Conference PaperPDF Available

Flexural Bending Behaviour of Built-up Glulam Box-section Beams

Authors:

Abstract and Figures

The rising popularity of engineered-wood products, such as glued-laminated timber (glulam), as an alternative to traditional sawn lumber encourages to fabricate glulam built-up sections that can expand the horizon of the use of this sustainable material in the construction of mid-and high-rise timber buildings. As a pilot investigation into the subject, five full-size built-up glulam box-section beam assemblies were experimentally examined under four-point flexural bending. Self-tapping screws were used in different patterns to form three beam assembly configurations. Each beam built-up section was made of four glulam panels, each of 44-mm thickness except the bottom flange panel that had 86-mm thickness. Experimental testing showed that reducing the spacing from 800 mm to 200 mm of the screws connecting the built-up section's top and bottom flange panels to the web panels increased the beam flexural bending strength by about 45%. While reducing the spacing from 200 mm to 100 mm only for the screws connecting the bottom flange to the web panels over a distance equal to one-third beam span length from each support, where the maximum shear stresses existed, increased the beam flexural bending strength by an additional 10%. .
Content may be subject to copyright.
FLEXURAL BENDING BEHAVIOUR OF BUILT-UP GLULAM BOX-
SECTION BEAMS
Nishant Verma
1
, Osama Salem
2
ABSTRACT: The rising popularity of engineered-wood products, such as glued-laminated timber (glulam), as an
alternative to traditional sawn lumber encourages to fabricate glulam built-up sections that can expand the horizon of
the use of this sustainable material in the construction of mid- and high-rise timber buildings. As a pilot investigation
into the subject, five full-size built-up glulam box-section beam assemblies were experimentally examined under four-
point flexural bending. Self-tapping screws were used in different patterns to form three beam assembly configurations.
Each beam built-up section was made of four glulam panels, each of 44-mm thickness except the bottom flange panel
that had 86-mm thickness. Experimental testing showed that reducing the spacing from 800 mm to 200 mm of the
screws connecting the built-up section’s top and bottom flange panels to the web panels increased the beam flexural
bending strength by about 45%. While reducing the spacing from 200 mm to 100 mm only for the screws connecting
the bottom flange to the web panels over a distance equal to one-third beam span length from each support, where the
maximum shear stresses existed, increased the beam flexural bending strength by an additional 10%.
.
KEYWORDS:
Glue
d
-
laminated timber
;
uilt
-
up
box
section;
Self
-
tapping
s
crews
;
Screw spacing;
Flexural bending
.
1 INTRODUCTION 123
Timber as a well-used construction material has several
advantages over other alternatives such as ease of
fabrication, lower cost and being environmental friendly
material. With the increasing trend of sustainable
building construction, timber has been gaining
considerable attention as a “green solution” for
construction. Commercial sawn lumber even though can
be easily attainable, it has size limitations which can
restrict designers when larger sections are required.
Thus, the development of engineered-wood products like
glued-laminated timber (glulam) and cross-laminated
timber (CLT) helped in encountering this problem. The
intensifying demand of engineered-wood products,
especially Glulam, compels the inclusion of design
methods for structural elements made of such relatively
high strength-to-weight ratio material in wood design
manuals. Unfortunately, the design of built-up section
beams made of engineered wood has not been
incorporated yet in most design manuals around the
world, including the Canadian Design Manual [1]. Thus,
this lack of design guidelines brings forth the demand of
developing acceptable methods to analyse and design
such built-up sections. Some research work has shown
that a built-up timber box-section beam can fail
1 Nishant Verma, M.Sc. Candidate, Dept. of Civil Eng.,
Lakehead University, Canada, nverma@lakeheadu.ca
2 Sam Salem, Ph.D., P.Eng., Associate Professor, Dept. of Civil
Eng., Lakehead University, Canada, sam.salem@lakeheadu.ca
prematurely at the bond between the section’s web and
flange panels [2]. Thus, to strengthen this bond, some
researchers used nails at small spacings to make the
beam’s built-up section behaves more in composite
action [3]. However, nails have low resistance to shear,
and thus they deform rapidly causing a considerable
decrease in the beam flexural bending strength.
Therefore, fasteners with high shear and withdrawal
strengths, such as self-tapping screws, are more practical
to use in order to enhance the composite action of timber
beams with built-up sections. A few European design
standards provide guidelines on the minimum and
maximum spacing between screws to be used in built-up
sections of structural timber members [4, 5]. A very few
researchers have explored the technique of utilizing self-
tapping screws to fabricate and strengthen timber beams
made of built-up sections [6]. Accordingly, there is lack
of good understanding of the flexural bending behaviour
of built-up section timber beams fabricated using screws
so that, they can be sufficiently implemented in building
construction. As a pilot investigation into the subject,
five full-size built-up glulam box-section beam
assemblies were fabricated using self-tapping screws,
and then experimentally examined under four-point
flexural bending till failure.
2 EXPERIMENTAL TESTING
PROGRAM
Through the preliminary analytical calculations for the
three screwed built-up glulam box-section beam
assembly configurations, summarised in Table 1, it was
observed that the flexural bending strength of the built-
up beam assembly increases by reducing the spacing of
screws connecting the beam’s bottom flange to the web
panels over a distance of one-third beam span length
from the supports. As the interface between the bottom
flange and the web panels lies in the proximity of the
cross section neutral axis, this region exerts excessive
shear stresses, thus causing the beam to exhibit lower
design load capacity. Therefore, the test assemblies were
loaded till failure to observe the failure modes and
ultimate load carrying capacities of each different
assembly. The beams’ flexural bending strengths were
determined by calculating the shear flow between the
beam’s bottom flange and the web panels in comparison
to the shear resistance capacity of the utilized screws [7],
as per Equations 1 and 2, respectively.
Shear Flow, q = (1)
Spacing of Screws =
(2)
Table 1: Summary of design load capacity of screwed built-up
beam assemblies
Assembly
No. Side Screw spacing
(mm) Design
Load
(kN)
Pilot Top 800 8.0
Bottom 800 5.7
Assembly 1 Top 200 32.0
Bottom 200 23.0
Assembly 2 Top 200 32.0
Bottom 100 (200 in the
middle 1/3 length) 46.0
2.1 MATERIALS
The five full-size glulam box-section beam assemblies
were built up using long self-tapping screws at various
spacings.
2.1.1 Glulam Panels
All glulam panels used to build the 222 mm X 327 mm
box-section beams were made of black spruce pine fir,
with stress grade of 24F-1.9E. The lamina used in those
glulam sections was of 38 mm X 50 mm cross-sectional
dimensions, which were finger-jointed and glued
together in horizontal layers. Outer laminas were sanded
to the designed width and depth of each panel. The
mechanical properties of the glulam panels in the
longitudinal direction, as provided by the glulam
manufacturer, are listed in Table 2 [8].
Table 2: Mechanical properties of glulam panels
Property Unit (MPa)
Comp. parallel to grain 33.0
Comp. perp. to grain 7.0
Tension parallel to grain 20.4
Modulus of Elasticity 13100
Density 560 (kg/m3)
2.1.2 Self-tapping Screws
Rugged structural self-tapping screws of lengths 100 mm
and 150 mm (4.0 and 6.0 inches) were used to connect
the top and bottom flanges to the web panels,
respectively. The screws were made of specially
hardened steel to provide higher torque, tensile and shear
strengths. In comparison to traditional screws, the
employed self-tapping screws had a special thread that
helped in enlarging the screw hole to allow easy
penetration in wood without the need for pre-drilling
holes, which in turn also increases their withdrawal
strength. Hence, adequately designed connection using
self-tapping screws is much stronger than conventional
screws. The mechanical properties of the utilized type of
self-tapping screws, as provided by the screw
manufacturer, are provided in Table 3 [9].
Table 3: Mechanical properties of structural screws
Property Unit (MPa)
Bending Strength 1175.0
Tensile Strength 1298.0
Shear Strength 881.0
2.2 FABRICATION AND TEST ASSEMBLIES
All full-size built-up glulam box-section beams were
fabricated by joining all four glulam panels together
using 8-mm diameter self-tapping screws. The pilot
built-up beam assembly had the largest centre-to-centre
screw spacing of 800 mm, while beam Assembly no. 1
used in Tests 1(A and B) had screws spaced at 200 mm
centre-to-centre connecting the top and bottom flanges to
the web panels. However, Assembly no. 2 used in Tests
2(A and B) had screws placed at 200 mm connecting the
top and bottom flange to the web panels, except over a
distance of one-third beam length from each support
where screws connecting the bottom flange to the web
panels were spaced at only 100 mm. Even though the
screws were equipped with what so-called Zip-Tip,
which allowed easy drawing of the screw without pre-
drilling, pilot holes were drilled so that the screws
followed a straight path through the glulam flange to the
web panels when inserted by an impact wrench. The
screw end distance is defined as the distance between the
edge of the beam to the centre of the first screw, which
in this case was kept at 50 mm as specified in (ETA
12/0062, 2012. The complete test matrix is shown in
Table 4 below.
Table 4: Test matrix for the five built-up glulam beam
experiments
Test No. Top spacing
(mm)
Bottom Spacing
(mm)
Pilot Test 800 800
Test 1A 200 200
Test 1B 200 200
Test 2A 200 100
Test 2B 200 100
2.3 TEST SETUP AND PROCEDURE
Each beam test assembly was simply supported over two
supports, 3000 mm apart, that were restrained to strong
steel bottom beam placed within a large universal testing
machine. Test assemblies were linearly loaded to failure
in order to assess their ultimate flexural bending
strengths as well as their different failure modes. One
draw-wire displacement transducer and six Linear
Variable Differential Transformers (LVDTs) were
attached to each test assembly to monitor the vertical
deflections at the beam mid span and near the supports,
as well as to detect the relative slip between the flange
and web panels. A schematic of a general test setup with
all displacement sensors layout is shown in Figures 1 (a)
and (b).
(a) Elevation of general Test setup
(b) Displacement sensors layout
Figure 1: A general built-up beam assembly test setup with
instrument layout
3 EXPERIMENTAL RESULTS
The measurements of LVDTs labelled T1 and T2 that
were installed near the left- and right-side supports,
respectively, were used to calculate the gradually-
increased beam end rotations. While LVDTs labelled T3
through T6 were used to measure the relative slips
between the beam’s flange and web panels. In this study,
the beam’s mid-span maximum vertical displacements,
measured using displacement transducer labelled T7,
were compared between the different test assemblies.
3.1 BEAM MID-SPAN DEFLECTIONS
The effect of varying screw spacings on the beam
assemblies’ mid-span deflections was observed and
studied in the pilot test, Tests 1(A and B), and Tests 2(A
and B), as shown in Figure 2. Analysis of the
measurements provided by the draw-wire displacement
transducer labelled T7 revealed that the beam
experienced a maximum deflection of 96.0 mm at only
64.0 kN load in the pilot test with 800 mm screw
spacings. Whereas for beam Assembly no. 1 in Tests
1(A and B), it took an average load of 96.0 kN for the
beam assembly to experience a maximum deflection of
100 mm and ultimately fail. Thus, increasing the beam
flexural bending strength by about 45%. While reducing
the spacing from 200 mm to 100 mm only for the screws
connecting the bottom flange panel to the web panels
over a distance equal to one-third beam span length from
each support, where additional shear stresses existed,
increased the beam flexural bending strength by an
additional 10%. Increasing the ultimate load capacity of
the built-up beam assembly to an average of 100.0 kN.
0
20
40
60
80
100
120
140
0 20 40 60 80 100 120
Total Load (kN)
Mid-Span Deflection (mm)
P-Test Test 1A
Test 1B Test 2A
Test 2B
Figure 2: Load vs. mid-span deflection of built-up beam
assemblies
Figure 3 shows a general built-up glulam box-section
beam assembly undergoing flexural bending during
testing.
Figure 3: A general built-up beam assembly undergoing
flexural bending
3.2 BEAM END ROTATIONS
The rotations at the beam ends were found to be in good
agreement with each other. Thus, the results of only one
side are presented in Figure 4 that illustrates the effect of
increasing the applied loads on the beam end rotations.
Like the beam mid-span deflection measurements, it was
observed that the beam end rotations increased in
proportion to the applied loads. The weakest beam test
assembly, which was the pilot test, experienced a
maximum rotation of 0.1 radians when experienced a
load of only 62.0 kN. Whereas, the strongest beam test
assemblies in Tests 2(A and B) sustained a maximum
load of about 98.0 kN right before reaching a 0.1 radians
rotation.
This shows that by reducing the screw spacing only at
the bottom face of the beam near supports considerably
assisted in increasing the flexural bending strength of the
built-up beam assembly, allowing the beam to resist
greater applied loads before reaching the failure
criterion, which was set at 0.1 radians in this case.
Figure 4: Load vs. beam end rotation
3.3 BEAM RELATIVE SLIPS
The connection between the top and bottom flanges and
the web panels of the built-up beam section plays a vital
role in the beam overall performance. Connection at the
flange-web interface must be able to resist the applied
load and capable of transferring the shear stresses from
one to another. Figures 5 and 6, illustrate the effect of
varying screw spacing on the relative slips between the
top and bottom flange and the web panels, respectively.
It was observed that like other characteristics, such as
beam mid-span deflections and end rotations, the relative
slips between the top and bottom flange and the web
panels were also impacted by the built-up beam
assembly configuration. From Figure 5, Test 1A
assembly showed a relative slip of 5.5 mm between the
top flange and the web panel after reaching a maximum
applied load of 90.0 kN; whereas, Test 2A assembly
experienced the same slip magnitude at a greater load of
96.0 kN.
After comparing the load-relative slip relationships of
Figure 5 with the corresponding relationships of Figure
6, it was observed that the relative slips at the interface
between the bottom flange and the web panels were
considerably greater than those occurred at the interface
between the top flange and the web panels. These
increased relative slips at the bottom interface can be
interpreted to the excessive shear stresses in the
proximity of the thicker bottom flange panel. This also
validates the results of the analytical calculations which
depicted that since the bottom interface is in the
proximity of the neutral axis, it will encounter more
substantial shear flow.
Figure 5: Load vs. relative slip between the top flange and web
panels of the built-up beam sections
Figure 6: Load vs. relative slip between the bottom flange and
web panels of the built-up beam sections
3.4 FAILURE MODES
Brittle failure modes, such as rolling shear and splitting,
were observed in the built-up section beam assemblies.
Through experimental testing, all failure symptoms such
as cracks and splits were marked in the order of their
occurrence until the ultimate failure occurred, and then
the test was terminated. As shown in Figure 7, the failure
caused by rolling shear occurred in the beam web panel.
Figure 7: Row shear failure in the beam’s web panel
Figure 8 shows the relative slip occurred between the top
and bottom flanges and the web panels due to the
excessive shear stresses that also resulted in yielding of
the screws. This yielding was prominent in the screws
closer to the beam ends near the supports, and as we
moved inwards towards the middle of the beam less
yielding was noticed, as illustrated in Figure 9.
Figure 8: Relative slips between the top and bottom flanges
and the web panels of a general built-up beam assembly
Figure 9: Yielding in the screws connected the bottom flange
and the web panels due to excessive shear stresses
3.5 SUMMARY OF EXPERIMENTAL RESULTS
Tables 5 and 6 summarise the most important
experimental outcomes of testing the five different built-
up glulam box-section beam assemblies.
Table 5: Maximum load and deflection values
Test No. Max. Load
(kN)
Max. Deflection
(mm)
Pilot Test 64.6 94.0
Test 1A 75.0 95.0
Test 1B 91.0 101.7
Test 2A 103.0 98.8
Test 2B 98.0 75.7
Table 6: Maximum rotation and relative slip values
Test No. Max.
Load
(kN)
Max.
Rotation
(radians)
Max. Slip (mm)
Up Down
Pilot Test 64.6 0.11 5.9 9.9
Test 1A 75.0 0.09 4.2 5.5
Test 1B 91.0 0.10 5.5 11.0
Test 2A 103.0 0.12 7.3 7.8
Test 2B 98.0 0.08 7.9 9.1
4 CONCLUSIONS AND
RECOMMENDATIONS
Based on the analysis of the experimental results of the
pilot investigation on the structural behaviour of five
full-size built-up glulam box-section beam assemblies
presented in this paper, a few deductions have been
made and are listed as follows:
Reducing the spacing from 200 mm to 100 mm of
the screws connecting the bottom flange to the
web panels increased the flexural bending
strength of the built-up beam assembly by about
10%;
The experimentally examined test assemblies
experienced varying flange-to-web relative slips
that caused various levels of screws’ yielding.
However, the pilot test assembly with the largest
screw spacing experienced the most relative slip
values at much lower load;
In all test assemblies, the web panels experienced
higher volume of cracks because of the high
rolling shear stresses developed in the proximity
of the interface between the bottom flange and the
web panels. This failure mode was followed by
the brittle failure developed in the bottom flange
when the beam’s ultimate load capacity was
attained;
Excessive shear stresses caused greater yielding
of the screws connecting the bottom flange to the
web panels compared to the yielding of the
counterpart screws at the top interface;
Although it was observed that decreasing the
spacing of the screws connecting the bottom
flange to the web panels increased the flexural
bending strength of the beam assembly, it was
also noticed that since the top flanges of the beam
assemblies encountered almost no damage; thus
the screw spacings at the top can be increased
which will help in further decreasing the cost of
production of such built-up glulam beam
assemblies;
Built-up glulam box-section beam assemblies
such as the ones experimentally examined in
this study will enable designers to utilize
lightweight yet strong glulam structural beams.
This will also open new opportunities in the
field of pre-fabricated construction and promote
the construction of mid- and high-rise timber
buildings in Canada.
ACKNOWLEDGEMENT
This research project was funded using NSERC-
Discovery Grant held by the second author, as well as in-
kind contribution by Nordic Structures Inc. The authors
would like to thank research assistant C. Hubbard for his
assistance in conducting the experiments of this research
project. Thanks, should also be extended to C. Hagstrom
and R. Timmon for their assistance in the Civil
Engineering’s Structures Laboratory at Lakehead
University.
REFERENCES
[1]. Canadian Wood Council: Wood Design Manual.
Ottawa, ON, Canada, 2015.
[2]. Hoger, C., Suntharavadivel, T.G., Duan, K. Failure
and analysis of composite timber beams with box
section, 2013.
[3]. Milner, H.R., Tan, H.H. Modelling deformation in
nailed, thin-webbed timber box beams. Computers
and Structures, 79(29-30): 2541-2546, 2001.
[4]. SFS intec AG. 2012. SFS self-tapping screws WR,
European Technical Approval. ETA-12/0062. SFS
intec AG, Schweiz, Germany.
[5]. Adolf Würth GmbH & Co. KG. 2013. Self-tapping
screws for use in timber constructions, European
Technical Approval. ETA-11/0190. Adolf Würth
GmbH & Co. KG, Deutschland.
[6]. Abukari, M.H. The performance of structural screws
in Canadian glulam. Master Disseritation, Dept. Of
Civil Engineering, McGill University Libraries,
2012.
[7]. Hibbeler, R.C. Statics and mechanics of materials.
Pearson Higher Education, 2017.
[8]. Nordic Engineered Wood. Nordic Lam. CCMCE
Evaluation Report. 13216-R. Nordic Structures Inc.,
2018, Quebec, Canada.
[9]. GRK Fasteners, A Division of Illinois Tool Works
Inc. Rugged Structural Screws. ICC-ES Evaluation
Report. ESR-2442. Bartlett. Illinois, 2017.
ResearchGate has not been able to resolve any citations for this publication.
Article
Full-text available
The goal of this project, which was part of the DARPA Balanced Technology Initiative on Computational Mechanics, was to develop a comprehensive approach to the numerical modeling of the mechanical behavior of materials. Particular areas for focus in the project were the inelastic deformation of highly anisotropic materials such as single crystals and textured polycrystals, as well as evolving microstructural damage in ceramics and ductile metals in both slow and high rate deformation processes. While the contract was awarded for a period of four years, it was actually funded only for the first year plus a few months into the second year. Nonetheless, significant progress can be reported as a direct result of this project.
Article
Thin-web box beams made by nailing plywood or other webs to solid timber flanges are an economic form of timber construction. Their analysis presents problems because of the significant shear and nail distortions that increase deflections. This paper demonstrates how to model such members for any load distribution in any type of structure (portal frames, continuous beams) using commercially available structural analysis software. The method is approximate but slightly conservative.
Failure and analysis of composite timber beams with box section
  • C Hoger
  • T G Suntharavadivel
  • K Duan
Hoger, C., Suntharavadivel, T.G., Duan, K. Failure and analysis of composite timber beams with box section, 2013.
SFS self-tapping screws WR, European Technical Approval
  • A G Sfs
SFS intec AG. 2012. SFS self-tapping screws WR, European Technical Approval. ETA-12/0062. SFS intec AG, Schweiz, Germany.
Self-tapping screws for use in timber constructions
  • Adolf Würth
  • Co
  • Kg
Adolf Würth GmbH & Co. KG. 2013. Self-tapping screws for use in timber constructions, European Technical Approval. ETA-11/0190. Adolf Würth GmbH & Co. KG, Deutschland.
The performance of structural screws in Canadian glulam
  • M H Abukari
Abukari, M.H. The performance of structural screws in Canadian glulam. Master Disseritation, Dept. Of Civil Engineering, McGill University Libraries, 2012.