Article

Bending Stiffness, Load-Bearing Capacity and

Flexural Rigidity of Slender Hybrid
Wood-Based Beams
Barbara Šubic 1, *, Gorazd Fajdiga 2 and Jože Lopatič 3
1 M SORA d.d., Trg Svobode 2, 4226 Žiri, Slovenia
2 Biotechnical Faculty, University of Ljubljana, Jamnikarjeva 101, 1000 Ljubljana, Slovenia;
Gorazd.Fajdiga@bf.uni-lj.si
3 Faculty of Civil and Geodetic Engineering, University of Ljubljana, Jamova 2, 1000 Ljubljana, Slovenia;
Joze.Lopatic@fgg.uni-lj.si
* Correspondence: Barbara.Subic@m-sora.si; Tel.: +386-31-541-681




Received: 19 October 2018; Accepted: 9 November 2018; Published: 13 November 2018


Abstract: Modern architecture suggests the use of opened spaces with large transparent envelope
surfaces. Therefore, windows of long widths and large heights are needed. In order to withstand the
wind loads, such wooden windows can be reinforced with stiffer materials, such as aluminium (Al),
glass-fibre reinforced polymer (GFRP), and carbon-fibre reinforced polymer (CFRP). The bending
stiffness, load-bearing capacity, and flexural rigidity of hybrid beams, reinforced with aluminium,
were compared through experimental analysis, using a four-point bending tests method, with those
of reference wooden beams. The largest increases in bending stiffness (29%–39%), load-bearing
capacity (33%–45%), and flexural rigidity (43%–50%) were observed in the case of the hybrid beams,
with the highest percentage of reinforcements (12.9%—six reinforcements in their tensile and six
reinforcements in their compressive zone). The results of the experiments confirmed the high potential
of using hybrid beams to produce large wooden windows, for different wind zones, worldwide.

Keywords: wood based composites; hybrid beams; bending stiffness; flexural rigidity; aluminium
reinforcements; wooden windows

1. Introduction
The share of wood-based windows (i.e., wooden and aluminium-wooden profiled windows) is
the largest (66%) among certified windows for low-energy houses, due to their high energy efficiency
and their beneficial life cycle assessment characteristics [1]. Polyvinyl chloride (PVC) and aluminium
windows have 19% and 13% shares, respectively. Contemporary architectural design of buildings
integrate wooden products in the building envelope as well as in the interior. One of the reasons for
this is that it was proven that the use of natural wood in the buildings has positive psychological,
emotional, and health impacts on the people living in such an environment [2]. Windows have an
important role (solar gains through glazing), effecting the energy efficiency of the building. Therefore,
there has been a gradual increase in the proportion of the transparent part of the building envelope,
over the last decade. Windows are more and more frequently placed from the bottom to the top of
individual storeys, with heights exceeding 3.0 m, or even over several storeys, with heights exceeding
5.0 m (an example of such building is presented in Figure 1).
Two most important window elements are window frame and window glazing. From a thermal
point of view, the window frame is a weak point of the low energy windows, since its thermal
transmittance varies from 0.8–1.6 W/m2 K, whereas thermal transmittance of the glazing is almost
half this value and varies from 0.5–0.8 W/m2 K [1]. To improve thermal transmittance of the window

Forests 2018, 9, 703; doi:10.3390/f9110703 www.mdpi.com/journal/forests

Forests 2018, 9, 703 2 of 16

we need to either improve thermal transmittance of the frame or increase the surface of the glazing,
compared to the whole window area. The architectural trends prefer the open view through windows,
with as little visible window frame as possible. Therefore, for this study small cross sectioned
(i.e., slender) profiles were analysed, to explore the limits of improving their bending stiffness and
strength,
Forests 2018,when being
9, x FOR PEERreinforced
REVIEW with stiffer materials. 2 of 17

Figure 1. An example of window elements with a height of 5.8 m.

In
TwoEurope, windowswindow
most important are mostly manufactured
elements are window from
framesoftwood
and window species—90%
glazing. From(e.g.,aNorway
thermal
point of view,abies
spruce—Picea the (L.) H. Karst.,
window framePine—Pinus
is a weak pointsylvestris L. and
of the lowSiberian
energy larch—Larix
windows, since sibirica
its Ledeb.),
thermal
whereas
transmittancehardwoods are seldom
varies from used—10%
0.8–1.6 W/m (e.g.,thermal
2K, whereas oak—Quercus spp. L.,ofRed
transmittance grandis—Eucalyptus
the glazing is almost half
grandis W. Hill ex Maiden, walnut—Juglans 2 regia
this value and varies from 0.5–0.8 W/m K [1]. To improve thermal transmittance ofThe
L., beech—Fagus sylvatica L.) [3]. the main
window reason
we
for
needthistoiseither
that softwoods (especially
improve thermal Norway spruce
transmittance of the and Pine,orwhich
frame in Europe
increase are a domestic
the surface wood
of the glazing,
species),
compared aretoeasy
thetowhole
processwindow
and have up toThe
area. 38%architectural
better thermal properties
trends preferthan thehardwoods
open view[4].through
This is
an important
windows, parameter
with as little when
visibletrying
window to comply
frame as with today’sTherefore,
possible. high thermal for efficiency
this studydemands
small crossfor
modern buildings.
sectioned (i.e., slender) profiles were analysed, to explore the limits of improving their bending
Overand
stiffness thestrength,
last twenty years,
when many
being studies have
reinforced with been
stifferperformed
materials.with regard to the flexural rigidity
and load-bearing
In Europe, windowscapacity ofare
hybrid
mostly beams; not only experimentally
manufactured from softwood [5–8], but also analytically
species—90% [9–14]
(e.g., Norway
and numerically [15–19]. In most of the mentioned studies much larger cross sections
spruce—Picea abies (L.) H. Karst., Pine—Pinus sylvestris L. and Siberian larch—Larix sibirica Ledeb.), than those used for
windows were investigated (e.g., up to 500 mm in height) [12,15]. Only
whereas hardwoods are seldom used—10% (e.g., oak—Quercus spp. L., Red grandis—Eucalyptus a few researchers have studied
the reinforcing
grandis W. Hilleffect on specimens
ex Maiden, with cross sections
walnut—Juglans regia L.,smaller
beech—Fagus mm × 100
than 100sylvatica L.) mm [14,19,20],
[3]. The where
main reason
limited space for reinforcements is available. In most cases the reinforcements
for this is that softwoods (especially Norway spruce and Pine, which in Europe are a domestic wood are visible at least on
one side of
species), arethe tested
easy specimens
to process and and
havemostly
up to 38%oriented
better horizontally [5,6,12]. than
thermal properties Windows are influencing
hardwoods [4]. This
the overall architectural appearance of the building from the outside,
is an important parameter when trying to comply with today’s high thermal efficiency demands as well as the interior design,
for
and therefore
modern buildings. no visible reinforcements are preferred on either side of the window.
Different
Over the reinforcing
last twentymaterials
years, manyhave studies
been used in many
have been studies
performedto improve the bending
with regard to the stiffness
flexural
and load-bearing capacity of wooden beams. Steel profiles [5,9,12],
rigidity and load-bearing capacity of hybrid beams; not only experimentally [5–8], but also glass-fibre reinforced polymer
profiles (GFRP
analytically (glass-fibre
[9–14] reinforced polymer)
and numerically [15–19]. In profiles)
most of [6,14,18], and carbon-fibre
the mentioned studies reinforced
much larger polymer
cross
profiles
sections (CFRP (carbon-fibre
than those reinforcedwere
used for windows polymer) profiles)(e.g.,
investigated [7,11,15,19]
up to 500 aremmthe in
most common,
height) but
[12,15]. the
Only
a few researchers have studied the reinforcing effect on specimens with cross sections smaller than
100 mm × 100 mm [14,19,20], where limited space for reinforcements is available. In most cases the
reinforcements are visible at least on one side of the tested specimens and mostly oriented
horizontally [5,6,12]. Windows are influencing the overall architectural appearance of the building
from the outside, as well as the interior design, and therefore no visible reinforcements are preferred
Forests 2018, 9, 703 3 of 16

effect of many others has also been investigated; e.g., basalt fibres [13], hemp, flax, basalt, and bamboo
fibre reinforcements [21], as well as concrete in combination with wood [22].
The aim of this study is to explore possibilities of using domestic wood species for technologically
challenging big-sized window elements, with high added value. For this reason, the Norway spruce,
as the basic wood species, was used in this study.
During the first stage of the study, the mechanical characteristics of the used materials were
defined. Slender hybrid beams were manufactured using aluminium reinforcements and four-point
bending tests were performed in order to compare the bending stiffness and load-bearing capacities of
beams having different percentages, orientations, and positions of the reinforcing material.
All decisions about the type of adhesive, the number and position of the reinforcements and the
choice of material were made taking into account not only the required mechanical characteristics but
also future production feasibility and cost-effectiveness.

2. Materials and Methods

Norway spruce was used as the basic material for all the beam specimens which were prepared
for this study. Beams were manufactured from lamellas with thicknesses from 8 to 28 mm, with a
measured averaged moisture content of 12% and an average density of 460 kg/m3 . The mechanical
characteristics of the spruce, that correspond to structural grade C30 [23], used in window production,
compared to those of the reinforcing materials, are presented in Table 1. The shear strength was
determined experimentally, according to standard EN 205 [24].

Table 1. Material characteristics.

Norway Spruce Aluminium GFRP CFRP

(kg/m3 )
Density $ 460 2700 2000 1420
Poisson’s ratio ν 0.37 0.30 0.27 0.3
Modulus of elasticity E0 (MPa) 12,000 70,000 38,000 120,000
Shear modulus G (MPa) 750 25,000 5000 4000
Tensile strength ft (MPa) 19 195 / /
Compressive strength fc (MPa) 24 / / /
Shear strength fv (MPa) 8.4 * / / /
Thermal conductivity λ (W/mK) 0.11 210 0.3 5
Price factors compared to wood (per m3 ) 1 28 71 180
* Experimentally defined value.

Three different types of reinforcing material were planned to be used later on within the wider
scope of this study; aluminium, GFRP, and CFRP profiles. The longitudinal modulus of elasticity
of the CFRP profiles, aluminium, and GFRP profiles compared to that of Norway spruce are higher
by a factor of 10.0, 5.8, and 3.2, respectively (Table 1). Thus, CFRP profiles should provide the best
improvements in bending stiffness and load-bearing capacity [10,14,25]. Despite this, we decided
to perform the first phase of the tests with aluminium reinforcements only, in order to define their
positioning and orientation effect on the observed parameters and to analyse the possibilities of
manufacturing window profiles with non-visible reinforcements. The advantages of aluminium are the
uniformity of its mechanical properties, its high tensile capacity, and the easy processing, accessibility,
and low costs of the material itself. On the other hand, its high thermal conductivity decreases the
low energy efficiency of the windows, so it would be expected that in a further step of the research it
would be replaced with GFRP or CFRP profiles.
The aluminium profiles used for the analysis were made of aluminium alloy 6060-T66 and
technical data are collected from technical sheets provided by the manufacturer [26]. For GFRP and
CFRP profiles the data were also obtained from technical data sheets from the manufacturers [27].
For this study pre-analysis of the adhesive selection for connection between aluminium and wood
was performed. Tensile shear strength of three two-component epoxy adhesives (2K-EP) and three
Forests 2018, 9, 703 4 of 16

two-component polyurethane adhesives (2K-PU) were defined according to EN 1465:2009 [28], with
tensile lap-joint tests on aluminium specimens. For each adhesive, 10 specimens were tested and the
shear strength was measured. The mean shear strength of the Novasil P-SP adhesive was the lowest
and of COSMO PU was the highest (Table 2).

Table 2. Adhesives characteristics and experimental results of their shear strength.

Mean Shear Standard Deviation of Technical Sheet Shear

Brand Name Type
Strength (MPa) Shear Strength (MPa) Strength (MPa)
Körapox 565/GB 2K-EP 9.9 0.96 24
Permabond ET515 2K-EP 4.4 0.53 10
COSMO EP-200.110 2K-EP 6.1 0.75 18
Körapur 790/30 2K-PU 8.9 1.13 18
Novasil P-SP 6944 2K-PU 4.0 0.96 /
COSMO PU-200.280 2K-PU 11.2 0.75 18

Due to its largest shear strength, it was decided to use the two-component polyurethane
adhesive—COSMO PU 200.280—for further experimental analysis. All of the mean shear strengths
defined through experimental analysis had a value that was lower compared to the declared ones
on the manufacturers’ technical sheets. Nevertheless, the shear strength of 11.2 MPa (of the selected
2K-PU adhesive), was higher than the shear strength of the wood itself (8.4 MPa), which should present
sufficient strength to avoid the cohesive failure of the joint.
In the second step the four different window beam sections (scantlings), designated types A, B,
C, and D, were selected for the experimental study (Figure 2). Norway spruce was used as the basic
material,
Forests 2018,and aluminium
9, x FOR as the reinforcing material.
PEER REVIEW 5 of 17

Type A Type B
84 84
22

22

5
68
24

68
24

80
81

Type C Type D
84 84
10 8 8 17 24
28

20

21
68
12

68

3 4

4 5

Figure 2.
Figure Cross sections
2. Cross sections of
of the
the type
type A,
A, B,
B, C,
C, and
and D
D specimens
specimens (measurements
(measurements in
in mm).
mm).

The position and orientation of the reinforcements in the specimens, B, C, and D, were defined,
taking into account number of different aspects:
 To improve bending stiffness, load-bearing capacity and flexural-rigidity the reinforcements
need to be positioned near tensile and compressive faces, as far away from the neutral axis as
possible [9,15,25,29].
Forests 2018, 9, 703 5 of 16

The position and orientation of the reinforcements in the specimens, B, C, and D, were defined,
taking into account number of different aspects:

• To improve bending stiffness, load-bearing capacity and flexural-rigidity the reinforcements

need to be positioned near tensile and compressive faces, as far away from the neutral axis as
possible [9,15,25,29].
• The moment of inertia of the cross section should be maximized by proper orientation of the
reinforcements and the provision of a sufficient number of them.
• An aesthetic requirement—no reinforcement should be visible when the window is completed.
• The dimensions of the reinforcements should correspond to those of standard products available
on the market, in order to decrease production costs.
• The production feasibility aspect—the basic thickness of wooden lamellas, to produce window
scantlings, is 24 mm.
• Cost-effectiveness of the hybrid beams.

The outside measurements of the specimens were prepared with the following dimensions:
width (w) × height (h) × length (Le ) = 84 mm × 68 mm × 1950 mm. Basic wooden beams for
specimens were manufactured from lamellas with thicknesses from 8 to 28 mm. The type A specimens
were unreinforced (i.e., reference specimens). The type B specimens had two horizontal reinforcements
(with width × height dimensions of 80 mm × 4 mm) placed in the pre-milled grooves of the contact
regions between the two neighbouring wooden lamellas. The type C specimens had twelve vertical
reinforcements (with width × height dimensions of 3 mm × 20 mm), which were located in each of the
two outside wooden lamellas (six in each lamella). Type D specimens were prepared with two vertical
reinforcements with dimensions of 68 × 4 mm. All the pre-milled grooves were in width and height 1
mm bigger than reinforcements, leaving 0.5 mm space for the adhesive around the reinforcements.
The adhesive was applied to the grooves and the reinforcements were then manually inserted into
them. Specimens C had the highest number of reinforcements with its highest area (12.6%), while
specimens D had smallest number of reinforcements with the smallest area (9.5%) (Table 3).

Table 3. Characteristics of the tested specimens.

Dimensions of
Number of Number of Areal Percentage of
Specimen Series Reinforcement b
Specimens (-) Reinforcements (-) Reinforcement (%)
× h (mm) h
A1–A5 5 S1 - - -
* B1–B5 4 S1 80 × 4 2 11.2
C1–C5 5 S1 3 × 20 12 12.6
A6–A10 5 S2 - - -
B6–B10 5 S2 80 × 4 2 11.2
C6–C10 5 S2 3 × 20 12 12.6
D1–D5 5 S2 4 × 68 2 9.5
* B4 was excluded from the analysis due to its different dimension of the cross section (84 mm × 65 mm).

A decision about the preparation of type D specimens was made after the first tests of Series S1 on
specimens A, B, and C had been completed, in order to be able to make additional comparisons of the
effect of different reinforcement orientations on the bending stiffness, load-bearing capacity, flexural
rigidity, and type of failure mechanism. The type D specimens did not fulfil the aesthetic requirement
for non-visibility of the reinforcements.
Two test fields, for four-point bending tests, were prepared for two series of hybrid specimens;
Series S1 (Figure 3) and Series S2 (Figure 4). The difference between them was in different spans L,
lengths of the middle areas l, and the distance of the applied load from the support a. The simply
supported beams of Series S1 were 1950 mm long and had a span of L = 1750 mm. Loads of P/2,
which were 750 mm apart, were applied at a distance a = 500 mm from the beam supports (Figure 3).
The simply supported beams of Series S2 were 1950 mm long and had a span of L = 1850 mm. In this
requirement for non-visibility of the reinforcements.
Two test fields, for four-point bending tests, were prepared for two series of hybrid specimens;
Series S1 (Figure 3) and Series S2 (Figure 4). The difference between them was in different spans L,
lengths of the middle areas l, and the distance of the applied load from the support a. The simply
supported beams of Series S1 were 1950 mm long and had a span of L = 1750 mm. Loads of P/2, which
Forests 2018, 9, 703
were 750 mm apart, were applied at a distance a = 500 mm from the beam supports (Figure 3).6 of 16
The
simply supported beams of Series S2 were 1950 mm long and had a span of L = 1850 mm. In this case
loads of P/2,
case loads ofwhich were 700
P/2, which weremm apart,
700 mmwere
apart,applied at a distance
were applied a = 625 mm
at a distance a = from
625 mmthe beam supports
from the beam
(Figure
supports4).(Figure
The difference between both
4). The difference seriesboth
between wasseries
that the
was distance
that thebetween
distancethe loads l the
between decreased
loads l
and the span
decreased andLthe increased, in the case
span L increased, of Series
in the case ofS2. With
Series S2.this
Withchange, the bending
this change, moment
the bending with
moment
specimens of Series S2 increased, at the same loading force, compared to specimens
with specimens of Series S2 increased, at the same loading force, compared to specimens of Series of Series S1. This
step wasstep
S1. This made wasin order
made to
in decrease the probability
order to decrease that the failure
the probability would
that the bewould
failure causedbe bycaused
shear strength,
by shear
prior the bending strength would be fully exploited.
strength, prior the bending strength would be fully exploited.

P/2 P/2
a = 500 mm l = 750 mm a = 500 mm

L = 1750 mm
1950 mm

ForestsFigure Schematic
2018, 9,3.x FOR diagram of the four-point bending test field configuration for the specimens of7 of 17
PEER REVIEW
Figure 3. Schematic diagram of the four-point bending test field configuration for the specimens of
Series S1.
Series S1.
P/2 P/2
a = 625 mm l = 700 mm a = 625 mm

L = 1850 mm
1950 mm

Figure 4. Schematic diagram of the four-point bending test field configuration for the specimens of
Figure 4. Schematic diagram of the four-point bending test field configuration for the specimens of
Series S2.
Series S2.
The four-point static bending tests, up to failure, were performed using a Roel Amsler HA 100
universal servo-hydraulic
The four-point testing machine
static bending tests, up(Zwick GmbH
to failure, were& performed
Co. KG, Ulm, Germany)
using (Figure HA
a Roel Amsler 5). 100
The flexural
universal loading tests
servo-hydraulic were
testing performed
machine (Zwickin GmbH
displacement (i.e.,Ulm,
& Co. KG, stroke) control (Figure
Germany) mode, with
5). an
actuator
The flexural loading tests were performed in displacement (i.e., stroke) control mode, withthe
movement rate of 0.1 mm/s, up to failure. In the region between the two applied loads an
bending movement
actuator moment was constant,
rate the shear
of 0.1 mm/s, up toforce wasIn
failure. zero,
the and thebetween
region maximum therelative deflection
two applied loadswas
the
caused bymoment
bending the bending moments the
was constant, only. On each
shear forcespecimen
was zero,the deflections
and the maximum were measured at two points
relative deflection was
using linear voltage displacement transducers (LVDTs) marked as IND in Figure
caused by the bending moments only. On each specimen the deflections were measured at two points6.
usingThe absolute
linear deflection
voltage w1 was
displacement calculated(LVDTs)
transducers as the average
markedvalue
as IND of in
theFigure
deflections
6. measured at
the midspan by the IND 1 and IND 2. All the physical properties (i.e., the displacements, the strains,
and the load P) were measured and recorded by a Dewesoft DEWE 2500 data acquisition system
(DEWESoft d.o.o., Trbovlje, Slovenija).
 universal servo-hydraulic testing machine (Zwick GmbH & Co. KG, Ulm, Germany) (Figure 5).
The flexural loading tests were performed in displacement (i.e., stroke) control mode, with an
actuator movement rate of 0.1 mm/s, up to failure. In the region between the two applied loads the
bending moment was constant, the shear force was zero, and the maximum relative deflection was
caused
Forests 2018,by the bending moments only. On each specimen the deflections were measured at two points
9, 703 7 of 16
using linear voltage displacement transducers (LVDTs) marked as IND in Figure 6.

Forests 2018, 9, xFigure

FOR PEER REVIEW
5. The testing equipment for the performance of the four-point bending tests. 8 of 17
Figure 5. The testing equipment for the performance of the four-point bending tests.
P

Le

L/2 IND 1
IND 2

Figure6.6.Basic
Figure Basicconfiguration of the
configuration of the test
testinstrumentation.
instrumentation.

Effective flexural
The absolute rigidityw(EI)
deflection eff was
1 was
calculated
calculated as theon the basis
average valueofofthe
theaverage absolute
deflections midspan
measured at
deflection w1 , by
the midspan using an expression
the IND 1 and IND 2. ofAll
technical mechanics
the physical (Equation
properties (i.e., the(1)) that corresponds
displacements, to the
the strains,
four-point bending
and the load test load
P) were arrangement.
measured The equation
and recorded neglects
by a Dewesoft a small
DEWE part
2500 of deflections
data acquisition caused
system by
shear stresses.
(DEWESoft d.o.o., Trbovlje, Slovenija).
Effective flexural rigidity (EI)
(EI)eff = 0.33
eff was Pu ·a·(3L2on−the
calculated 4a2basis
)/48·w of1 the average absolute midspan(1)
deflection w1, using an expression of technical mechanics (Equation (1)) that corresponds to the four-
The ultimate load (Pu ) is the load where the failure occurred, a is the distance between the support
point bending test load arrangement. The equation neglects a small part of deflections caused by
and the load, L is the span length and w1 is the absolute midspan deflection.
shear stresses.
3. Results (EI)eff = 0.33 Pu·a·(3L2 − 4a2)/48·w1 (1)
The
Thebending stiffness
ultimate load (P was
u) iscalculated for eachthe
the load where specimen type at three
failure occurred, a is different loading
the distance between stages:
theat a
support
load and the load,
corresponding L is the
to 33% span
of the length and
ultimate loadw(0.33
1 is the
Puabsolute
); at themidspan deflection.
limit of proportionality (Pp ), which
represents the end of the region of linear behaviour; and at the ultimate load (Pu ) (Table 4). Beside
the3.corresponding
Results load, the deflection occurring at midspan (w1,i ) and the bending stiffness (k1,i ) are
given atTheeach loading step (iwas
bending stiffness = 1/3 for the loading
calculated for each step 0.33 Ptype
specimen u , i =atpthree
for the loading
different step Pstages:
loading p and iat= u
forathe
loadloading step Pu ).toThe
corresponding 33%bending stiffness
of the ultimate was
load calculated
(0.33 Pu); at the aslimit
k1,i =ofPproportionality
i /w1 ,i . The flexural rigidity
(Pp), which
was calculated
represents thefrom
end of the
thedisplacement
region of linearw1behaviour;
, and wasand the at
highest at C specimens
the ultimate and the
load (Pu) (Table lowestthe
4). Beside at A
corresponding
specimens (Table 4).load, the deflection occurring at midspan (w 1,i) and the bending stiffness (k 1,i) are given
at each loading step (i = 1/3 for the loading step 0.33 Pu, i = p for the loading step Pp and i = u for the
loading step Pu). The bending stiffness was calculated as k1,i = Pi/w1,i. The flexural rigidity was
calculated from the displacement w1, and was the highest at C specimens and the lowest at A
specimens (Table 4).
Forests 2018, 9, 703 8 of 16

Table 4. Summary of the experimental results of the specimens of both series of tests (i.e., S1 and S2). The loads (Pi ) and midspan deflections (w1,i ) are presented as
averaged measured values of all specimens, of the same type and series, and the bending stiffness was calculated as ki = Pi /wi . In the last six columns comparisons of
the different loads (Pi ), the midspan deflections (w1,i ), the bending stiffnesses, and flexural rigidity are presented.

Flexural
At 33% of Ultimate Load At the Proportional Limit At Ultimate Load Comparisons
Rigidity
Specimens P1/3 w1,1/3 k1,1/3 Pp w1,p k1,p Pu w1,u k1,u Pu,(B,C,D) /Pu,A Pp /Pu w1,p /w1,u k1,p /k1,u k1,1/3 /k1,u (EI)eff
No. (kN) (mm) (kN/mm) (kN) (mm) (kN/mm) (kN) (mm) (kN/mm) (-) (-) (-) (-) (-) (kN*m2 )
A1–A5 6.2 19.5 0.32 12.7 39.4 0.32 18.6 67.4 0.30 - 0.68 0.58 1.09 1.13 27.5
Series
B1–B5 * 5.7 12.1 0.48 14.3 30.9 0.46 17.1 37.5 0.46 0.92 0.84 0.82 1.01 1.04 38.7
S1
C1–C5 8.2 17.0 0.47 15.9 33.2 0.48 24.7 58.2 0.44 1.33 0.64 0.57 1.09 1.10 41.2
A6–A10 4.8 20.0 0.24 10.3 43.3 0.24 14.5 78.6 0.19 - 0.71 0.55 1.26 1.29 27.3
Series B6–B10 5.7 17.3 0.33 14.0 42.5 0.33 17.2 54.6 0.32 1.18 0.82 0.78 1.04 1.04 37.5
S2 C6–C10 7.0 20.8 0.34 13.9 40.6 0.34 21.1 80.4 0.27 1.45 0.66 0.51 1.28 1.27 39
D1–D5 6.5 20.0 0.32 11.8 36.2 0.33 19.5 81.7 0.24 1.35 0.60 0.44 1.35 1.34 37.6
* B4 was excluded from further analysis due to the change in cross section dimension.
Forests 2018, 9, x FOR PEER REVIEW 9 of 17
Forests2018,
Forests 2018,9,9,x703
FOR PEER REVIEW 99ofof17
16

The results show that the ultimate load (Pu) and bending stiffness (ki) were the highest in the
The results show that the ultimate load (Pu) and bending stiffness (ki) were the highest in the
case of theresults
The type Cshowbeamthat
specimens, which
the ultimate had(Pthe
load highest
u ) and percentage
bending of reinforcements,
stiffness whereas
(ki ) were the highest it
in the
case of the type C beam specimens, which had the highest percentage of reinforcements, whereas it
was
casethe lowest,
of the typeas C expected, in the case
beam specimens, of the
which had type
theA beam specimens,
highest percentage which had no reinforcement
of reinforcements, whereas it
was the lowest, as expected, in the case of the type A beam specimens, which had no reinforcement
(Figure
was the7). The comparisons
lowest, as expected,ofinthe
thebending stiffnesses
case of the type A k 1,p and
beam k1,u show the
specimens, plasticity
which had nocapacity of the
reinforcement
(Figure 7). The comparisons of the bending stiffnesses k1,p and k1,u show the plasticity capacity of the
tested
(Figurebeams.
7). TheIncomparisons
the case of the type
of the B beamstiffnesses
bending specimens, both
k1,p andparameters
k1,u show thehave almost capacity
plasticity equal values,
of the
tested beams. In the case of the type B beam specimens, both parameters have almost equal values,
which
tested corresponds
beams. In thetocase an of
almost linear
the type load‐deflection
B beam specimens, relationship up tohave
both parameters failure (Figure
almost equal7). The
values,
which corresponds to an almost linear load‐deflection relationship up to failure (Figure 7). The
biggest difference between
which corresponds these linear
to an almost two parameters (i.e., 34%)
load-deflection occurred
relationship upin
tothe case(Figure
failure of the beams
7). TheD1–D5,
biggest
biggest difference between these two parameters (i.e., 34%) occurred in the case of the beams D1–D5,
which indicates
difference theirthese
between nonlinear behaviour (i.e.,
two parameters between
34%)the limit ofinthe
occurred theproportionality
case of the beams andD1–D5,
the ultimate
which
which indicates their nonlinear behaviour between the limit of the proportionality and the ultimate
load.
indicates their nonlinear behaviour between the limit of the proportionality and the ultimate load.
load.

Figure 7. Ultimate load-bearing capacity and stiffness coefficient of the tested specimens of types A,
Ultimateload-bearing
Figure7.7.Ultimate
Figure load-bearing capacity
capacity and stiffness coefficient
coefficient of
ofthe
thetested
testedspecimens
specimensofoftypes
typesA,A,B,
B, C, and D.
C, and D.
B, C, and D.
The
The average values of
average values of themeasured
measured midspan deflections were comparedat anat an applied load
The average values ofthe
the measuredmidspan
midspan deflections
deflectionswere
werecompared
compared atapplied load equal
an applied load
equal to P =
to P =to10PkN,10 kN, where the load‐deflection relationships are still linear (Figure 8).
equal = 10where the load-deflection
kN, where relationships
the load‐deflection are still
relationships arelinear (Figure
still linear 8). 8).
(Figure

Averagemidspan
Figure8.8.Average
Figure midspandeflection
deflection (w
(w11))of
ofthe
thebeam
beamspecimens
specimensA,A,B,
B,C,
C,DDat
atthe
theload
loadPP==1010kN
kNand
and
Figure
the 8. Average
bending midspan
stiffness deflection
improvement (wof
(%) 1) the
of the beamofspecimens
beams types B, A,DB,
C, C, D at the
compared to load
the P = 10
type A kN and
beams.
the bending stiffness improvement (%) of the beams of types B, C, D compared to the type A beams.
the bending stiffness improvement (%) of the beams of types B, C, D compared to the type A beams.
 Forests 2018, 9, 703 10 of 16
Forests 2018, 9, x FOR PEER REVIEW 10 of 17

The
Theaverage
averagemeasured
measuredmidspan midspandeflections
deflections(w (w11—calculated
—calculatedas asthe
themean
meanvalues
valuesofofIND IND11and and2)2)
of
ofthe
thebeam
beamspecimens
specimensof oftypes
typesA, A,B,B,and
andC, C,ofofSeries
SeriesS1, S1,were
were32.232.2mm,
mm,21.5 21.5mm,
mm,and and19.8
19.8mm,mm,
respectively (Figure 8). The bending stiffness improvements of the beam
respectively (Figure 8). The bending stiffness improvements of the beam specimens of types B and specimens of types B and C
when compared to the type A reference specimens, were 33% and
C when compared to the type A reference specimens, were 33% and 39%, respectively. The average39%, respectively. The average
midspan
midspandeflections
deflectionsofofthe theSeries
SeriesS2S2beams
beamsofoftypes
typesA, A,B,B,CCand andDDamounted
amountedtoto42.0 42.0mm,
mm,30.130.1mm,mm,
30.0
30.0 mm, and 30.6 mm, respectively. The improvement in bending stiffness of the hybrid beamsof
mm, and 30.6 mm, respectively. The improvement in bending stiffness of the hybrid beams of
types
typesB,B,C,
C, and
and D, D, compared
compared to to that
that of
of type
type AA beams,
beams,amounted
amountedtoto28%, 28%,29%,
29%,andand27%,
27%,respectively.
respectively. In
In both
both series
series of tests,
of tests, the midspan
the midspan deflection
deflection of the reinforced
of the reinforced hybridwas
hybrid beams beams was decreased
decreased compared
compared to the reference specimens of type A. The largest standard
to the reference specimens of type A. The largest standard deviation of the midspan deflection deviation of the midspan was
deflection
observed withwas observed
the type Awith the type
beams, in theAcase
beams, in the
of both caseThis
series. of both
wasseries.
expected,Thissince
was these
expected, since
specimens
these
were specimens
made solelywere madewhich
of wood, solelyisof wood, which is a non-homogeneous
a non-homogeneous material,
material, with significant with significant
variations in density
variations in density and
and different grain orientations. different grain orientations.
On
Onaverage,
average,the thehybrid
hybridbeamsbeamshavehaveimproved
improvedload-bearing
load-bearingcapacity
capacitycompared
comparedto tothe
thereference
reference
beams (Figure 9). Only the type B hybrid beams, of Series S1, had 8%
beams (Figure 9). Only the type B hybrid beams, of Series S1, had 8% lower average load-bearing lower average load-bearing
capacity
capacitythan
thanthe thereference
referencetype typeA Aspecimens.
specimens.The Thereason
reasonfor forthis
thislies
liesin
inthe
thehigh
highshear
shearstresses,
stresses,which
which
caused a shear failure (Figure 10b), before the flexural load-bearing capacity
caused a shear failure (Figure 10b), before the flexural load-bearing capacity was reached. The failure was reached. The failure
happened
happenedin inthe
the adhesion
adhesion layerlayer between
between the wood and
the wood and thethe reinforcement.
reinforcement.The Thetype
typeCCspecimens
specimenshad hada
aload-bearing
load-bearingcapacity
capacitywhichwhichwas was33%33%greater
greaterthan
thanthat
thatofofthe
thetype
typeAAreference
reference specimens
specimens (Figure
(Figure 9).
9). The load-bearing capacity of the Series S2 tested specimens of types A, B,
The load-bearing capacity of the Series S2 tested specimens of types A, B, C, and D amounted to 14.5 kN, C, and D amounted to
14.5
17.2 kN, 17.2 kN,
kN, 21.1 kN,and 21.119.5
kN,kN, and 19.5 kN, respectively
respectively (Figure 9). The (Figure 9). The capacity
load-bearing load-bearing capacity in
was increased was
all
increased in all cases of the reinforced beams by more than 18%. The highest
cases of the reinforced beams by more than 18%. The highest load-bearing capacity was observed in load-bearing capacity
was observed
the case of theintypethe C case of the
hybrid type C
beams. It hybrid beams.
was higher than It the
wasload-bearing
higher than capacity
the load-bearing
of the type capacity of
A, B, and
the type A, B, and D beams by 45%, 23%, and 8%, respectively. Type A,
D beams by 45%, 23%, and 8%, respectively. Type A, C, and D specimens had a typical tensile failure C, and D specimens had a
typical
and typetensile failure and
B specimens type Bshear
a typical specimens
failurea(Figure
typical 10).
shear failure (Figure 10).

Figure 9. Average load-bearing capacity (Pu ) of the beams of type A, B, C, and D, and the improvement
Figure 9. Average load-bearing capacity (Pu) of the beams of type A, B, C, and D, and the improvement
in this capacity of the beams of types B, C, and D compared to the type A beams.
in this capacity of the beams of types B, C, and D compared to the type A beams.
Forests 2018, 9, x FOR PEER REVIEW 11 of 17

Forests 2018, 9, 703 11 of 16

Forests 2018, 9, x FOR PEER REVIEW 11 of 17

(a) tensile failure (b) shear failure

Figure 10. Typical failure mechanisms of specimen types A, C, D (a) and specimen type B (b).
(a) tensile failure (b) shear failure
Effective
Figure flexural rigidity
10. Typical (EI)
failure eff was calculated
mechanisms of specimen
specimenon types
the basis
types A, C, of
A, C, D theand
D (a)
(a) average
and specimen
specimenabsolute
type midspan
type BB (b).
(b).
deflection w1, using a well-known expression of technical mechanics (Equation (1)) that corresponds
Effective
Effective flexural
to the four-point bending
flexural rigidity
test load
rigidity (EI) was calculated
(EI)arrangement.
eff was
eff calculated on the
the basis
The equation
on of
of the
neglects
basis thea average
small part
average absolute midspan
of deflections
absolute midspan
deflection
caused by w
shear , using a
stresses.well-known expression of technical mechanics (Equation
deflection w1, using a well-known expression of technical mechanics (Equation (1)) that corresponds
1 (1)) that corresponds
to
to the
the four-point
four-point bending
bending test
test load arrangement. The equation
equation neglects
neglects a a small
small part
part ofof deflections
(EI)load arrangement. The
eff = 0.33 Pu·a·(3L2 − 4a2)/48·w1
deflections
(2)
caused
caused by by shear
shear stresses.
stresses.
The
The hybrid
hybridbeams beamsofoftype
typeCChad hadthe the highest
highest flexural rigidity,followed
flexural rigidity, followedby bythe
thebeams
beamsofoftype typeB,B,
the
the beams
beams of of type
type D, D, and
and lastly
lastly (EI)
thethe
beams eff = 0.33 Pu·a·(3L2 − 4a2)/48·w1
beams of type
of type A (Figure
A (Figure 11). 11).
For For
SeriesSeries
S1 theS1flexural
the flexural rigidity
rigidity (2)
of B
of type
type
andBCThe
and C beams
beams
hybrid was higher
wasbeams
higherofintype inCcomparison
comparisonhad the with with
beams
highest beams
type type
flexural byA41%by followed
Arigidity, 41% and 50%,
and 50%, respectively.
respectively.
by the beams For ForB,
Series
of type
Series
S2, S2, the flexural rigidity of type B, C, and D beams was higher in comparison
the beams of type D, and lastly the beams of type A (Figure 11). For Series S1 the flexural rigidityfor
the flexural rigidity of type B, C, and D beams was higher in comparison with with
beam beamtype type
A of
A 37%,
for 37%,
type 43%,
B and 43%,
and and
38%,38%,
C beams wasrespectively.
respectively.
higher in comparison with beams type A by 41% and 50%, respectively. For
Series S2, the flexural rigidity of type B, C, and D beams was higher in comparison with beam type
A for 37%, 43%, and 38%, respectively.

Figure
Figure 11.11. Effective
Effective flexural
flexural rigidity
rigidity of of type
type A,A,
B, B,
C,C, and
and DD specimens.
specimens.

DueDuetoto thethe high

high ratio
ratio between
between thethe modulus
modulus ofof elasticity
elasticity (E(E 0 ) and the shear modulus (G) of wood
0) and the shear modulus (G) of wood
Figure 11. Effective flexural rigidity of type A, B, C, and D specimens.
(i.e., E
(i.e., E0/G /G = 16, Table 1) and a relatively high span
0 = 16, Table 1) and a relatively high span and specimen height and specimen height
(L/h) (L/h)
ratioratio (which
(which amounts
amounts to
to L/h
L/h = 25.7 = 25.7 and L/h = 27.2 for Series S1 and Series S2, respectively), the deflection caused by shear
Dueand to theL/h = 27.2
high for
ratio Series S1
between theand Seriesof
modulus S2, respectively),
elasticity (E0) and thethedeflection
shear moduluscaused(G) byofshear
wood
stressesshould
stresses shouldnot notbebeneglected.
neglected.Kretschmann
Kretschmann [30]
[30] suggested
suggested that
that thethe flexural
flexural rigidity
rigidity should
should bebe
(i.e., E0/G = 16, Table 1) and a relatively high span and specimen height (L/h) ratio (which amounts to
increased
increased by
by and 10%
10% L/h if the deflection
if the= deflection due
due to to the shear stresses is neglected, as it is in the case of Equation
L/h = 25.7 27.2 for Series S1theandshear stresses
Series is neglected, the
S2, respectively), as itdeflection
is in the case of Equation
caused by shear
(1).
(1). TheThe results
results ofof a study
anot
study reported byby Eierle and Bös [31] confirmed, that
thethe shear deflection depends
stresses should be reported
neglected. Eierle and
Kretschmann Bös [31]
[30] confirmed,
suggested that
that the shear
flexural deflection
rigidity depends
should be
on the
onincreased length
the length and height ratio (i.e., L/h) of the beam, and on the ratio between the shear and elastic
byand
10%height ratio (i.e., L/h)
if the deflection due to ofthe
theshear
beam,stresses
and onisthe ratio between
neglected, as it is the shear
in the caseand elastic
of Equation
moduli
moduli (i.e.,
(i.e., G/E
G/E0of ).
). 0The The smaller
smaller this
this ratioratio is,
is, the the larger
larger is
is the the effect of
effect of shear shear stresses
stresses on the deflection.
(1). The results a study reported by Eierle and Bös [31] confirmed, that the shearon the deflection.
deflection depends
on the length and height ratio (i.e., L/h) of the beam, and on the ratio between the shear and elastic
moduli (i.e., G/E0). The smaller this ratio is, the larger is the effect of shear stresses on the deflection.
 Forests 2018, 9, x FOR PEER REVIEW 12 of 17
Forests 2018, 9, 703 12 of 16
Forests 2018, 9, x FOR PEER REVIEW 12 of 17
At rectangular shaped simply supported beams with L/h ratio above 25, the shear deflection should
At
At rectangular
notrectangular shaped
represent more
shaped simply
than supported
2% ofsupported
simply beams with
the total deflection.
beams with L/h
L/h ratio
ratio above
above 25,25, the
the shear
shear deflection
deflection should
should
not represent
Midspanmore
not represent more than
deflection 2%
than 2%(w1ofof the
) was total deflection.
the measured in dependence of the load (P), for all beam types of Series
total deflection.
Midspan
S1 and deflection
Seriesdeflection
Midspan S2 (Figures(w12
(w 1) was
andmeasured in dependence
13). The angle of the
of the linear partload (P), curves
of the for all beam
from types of Seriesis
the abscissa
1 ) was measured in dependence of the load (P), for all beam types of Series
S1 and
defining
S1 Series S2 (Figures
the bending
and Series 12
stiffness;
S2 (Figures and
12 andthe 13). The
13).higher angle of
the angle
The angle the linear
of thethe part
higher
linear parttheof the curves
of bending
the curves from
stiffness.the
from the abscissa is
abscissa
Load-bearing is
defining
defining the
capacity the bending
is defined
bending stiffness;
bystiffness;
the load at the higher
thewhich
higher the
thethe angle
curve
angle the
ends higher
thefrom the
thethe
higher bending
exception
bending stiffness.
of stiffness. Load-bearing
specimensLoad-bearing
marked with
capacity
capacity is
is defined
* sign, which defined bybythe
theload
are specimens, load atat
wherewhich thethe
the full
which curve
range ends
of at
curve from
least
ends the exception
one
from LVDT wasofreached,
the exception specimens marked
before
of specimens the with
failure
marked
*with
sign, which
occurred. are specimens, where the full range of at least one LVDT was reached,
* sign, which are specimens, where the full range of at least one LVDT was reached, before the before the failure
occurred.
failure occurred.

Figure 12. Load‐midspan deflection relationship of Series S1. * B4 was excluded from further analysis
Figure
Figure 12.
due to12. Load-midspan
theLoad‐midspan deflection
change in crossdeflection relationship of
relationship
section dimension. of Series
Series S1.
S1. ** B4
B4 was
was excluded
excluded from
from further
further analysis
analysis
due to the change in cross section dimension.
due to the change in cross section dimension.

Figure 13. Load-midspan deflection relationship of Series S2.

Figure 13. Load‐midspan deflection relationship of Series S2.
Figure 13. Load‐midspan deflection relationship of Series S2.
Forests 2018, 9, 703 13 of 16

Type A specimens had the smallest bending stiffness and load-bearing capacity but showing the
highest dispersion of the results at the same time (Figure 12). The highest values (bending stiffness
and ultimate load) were measured at type C specimens. The ultimate load was recorded and taken
into consideration in the analysis of the results in Table 4.
In both figures (Figures 12 and 13) the same deflection range and load range had been used in
order to be able to directly compare also the behaviour of specimens being tested on two different
testing fields. With the increase of the span L and the increase of the distance between the support
and the load a, the bending moment has increased (at the same load level) at specimens of Series S2,
compared to specimens of Series S1. The increase of the bending moment by 25% in Series S2, has
decreased the ultimate load Pu (Series S2) for only 12% compared to ultimate load of specimens of
Series S1. The reason lies in a prevailing failure mechanism, which was with Series S2 specimens a
typical tensile failure.

4. Discussion
Aluminium reinforcements used in wooden beams have improved bending stiffness,
load-bearing capacity and flexural rigidity. The experiments showed almost linear dependency
between reinforcement-wood ratio and bending stiffness and nonlinear dependency between
reinforcement-wood ratio and load-bearing capacity. This was proven also by Kim and Harries [16],
who showed that there is a limit of the reinforcement/wood ratio, beyond which the load-bearing
capacity can no longer increase. The reason for the smallest increase of load-bearing capacity with
specimens B, lies in the high number of shear failure mechanisms (89% of all B specimens tested
have failed in shear). Borgin et al. [32] have described the positive effect of the knots in the wood
on the shear stress resistance. In the window industry all wood must be free of knots, so the shear
stress depends on the wood shear properties only. The shear strength of the specimens could be
additionally increased with different wood species in the middle part of the cross section [33], but the
most effective it is with the additional reinforcements. There are not many attempts to use aluminium
as the reinforcing material. Due to higher moduli of elasticity and better thermal properties, steel
is used much more often, [5,9,12]. Nevertheless, in our experimental analysis, it was proven that
aluminium is a good material to be used in a first phase of the wood hybrid beam analysis, to define
the positioning, orientation of the reinforcements, and the selection of the proper adhesives. One of the
biggest advantages of this approach is the ability to process aluminium-wooden hybrid beams with
standard machinery for window manufacturing. During the shear strength tests of the adhesives it was
observed that surface treatment of the aluminum is essential in order to achieve the adequate adhesion
between adhesive and aluminium. The importance of surface treatment of the reinforcement was
also declared by Jasienko and Nowak [5], who sanded the steel surface before adhesion. For proper
adhesion, aluminium embedded in our specimens was anodized.
The presented study was made as a starting point for further analysis, more focused on the
reinforcements in window profiled elements. Mullion profile (middle window profile in the double
sashed window, Figure 14) is a combination of one of the most frequently used window profiles and
one of the most exposed window elements to wind load.
From the four-point bending tests results and observed failure mechanisms it can be concluded,
that the optimal reinforcement arrangement would be a combination of the reinforcement used in the
type C and D specimens. With reinforcements positioned in both, the tensile and compressive zones,
the tensile and compressive stress strength of the beams (as in the case of the type C specimens) can be
increased. A vertically positioned reinforcement over the almost whole cross sectional height (as in the
case of the type D specimens) prevents a shear failure before the tensile and compressive stress limits
are reached. A preposition of such a window hybrid beam is presented in Figure 15.
Forests 2018, 9, x FOR PEER REVIEW 14 of 17

Forests 2018,9,9,703
Forests2018, x FOR PEER REVIEW 14
14ofof16
17
105
105

24 24 22 22
68 68

37
37
190
190
Figure 14. Mullion‐middle fixed element of the double sashed window (measurements in mm).
Figure 14. Mullion-middle fixed element of the double sashed window (measurements in mm).
Figure 14. Mullion‐middle fixed element of the double sashed window (measurements in mm).
80
80

2 8 8
41 41
2
10 10

25
Figure 15.
25
Example of window profile with horizontally and vertically positioned reinforcements
Figure 15. Example
(measurements of window profile with horizontally and vertically positioned reinforcements
in mm).
(measurements in mm).
Figure 15. Example of window profile with horizontally and vertically positioned reinforcements
But on the other
(measurements in hand,
mm). a vertical metal reinforcement through the whole specimen height can
lead But on the transmission
to thermal other hand, aand vertical metaldecrease
therefore reinforcement
thermalthrough
efficiencytheofwhole specimen
the window height
or even can
worse,
lead to
can lead thermal
Butto ona the transmission
localother
decrease and therefore
hand,ofainner decrease
temperature
vertical thermal efficiency
that can further
metal reinforcement through of
on lead the window
to condensation
the whole or even worse,
and mold
specimen height can
can
growthlead to
[34].a local
If we decrease
could of
achieve inner
the temperature
proper bending that can further
stiffness of the on lead
window to condensation
elements
lead to thermal transmission and therefore decrease thermal efficiency of the window or even worse, with and mold
aluminium
growth
can lead[34].
reinforcementsto a If wedecrease
that
local could achieve
would not the
go through
of inner proper bending
the whole
temperature stiffness
height
that can of the
further onof the towindow
profile,
lead this wouldelements
condensation be and with
the most
mold
aluminium
cost-effective reinforcements
and applicable that would
solution innot go
window through the
production. whole height
growth [34]. If we could achieve the proper bending stiffness of the window elements with of the profile, this would be
the most
aluminium cost-effective
In further analysis,and
reinforcements theapplicable
mullion
that would solution
profile willin
not go window
be
through theproduction.
investigated
whole toheight
defineof maximal possible
the profile, bending
this would be
In
stiffness further
the mostand analysis,
load-bearing
cost-effective the
and mullion
improvements,profile will be
considering
applicable solution investigated
positioning,
in window to define maximal possible bending
orientation, material, and number
production.
stiffness and load-bearing
of reinforcements.
In further improvements,
analysis, the mullion profile considering positioning,toorientation,
will be investigated define maximal material, and number
possible bending
ofstiffness
reinforcements.
and load-bearing improvements, considering positioning, orientation, material, and number
5. Conclusions
of reinforcements.
5. Conclusions
An experimental analysis of the effect of reinforcements on the bending stiffness, load-bearing
5. Conclusions
An experimental
capacity, analysisofof
and flexural rigidity the effect
small of reinforcements
cross section hybrid window on the bending
beams, stiffness,
reinforced load-bearing
with aluminium
capacity,
reinforcementsand flexural
An experimental rigidity
was analysed.
analysis Theof small
of results cross
show
the effect section
of that hybrid
all three analysed
reinforcements window beams,
on the parameters reinforced
were the
bending stiffness, with
best in
load-bearingthe
capacity, and flexural rigidity of small cross section hybrid window beams, reinforced with
Forests 2018, 9, 703 15 of 16

case of the type C specimens with 12 (3 × 20 mm) vertical reinforcements in two rows. The type B
specimens had a very low load-bearing capacity, due to their low resistance to shear stress, whereas
the type D beams never failed due to a lack of shear strength. The shape of the window profile and the
results of this analysis indicate that the reinforcement should be arranged by combining the reinforcing
methods used in the type C and type D beam specimens. Based on the results of this study, further
numerical and experimental analyses will need to be performed, using real shaped window profiles
(Figure 15).

Author Contributions: B.Š., J.L. and G.F. conceived and designed experiments; B.Š. and J.L. performed
the experiments; B.Š. analysed the data; J.L. and G.F. conducted the validation and formal analysis; B.Š.,
writing—original draft preparation; J.L. and G.F., writing—review and editing.
Funding: This work was supported by Ministry of Education, Science and Sport of the Republic of Slovenia,
and by the European Regional Development Fund, European Commission (project TIGR4smart, grant number
5441-1/2016/116) and by Ministry of Education, Science and Sport of the Republic of Slovenia within the
framework of the Program P2-0182.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. PHI Certified Component Database. Available online: https://database.passivehouse.com/en/components/
list/window (accessed on 30 September 2018).
2. Jiménez, P.; Dunkl, A.; Eibel, K.; Denk, E.; Grote, V.; Kelz, C.; Moser, M. Wood or Laminate? Psychological
Research of Customer Expectations. Forests 2016, 7, 275. [CrossRef]
3. Verband Fenster + Fassade, G.F. und H.V. Holzarten für den Fensterbau—Teil 1: Eigenschaften, Holzarttabellen;
Verband Fenster + Fassade, G.F. und H.V: Frankfurt am Main, Germany, 2018.
4. ISO 10077-2: 2017. Thermal Performance of Windows, Doors and Shutters—Calculation of Thermal
Transmittance—Part 2: Numerical Method for Frames; ISO: Geneva, Switzerland, 2017.
5. Jasieńko, J.; Nowak, T.P. Solid timber beams strengthened with steel plates—Experimental studies.
Constr. Build. Mater. 2014, 63, 81–88. [CrossRef]
6. Alhayek, H.; Svecova, D. Flexural Stiffness and Strength of GFRP-Reinforced Timber Beams. J. Compos. Constr.
2012, 16, 245–252. [CrossRef]
7. Andor, K.; Lengyel, A.; Polgár, R.; Fodor, T.; Karácsonyi, Z. Experimental and statistical analysis of spruce
timber beams reinforced with CFRP fabric. Constr. Build. Mater. 2015, 99, 200–207. [CrossRef]
8. Raftery, G.M.; Kelly, F. Basalt FRP rods for reinforcement and repair of timber. Compos. Part B Eng. 2015, 70,
9–19. [CrossRef]
9. Borri, A.; Corradi, M. Strengthening of timber beams with high strength steel cords. Compos. Part B Eng.
2011, 42, 1480–1491. [CrossRef]
10. Fiorelli, J.; Dias, A.A. Analysis of the strength and stiffness of timber beams reinforced with carbon fiber and
glass fiber. Mater. Res. 2003, 6, 193–202. [CrossRef]
11. Premrov, M.; Dobrila, P. Experimental analysis of timber–concrete composite beam strengthened with carbon
fibres. Constr. Build. Mater. 2012, 37, 499–506. [CrossRef]
12. Winter, W.; Tavoussi, K.; Pixner, T.; Parada, F.R. Timber-Steel-Hybrid Beams for Multi-Storey Buildings.
In Proceedings of the World Conference on Timber Engineering, Auckland, New Zealand, 15–19 July 2012;
Volume 2012.
13. Thorhallsson, E.R.; Hinriksson, G.I.; Snæbjörnsson, J.T. Strength and stiffness of glulam beams reinforced
with glass and basalt fibres. Compos. Part B Eng. 2017, 115, 300–307. [CrossRef]
14. Nadir, Y.; Nagarajan, P.; Ameen, M.; Arif, M. Flexural stiffness and strength enhancement of horizontally
glued laminated wood beams with GFRP and CFRP composite sheets. Constr. Build. Mater. 2016, 112,
547–555. [CrossRef]
15. Valipour, H.R.; Crews, K. Efficient finite element modelling of timber beams strengthened with bonded fibre
reinforced polymers. Constr. Build. Mater. 2011, 25, 3291–3300. [CrossRef]
16. Kim, Y.J.; Harries, K.A. Modeling of timber beams strengthened with various CFRP composites. Eng. Struct.
2010, 32, 3225–3234. [CrossRef]
Forests 2018, 9, 703 16 of 16

17. Taheri, F.; Nagaraj, M.; Khosravi, P. Buckling response of glue-laminated columns reinforced with
fiber-reinforced plastic sheets. Compos. Struct. 2009, 88, 481–490. [CrossRef]
18. Raftery, G.M.; Harte, A.M. Nonlinear numerical modelling of FRP reinforced glued laminated timber.
Compos. Part B Eng. 2013, 52, 40–50. [CrossRef]
19. de Jesus, A.M.P.; Pinto, J.M.T.; Morais, J.J.L. Analysis of solid wood beams strengthened with CFRP laminates
of distinct lengths. Constr. Build. Mater. 2012, 35, 817–828. [CrossRef]
20. Khelifa, M.; Lahouar, M.A.; Celzard, A. Flexural strengthening of finger-jointed Spruce timber beams with
CFRP. J. Adhes. Sci. Technol. 2015, 29, 2104–2116. [CrossRef]
21. Borri, A.; Corradi, M.; Speranzini, E. Reinforcement of wood with natural fibers. Compos. Part B Eng. 2013,
53, 1–8. [CrossRef]
22. Ferrier, E.; Agbossou, A.; Michel, L. Mechanical behaviour of ultra-high-performance fibrous-concrete wood
panels reinforced by FRP bars. Compos. Part B Eng. 2014, 60, 663–672. [CrossRef]
23. EN 338:2016. Structural Timber—Strength Classes; BSI: London, UK, 2016.
24. EN 205:2016. Adhesives—Wood Adhesives for Non-Structural Applications—Determination of Tensile Shear Strength
of Lap Joints; BSI: London, UK, 2016.
25. Yang, Y.; Liu, J.; Xiong, G. Flexural behavior of wood beams strengthened with HFRP. Constr. Build. Mater.
2013, 43, 118–124. [CrossRef]
26. Voßedelstahlhandel GmbH & Co. KG. Technical Datasheet for Aluminium Alloy—6060-T66; Voßedelstahlhandel
GmbH & Co. KG: Neu Wulmstorf, Germany, 2014.
27. Exel Composites Benefites of Composites. Available online: http://www.exelcomposites.com/en-us/
english/composites/composites/ (accessed on 9 March 2017).
28. EN 1465:2009. Adhesives—Determination of Tensile Lap-Shear Strength of Bonded Assemblies; BSI: London,
UK, 2009.
29. De Luca, V.; Marano, C. Prestressed glulam timbers reinforced with steel bars. Constr. Build. Mater. 2012, 30,
206–217. [CrossRef]
30. Kretschmann, E.D. Mechanical Properties of Wood; General Technical Report FPL-GTR-190; Forest Products
Laboratory, United States Department of Agriculture: Washington, DC, USA, 2010; p. 46.
31. Eierle, B.; Bös, B. Schubverformungen von Stabtragwerken in der praktischen Anwendung. Bautechnik 2013,
90, 747–752. [CrossRef]
32. Borgin, K.B.; Loedolff, G.F.; Asce, M.; Saunders, G.R. Laminated wood beams reinforced with steel strips.
Proc. Am. Soc. Civ. Eng. 1968, 94, 1681–1706.
33. Tomasi, R.; Parisi, M.A.; Piazza, M. Ductile Design of Glued-Laminated Timber Beams. Pract. Period. Struct.
Des. Constr. 2009, 14, 113–122. [CrossRef]
34. Ruediger, D.; Jehl, W.; Koos, F.; Seggebruch, F.; Strasser, G.; Detlef, T.; Wiegand, L.; Zimmermann, H.-H.
Wärmetechnische Anforderungen an Baukörperanschlüsse für Fenster; Verband der Fenster-und Fassadenhersteller
e.V. Gutgemeinschaften Fenster und Haustüren: Frankfurt am Main, Germany, 2016; Volume 2001.

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).