-
An innovative prefabricated timber-concretecomposite system
Roberto Crocetti, Tiziano Sartori, Roberto Tomasi, Jose L. F.
Cabo
1 Abstract
A novel type of timber-concrete composite floor, consisting of
longitudinal glulambeams with a fibre reinforced concrete (FRC)
slab on the top is proposed. In orderto check some relevant
mechanical properties of such a floor, full-scale laboratorytests
along with numerical analyses were carried out. The shear connector
systemused in the investigation consisted of self-tapping screws
driven at an angle of 45◦ tothe grain direction of the glulam
beams. The manufacture of the structure occurredaccording to the
following steps: (a) the screws were inserted on the top of
theglulam beams; (b) the beams were rotated 180◦ about the
longitudinal axis andplaced in a concrete formwork; (c) the FRC was
cast into the formwork; (d) aftercuring of the FRC, the composite
floor was again rotated 180◦ about the longitudinalaxis into its
right position, i.e. with the FRC slab on the top side. Long term
tests andquasi-static bending tests were performed. It was found
that the proposed connectionsystem showed a very high degree of
composite action both during the long-termtesting and at load
levels close to the failure load. Furthermore, the assembly of
theprefabricated timber-concrete composite system revealed to be
very fast and easy.
Roberto CrocettiDepartment of Structural Engineering, Lund
University, Sweden, e-mail:[email protected]
Tiziano SartoriDepartment of civil, environmental and mechanical
engineering, University of Trento, , Italye-mail:
[email protected]
Roberto TomasiDepartment of civil, environmental and mechanical
engineering, University of Trento, , Italye-mail:
[email protected]
Jose L. F. CaboETS of Architecture, Polytechnic University of
Madrid (UPM), Spain e-mail: [email protected]
1
-
2 Roberto Crocetti, Tiziano Sartori, Roberto Tomasi, Jose L. F.
Cabo
2 Introduction
Timber-concrete composite structure consists of timber beams
effectively intercon-nected to a concrete slab cast on top of the
timber members. Most of the studiesperformed to date have focused
on composite systems where wet ordinary concretewas cast on top of
timber beams with mounted shear connectors. Even though suchsystems
have proven to perform very well from the point of view of statics
and dy-namics, in-situ concrete casting has some clear
disadvantages, for example, wasteof time due to concrete curing,
low stiffness and high creep, concrete shrinkage ef-fects on the
composite beam, high cost of cast-in-situ concrete slabs, etc.
Recently,composite systems where the concrete slab is prefabricated
off-site with shear con-nectors already embedded and then connected
to the timber beams on site have beeninvestigated ([4] and [7]).
The research presented herein focus on the use of com-posite
structure with very high prefabrication level, good performance and
shortconstruction time. Such composite structures are ”floor
modules” consisting of twoglulam beams with a a concrete slab on
the top.
3 Materials and methods
In order to investigate the behaviour of the proposed composite
system, three full-scale floors were built at the laboratory of
Structural Engineering, Lund University.The main dimensions of the
floor system are reported in Table 1.
Table 1 Geometry of the A, B and C composite system (dimensions
in [mm]). See also Fig. 1.Span (l) Slab width Slab thickness(h1)
Beam width (b2) Beam depth (h2) Beam spacing (i)
7200 800 50 115 360 585
Fig. 1 Geometry of the floor system
The geometry of the three tested floors was nominally identical.
The timber usedfor the manufacture of the floors was glulam GL30c.
The moisture content of the
-
An innovative prefabricated timber-concrete composite system
3
beams was approximately 12%. For the production of 1 m3 of fiber
reinforced con-crete, 45 kg of steel fibres and 480 kg of cement
were used, which gave a meanvalue of compression strength fc = 51
MPa for the concrete. In the following text,the three tested
specimens will be referred to: A, B and C. Specimens A and B
weretested on short term bending, whilst specimen C was tested on
long term bending.During the short-term bending tests, the load was
applied by an actuator in a dis-placement controlled manner. The
load was distributed on four lines perpendicularto the longitudinal
direction of the floor in order to induce stresses and
deformationsin the floor similar to those induced by a uniformly
distributed load q, see Fig. 2.
The total load applied to the specimen, the mid-span deflection,
and the relativeslip between slab and beam at the supports were
continuously measured during test-ing. For the long-term bending
test, a uniformly distributed load of 1 kNm2 was appliedon the slab
by means of sacks of cement. The mid-span deflection was
measuredover time in order to investigate the creep effects of both
timber and concrete.
3.1 Shear connectors and manufacture of the floor systems
In order to achieve composite action between the timber beams
and the concreteslab, self-tapping screws with dimensions d = 11mm
and l = 250mm were driveninto the timber beams before the concrete
was cast. The screws were driven at anangle of 45◦ to the
longitudinal directions with a spacing of 200 mm close to the
Fig. 2 Test setup
-
4 Roberto Crocetti, Tiziano Sartori, Roberto Tomasi, Jose L. F.
Cabo
supports and 300 mm in the middle part of the floor respectively
(Fig. 3). The mainfunction of the inclined screws is to transfer
the shear force from the concrete slabto the timber beam both by
shear in the direction parallel to the slip interface andmainly by
tension in the direction of the screw axis. The ultimate tensile
strength ofthe screws was approximately fu = 1250MPa.
Fig. 3 Screw positions
The screws were inserted on the top of the glulam beams, then
the beams wererotated 180◦ about the longitudinal axis and placed
in a concrete formwork. TheFRC was cast into the formwork, see Fig.
4. After curing of the FRC, the compositefloor was again rotated
180◦ about the longitudinal axis into its right position, i.e.with
the FRC slab on the top side.
4 Test
4.1 Preliminary tests
Preliminary tests were performed in order to obtain modulus of
elasticity of timberbeam, strength of concrete, withdrawal
resistance of screw to concrete connectionand strength of steel
screws.Standard compression tests were carried out on three
concrete cubes. The geometryand the compression strength of the
tested specimens are resumed in Table 2.
Screws with different penetration length were inserted in the
concrete cubes. Thevalue of withdrawal tests on these specimens are
reported in Table 3.
Non destructive bending tests were performed on two timber beams
in order toestimate the modulus of elasticity. The results are
reported in Table 4.
-
An innovative prefabricated timber-concrete composite system
5
(a)
(b)
Fig. 4 (a) Timber beam with inserted screws- (b) ”upside down”
floor system (including the form-work) directly after concrete
casting
Table 2 Geometry and compression strength of the tested concrete
cubesID Side Side Depth Compression
strengtha b h σm
# [mm] [mm] [mm] [N/mm2]A 150 150 150 51,11B 150 150 150 50,89C
150 150 150 51,11
Table 3 Ultimate withdrawal capacity of the screws inserted into
the concrete with different pen-etration lengths.
ID specimen penetration lengths [mm] Ultimate tensile capacity
[kN]
A 50 27,49B 50 19,75C 50 23,90
Mean value 23,71A 75 40,03B 75 40,04C 75 38,32
Mean value 39,46A 100 40,68B 100 42,92C 100 40,58
Mean value 41,39
-
6 Roberto Crocetti, Tiziano Sartori, Roberto Tomasi, Jose L. F.
Cabo
Table 4 Modulus of elasticity and density of timber beamsID
Lenght [mm] Depth [mm] Width [mm] ρm [kg/m3] Em [MPa]A 7198 355 112
477 12480B 6595 356 111 449 12189
4.2 Short-term bending tests
Load-deflection curves and load-slip curves are shown in Fig. 5
and Fig. 6 respec-tively.
0 20 40 60 800
20
40
60
80
Deflection at mid-span f (mm)
Equ
ival
entu
nifo
rmly
dist
ribu
ted
load
q(k
N m2
)
Floor AFloor BEImaxEImin
Fig. 5 Equivalent uniformly distributed load vs mid span
deflection
The curves in Fig. 5 show the relationship between the
equivalent uniformly dis-tributed load q (i.e. the total load
applied divided by the slab area) and the deflectionf at mid-span.
As it ca be observed, the behaviour is linear up to a load level of
ap-proximately 80 kNm2 , which is well above the design load used
in design of commonfloor structures. The stiffness of both
composite floors shows, after a slightly non-linear initial part, a
constant trend up to the failure of one of the two timber beams.The
curve of Fig. 6 shows the slip at the support related to the
equivalent distributedapplied load. Failure of floor A occurred at
a load q ' 84 kNm2 , with the propagationin one of the two beams of
two large cracks in the direction parallel to the grain.The failure
of the floor type B, on the other hand can be attributed to local
failure ofa finger joint of the lowest lamination located close to
mid span of one of the twobeams. The collapse of floor A occurred
at q=80 kNm2 firstly due to bending failure ata finger joint in one
beam and secondly due to a shear failure located along a
linerunning through the tips of the screws used as shear
connectors.
-
An innovative prefabricated timber-concrete composite system
7
0 20 40 60 800
0.5
1
1.5
2
2.5
Equivalent uniformly distributed load q ( kNm2 )
Slip
atth
esu
ppor
t(m
m)
Floor AFloor B
Fig. 6 Load slip deflection vs equivalent uniformly distributed
load
4.3 Long term bending test
For the long-term bending test, a uniformly distributed load of
1 kNm2 was appliedon the slab by means of sacks of cement(see Fig.
7). The purpose of the long-termtest was to investigate the
time-dependent behaviour of the prefabricated timber-concrete
composite system. The long-term test results for the specimen B is
pre-sented in Fig. 8 in terms of time vs mid-span deflection. The
variables monitoredduring the entire test were the mid-span
deflection through 2 inductive transducers,positioned at the
mid-span of each glulam beam.
Fig. 7 Long term test setup
-
8 Roberto Crocetti, Tiziano Sartori, Roberto Tomasi, Jose L. F.
Cabo
Fig. 8 Increase in mid-span deflection of the floor C with
time
Also the temperature and the humidity in the laboratory was
continuously moni-tored. The value of these parameters are
presented in Fig. 9.
Fig. 9 Relative humidity and temperature observed in the
laboratory during the period 01-02-2013/ 30-11-2013
-
An innovative prefabricated timber-concrete composite system
9
As it can be seen in Fig. 8 the mid-span deflection increase to
about three timesthe instantaneous deflection after three months.
In March the floor has been down-loaded for two days. When the
floor was reloaded it achieved again the same levelof deflection as
before unloading. This operation was performed again in the
firstdays of May and similar results were obtained.
5 Efficiency of the composite beams
Deformations at the shear connectors generate horizontal
movement, i.e. slip at theinterface between concrete and timber.
Such a behaviour is due to as partial com-posite action and, as the
slip increases it reduces the efficiency of the cross section.The
efficiency of a shear connection for a composite beam can be
estimated usingthe following equation, see [9] and [10].
η =EIreal−EIminEImax−EImin
(1)
where η is the efficiency, EImax is the bending stiffness of the
floor with fullcomposite action, EImin is the bending stiffness of
the floor with no composite actionand EIreal is the actual bending
stiffness of the floor. At load levels comparable tothose at the
serviceability limit state (i.e. 1 kNm2 ) the efficiency η is
approximately1.0. The efficiency at a load of 20 kNm2 or more
remains constant,i.e. η ' 0.85.
6 Conclusions
This paper presents the main results of a research project
conducted on a novelprefabricated timber-concrete composite system.
The system provides several ad-vantages compared to cast in-situ
concrete slabs, e.g. reduced time of constructionand considerable
reduction of the effects of concrete shrinkage. Experimental
testswere carried out on three 7.2 m long strip floor specimens to
investigate on stiff-ness and the strength and stiffness of the
prefabricated systems, of which two teststo failure and one
long-time test. The principal observations from the
experimentalinvestigations are:
• The tested system showed considerably higher stiffness and
strength propertiesthan a similar system with concrete deck not
able to transfer shear stress
• The load carrying capacity was very high. The equivalent
uniformly distributeload at failure was approximately 80 kN/m2,
which is considerably larger thancommon designs load for floor
structures.
• The stiffness of the system was also very high. This depends
primarily on theability of the shear connectors to transmit shear
without (or with minor) slipping.In the tested specimens the
efficiency of the system was approximately 1 for
-
10 Roberto Crocetti, Tiziano Sartori, Roberto Tomasi, Jose L. F.
Cabo
loads well above common design loads. For extremely high loads
(q≥ 20kN/m2)the efficiency of the system was approximately 0.85,
which is also very highcompared to similar composite system with
more traditional shear connectors,i.e. screws or bars inserted
perpendicularly to the plane of the slab.
• The instantaneous deflection of the floor increased roughly by
a factor 3 after aperiod of approximately 7-8 months. This
relatively large increase in deflectionis believed to be due mainly
to the creep of the concrete slab and of the timberbeam and - in
some minor extent - to the long-term deformation of the
shearconnectors.
• Last but not least, the easiness of manufacture of the
proposed system shouldnot be underestimated, since it allows for a
quick construction whit a reducedpossibility of human errors.
7 Acknowledgements
The authors wish to gratefully acknowledge the Mr. Franco Moar
who has per-formed his Master’s thesis on this topic. The timber
material was supplied by theglulam mill Moelven Töreboda AB,
Treboda, Sweden. The screws were suppliedby Rotho Blaas srl,
Cortaccia, Italy. The fibers for the FR concrete were supplied
byBekaert Svenska A.B. All the suppliers are kindly
acknowledged.
References
1. Bathon, L. A., Bletz, O. and Bahmer R. Concrete bearings a
new design approach in wood-concrete-composite applications.
Proceedings of World conference timber engineering, Port-land -
Oregon-USA (2006).
2. Blass, H. J., Bejtka I. Screws with continuous threads in
timber connections. InternationalRILEM Symposium on Joints in
Timber Structures: 193 - 201.
3. Moar, F. Prefabricated timber-concrete composite system
Master thesis, Lunds TekniskaHgskola, Lunds Universitet (2012).
4. Crocetti R., Sartori T., Flansbjer M.. Timber-Concrete
Composite Structures with Prefabri-cated FRC Slab, Proceedings of
World conference timber engineering, Riva del Garda,
Italy(2010).
5. Gutkowski, R., Brown, K., Shigidi, A., Natterer, J.
Laboratory tests of composite wood con-crete beams, Construction
and Building Materials, 22 (6), 1059-1066 (2008).
6. Lukaszewska E. Development of Prefabricated Timber-Concrete
Composite Floors. PhD the-sis, Department of Civil, Mining and
Environmental Engineering Division of Structural En-gineering, Lule
(2009).
7. Lukaszewska, E., Johnsson, H. and Sthen L. Connections for
Prefabricated Timber ConcreteComposite Systems. Proceedings of
World conference timber engineering, Portland - Oregon-USA .
(2006).
8. A. Sjöström, J. N.-Montero, D. Bard, R. Crocetti Vibratory
investigation of a fiber reinforcedconcrete floor supported by
wooden beams: part I. Proceedings of Joint Baltic-Nordic Acous-tics
Meeting, Odense - Denmark. (2012).
-
An innovative prefabricated timber-concrete composite system
11
9. M. Piazza, G. Turrini Sulle strutture composte legno-legno.
Proceedings Italian Workshop onComposite Structures, Department of
Mechanical and Structural Engineering, University ofTrento, Villa
Madruzzo, Trento, Italy, 1993: pp. 349-370.
10. M. Piazza Restoration of timber floors via a composite
timber-timber solution. RILEM Work-shop Timber: a Structural
Material from the Past to the Future, Trento, 1994: pp.
167-187.