Construction and Building Materials 37 (2012) 499506

Contents lists available at SciVerse ScienceDirect

Construction and Building Materials

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

Experimental analysis of timberconcrete composite beam strengthened with

carbon bres q
Miroslav Premrov , Peter Dobrila
University of Maribor, Faculty of Civil Engineering, Smetanova ul. 17, SI-2000 Maribor, Slovenia

h i g h l i g h t s

" We experimentally tested elements composed of a concrete plate connected to a timber beam.
" The beam is strengthened at the bottom of the tensile side with a CFRP strip.
" The measured results are compared with numerical results.
" Findings are important for the renovation principles of the old residential houses with timber oors.

a r t i c l e i n f o a b s t r a c t

Article history: Reconstruction of old residential buildings is most often accompanied with a tendency to preserve the
Received 3 February 2012 original features of the building, which leads to subsequent preservation of the oor elements. The ageing
Received in revised form 24 July 2012 of the material and the consequent lower stiffness and load bearing capacity of the oor element along
Accepted 4 August 2012
with the occasional change in the purpose of the facility and the consequent increase in the load require
Available online 6 September 2012
structural reinforcement of the oor elements. A possible solution lies in replacing the classic timber
planks with a concrete slab and in the corresponding strengthening of the timber beam in the tensile
Keywords:
zone. Therefore, the paper presents an experimental study performed on composite beam test samples
Timber
Reconstruction
composed of a concrete plate connected to a timber beam with dowel-type fasteners. Since the tension
Composite beam failure of the timber beam is the decisive resisting criterion, the beam is strengthened at the bottom of
Reinforced timber structures the tensile side with a carbon bre-reinforced polymer strip. The measured results are compared with
Experimental study numerical results obtained by an already developed analytical model [1] based on Mohlers simplied
formulation and on European standards for timber, concrete and steel structures. Since the experimental
results demonstrate a relatively low deviation and additionally a good agreement with the numerical
ones, the usage of the presented analytical model can be recommended to predict proper stiffness and
load bearing capacity for these types of composite oor structures.
2012 Elsevier Ltd. All rights reserved.

1. Introduction heating, cooling and lighting, reduction of green house gas (GHG)
emissions, are strongly recommended.
The present times, characterised by specic circumstances in Being a natural raw material timber helps the environment by
the sphere of climate change, witness an intensive focus of the sci- absorbing and storing CO2 while it grows and therefore plays an
ences of civil engineering and architecture on searching for ecolog- important role in reduction of the CO2 emissions [2]. On the other
ical solutions and construction methods that would allow for hand, brick and concrete industries are responsible for about 10%
higher energy efciency and, consequently, for reduced environ- of the global CO2 emissions. Moreover, in spite of the smaller wall
mental burdening. Due to the fact that buildings represent one of thickness timber buildings have better thermal properties than
the largest energy consumers and greenhouse gas emitters, energy those built by using conventional brick or concrete construction
saving strategies related to buildings, such as the use of methods. In comparison with other types of buildings, the
eco-friendly building materials, reduction of energy demand for energy-efcient properties of timber-frame buildings are excellent
not only because well insulated buildings use less energy for heat-
ing, which is environmentally friendly, but also due to a comfort-
q
In memory of Mr. Matjaz Tajnik, victim of a tragic death on 17 May 2011, who able indoor climate of timber-frame houses [3]. Considering the
carried out most of the presented experimental work.
growing importance of energy-efcient building methods, timber
Corresponding author. Tel.: +386 2 22 94 303; fax: +386 2 25 24 179.
construction will play an increasingly important role in the future.
E-mail address: miroslav.premrov@uni-mb.si (M. Premrov).

0950-0618/$ - see front matter 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.conbuildmat.2012.08.005
500 M. Premrov, P. Dobrila / Construction and Building Materials 37 (2012) 499506

The use of timber in construction is gaining ever more support, We know that the chosen materials of reconstruction (concrete,
especially in regions with vast forest resources since it reduces CFRP strips and the adhesive) are less eco-friendly than timber.
the energy demand for transport if the building material is avail- Especially, the exposure of the CFRP and the dust due to the ageing
able from the local area, [2,4]. Respecting all given facts, timber in the closed space may be seriously harmful to user of the build-
as a material for load bearing construction represents a future chal- ing. Especially, the exposure of the CFRP and the dust due to the
lenge for residential and public buildings. ageing in the closed space may be seriously harmful to user of
Energy-efcient construction has become an unavoidable fact. the building. Of course, according to the given facts, the strength-
Unfortunately, we frequently tend to forget that new residential ened oor elements are less eco-friendly than the original ele-
buildings, which we mostly lay focus on, represent only a minor ments with the timber planks, but with this subsequent
part of the existing housing stock. More attention should be there- preservation of the timber beams we still can keep the original fea-
fore drawn to the renovation of older buildings the biggest air tures of the building. Additionally, the costs of employing CFRP are
pollutants, whose energy related properties no longer satisfy the in comparison with other strengthening materials at the moment
requirements of the existing legislation. It is thus most vital to con- rather high. However, the main advantages of using CFRP strips
sider not only energy-efcient new buildings but to attend to en- in particular compared to other optional materials of strengthening
ergy-related refurbishment of older residential buildings and in this case (for example steel plates) are their corrosion resistance,
houses as well. Such undertaking should imply the renovation of light weight and exibility, which allow convenient and easy
the buildings envelope, which could be realized by adding ther- transport to the place of erection. Continuously decreasing prices
mal-insulation materials and replacing the doors and windows of these materials make the new technology more economical
[5]. On the other hand, we ought to strive for the maximum pres- and interesting. Therefore the use of CFRP for the repair and
ervation of the original features of the building and for the subse- strengthening of timber elements opens new perspectives for tim-
quent preservation of the oor elements. ber structures design.
Since the oor elements in old residential buildings usually con-
sist of timber beams with planks simply connected to the beams, 2.1. Reconstruction with a concrete slab
the renovation or reconstruction process usually brings about cer-
tain construction-related problems, such as a change in the pur- Using the existing timber oor we can develop an efcient com-
pose of the facility and the consequent increase in the load on posite system made of timber members in the tensile zone, a con-
the oor elements, a decrease in the strength due to the ageing crete layer in the compression zone and a timberconcrete
of the material and the subsequent lower load bearing capacity connection between them. The benets of such reconstruction-
of the oor element. The oor elements in such cases therefore strengthening procedures, sometimes applicable even to newly
have to be adequately reconstructed and strengthened, which is constructed buildings, are increased stiffness and load bearing
more specically presented in two possible steps described in Sec- resistance, better sound insulation and re resistance as well as
tion 2 of this paper including a summary of the similar previous cost and environment related advantages gained when the existing
studies. The most important facts about the structural behaviour supporting timber structure is used as framework.
of such reconstructed and strengthened composite oor elements The behaviour of the described oor construction depends on
are given in Section 3. Experimental analysis performed on test the connection between the timber beam and the concrete slab,
samples with different types of CFRP strengthening is presented i.e. on the fasteners used. Owing to the negative impact humidity
in Section 4. The measured results are additionally supported has on the strength of timber, it is of outmost importance to install
and compared with the already presented numerical results in [1]. hydro-isolation in the at area between the concrete slab and the
timber beam in order to prevent humidity invasion from the rst to
2. Reconstruction of old timber oors the latter. Afterwards, the timber beam equipped with hydro-isola-
tion undergoes the process of installation of the steel dowels which
The oors in old residential buildings usually consist of timber connect the concrete slab to the timber beam (Fig. 2a). Next, a min-
beams placed in the tensile zone and of timber planks simply con- imum reinforced concrete slab is concreted onto the timber beams
nected to the timber beams. In order to increase the bending and with the installed steel dowels in the following way: the timber
shear resistance of the oor elements, two main successive steps beam with steel dowels is simply laid onto the concrete plate
of reconstruction are to be taken, as is schematically presented in (Fig. 2b).
Fig. 1: The experiments showed that the use of steel bre reinforced
concrete SFRC displays better characteristics than the use of
the timber planks are removed and replaced with the concrete the classic concrete, regardless of the type of fasteners connecting
slab, the concrete slab and the timber beam. The experimental study
the timber beams are reinforced with carbon bre-reinforced involving the use of steel bre reinforced concrete by Holschem-
polymer (CFRP) strips glued to the bottom of the beam. acher et al. [6] demonstrated a 27% higher nal load bearing capac-
ity of the fasteners and a few times higher ductility, in comparison
The rst step of merely replacing the timber plank with the con- with the classically reinforced concrete. Practical examples of this
crete slab (Fig. 1a) involves three possible failure mechanisms of kind of reconstruction procedure performed on the existing timber
such a composite oor element under the bending load: compres- oor can be found in many European cities (e.g. Leipzig, Tbingen,
sive failure of the concrete slab, shear failure of the fasteners in the etc.), as presented by Holschemacher et al. [6], Kenel [7] and
concretetimber connecting area and bending failure in the tensile Schanzlin [8].
area of the timber beam. In case the last failure mechanism has a Since the time-dependent effects have a big inuence on rear-
decisive role, the oor element needs further reinforcement com- rangement of stresses in a section composed of materials with dif-
prising only that of the timber beam, which is performed by gluing ferent creep behaviour, as it is in our case, a special attention
CFRP strips to the bottom-tensile side of the beam, as shown sche- should also be dedicated to this problem. The problem of time-
matically in Fig. 1b. It is important to stress that the second step of dependent effects presented in the connection between the con-
strengthening is signicant only in case when the tensile strength crete plate and the timber beam are numerically discussed in the
criterion in the timber beam is decisive for the failure and when study Tajnik et al. [1], where it is approximated with the effective
there is no risk of other two possible types of failure. slip modulus Kser,n to take into account a timber nal creep
 M. Premrov, P. Dobrila / Construction and Building Materials 37 (2012) 499506 501

Fig. 1. Types of reconstruction of old timber oors with a concrete slab (a) and additionally with CFRP strips (b).

Fig. 2. (a) Installation of the steel dowels onto the timber beam. (b) Laying of the timber beam onto the concrete slab.

coefcient, a factor for a quasi permanent combination and a con- Since the tensile strength of timber is similar to its compressive
crete creep coefcient. It is important that even in ultimate limit strength, strengthening applications of bre reinforced polymers
state calculations creep should not be neglected as conrmed by (FRPs) in timber structures have not been so frequently used as
the research of Kavaliauskas et al. [9], Takac et al. [10] and Bob in concrete or masonry structures. The main advantages of FRP in
and Bob [11]. Ceccotti et al. [12] also recommends push-out tests comparison to other materials, for example steel plates, are their
to obtain realistic values of the connection stiffness and empha- corrosion resistance, light weight and exibility allowing conve-
sises the signicance of time dependent effects in an analysis of nient and easy transport to the place of erection. The use of
such structures. (HSF) high strength bre and CFRP for the reconstruction and
strengthening of timber elements opens new perspectives for a
2.2. Additional strengthening with CFRP strips timber structure design, particularly as it does not affect the
appearance of the timber. Dagher and Breton [14] reinforced lam-
In many cases the above described strengthening concept is not inated timber beams in the tensile area using FRP lamellas. The test
sufcient. It is therefore recommendable to use additional carbon results showed an essential increase in bending resistance. The test
bre-reinforced polymer (CFRP) strip at the bottom-tension side results using carbon bres in laminated beams are presented also
of the timber beam to gain higher bending resistance and stiffness by Bergmeister and Luggin [15]. Stevens and Criner [16] conducted
of the oor in buildings or decks on bridges, as shown schemati- an economic analysis of bre-reinforced polymer (FRP) glulam
cally in Figs. 1b and 3. Additional strengthening of the tensile zone beams. The results showed practical applicability of FRP reinforced
with CFRP strips can increase the bearing load capacity in bending elements, especially for bridges of greater spans, where beam
of the beam by approximately 15% [13], or even more depending dimensions can be substantially reduced using the FRP solution
on the type of CFRP strip, which consequently leads to an even bet- presented.
ter use of each component of the composite section. Composite reinforcement on sawn timber elements is less com-
It is rather important, from the purely technological point of mon in literature although many applications exist, especially for
view, to glue the CFRP strips onto a clean, previously prepared sur- bending reinforcement. Timber beams reinforced with a layer of
face at the bottom of the timber beams and thus ensure a com- high-modulus composite material may be analysed using a trans-
pletely stiff connection (Fig. 4). formed section of equivalent wood ([17]). Johns and Racin [18]
502 M. Premrov, P. Dobrila / Construction and Building Materials 37 (2012) 499506

Fig. 3. Reinforcement by gluing CFRP strips on the tensile edge of the beam.

Fig. 4. Gluing of CFRP strips to the timber frame.

demonstrated their experimental studies using glass bres to rein- contribution of a potential tension area in the concrete slab is
force sawn timber sections. In the study of Dourado et al. [19] the neglected in the load bearing capacity of the beam cross-
inuence of patch thickness and adhesive lleting was analysed section,
experimentally and numerically. The main conclusion was that slip modulus is taken in the elastic region for the serviceability
repairing leads to a remarkable gain in the bearing capacity of limit state (Kser) and in the plastic region for the ultimate limit
the damaged beam a simple bonding repair with a patch of state (Ku). The secant slip modulus can be determined in prac-
2.0 mm thickness revealed a considerable increase of 62% in the tice from push out tests according to Werner [23] and EN
bearing load. However, adhesive lleting did not induce a clear in- 26891 [24] which determine Kser at 0.4Fultimate and Ku at 0.6Fulti-
crease in the load bearing capacity. Investigation results for bre mate. Ceccotti et al. [12] additionally recommend the use of
reinforced hollow wood beams are presented by Kent and Tingley Kcolaps at 0.8Fultimate for experimental collapse loads.
[20]. They show that a glass-aramid reinforced plastic (GARP) rein-
forcement increased the average strength and stiffness of the The effective bending stiffness (EIy)eff in accordance with EN
beams, compared to the non-reinforced control samples, by 22% 1995-1-1 [22] can be written in the form:
and 5%, respectively. ! !
3 3
hc  b c ht  bt
EIy eff Ec  cct  Ac  z2c Et  At  z2t
12 12
3. Basic design model
Ef  Af  z2f 1
Structural behaviour of timberconcrete composite members,
governed by the shear connection between timber and concrete, where Ec is the mean modulus of elasticity of concrete, Et the mean
can be predicted by the elasto-plastic model presented by Frangi modulus of elasticity of timber and Ef is the mean modulus of elas-
and Fontana [21] or as a simplied elastic model appropriate for ticity of the carbon strip. As shown in Fig. 5 zi is the distance from
everyday engineering practice. For design and parametric study overall neutral axis to the centre of gravity of each sub-component,
purposes, a simplied design method for mechanically jointed ele- Ac, Af, At are cross section areas of each sub-component and hc, hf, ht
ments (Moehlers formulation) according to Annex B of EN 1995-1- are thicknesses of each sub-component. The stiffness coefcient
1 [22] has been implemented. The expression of the so-called c- (cct) in the connecting area between concrete and timber is calcu-
method is used in equations with the following fundamental lated according to [22] in the form of:
assumptions:
1
cct 2
1p
2 E A S
c c i
bernoullis hypothesis is valid for each sub-component (Fig. 5), Kl2eff
material behaviour of all sub-components is linear elastic
(Fig. 5), where leff is the effective length of the composite beam and si is the
the distances between the dowels are constant along the beam effective spacing between the fasteners. The stiffness coefcient in
(Fig. 3). the plane between the carbon strip and the timber web, which are
glued together, is considered as fully connected, therefore cf = 1.0.
Additionally, two important further assumptions are considered A normal stress in the composite section for each sub-component
in the proposed model: can be obtained according to the given assumptions in the form of:
 M. Premrov, P. Dobrila / Construction and Building Materials 37 (2012) 499506 503

Fig. 5. Composite cross-section with a diagram of normal stress distribution.

M y  Ei practical importance of information which the measured values can provide. The
rd  ci  zi  Dz 3 test specimen dimensions and the fasteners arrangement density underwent addi-
EIy eff
tional careful selection based on the numerical results of the previously described
where My is the bending moment acting on the composite section, mathematical model. The aim was to ensure previously mentioned failure of the
Dz the distance from the centre of gravity of each sub-component to timber beam and after that the failure of the CFRP strips, which was the only sen-
sible step to take in order to test the strengthened CFRP specimen and especially the
the edge (or any bre) of each sub-component and Ei is the modulus
inuence of the CFRP reinforcement in order to recognize practical advantages of
of elasticity of each sub-component. The considered linear stress using such kind of reinforcing.
distribution is schematically presented in Fig. 5. The aim was to ensure previously mentioned failure of the timber beam and the
CFRP strips, which was the only sensible step to take in order to test the strength-
4. Experimental analysis ened CFRP specimen.
The production of the tested beams was performed in three successive steps: (a)
A detailed numerical analysis based on the mathematical model in Section 3 is production of the glulam beam of the quality GL34h with the constant height of ht
presented in Tajnik et al. [1]. A selected set of results will serve the purpose of com- = 250 mm and width of bt = 160 mm, drilling of holes for the dowels, gluing of the
parison with the measured values obtained through the experimental analysis, hydroinsulation and insertion of the steel dowels of diameter d = 16 mm into the
which will be the only relevant factor indicating the suitability of choice of the pre- timber beam (Fig. 2a). The dowels were placed in two parallel rows at the distance
sented mathematical model. The decision to seek experimental conrmation of the of 50 mm between the rows (Fig. 6); (b) on the timber beam with already inserted
model additionally rests on the fact that the sphere of science sees the appearance dowels the concrete slab of a constant thickness of hc = 60 mm and compressive
of numerous articles denying the suitability of using the c-method which is based strength of 30 MPa with a reinforcement S500 placed in the centre of the slab
on Eurocode 5 [22]. Moreover, studies like de Goes and Junior [25] tend to appear and with a minimum cross-section of As = 1.39 cm2/m (Fig. 6) was produced. The
and conrm this method based on experimental measurements and nite element timber beam with steel dowels is simply laid onto the concrete slab (Fig. 2b); (c)
models (FEMs) calculations on road bridges. nally, the CFRP strips of the thickness of 1.2 mm and with the width of 150 mm
which was constant in the whole length of the tested beam were glued to the bot-
4.1. Test conguration tom side of the timber beam (Fig. 4).
A risk of shear failure in the selected specimen does not exist in practice. Predic-
The longitudinal section and the cross section of the analysed four test speci- tion resting on numerical results gained through the analytical calculation model
mens P1, P2, P3 and P4 are presented in Fig. 6. The dimensions were determined estimates that shear failure occurs only at 220% overload above bending failure in
according to the simplication for production, suitability for transportation and the tensile area of the timber beam. To suit the failure a minimum possible

Fig. 6. Geometry of the test specimens.

504 M. Premrov, P. Dobrila / Construction and Building Materials 37 (2012) 499506

Table 1
Properties of the materials used.

Concrete C30/37 Timber GL34h SikaCarboDur-H514

Em (MPa) 31,939.00 11,600.00 300,000.00
fm,k (MPa) 34.00
ft,0,k (MPa) 23.50 1300.00
fc,0,k (MPa) 30.00 24.00
qk (kg/m3) 430.00
qm (kg/m3) 2400.00 516.00 1600.00

Em mean value of modulus of elasticity.

fm,k characteristic bending strength.
ft,0,k characteristic tensile strength (for timber parallel to grain).
fc,0,k characteristic compressive strength (for timber parallel to grain).
qk characteristic density.
qm mean density.

regulations-abiding installation of the fasteners in the longitudinal range of 80 mm

was selected which subsequently ensures high stiffness and load bearing capacity
of the connection. Fig. 7. Installation of the measuring instruments.
For the purpose of the testing we deliberately decided for a slightly atypical
three-point bending test (although it is common practice to conduct a four-point 4.2. Test results
bending test), with the load being concentrated in the middle of the 4.5 m span.
Such load distribution does in fact bring a risk of partial bending-shear failure The ultimate failure force (Fu), the maximal cantilever bending deection (w)
but it simultaneously offers far more reliable results about the impact the stiffness under the acting force (F) and the slip (D) between the concrete slab and the timber
of the fasteners has due to constant distribution of the shear ux across the entire beam were all measured (Fig. 7). For the point of view of statistics, the following
shear plane between the concrete slab and the timber beam. For the same reason nding is of signicant importance: similar load bearing and stiffness characteris-
we opted for constant distribution of the dowels across the entire shear plane along tics were demonstrated by all specimens with the exception of one which showed
the timber beam. An additional reason was to avoid potentially unreal results in a certain deviation in the load bearing capacity due to the expected higher deviation
case of using the equation for effective distance between the fasteners (sef) follow- of the average timber strength. The specimen in question remained part of the anal-
ing Eurocode 5 [22]. We tested four identical specimens. ysis since it displayed a suitable stiffness level.
Material properties for the timber of quality GL24h are taken from EN1194 [26] Fig. 8 shows steps of one of the specimens failure where the rst evident failure
and for the concrete from Eurocode 2 [27]. The values for the CFRP strips of SikaCar- was that of the timber beam (Fig. 8a), followed by the failure of the CFRP reinforce-
boDur-H514 are taken from Sika [28]. All material properties are listed in Table 1. ment (Fig. 8b) and nally by the total failure of the specimen that of the concrete

Fig. 8. Tensile failure of the timber beam (a), failure of the CFRP reinforcement (b) and total failure of the specimen (c).

Fig. 9. Fw diagram of the test samples and the calculated numerical results.
 M. Premrov, P. Dobrila / Construction and Building Materials 37 (2012) 499506 505

Table 2
Measured test results and the calculated numerical results.

Test specimen Failure mode: timber Deection by force Failure mode: concrete Force by w = 15 mm
in tension Fu,timber (kN) Fu,timberw (mm) in compression Fu,concrete (kN) (kN)
P1 133.10 30.47 144.11 66.79
P2 140.48 38.90 152.70 60.16
P3 122.01 31.34 143.54 62.47
P4 126.32 30.05 148.23 68.97
Numerical model [1] 125.20 30.79 147.67 67.20

slab, which classies the failure as brittle, i.e. non-ductile failure (Fig. 8c). The fas- of timber elements opens new perspectives for the timber struc-
teners, i.e. the connection between the concrete slab and the timber beam, were not
tures design. Continuing reduction in the price of these materials
at risk, according to expectations.
The measured results for the failure force in tensile area of the timber beam
make the new technology more economical and interesting. On
(Fu,timber), the force of the total failure of the specimen that of the concrete slab the other hand, in comparison with traditional reinforcement the
(Fu,concrete) and the measured deection by the force Fu,timber are presented in use of bre composites in timber buildings calls for experience
Table 2. Additionally, the measured value of the force by a known deection of and higher quality of construction works. This leads to a conclusion
w = 15 mm is presented to determine the bending stiffness of the test sample.
that the presented composite oor system, consisting of the con-
The calculated results using the proposed design model from Section 3 are given
to compare the measured and the numerical values. crete slab and the timber beam reinforced with CFRP, means a
For information only, the calculated load-bearing capacity of the un-strength- highly interesting potential way of increasing the load bearing
ened element with the concrete slab can be calculated from [1] and it is Fu,timber,- capacity of old timber oors in the future. At the same time this
uns. = 99.36 kN. The load-bearing capacity of the original structure composed of
method preserves the original features of the existing buildings.
the timber beam and timber planks placed perpendicular to the grains of the beam
can be calculated considering only the contribution of the timber beam and it is
The presented experimental study certainly reached its purpose
Fu,timber,orig. = 50.37 kN. These results for the increasing in the bending resistance since it leads to a conclusion based on the analysed results and
are close to the gains in a load-bearing capacity obtained by the experiments with their relatively low deviation that it largely conrms, in spite
similar strengthening concepts in Ajdukiewicz and Brol [29]. of the limited number of samples, the adopted analytical calcula-
Fig. 9 shows Fw diagrams of all four specimens up to the point of failure. For
tion model. We can thus claim that the analytical calculation mod-
the purpose of comparison the diagrams contain additionally drawn calculation
stiffness which was calculated according to the mathematical model in Section 3. el, meticulously presented and analysed in [1], proves to be most
Further items added to the scheme of the diagrams predicted upon the model are suitable for the calculations of the load bearing capacity and the
(horizontal lines): the design value of failure in the tensile area of the timber beam calculation or prediction of deformation of the oor constructions
(ULS-timber), the provisional design load bearing capacity of the connection, i.e. of
in question. Extensive numerical parametric analyses, carried out
the fasteners (ULS-connection) and the provisional load bearing capacity according
to the limit value for instantaneous deection L/300 = 15 mm (SLS-deection). The
in [1] with variations including timber quality, the height of the
interrupted red line represents a clearly visible leap appearing in the mathematical timber beam and the installation of the fasteners in the connection
model, at the crossing from the fasteners shift module in the range of serviceability showed that the contribution of the CFRP strips in the form of addi-
limit state SLS (Kser) into the fasteners shift module in the range of ultimate limit tional strengthening can improve the bending resistance up to 26%
state ULS (Ku) at the approximate force of 96 kN. It needs to be stressed that the
and in the bending stiffness up to 18%.
fasteners stiffness and that of the timber beam should be taken into separate con-
sideration despite their connection through Eqs. (1) and (2), since the rst necessary We would like to emphasise that the presented concept of rein-
step is to dene a borderline between serviceability limit state (SLS) and ultimate forcing is recommended only for strengthening old timber oor
limit state (ULS) for the fasteners with respect to the failure load bearing capacity structures to assure a higher load-bearing capacity. In new timber
by Johansens terms. It means dening the limit value up to which the connecting oor elements a higher load-bearing capacity can be easily
areas behaviour has linearly-elastic features, by the shift module for SLS (Kser), and
the limit value at which the connecting areas behaviour is non-linear, by the shift
achieved by using timber beams of higher dimensions or by using
module for ULS (Ku). Attention should also be paid to the fact that the above limits a timber of a higher quality.
do not necessarily coincide with those of the composite timber beam as whole, un-
less the decisive role in the timber beam failure is seen in the connection itself or in
the fasteners between the slab and the timber beam. This nding is highly relevant
References
to the designers decision and his preference for non-ductile bending failure in the
tensile area of the timber beam or the ductile yielding failure of the fasteners. The
[1] Tajnik M, Premrov M, Dobrila P, Bedenik B. Parametric study of composite T-
failure of the timber beam in tension was practically decisive, there was (especially
beam. Proc ICE Struct Build 2011;164(5):34553.
by test samples P3 and P4) practically no post-breaking stability of the element [2] Natterer J. Massivholz-konstruktionen: Herausforderung fr eine nachhaltige
after that, as it can be seen from Fig. 9. The deections rapidly increased and they kobilanz. Building with timber, challenge and opportunity for Slovenia.
were not under control anymore. University of Ljubljana, Biotechnical Faculty, Ljubljana; 2009.
As seen in the diagrams of the test samples in Fig. 9, we decided for the non- [3] Gold S, Rubik F. Consumer attitudes towards timber as a construction material
ductile failure mechanism in the bottom tensile area of the timber beam. The pre- and towards timber frame houses selected ndings of a representative
sented measured results show a relatively small discrepancy in the four samples, survey among the German population. J Cleaner Prod 2009;17:3039.
which can justify a relatively small number of the chosen identical samples. With [4] Joseph P, Tretsiakova-McNally S. Sustainable non-metallic building materials.
regard to dening the ultimate failure load bearing capacity, the diagram further- Sustainability 2010;2:40027.
more shows a good agreement between the results gained through the calculation [5] Zegarac Leskovar V, Premrov M. Design approach for the optimal model of an
analytical model [1] and the results of the experimental specimens P1P4, with the energy-efcient timber building with a focus on the south-oriented window
maximum deviation between sample P1 and the model being 8.2%. area. V: Internationaler Kongress E-Nova 2011, 24. und 25. November 2011,
Pinkafeld. Nachhaltige Gebude: Innovative Technologien & Konzepte fr
In addition, our ndings reveal good agreement in predicting deformation or
Gebude der Zukunft. Band 15. Pinkafeld: Fachhochschulstudiengnge
the stiffness of such constructions since the difference between the calculation-pre-
Burgenland, Forschungs- und Studienzentrum Pinkafeld.
dicted value and the average measured value is as little as 30.36 mm/
[6] Holschemacher K, Klotz S, Weibe D. Application of steel bre reinforced
30.00 mm = 1.012% or 1.2%, where shifts located at areas of the predicted failure concrete for timberconcrete composite constructions. Lacer 2002;7:16170.
were observed. The serviceability limit state range (L/300 = 15 mm), where predict- [7] Kenel A. Zur Berechnung von Holz/Beton-Verbundkonstruktionen. Abteilung
ing deformation becomes even more important from the designers perspective, Holz, Arbeitsbericht 115/39, EMPA Dubendorf; 2000.
also shows the average deviation of 15.75 mm/15.00 mm = 1.05% or 5.0%. [8] Schanzlin S. Zum langzeitverhalten von brettstapel-beton-
verbunddecken. Stuttgart: Institut fur Konstruktion und Entwurf; 2003.
[9] Kavaliauskas S, Kvedaras A, Gurknys K. Evaluation of long-term behaviour of
5. Conclusion composite timberconcrete structures according to EC. Ukio technologinis ir
ekonominis vystymas 2005;XI(4):2926.
[10] Takac S, Matoevic , Bogoicevic P. Rheological research of sliding modulus of
The use of high strength bres (HSFs) and carbon-bre rein- the woodconcrete connection. In: IABSE conference Lahti 2001 innovative
forced polymers (CFRPs) for the reconstruction and strengthening wooden structures and bridges. IABSE-AIPC-IVBH, vol. 85; 2001. p. 495500.
506 M. Premrov, P. Dobrila / Construction and Building Materials 37 (2012) 499506

[11] Bob L, Bob C. Researches regarding the behaviour of the composite wood- [20] Kent S, Tingley D. Structural evaluation of ber reinforced hollow wood beams.
concrete oors. In: IABSE conference Lahti 2001 innovative wooden In: IABSE conference Lahti 2001 innovative wooden structures and bridges.
structures and bridges. IABSE-AIPC-IVBH, vol. 85; 2001. p. 5316. IABSE-AIPC-IVBH, vol. 85; 2001. p. 3672.
[12] Ceccotti A, Fragiocomo M, Giordano S. Long-term and collapse tests on a [21] Frangi A, Fontana M. Elasto-plastic model for timberconcrete composite
timberconcrete composite beam with glued-in connection. Mater Struct- beams with ductile connection. Struct Eng Int 2003;13(1):4757.
Springer Netherlands 2007;40(1):1525. [22] CEN European Committee for Standardization. Eurocode 5: Design of timber
[13] Tajnik M, Dobrila P, Premrov M. Analysis of composite T beam composed of structures. General common rules and rules for buildings, EN 1995-1-1. CEN,
timber, concrete and carbon strip. WSEAS Trans Appl Theor Mech Brussels; 2006.
2007;2(9):18190. [23] Werner H. Holzbau Handbuch, Reihe 1. Arbeitsgemeinschaft Holz e.V.,
[14] Dagher HJ, Breton J. Creep behaviour of FRP-reinforced glulam beams. In: Dusseldorf; 1997.
Proceedings of 5th world conference on timber engineering. Lousanne, vol. 2; [24] CEN European Committee for Standardization. Timber structures. Joints made
1998. with mechanical fasteners. General principles for the determination of
[15] Bergmeister K, Luggin W. Innovative strengthening of timber structures using strength and deformation characteristic, EN 26891. CEN, Brussels; 1991.
carbon bres. In: IABSE conference Lahti 2001 innovative wooden structures [25] de Goes JLN, Junior CC. Field performance of a composite log-concrete bridge.
and bridges. IABSE-AIPC-IVBH, vol. 85; 2001. p. 3616. Struct Eng Int 2008;18(1):7983.
[16] Stevens ND, Criner GK. Economic analysis of ber-reinforced polymer wood [26] CEN European Committee for Standardization. EN 1194. Glued laminated
beams. University of Maine; 2000. timber Strength classes and determination of characteristic values. Brussels;
[17] Johns KC, Lacroix S. Composite reinforcement of timber in bending. Can J Civ 1999.
Eng 2000;27(5):899906. [27] CEN European Committee for Standardization. Eurocode 2: Design of concrete
[18] Johns KC, Racin P. Composite reinforcement of timber in bending. In: structures. General common rules and rules for buildings, EN 1992-1-1. CEN,
Conference Lahti 2001 innovative wooden structures and bridges. IABSE- Brussels; 2002.
AIPC-IVBH, vol. 85; 2001. p. 5494. [28] Sika. Sicher bauen mit System. Technische Merkbltter. Ausgabe 5; 2003.
[19] Dourado N, Pereira FAM, de Moura MFSF, Morais JJL. Repairing wood beams [29] Ajdukiewicz A, Brol J. Modern strengthening methods of old timber structures.
under bending using carbonepoxy composites. Eng Struct 2012;34: In: IABSE conference Lahti 2001 innovative wooden structures and bridges.
34250. IABSE-AIPC-IVBH, vol. 85; 2001. p. 47984.