 2.61 m/s2, 0.5% is the drift limit for DLS.
were carried out for the strengthened configurations of the 5- and 7-sto­ The comparison in Fig. 11 for the 5-storey building confirms the
rey Building A. The optimized configurations are marked with a star; the substantial decrease of fundamental periods of vibration and the in­
additional configurations are therefore: A5NP + T*; A5X + T*; A5T*; crease of tensile forces for the strengthened shear walls. The improve­
A7NP + T*; A7X + T*; A7T*. The cross-section areas of the ties resulting ment is evident in terms of DLS verification, confirming that it is fulfilled
from the design and the relative working rates are listed in Table 5 and for all the strengthened configurations of the 5- and 7-storey shear walls.
compared with the non-optimized counterparts. It is worth noting that On the contrary, δ is higher than 1 for the A5H configuration, confirming
the optimized design resulted in an important reduction of the cross- that the DLS verification is not fulfilled and stiffening of the shear wall is
section of the ties for the upper storeys. Conversely, the number of the needed. Finally, the results for the optimized configurations are
other connection elements (nailed plates, angle brackets and X-RAD), confirmed to be similar to the non-optimized ones, with slight decrease
remained unchanged with respect to those of Table 4 and negligible of lateral stiffness and consequent increase of inter-storey drift.
variations occurred in terms of periods (below 12%) and therefore in
forces acting in connections (below 12%), as can be seen in Table 6.
Finally, the decrease in the overall stiffness of the optimized structures 5.2. Evaluation of the actual capacity with non-linear static analyses
resulted in a negligible increase of inter-storey drifts (below 15%) and
top displacements (below 12%). The actual mechanical behaviour of the proposed innovative systems
A final comparison among all the analysed configurations for the 5- and their ultimate capacities have been assessed and compared with
those of traditional systems performing non-linear static analyses. In

Fig. 11. Period, force and drift index for the 5- and 7-storey configurations of Building A. (a) 5 storeys; (b) 7 storeys.

8
S. Pacchioli et al. Journal of Building Engineering 40 (2021) 102334

Table 7
Results of non-linear static analyses for triangular deformed shape.
Kel Kpl Ductility Maximum inter- Failure
(kN/ (kN/ storey drift at failure mode
mm) mm) (‰)

A5H 1.47 0.8 1.6 7.1 Hold-

downs
A5X 1.5 0.77 1.65 6.2 X-RAD
A5NP 3.54 1.5 1.25 5.4 Angle
+T brackets
A5X + 2.9 1.6 1.35 5.3 X-RAD
T
A5T 4.6 2.9 1.15 5.1 Angle
brackets

Fig. 12. Base shear force vs. horizontal top displacement curves for tradi­
tional systems. Table 8
Results of non-linear static analyses for constant deformed shape.
Kel Kpl Ductility Maximum inter- Failure
(kN/ (kN/ storey drift at failure mode
mm) mm) (‰)

A5H 1.93 1.17 1.7 7.5 Hold-

downs
A5X 1.95 1.14 1.8 6.6 X-RAD
A5NP 4.04 1.5 1.34 5.7 Angle
+T brackets
A5X + 3.35 1.6 1.48 5.5 X-RAD
T
A5T 5.1 3.1 1.15 5.4 Angle
brackets

to failure of the hold-down connections, which represent the weakest

Fig. 13. Base shear force vs. horizontal top displacement curves for the con­ elements. On the contrary, the wall systems equipped with ties show a
figurations with steel ties. clear transition between elastic and post-elastic phase, allowing the
exploitation of the ductile phase of the shear resistant connection ele­
ments. The results in terms of global elastic (Kel) and post-elastic (Kpl)
stiffness, ductility, maximum inter-storey drift at failure and actual
detail, the 5-storey configurations of building A have been studied (A5H,
failure mode are listed in Tables 7 and 8.
A5X, A5NP + T, A5X + T, A5T) applying two imposed deformed shapes,
The obtained results highlight that all the configurations with steel
i.e., two horizontal force patterns per floor: one providing the same
ties show stiffness much greater than configurations without steel ties
forces for each floor (constant distribution) and the other proportional to
(about two-times greater). The displacement capacity showed by all the
the height of each floor (perfect triangular distribution). It is expected
configurations is between 60 and 80 mm, except for the case with steel
that the actual dynamic response can be within the two capacity curves
ties only, which reaches about 40 mm at failure. The ductility values
(base shear force vs. horizontal top displacement) obtained with the two
obtained for all the investigated configurations are anyway limited, also
force patterns. The models described in Fig. 4 were adapted for the non-
for the case of CLT systems assembled with traditional high-capacity
linear analyses, modifying the linear elements representing connections
connectors, according to similar observations in the literature [41].
(hold-downs, nailed plates, angle brackets and X-RAD) with non-linear
This demonstrates that the design of connections with a linear approach
elements, according to an elastic-plastic behaviour reproducing the
is adequate to reach a balanced distribution of connections in the shear
experimental behaviour available in Refs. [4,29,37] and adding
wall. Finally, Fig. 14 illustrates conceptually the overall deformation
compression-only elements to simulate the contact between panels at
distribution along the height of the wall. The weak link, where the first
each floor level and between panel and foundation. Friction effects were
failure occurs, is highlighted with a circle, according to the position of
not considered in the model, according to Franco et al. (2019) [40]. The
connections in Fig. 4. The force pattern proportional to the height of
non-linear static analyses returned the base shear force vs. top
each floor (perfect triangular distribution) is shown since deformation is
displacement curves shown in Fig. 12 and Fig. 13. These curves stop at
similar to the other one. It can be noted that the configurations without
the achievement of failure in a connection element, according to the
steel ties are characterized by a prevalent rocking behaviour. On the
failure modes listed in Table 7 and Table 8. A limited ductile behaviour
contrary, in the configurations strengthened with steel ties, the sliding
can be noted, regardless of the force pattern. In detail, the wall system
contribution for each storey is evident.
equipped with the traditional connections behaves almost elastically up

9
S. Pacchioli et al. Journal of Building Engineering 40 (2021) 102334

Fig. 14. Deformation shape of the shear walls for the force pattern proportional to the height of each floor (perfect triangular distribution). (a) A5H; (b) A5X; (c)
A5NP + T; (d) A5X + T; (e) A5T.

6. Conclusions confirmed this conclusion and have demonstrated that the proposed use
of steel ties to realize high-strength shear walls can address both these
Three earthquake-resistant systems for timber shear-wall structures issues, improving the feasibility of multi-storey CLT buildings in high-
have been presented, based on the use of vertical steel ties as alternative seismicity areas. This strategy makes also possible a reduction of the
or in addition to traditional nailed plates or screwed connections. The number of earthquake-resistant walls and connections in the building.
proposed technologies are intended for multi-storey CLT buildings and The parametric analyses have shown that the spectral accelerations
have the aim of withstanding high tensile forces due to rocking of the increase by adding the ties to the structure, due to an increase in the
timber panels, transmitting them directly from each storey level to the global stiffness of the shear walls and consequent decrease of periods of
concrete foundation. At the same time, the ties stiffen the structure, vibration. Nevertheless, the tensile forces in the connections signifi­
reducing lateral displacements and inter-storey drifts. The fastening of cantly decrease, being partially or totally resisted by the steel ties,
the ties to the panels at each storey level can be realized with traditional depending on the fastening strategy adopted. Results in terms of top
nailed plates or screwed connections. As an alternative to the use of displacements and inter-storey drifts have demonstrated that the use of
dowel-type fasteners, the ties can be directly secured to thick steel steel ties improves significantly the stiffness and the elastic response of
plates, placed in direct contact to the top narrow side of each wall panel. the building, with clear advantages in complying with requirements for
Previous studies available in the literature proved that the main is­ damage limitation state.
sues of multi-storey CLT buildings subjected to seismic action are the Finally, the reduction of the cross-section area of the ties from the
limited lateral stiffness and the high tensile forces concentrated in hold- foundation to the top of the building according to the seismic demand
down connections, which become the critical components of the build­ leads to the optimization of the structural material without affecting the
ing. The parametric dynamic analyses presented in this work have response of the building.

10
S. Pacchioli et al. Journal of Building Engineering 40 (2021) 102334

The earthquake-resistant systems presented in this work have proven [13] X. Sun, M. He, Z. Li, F. Lam, Seismic performance of energy-dissipating post-
tensioned CLT shear wall structures II: dynamic analysis and dissipater
a good seismic performance, making them worthy of further de­
comparison, Soil Dynam. Earthq. Eng. 130 (2020) 105980, https://doi.org/
velopments and research. The non-linear static analyses have demon­ 10.1016/j.soildyn.2019.105980.
strated the actual response of the analysed techniques, which ensure a [14] D.S. Pilon, A. Palermo, F. Sarti, A. Salenikovich, Benefits of multiple rocking
significant increase of global stiffness without relevant reduction of segments for CLT and LVL Pres-Lam wall systems, Soil Dynam. Earthq. Eng. 117
(2019) 234–244, https://doi.org/10.1016/j.soildyn.2018.11.026.
ductility and displacement capacity when the ties are nailed or screwed [15] A.W. Wilson, C.J. Motter, A.R. Phillips, J.D. Dolan, Modeling techniques for post-
to the CLT panels. tensioned cross-laminated timber rocking walls, Eng. Struct. 195 (2019) 299–308,
The results presented in this work only apply to the buildings ana­ https://doi.org/10.1016/j.engstruct.2019.06.011.
[16] X. Sun, M. He, Z. Li, F. Lam, Seismic performance of energy-dissipating post-
lysed as many changes in the building geometry, occupancy type, etc. tensioned CLT shear wall structures I: shear wall modeling and design procedure,
may affect the results. Additional analyses can be performed to inves­ Soil Dynam. Earthq. Eng. 131 (2020) 106022, https://doi.org/10.1016/j.
tigate the possible coupling of the proposed systems with ductile de­ soildyn.2019.106022.
[17] A. Hashemi, P. Zarnani, R. Masoudnia, P. Quenneville, Seismic resistant rocking
vices, able to dissipate energy, in order to reduce the inertial forces coupled walls with innovative Resilient Slip Friction (RSF) joints, J. Constr. Steel
introduced in the steel ties at each floor level. Res. 129 (2017) 215–226, https://doi.org/10.1016/j.jcsr.2016.11.016.
[18] J. Liu, F. Lam, Numerical simulation for the seismic behaviour of mid-rise CLT
shear walls with coupling beams, in: WCTE 2014 - World Conf Timber Eng Proc,
Author contribution 2014, pp. 2–3.
[19] A. Baird, T. Smith, A. Palermo, S. Pampanin, Experimental and numerical study of
The authors contributed equally to the research, analysis of results U-shape flexural plate (UFP) dissipators, in: NZSEE Conf, 2014.
[20] M. Latour, G. Rizzano, Cyclic behavior and modeling of a dissipative connector for
and production of the paper.
cross-laminated timber panel buildings, J. Earthq. Eng. 19 (2015) 137–171,
https://doi.org/10.1080/13632469.2014.948645.
Declaration of competing interest [21] T. Schmidt, H.J. Blass, Dissipative joints for CLT shear walls, Int Netw Timber Eng
Res (2017). Meeting fifty, Kyoto, Japan, paper no. 50-15-2.
[22] W.Y. Loo, C. Kun, P. Quenneville, N. Chouw, Experimental testing of a rocking
The authors declare that they have no known competing financial timber shear wall with slip-friction connectors, Earthq. Eng. Struct. Dynam. 43
interests or personal relationships that could have appeared to influence (2014) 1621–1639, https://doi.org/10.1002/eqe.2413.
the work reported in this paper. [23] R. Scotta, L. Marchi, D. Trutalli, L. Pozza, X-bracket, A High-Ductility and
Dissipative Connection for Earthquake-Resistant Cross-Laminated Timber
Structures, Research Report), 2019.
Acknowledgements [24] R. Scotta, L. Marchi, D. Trutalli, L. Pozza, A dissipative connector for CLT
buildings: concept, design and testing, Materials 9 (2016) 139, https://doi.org/
10.3390/ma9030139.
The authors would like to thank STP s.r.l. company for founding the [25] F. Sarti, A. Palermo, S. Pampanin, Quasi-static cyclic testing of two-thirds scale
research and Dr. Francesca Paoloni for the cooperation in the graphic unbonded posttensioned rocking dissipative timber walls, J. Struct. Eng. 142
representations of the technology. (2015) 1–14, https://doi.org/10.1061/(ASCE)ST.1943-541X.0001291.
[26] CEN. EN 1998-1, Eurocode 8: Design of Structures for Earthquake Resistance -Part
1: General Rules, Seismic Actions and Rules for Buildings, CEN, Brussels, 2013.
References [27] Mit. D.M. 17.01.18, Aggiornamento delle «Norme tecniche per le costruzioni»,
2018 (in italian).
[1] R. Brandner, G. Flatscher, A. Ringhofer, G. Schickhofer, A. Thiel, Cross laminated [28] G.C. Foliente, Wood products: performance in seismic and high wind events, in:
timber (CLT): overview and development, Eur J Wood Wood Prod 74 (2016) Encycl. Mater. Sci. Technol., Elsevier, 2001, pp. 9707–9712, https://doi.org/
331–351, https://doi.org/10.1007/s00107-015-0999-5. 10.1016/B0-08-043152-6/01762-9.
[2] M. Izzi, D. Casagrande, S. Bezzi, D. Pasca, M. Follesa, R. Tomasi, Seismic behaviour [29] A. Polastri, I. Giongo, A. Angeli, R. Brandner, Mechanical characterization of a pre-
of Cross-Laminated Timber structures: a state-of-the-art review, Eng. Struct. 170 fabricated connection system for cross laminated timber structures in seismic
(2018) 42–52, https://doi.org/10.1016/J.ENGSTRUCT.2018.05.060. regions, Eng. Struct. 167 (2018) 705–715, https://doi.org/10.1016/j.
[3] D. Trutalli, L. Marchi, R. Scotta, L. Pozza, Capacity design of traditional and engstruct.2017.12.022.
innovative ductile connections for earthquake-resistant CLT structures, Bull. [30] CEN. EN 1990, Eurocode - Basis of Structural Design, CEN, Brussels, 2005.
Earthq. Eng. 17 (2019) 2115–2136, https://doi.org/10.1007/s10518-018-00536- [31] M. Izzi, A. Polastri, M. Fragiacomo, Modelling the mechanical behaviour of typical
6. wall-to-floor connection systems for cross-laminated timber structures, Eng. Struct.
[4] A. Polastri, M. Izzi, L. Pozza, C. Loss, I. Smith, Seismic analysis of multi-storey 162 (2018) 270–282, https://doi.org/10.1016/j.engstruct.2018.02.045.
timber buildings braced with a CLT core and perimeter shear-walls, Bull. Earthq. [32] Computers and Structures I, 2018. SAP2000 v20.
Eng. 17 (2019) 1009–1028, https://doi.org/10.1007/s10518-018-0467-9. [33] A. Polastri, L. Pozza, Proposal for a standardized design and modeling procedure of
[5] A. Ceccotti, C. Sandhaas, M. Okabe, M. Yasumura, C. Minowa, N. Kawai, SOFIE tall CLT buildings, Int J Qual Res 10 (2016) 607–624, https://doi.org/10.18421/
project - 3D shaking table test on a seven-storey full-scale cross-laminated timber IJQR10.03-12.
building, Earthq. Eng. Struct. Dynam. 42 (2013) 2003–2021, https://doi.org/ [34] European Technical Assessment. ETA-11/0086, Three-dimensional Nailing Plate
10.1002/eqe.2309. (Angle Brackets and Hold-Downs for Timber-To-Timber or Timber-To-Concrete or
[6] M. Stepinac, I. Šušteršič, I. Gavrić, V. Rajčić, Seismic design of timber buildings: Steel Connections), Nordhavn, Danmark, 2015.
highlighted challenges and future trends, Appl. Sci. 10 (2020) 1380, https://doi. [35] European Technical Assessment. ETA-11/0496, Three-dimensional Nailing Plate
org/10.3390/app10041380. (Angle Bracket for Timber-To-Timber or Timber-To-Concrete or Steel
[7] D. Ugalde, J.L. Almazán, H. Santa María, P. Guindos, Seismic protection Connections), Nordhavn, Danmark, 2018.
technologies for timber structures: a review, Eur J Wood Wood Prod 77 (2019) [36] European Technical Assessment, ETA-15/0632, Three-dimensional Nailing Plate,
173–194, https://doi.org/10.1007/s00107-019-01389-9. 2015. Vienna, Austria.
[8] J.W.G. Van De Kuilen, A. Ceccotti, Z. Xia, M. He, Very tall wooden buildings with [37] M. Izzi, G. Flatscher, M. Fragiacomo, G. Schickhofer, Experimental investigations
cross laminated timber, Procedia Eng 14 (2011) 1621–1628, https://doi.org/ and design provisions of steel-to-timber joints with annular-ringed shank nails for
10.1016/j.proeng.2011.07.204. Cross-Laminated Timber structures, Construct. Build. Mater. 122 (2016) 446–457,
[9] S. Tesfamariam, S.F. Stiemer, C. Dickof, M.A. Bezabeh, Seismic vulnerability https://doi.org/10.1016/j.conbuildmat.2016.06.072.
assessment of hybrid steel-timber structure: steel moment-resisting frames with [38] CEN. EN 1993-1-1, Eurocode 3: Design of Steel Structures - Part 1-1: General Rules
CLT infill, J. Earthq. Eng. 18 (2014) 929–944, https://doi.org/10.1080/ and Rules for Buildings, 2015.
13632469.2014.916240. [39] D. Casagrande, S. Pacchioli, A. Polastri, L. Pozza, Influence of the rocking behavior
[10] X. Zhang, M. Fairhurst, T. Tannert, Ductility estimation for a novel timber–steel of shearwalls on the fundamental period of CLT structures, Earthq. Eng. Struct.
hybrid system, J. Struct. Eng. 142 (2016) E4015001, https://doi.org/10.1061/ Dynam. (2021), https://doi.org/10.1002/eqe.3420.
(ASCE)ST.1943-541X.0001296. [40] L. Franco, L. Pozza, A. Saetta, M. Savoia, D. Talledo, Strategies for structural
[11] S. Pei, J.W. van de Lindt, A.R. Barbosa, J.W. Berman, E. McDonnell, J. Daniel modelling of CLT panels under cyclic loading conditions, Eng. Struct. 198 (2019)
Dolan, et al., Experimental seismic response of a resilient 2-story mass-timber 109476, https://doi.org/10.1016/j.engstruct.2019.109476.
building with post-tensioned rocking walls, J. Struct. Eng. 145 (2019), 04019120, [41] D. Casagrande, A. Polastri, T. Sartori, C. Loss, M. Chiodega, Experimental
https://doi.org/10.1061/(ASCE)ST.1943-541X.0002382. campaign for the mechanical characterization of connection systems in the seismic
[12] D. Moroder, T. Smith, A. Dunbar, S. Pampanin, A. Buchanan, Seismic testing of design of timber buildings, in: WCTE 2016 - World Conf, Timber Eng., 2016.
post-tensioned Pres-Lam core walls using cross laminated timber, Eng. Struct. 167
(2018) 639–654, https://doi.org/10.1016/j.engstruct.2018.02.075.

11