Content uploaded by Cristiano Loss
Author content
All content in this area was uploaded by Cristiano Loss on Sep 01, 2014
Content may be subject to copyright.
Experimental Tests of Cross-Laminated Timber Floors to be used in
Timber-Steel Hybrid Structures
Cristiano Loss1, Maurizio Piazza2, Riccardo Zandonini3
ABSTRACT: Hybrid structural systems assembled connecting steel elements and cross-laminated timber panels (CLT) can
be a valid alternative to traditional systems in the construction of residential buildings. Such systems can combine the
industrialized construction technology typical of steel systems with the advantages offered by CLT panels, namely lightness
and geometric stability. Moreover, CLT panels are timber-based products, and wood is recognized as an eco-friendly and
eco-compatible material. In hybrid structural systems, the seismic-resistant capacity of the structure can be achieved by
ensuring an adequate transmission of actions among the resistant elements, namely plain timber panels (floor and wall) and
steel frame elements (beams and columns). Specifically, the interaction between the steel frame and the wood panels shall
ensure both horizontal and vertical bracing to floors and walls, respectively. The work presented hereafter concerns the
study of the connections to be used among the individual building components of the horizontal elements, with the aim of
developing an effective collaboration among the materials, maximizing the level of prefabrication and industrialization of
the final components. In particular, the preliminary results of the experimental tests carried out on full-scale steel-to-timber
floor specimens, loaded by in-plane actions, will be presented.
KEYWORDS: Floor diaphragm, Hybrid Structures, In‐plane Diaphragm Stiffness, Cross Laminated Timber, CLT,
Connections
1 INTRODUCTION
123
Hybrid structures are structural systems assembled through
the connection and mutual interaction of different
structural materials. The combination of materials is
defined for different levels of collaboration, whether they
are referred to the construction element or to parts of the
structure. The goal is to obtain a final highly engineered
product/system, resorting to the highest resistant properties
of each material. More generally, the hybrid systems, if
well designed and engineered, can offer benefits -in terms
of load bearing and deformation capacity, seismic and
dynamic resistance, as well as fire resistance- higher than
those offered by each material used in non-collaborating
systems.
Hybrid structures are rapidly increasing in many
industrialized countries, particularly Canada, New
1
Cristiano Loss, Department of Civil, Environmental and
Mechanical Engineering, University of Trento, via Mesiano 77, I-
38123 Trento, Italy. E-mail: cristiano.loss@unitn.it
2
Maurizio Piazza, Department of Civil, Environmental and
Mechanical Engineering, University of Trento, via Mesiano 77, I-
38123 Trento, Italy. E-mail: maurizio.piazza@unitn.it
3
Riccardo Zandonini, Department of Civil, Environmental and
Mechanical Engineering, University of Trento, via Mesiano 77, I-
38123 Trento, Italy. E-mail: riccardo.zandonini@unitn.it
Zealand, USA, Japan and Europe. However, the current
knowledge is still limited and research activities are still in
progress. Within the field of hybrid construction systems
in which the timber is the main collaborating structural
material, the study primarily concerns medium-high rise
buildings [1, 2], as well as the definition of solutions for
the construction of high-rise buildings [3].
This work aims at studying the seismic behaviour of mixed
steel-timber systems, with particular reference to hybrid
structures assembled with steel frames and CLT panels.
Particularly, the system presented hereafter is designed so
that the seismic actions on each floor are transferred to the
ground firstly through the floor diaphragms - braced with
mechanically joined CLT panels - and then to the vertical
walls which act as a stabilizing element. The vertical
bracing elements are frames with stiff steel elements or,
alternatively, with X-shaped steel diagonals or properly
connected CLT panels. The transfer of forces from the
floor diaphragms to the vertical bracing is ensured by
accurately designed and correctly installed connections.
The floor plays a key role regarding the in-plane forces
distribution on the individual vertical bracing elements.
According to the stiffness and strength of the horizontal
elements, there can be two limit behaviours: flexible
diaphragm or stiff diaphragm. We are referring to the
distribution of the in-plane horizontal forces according to
the area of influence (tributary area) or to the bracing
stiffnesses, respectively.
This work aims at investigating the behaviour - expressed
in terms of in-plane deformation and bearing capacity - of
floors assembled through mechanically coupled steel
beams and cross-laminated timber panels (CLT). Within
this field, some engineered connection solutions, designed
to allow both a quick assembly on site and a possible
prefabrication in the workshop, will be presented. The
paper includes some of the load-slip curves directly
measured from the tests carried out on full-scale floor
specimens. The tests have been performed at the Materials
and Structures Test Laboratory of the University of Trento.
The paper describes (i) the building construction system
and its operation principle (Section 2), (ii) the expected
behaviour of the floor, mainly compared to the role and
structural hierarchy of the different connections used
(Section 3), (iii) the experimental tests carried out (Section
4), (iv) some experimental results (Section 5). Finally,
some discussions are reported in Section 6, regarding the
applicability of the proposed solutions.
2 HYBRID SYSTEM CONSTRUCTION
2.1 FLEXIBLE SOCIAL-HOUSING BUILDINGS
This paper deals with some aspects regarding the definition
and development of a new construction system for the
social housing market. The system is designed to make
different materials - firstly steel and timber - structurally
collaborate, in order to develop a construction system with
marked performance and architectonic flexibility features.
The structure is required to have some features: modularity
and easy assembly, flexibility and architectonic
disposition, also regarding the changing trends in housing
needs for the period of the building life cycle. In particular,
the technologies based on a “dry assembly” of construction
elements made with low environmental impact materials
(referring to the LCA) are more and more requested. A
structure with a steel bearing frame (beams, columns
and/or bracing elements) and horizontal elements
composed with CLT panels can then represent a valid
solution, mainly if compared to the building technologies
that use traditional materials.
2.2 STRUCTURAL SYSTEM DESCRIPTION
The reference hybrid system for this study is characterized
by a regular structural grid (Figure 1), standardized enough
and repeatable in space, with steel linear and timber
bidimensional elements. Among the possible structural
steel-timber configurations, we chose the technology based
on pendular frames with concentrical bracing X-shaped
elements in both directions.
The floors are assembled with steel beams forming a grid
structure, which support the CLT panels, properly
connected to ensure both the in-plane and out-of-plane
diaphragm behaviour to withstand the horizontal forces
and vertical loads, respectively.
Figure 1: 3D axonometric section of the building with
corresponding structural plan of the type plane
Figure 2: Mounting scheme of CLT panels and distribution
in-plane layout
In a more advanced stage of the research, the vertical
elements made of CLT panels - or alternatively of walls
made of CLT panels inserted into the steel frames - tied to
the floors in the two main directions will be also
considered. The structure as a whole will have a box
behaviour, mainly to ensure the stability and to withstand
to the horizontal forces.
Figure 2 describes the structural plan of a generic
intermediate storey for a prototype building, with a detail
Primary
beams Secondary
beams Bracing
systems Columns
of the structural arrangement of the bidimensional CLT
elements. The 3-dimensional behaviour towards the
horizontal forces is governed by the interaction of the
vertical bracing elements with the in-plane floors. The
paper will focus only on the floor behaviour, paying
attention to the diaphragm behaviour as well as to the
design hypothesis in the case of an infinitely stiff plane.
3 FLOORS BEHAVIOUR
3.1 DESCRIPTION OF FLOORS
With reference to the prototype building of Figure 1, in
relation to the possible location and distribution of the
vertical bracing in the two main directions, it is possible to
isolate a portion of the floor in order to maximize the
effects of the in-plane element actions. This floor
configuration has a rectangular shape with sides equal to 5
x 12 m, and is bound on both ends by a continuous and
linear 5 m bracing system. The floor is composed of a steel
frame with hot-rolled steel sections (HE-shaped profile
section).
(a)
(b)
(c)
(a) pair of cross full thread screws
with 45° inclination
(b) pair of cross full thread screws
with 45° inclination in space
(c) double surface stripes with self-
tapping screws
Figure 3: Joints for floor-floor edges of CLT panels
The steel elements of the frame are mutually connected by
means of metal bolted brackets. Some secondary steel
beams are fixed inside the frame and are arranged at a
constant pitch. Some CLT panels are placed over the frame
and mechanically joined to the beams with the use of
modern connecting devices. The connection solutions the
authors consider as more suitable to be used are those
deriving from the technology of timber/timber-concrete
and/or steel construction systems. The individual CLT
panels, with dimensions limited by production and
transport needs, are joined by contact and at the head
through pin-type mechanical connections, such as crossed
screws arranged at a constant pitch (Figure 3a and 3b), or
by using timber-based material, mechanically fastened
with self-tapping screws or nails (Figure 3c).
3.2 IN- AND OUT-OF-PLANE BEHAVIOUR
The operating principle of the floor is similar to an in-
plane timber diaphragm used in lightweight wood-frame
construction systems. The bracing function is ensured by
the beam-panels connections. The resistant mechanism is
activated as a result of deformations in the connectors,
placed at the edges of the floor grid to join the CLT panels
to the steel elements, as well as in the end- and side-joints
among the panels (Figure 4). With loads applied to the
floor plane, the secondary beams perform a stabilizing
function and prevent a possible instability of CLT panels.
Figure 4: In-plane behaviour of hybrid steel-CLT floors
Compared to the vertical loads, the transfer of actions is
carried out directly by contact among the steel primary and
secondary beams and the CLT panels. However, in order
to develop an effective and competitive system compared
to other traditional building solutions (for example steel-
concrete floors), the connections must also ensure an out-
of-plane collaboration (Figure 5).
As far as panel-panel connections are concerned, the
technology used in timber buildings offers effective
construction solutions both for the transfer of in-plane and
out-of-plane shear actions. Nevertheless, it is possible to
rely on proven solutions, supported by adequate laboratory
experiments: we can make reference to the research results
reported in Blass et al.'s [4], Mohammad et al.'s [5],
“BSPhandbuch” manual [6]. In any case, the use of
recognized Standard documents, such as DIN 1052:2008
or Eurocode 5:2009, allows to easily design the panel-to-
panel joints through simple calculation models. The
models are defined both for shear and tensile stressed
connectors, namely screws. For combined shear-tensile
actions, it is possible to apply the analytical model
proposed by Tomasi et al. [7].
The real innovation of the proposed hybrid system is the
techniques of connection among the steel structure and the
CLT panels. As a matter of fact, the current situation is
still too limited to rely on efficient and proven solutions, or
on the experience of new buildings.
The paper presents some connection technologies derived
from the current practice applied to composite timber-
Chord member
Intermediate
stiffer
Panel-to-beam
connection
Panel-to-panel
connection
CLT panel
End support
member
Chord member
Intermediate
stiffer
Panel-to-beam
connection
Panel-to-panel
connection
CLT panel
End support
member
Chord member
Intermediate
stiffer
Panel-to-beam
connection
Panel-to-panel
connection
CLT panel
End support
member
concrete [8] and steel-concrete [9] structures. The solutions
have been designed and engineered considering in
particular some technical aspects connected to the
mechanical behaviour of the materials, the installation
tolerances on site and the composite system stability. In
addition, the most practical aspects referred to the
mounting sequence and the cost were considered.
Figure 5: Out-of-plane behaviour of hybrid steel-CLT floors
3.3 PRE-DESIGN OF FLOORS
An extended experimental campaign has been carried out
in order to assess the connection behaviour among CLT
panels and the steel structure of the floor. With reference
to the structural system described above (Figures 1 and 2),
a floor module has been firstly isolated, and then designed.
The dimensions (cross sections) of the steel elements and
the CLT panels have been calculated according to the
loads acting on the building. In this stage, the floor has
been considered as collaborating exclusively in the plane
(diaphragm behaviour). The mechanical connections
arranged among the individual panels and between them
and the side beams have been dimensioned so as to ensure
a stiff behaviour, according to the calculation principles
contained in the American ASCE 7-05:2005 standard. In
this stage the in-plane stiff behaviour of floors is evaluated
neglecting the stiffness of connections used to join the
floor elements to the vertical bracing systems. When
dimensioning the connections, some simplified models
have been used, assuming some failure mechanisms of the
materials.
4 EXPERIMENTAL TESTS
The research activity carried out to define the floor
elements focused mainly on the experimental
characterization of the mechanical behaviour of the
connecting elements between steel beams and CLT panels.
The experimental campaign includes tests performed for
some typical floor configurations that differ in number,
type and geometry of connectors. We referred to push-out
(push and pull) tests, carried out controlling force and
displacement, and under monotonic, cyclic or oligo-cyclic
loads, according to the relevant international standard for
testing (EN 26891:1991, EN 12512:2001 and EN
1994:2004). The tests aim at measuring the behaviour
curve, force-slip (F-), for the different connection devices
used. Figure 6 shows one of the specimens after the
assembly, highlighting the classical test configuration with
two shear planes.
Figure 6: Front (Left) and lateral (Right) view of one test
specimen
4.1 TEST SPECIMENS
The elements that constitute the specimens, hot-rolled HE
shaped steel beams and CLT panels, have a fixed
geometry, while the connectors used vary from a specimen
to the other (Figure 7). For each specimen a different type
of connection has been defined and dimensioned. There
are “dry” mechanical devices (A-type), connections that
use an epoxy resin (B-type) and mixed mechanical devices
which combine the resistant properties of the resin and the
metallic connectors (C-type).
In this stage of the research, 9 different connection
typologies have been considered. Each connection has
been designed to ease the assembly operations of the
elements, in particular considering the installation
tolerances connected to the materials and the construction
system. The assembly method of the specimens has been
studied in detail in order to both respect the test principles
and reproduce the real situation in the worksite. Then a
mounting bench has been assembled (Figure 8), and all the
steps needed to avoid offsets and backlashes among the
coupled elements have been taken.
4.2 MATERIALS
The specimens have been assembled joining the certified
materials in compliance with the relevant product
standards, as requested by the new Italian building code. In
addition, some characterization tests of the materials have
been carried out so as to evaluate the effective mechanical
properties, here not included for the sake of brevity.
4.2.1 CLT
The CLT panels, provided by a local factory, were built
with 5 layers of C24 timber boards (according to EN
338:2009). The thickness of CLT panels is determined by
the design of floors for vertical loads. The grain direction
of the outer layers is oriented in the direction of the steel
beams. All specimens were assembled with panels of
Semi-rigid
connection Steel Beam
CLT panel
nominal thickness equal to 100 mm (CLT panel layup in
mm: 20/20/20/20/20). Density and moisture content were
measured on all CLT specimens, giving a medium value of
m=453 kg/m3 and an average moisture content of 11.7 %.
Figure 7: Drawing and geometry of a specimen
Figure 8: Mounting bench of the specimens: moment
related to a mounting stage
4.2.2 Steel member
The hot-rolled H-beams are made of mild steel S275
according to EN 10025:2004. The beams have been
processed and reinforced on one edge, welding a pre-
arranged flange with adequate holes to hook to the test
machine and to handle it. The support elements, ribs and
transversal stiffeners have been assembled starting from a
plate of the same material. The position of the support and
anchoring elements to the test setup has been studied to
ease the specimen positioning operations and consequent
instrumentation.
4.2.3 Connectors for shear resistance
The connection elements used for the assembly of the
specimens are devices engineered so as to make the most
of the currently used techniques provided for timber or
mixed timber-concrete construction systems, as well as for
steel-concrete collaborating systems. More in detail, for
the A category defined above, 5 solutions have been
studied and designed. For each solution, a different
installation procedure is scheduled, varying not only for
the equipment required but also for the different processing
time in the workshop and at the construction site,
respectively. In particular, we can find the following
configurations:
- A-1, threaded bar welded to the upper flange of
the beam and connected to the CLT panel by
means of a cold-shaped plate, screwed to the CLT
panel through full thread self-tapping screws with
a 45° inclination;
- A-II, threaded bar welded to the upper flange of
the beam and connected to the CLT panel through
a steel stripe, screwed to the CLT panel through
self-tapping screws inserted transversally to the
element;
- A-III, threaded bar welded to the upper flange of
the beam and connected to the CLT panel through
a thick perforated disc, screwed to the CLT panel
through full thread self-tapping screws with a 45°
inclination and radially arranged;
- A-IV, cold-shaped steel elements welded on both
sides of the beam flange and connected to the
CLT panel through full thread self-tapping screws
with a 45° inclination upward;
- A-V, thin L-shaped profiles welded on both sides
of the beam flange and connected to the CLT
panel through self-tapping screws inserted
transversally to the element;
Similarly, the B category includes 3 structural connection
solutions:
- B-I, threaded bar welded to the upper flange of
the beam and connected to the CLT panel through
the application of epoxy resin with aggregate;
- B -II, half-threaded bar welded to the upper flange
of the beam and connected to the CLT panel
through the application of epoxy resin with
aggregate;
- B-III, steel perforated blade, welded to the upper
flange of the beam and connected to the CLT
panel through the application of epoxy resin with
aggregate.
Finally, the C category is defined with the following
configuration:
- C-I, threaded bar welded to the upper flange of
the beam and connected to the CLT panel by
means of a cold-shaped plate, screwed to the CLT
panel through full-thread self-drilling screws with
a 45° inclination. In addition, the hole between
the bar and the panel is filled with epoxy resin
and aggregate.
For all the configurations described above, the CLT panels
have been properly arranged, performing, if necessary,
some holes or openings. In those cases which provide for
threaded bars on the upper flange of the profile, the
welding operation is carried out by means of an electric
gun, similarly to the technique of Nelson-type studs
installation. We point out here that the epoxy grout resin
used to fill the volume in holes between timber and steel is
a very common and cheap product. Other structural
Geometry
Drawing
CLT panels
(E1)
Steel beam
(E2)
950* 152*
800*
800*
AA
B
B
100*
A-A B-B
(E1)
(E2)
* Dimensions in mm
100*
adhesives with better mechanical properties are available
and should be used, as expected in a later stage of the
research.
Being an industrial development research study, specific
details -the dimensions of the connectors, the pitch and the
mechanical features- are omitted. Hence, the paper focuses
on the main features of the responses of the different types
of connectors, also in a comparative perspective.
4.3 SETUP AND SPECIMEN INSTRUMENTATION
The setup has been designed to carry out push-out (push
and pull) tests with a configuration based on a monotonic
and cyclic load and two shear planes (Figure 9a). The test
setup is equipped with a base and two equal components,
each formed from reinforced steel elements. The base and
components are joined to the reaction frame through
preloaded hardened bolts tightened by a pre-established
torque. The setup has been designed to withstand a
maximum traction force equal to 300 kN without
undergoing significant deformations. With this load
configuration, no excessive deformations - which could
influence the test data - are recorded and the deflection
measured during the acceptance tests of the system has
been maintained always below 1 mm. In any case, the
setup has been designed to withstand a maximum load of
500 kN.
The test setup has been inserted into a self-balancing
reaction frame (Figure 9b) and is tied to the base by means
of high resistant bolts, pre-loaded in order to minimize the
effects induced by the starting hole-bolt initial gaps. An
hydraulic electro-assisted actuator with push features equal
to +1000 kN and pull features equal to -600 kN is used in
the test. The actuator is tied on top of the reaction frame
and of the specimen, on the other side. The whole test
machine has been designed to generate shear forces in the
connections in the direction of the load application.
However, the lateral displacement among the elements has
been monitored, since it could arise due to non-expected
transversal forces or to a possible separation of the
components.
For the data acquisition, two Quantum X -HBM®- power
units, equipped with eight acquisition channels each one,
have been used. A sample rate equal to 5 Hz has been used
during the data acquisition. A measurement device has
been assigned to each channel. In particular, the
instruments are placed on the specimen as indicated in
Figure 9c. The instruments installed on the specimen are
strain gauge linear displacement transducers -LDT- AEP
50 and 100 (50 and 100 mm stroke, respectively) and they
detect the relative displacements among the elements. The
remaining three acquisition channels are connected to the
jack and to an instrument on the ground. The acquisition
channels connected to the jack allow to record both the
absolute displacement and the reached/imposed load level.
In any case, the sliding between the beam and the CLT
panels is obtained as the average value of four instruments
(Figure 9c). Similarly, the transversal displacement is the
mean of two instruments and considering the two ends -
high and low - of the panel. In this way it is possible to
derive any rotation of the elements for each of the two
main directions.
4.4 TESTING METHOD
Totally 18 experimental tests have been carried out,
applying defined load protocols. In particular, we identify
two groups of tests, carried out according to different
times. Each group is made of nine specimens assembled
according to the configurations described above. For the
first group, monotonic load-controlling tests have been
performed, whereas for the second group combined cyclic
and oligo-cyclic tests have been carried out.
The test methods have been defined respecting the
reference standard principles. The monotonic tests have
been carried out controlling the displacement, at a constant
speed of 0.05 mm/s, until reaching the failure load, here
conventionally considered as 50% of the maximum
registered load. The load protocol has been modified if
compared to what indicated in EN 26891:1991, but in any
case the test principle has been respected.
The combined cyclic-oligo-cyclic test has been carried out
comparing the protocols provided for by EN 1994:2004
(Eurocode 4) and EN 12512:2001.
In particular, the first two adjustment cycles (0.25vy and
0.5vy), provided for by EN 12512:2001 have been
removed, with a load principle controlling the force,
composed by incremental load cycles repeated 5 times in
the interval 0.05÷0.4 FM. The load protocol is the one
indicated in Figure 10, in which the yield sliding (vy) and
the maximum load (FM) have been estimated by the
monotonic tests. The protocol has been defined to simulate
the behaviour of a floor in operating state and is interested
by in-plane forces that bring it in the inelastic field of
deformation.
The load protocol of Figure 10 applied and specified for
each specimen has been modified to consider the effect
induced by the differential displacements generated due to
the deformability of the test setup and the reaction frame.
As a matter of fact, the deformation is partially related to
the elements softness, and partially to the hole-bolt gaps.
5 OUTCOMES
Only some numerical results and load-slip behaviour
curves (F-
) of the different floor connections tested are
reported here. Figure 11 includes and compares the
monotonic curves F-
, whereas Figure 12 separately
reports the cyclic-oligo-cyclic curves and the related
monotonic curves.
6 DISCUSSION OF RESULTS
This section briefly reports the main data obtained from
the experimental tests in order to assess the structural
efficiency of the connections.
The trends in Figures 11 and 12, together with the data of
Table 1, allow to judge the possibility of creating
composed steel-timber collaborating systems for
composite stress in-plane and out-of-plane actions (Figures
4 and 5, respectively). Some parameters in the table and
charts are expressed as ratio between the actual and the
reference values, the latter here named Fref, I and Fref, II,
respectively. The reference force Fref,I is the maximum
value measured from specimen B-III after the monotonic
test. Fref,II is simply derived as the ratio between Fref,I and
the number of connectors in the specimen B-III.
(a)
(b)
(c)
Instrumentation locations
LDT-1 Back left side
LDT-2 Back right side
LDT-3 Front left side
LDT-4 Front right side
LDT-5 Up-left lateral side
LDT-6 Down-left lateral side
LDT-7 Up-right lateral side
LDT-8 Down-right lateral side
*LDT, strain gauge linear displacement transducers
Figure 9: (a) Views of the test setup; (b) Test machine and
related reaction frame; (c) Specimen instruments
Figure 10: Load protocol
Table 1: Main test results for the different specimens (the
values are referred to a single connector)
ID
k/Fref,II
Fy
y
FM/Fref,
II
(FM)
n
1/mm
mm
mm
A-I
0.012
0.10
7.09
0.23
44.21
0.03
4
A-II
0.004
0.10
18.32
0.24
31.22
0.02
6
A-III
0.007
0.10
12.03
0.26
29.79
0.02
2
A-IV
0.067
0.12
1.67
0.14
8.44
0.24
4
A-V
0.019
0.04
1.86
0.05
12.47
0.23
12
B-I
0.307
0.17
0.53
0.17
4.36
0.71
6
B-II
0.207
0.14
0.68
0.18
13.62
0.62
6
B-III
2.290
0.91
0.40
1.00
9.29
0.84
2
C
0.249
0.17
0.71
0.18
10.24
0.42
4
k, stiffness according to EN 26891 | Fy, yield force |
y, yield slip | FM,
maximum load |
(FM), slip at maximum load |
, composite efficiency | n,
number of connector in a specimen | Fref,II, reference force defined as the
maximum force measured from tests ( here in specimen B-III).
With specific attention to the monotonic and envelope
curve of each specimen, Figures 12 shows that the
behaviour is basically ductile, even if the dissipative
capacity varies considerably in particular in relation to the
material wear. A delicate matter is represented by the
configurations that use epoxy resin: configurations B-I, II,
III and C-I. In fact, there is an abnormal deviation among
the monotonic curves and the envelope of the cyclic
curves. The trend is reversed compared to the expected
situation and it shows a bearing capacity higher in the
cyclic case than in the monotonic one. This phenomenon is
compromised by the process of specimen production. The
specimens are produced in different periods of time and
using different mixes, although coming from the same
supplier. Specific analyses shall be therefore carried out in
order to assess the actual variability of the bonding
material, even considered the processing (mixing and
pouring).
Set up
Frame reaction system
for testing used
Specimen
Hydraulic jack
LDT-1 LDT-2
LDT-3 LDT-4
LDT-5
LDT-6
LDT-7
LDT-8
Left side Right side
Back side
Front side
t, time
6
y
Displacement control
5 40% FM
, slip
4
y
2
y
1
y
0.75
y
Force control
Figure 11: Force-slip curves referred to each specimen and connector
A-I A-II A-III
A-IV A-V B-I
B-II B-III C-I
OBSERVED FAILURE MECHANISM
Breakage of steel bars: A-I, A-II, A-III, B-I, B-
II, B-III
Failure of epoxy resin layers: B-III
Failure in screws heads: A-V
Failure in screws heads and in steel parts: A-IV
F/Fref,I , ratio between the actual force (F) and the
reference force (Fref,I )
, slip obtained as mean of four measures
F/Fref,II , ratio between the actual force (F) and the
reference force (Fref,II )
Reference force is defined as the maximum force
measured from tests, here in specimen B-III.
Fref,I refers to the specimen B-III, while Fref,II is
specified for the single connector of such specimen.
0
0.2
0.4
0.6
0.8
1
1.2
010 20 30 40 50
Force ratio F/Fref,I [kN/kN]
Slip
[mm]
0
0.2
0.4
0.6
0.8
1
1.2
010 20 30 40 50
Force ratio F/Fref,II [kN/kN]
Slip
[mm]
EN 26891
EN 12512
EN 26891
EN 12512
(B) For single connector
(A) For specimen
Figure 12: Monotonic and cyclic response curves measured for each specimen (X axes: slip in mm; Y axes: ratio
between the actual force F and the reference Force Fref,I)
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
-10 -5 0 5 10
-1.6
-1.2
-0.8
-0.4
0
0.4
0.8
1.2
1.6
-20 -10 0 10 20
-0.6
-0.45
-0.3
-0.15
0
0.15
0.3
0.45
0.6
-10 -5 0 5 10
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
-5 -2.5 0 2.5 5
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
-20 -10 0 10 20
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
-20 -10 0 10 20
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
-30 -20 -10 0 10 20 30
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
-30 -20 -10 0 10 20 30
A-II
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
-30 -20 -10 0 10 20 30
A-I Monotonic
Cyclic Monotonic
Cyclic Monotonic
Cyclic
A-III
A-V
A-IV Monotonic
Cyclic
Monotonic
Cyclic Monotonic
Cyclic
B-I
B-IIIB-II Monotonic
Cyclic Monotonic
Cyclic Monotonic
Cyclic
C-I
In the out-of-plane behaviour perspective, from the
monotonic tests, we observe that no specimen was
interested by brittle failure mechanisms until the slip is
below or equal to approximately 10 mm. A more detailed
analysis regarding the single connector highlights the
impossibility of producing a composite steel-timber section
in bending for the solutions with mechanical “dry”
connections (A-type), except for case A-IV. For this last
case, the registered efficiency ≈0.24 can be increased
reducing/halving the pitch. Even the yield slip can be
considered compatible with the deflection limits related to
the service conditions [10]. The solutions that seem to be
more adequate for a composite solution are those with
epoxy resin, B-I, II and III, while the solution C-I doesn't
show big benefits. With particular reference to Figure 11,
it is to be noted that, regardless the types, in the
connections that provide for a bar welded on the beam, the
maximum bearing capacity is limited by the shear failure
of the same, both in the dry case and with epoxy resin.
If we only consider the diaphragm in-plane behaviour, it is
clear that the connections will have to ensure a resistant
capacity and stiffness so as to transfer the seismic actions
of the storeys to the ground, minimizing the corresponding
deformations. The solutions adequate for the purpose are
those that use epoxy resin (proper stiffness), even though
the mechanic solution A-IV is still valid. In any case, the
most outstanding solution from a performance point of
view is the configuration B-III. Another important aspect
can be observed for configuration A-III, with an increase
of the characteristic bearing capacity of the connector
induced by a favourable constraint condition. The increase
of resistance (FM) can also reach values near 10%, for
example if case A-I is compared to the similar
configuration A-III.
If we pay attention to the cyclic tests, another important
evaluation regards the initial permanent slip induced by the
variation of the operating loads (situation reproduced
through the load protocol of Eurocode 4). In cases A-I, A-
II and A-IV, in fact, permanent slip values equal to 2.5,
13.7 and 0.25 mm, respectively, are registered. Such
values are not allowable when an effective collaboration
among materials is requested. On the contrary, in the other
cases the permanent initial slip is negligible as should be
expected.
On the basis of these and other considerations, not
included for the sake of brevity, new experimental tests are
in progress on redefined and optimized connection
configurations.
ACKNOWLEDGEMENT
The authors gratefully acknowledge the ‘Premetal Spa’
factory for financing this study, within the research
program supported by the Autonomous Province of Trento,
as well as the laboratory team for their help during the
tests. The authors wish to thank also former students Luigi
Farinati and Alessio Chesani for their valuable
contribution to the research. Finally, a special thanks to the
laboratory technician Alfredo Pojer for his appreciated
work.
REFERENCES
[1] C. Dickof, S.F. Stiemer, and S. Tesfamariam. Wood-
steel hybrid seismic force resisting systems: seismic
ductility. In: 12th World Conference on Timber
Engineering (Session 49, Future trends 3), pages 199-
200, 2012.
[2] Ma. Zhog, and He. Minjuan. Experimental analysis of
timber diaphragm’s capacity on transferring horizontal
loads in timber-steel hybrid structure. In: 12th World
Conference on Timber Engineering (Strength and
Serviceability, Extreme events), pages 205-210, 2012.
[3] J. Weckendorf, and I. Smith. Multi-functional
interface concept for high-rise hybrid building systems
with structural timber. In: 12th World Conference on
Timber Engineering, (Session 20, Design 8), pages
192-197, 2012.
[4] H.J. Blass, I. Bejtka, and T. Uibel: Tragfähigkeit von
Verbin dungen mit selbstbohrenden Holzschrauben
mit Vollgewinde, Karlsruher Berichte zum
Ingenieurholzbau, Band 4, Lehrstuhl fur
Ingenieurholzbau und Baukonstruktionen,
Universitatsverlag Karlsuhe, ISSN 1860-093X, ISBN
3-86644-034-0, 2006 (German).
[5] M. Mohammad, B. Douglas, D. Rammer, and S.E.
Pryor: Connections in cross-laminated timber
buildings. In: E. Karacabeyli and B. Douglas, editors,
CLT handbook-U.S. edition, pages 177–230. FP
Innovations, Pointe-Claire, QC, 2011.
[6] G. Schickhofer, T. Bogensperger, and T.
Moosbrugger, editors: BSPhandbuch. Verlag der
Technischen Universität Graz, ISBN 978-3-85125-
076-3, 2009 (German).
[7] R. Tomasi, A. Crosatti, and M. Piazza. Theoretical and
experimental analysis of timber-to-timber joints
connected with inclined crews. Construction and
building materials, 24(9):1560-1571, 2010.
[8] D. Yeoh, M. Fragiacomo, M. De Franceschi, and K.
H. Boon: State of the art on timber-concrete composite
structures: literature review, Journal of Structural
Engineering, 137(10): 1085–1095, 2011.
[9] R. Eligehausen, W. Fuchs, P. Grosser, and G.
Genesio: Connections between steel and concrete. In:
The 2nd International Symposium, University of
Stuttgart, September 4th - 7th, Stuttgart, Germany,
2007.
[10] M. Piazza, R. Tomasi, and R. Modena: Strutture in
Legno, Hoepli, Milan, Italy, 2005 (Italian).
