Content uploaded by M. Shahria Alam
Author content
All content in this area was uploaded by M. Shahria Alam on Jul 26, 2020
Content may be subject to copyright.
Reinforced Wood I-Joists with Web Openings
Md Shahnewaz1; M. Shahidul Islam2; Moein Ahmadipour3;
Thomas Tannert, M.ASCE4; and M. Shahria Alam, M.ASCE5
Abstract: Prefabricated wood I-joists with web openings are commonly used in light-frame wood construction. The capacity and the
failure pattern of such I-joists in the presence of a circular web opening were experimentally investigated on 100 specimens with various
sizes and locations of web openings and two different span lengths of 3.66 m (12 ft) and 6.10 m (20 ft). The control I-joists, i.e., I-joists
without any opening, failed in flexure in the midspan, whereas most of the I-joists with an opening failed in a brittle and sudden shear mode.
The presence of an opening reduced the capacity up to 54% compared to I-joists without web openings depending on the size and location
of the opening. Subsequently, to prevent brittle failure and improve capacity, the I-joists were reinforced using two perpendicular OSB
collar layers with a variation in the reinforcement length. The effectiveness of reinforcing the web around openings was investigated on
another 100 specimens. After reinforcement, brittle premature failure of I-joists was prevented with an increase in capacity up to 27% com-
pared to the I-joists with openings. Analytical models to calculate the capacity of unreinforced and reinforced I-joists with openings are
proposed, validated with results from previous research, which prove to be more accurate compared to existing models from the literature.
DOI: 10.1061/(ASCE)ST.1943-541X.0001747.© 2017 American Society of Civil Engineers.
Author keywords: Wood structures; Experimentation; Reinforcement; Regression analysis.
Introduction
Timber as a Construction Material
Timber is a widely used renewable material that exhibits several
favorable characteristics, such as high strength-to-weight ratio,
low carbon footprint, and good insulation properties. Solid timber
as a natural material, however, exhibits very large variability in its
properties and quality and can only be used to produce linear mem-
bers. Historical timber structures were often characterized by ele-
ments that were limited both in their cross-sectional dimensions
and length by the dimensions of the existing trees in the surround-
ing area (Dietsch and Tannert 2015). To meet the engineering and
architectural desires to utilize the sustainable features of timber in
homogeneous and planar elements, timber can be broken down into
sequentially smaller fractions that can then be reassembled into
glued composite members, labeled engineered wood products
(EWPs) (Vallée et al. 2017). Depending on the wood fraction used
as raw material, grain orientations, glue application, and manufac-
turing processes, EWPs can be classified into lumber-based such
as glued laminated timber, veneer-based such as plywood or
laminated-veneer lumber (LVL), strand-based such as oriented-
strand board (OSB) or laminated-strand lumber (LSL), and particle-
based products such as Fibreboard.
Composite I-Joists
EWPs are combined to produce prefabricated composite structural
members such as wooden I-joists, which are popular in light-frame
construction as floor and roof joists because of their high strength
and stiffness, low weight, dimensional stability, and low cost in
comparison to solid timber (American Forest & Paper Association
2001). The flanges and the web are glued together to form an
I-shaped cross section that can save 50% of wood fiber compared
to solid lumber beams (Leichti and Tang 1983;Islam et al. 2011).
Composite I-joists often consist of flanges made of LVL or LSL
with the web made of OSB or plywood, where the flanges and
webs are designed to carry moment and shear forces, respectively,
and the stresses between the flanges and web are transmitted
through the flange-web glue line. Early studies on wood I-joists
(Fergus 1979;Hilson and Rodd 1984;Samson 1983;Leichti
and Tang 1983,1986;Leichti et al. 1990b) focused on determining
the influence of the flange and web materials on the capacity, stiff-
ness, and stability and provided the groundwork for the widespread
structural application of I-joists.
Modern wood I-joists are proprietary products with producers
providing their specific design values after conducting tests accord-
ing to ASTM D5055 (ASTM 2013) and WIJMA (1999). The de-
sign criteria for prefabricated composite I-joists include (1) bending
resistance, governed by the flanges; (2) shear resistance, governed
by the web; (3) deflection limits for live and dead loads; (4) bearing
deformation at supports; (5) span-to-height ratio to prevent web
instability, especially when web holes are present; and (6) bracing
for lateral stability. The manufacturers furthermore list their limi-
tations with respect to concentrated loads and web openings.
1Graduate Research Assistant, Dept. of Civil Engineering, Univ. of
British Columbia, Vancouver, BC, Canada V6T 1Z4. E-mail: md
.shahnewaz@alumni.ubc.ca
2Graduate Research Assistant, School of Engineering, Univ. of British
Columbia, Kelowna, BC, Canada V1V 1V7. E-mail: shahidul.islam@
ubc.ca
3Graduate Research Assistant, School of Engineering, Univ. of British
Columbia, Kelowna, BC, Canada V1V 1V7. E-mail: moein.ahmadipour@
ubc.ca
4Associate Professor, Dept. of Wood Science and Civil Engineering,
Univ. of British Columbia, Vancouver, BC, Canada V6T 1Z4. E-mail:
thomas.tannert@ubc.ca
5Associate Professor, School of Engineering, Univ. of British
Columbia, Kelowna, BC, Canada V1V 1V7 (corresponding author).
E-mail: shahria.alam@ubc.ca
Note. This manuscript was submitted on October 14, 2015; approved on
November 4, 2016; published online on February 13, 2017. Discussion
period open until July 13, 2017; separate discussions must be submitted
for individual papers. This paper is part of the Journal of Structural En-
gineering, © ASCE, ISSN 0733-9445.
© ASCE 04017022-1 J. Struct. Eng.
J. Struct. Eng., -1--1
Downloaded from ascelibrary.org by University Of British Columbia on 02/14/17. Copyright ASCE. For personal use only; all rights reserved.
Openings in I-Joists
During construction, openings are often introduced to the webs of
I-joists for passage of service ducts, plumbing, and wiring (Islam
et al. 2015). Openings allow builders to hide the utility services and
reduce the floor height, but the presence of the web openings leads
to reductions in stiffness and capacity. Such reductions are most
critical in the cases of web openings located close to supports
and may cause I-joists to fail in premature sudden and brittle shear
failure. The current edition of the Canadian Standard for Engineer-
ing Design in Wood [CSAO86 (CSA 2014)] provides no guidance
for such openings in I-joists and the National Design Specification
for Wood Construction in the United States (NDS 2015) recom-
mends manufacturer specifications for I-joists with openings.
Some previous research evaluated the failure modes and capac-
ity reduction of wood I-joists with web openings. Morris et al.
(1995) summarized three failure modes as web fracture, web buck-
ling, and debonding of the web-flange adhesive joint. Fergus
(1979) studied the effect of circular openings on moment-governed
7.3-m-long I-joists and shear-governed 2.4-m-long I-joists and
found no significant change in stiffness with a web removal of
up to 70% of total height. This finding, however, was limited
for the specific location of the web opening in the moment critical
I-joists close to support and close to midspan in the shear critical
I-joists. On the contrary, Maley (1987) and Wang and Cheng (1995)
reported that openings do reduce stiffness and shear capacity. Wang
and Cheng (1995) investigated 2.8–3.6-m-long I-joists with rectan-
gular web openings of 33–100% web height placed at a distance of
0.5–1.0 m from the support and observed that the shear strength
was reduced up to 79% when the opening height was equal to
the height of the web. No significant change occurred for opening
heights of 33% of web height.
Leichti et al. (1990a,b) reviewed previous studies on I-beams
with openings and reported that the restriction on the web opening
position depends on types and sizes of openings. They concluded
that rectangular openings are more restrictive than circular open-
ings (because of the stress concentration at corners) and that a
38-mm opening can be placed anywhere in the web, but larger
openings require specified minimum distances from supports
and flange edges.
Afzal et al. (2006) performed tests on wood I-joists with circular
and square openings. The I-joists were 302 and 406 mm deep and
the opening size was varied up to 100% of web height. While the
opening size-to-web depth and the span-to-depth ratio both affected
the capacity, the type of opening (circular/square) was found to be
insignificant. Zhu et al. (2005) investigated the failure load of wood
I-joists with and without web openings and observed that capacity
decreases linearly with opening size, while the location of the
opening has little effect on the reduction of capacity. They proposed
an empirical formula to calculate the capacity of I-joists with
openings:
Pu¼36.4−25.9d
hwð1Þ
where Pu= capacity of I-joists with opening (kN); d= diameter of
circular opening; and hw= web height. Zhu et al. (2007) developed
additional models to predict I-joist failure considering material
nonlinearity and crack propagation and achieved a better fit with
experimental results. Pirzada et al. (2008) developed another me-
chanics-based method to predict the capacity of wood I-joists with
circular web holes:
Pfailure ¼2ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
x0EGc
2π
rσt2ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ðpkÞπx0
p−ðpkÞπ−1
Papplied ð2Þ
where Pfailure = load causing fracture in web; Papplied = applied load;
p= web hole ratio; k= factor related to I-joist depth with a circular
hole; x0= length parameter related to characteristic material proper-
ties; E= equivalent modulus of the elasticity of the web; and
Gc= critical fracture energy of web.
Guan and Zhu (2004) performed finite-element analyses (FEA)
to predict the behavior of wood I-joists with openings such that the
opening sizes were varied from one-quarter to three-quarters of the
height of the I-joists. They observed that the predicted capacity for
I-joists with circular openings was 20% higher than that of I-joists
with rectangular openings. In their more recent study, Guan and
Zhu (2009) developed anisotropic elastoplastic constitutive models
that are able to identify the location of an initial crack, the growth of
the crack, the stress states, and the capacity of the wooden I-joist.
While previous FEA attempts provided some valuable insight,
further experiments are deemed necessary before predictive FEA
models can reliably be used for the investigation of I-joists with
openings.
Islam et al. (2015) tested nearly 100 I-joist specimens with
flange notches at different locations under four-point bending tests.
The results show a reduction of load-carrying capacity even up to
80% compared to an uncut I-joist. As the notch location moves
away from the support toward midspan, the flexural capacity de-
creases significantly. In most cases, the specimens experienced
combined flexural-shear failure.
Reinforcing of I-Joists
To increase the flexural and shear capacity of timber beams, several
techniques can be used, such as attaching metal, solid timber or
EWP plates, or fiber-reinforced-polymer (FRP) sheets either by
mechanical means or adhesive bonding (Franke et al. 2015).
Previous studies performed on wood I-joists with openings mostly
focused on the change in capacity caused by the openings and only
very few studies investigated reinforcing techniques. Morrissey
et al. (2009) investigated reinforced I-joists with steel angles at-
tached to both sides of the web and the flange above and below
the openings and obtained an increase in capacity up to 39%.
Polocoser et al. (2013) reinforced wood I-joists around the open-
ings with U-shaped LSL and OSB patches and OSB collars. They
performed tests on 356- and 406-mm-deep I-joists with spans
of 2.4 and 4.9 m and openings of up to 63% of the joist depth,
which reduced the capacity up to 58%. After retrofitting, some
of the specimens regained the capacity of the original joists. Among
the three different techniques, the OSB collar was found to be most
effective.
Other studies focused on reinforcing solid timber and glulam
beams using carbon-FRP (Nowak et al. 2013;Borri et al. 2005;
Li et al. 2009) or glass-FRP (Raftery and Harte 2011), but very
few studies, e.g., Hallström (1996), investigated glass-FRP
reinforcement of glulam beams with both circular and rectangular
openings (finding that shear capacity improved up to 100% com-
pared to unreinforced beams). No study on using FRP to reinforce
wooden I-joists has been reported in the literature so far.
Objective
Placing web openings in I-joists is a common practice that can lead
to significant reduction in stiffness and capacity, but that—if not
appropriately considered in design—may cause excessive deflec-
tions and premature brittle failure of the element and possibly
the structure. Practitioners, however, have limited design guidance
to calculate the reduction in capacity and stiffness of I-joists with
web openings. Guidance on how to improve capacity and stiffness
© ASCE 04017022-2 J. Struct. Eng.
J. Struct. Eng., -1--1
Downloaded from ascelibrary.org by University Of British Columbia on 02/14/17. Copyright ASCE. For personal use only; all rights reserved.
using reinforcements is also limited. The project was initiated by
demand from the light-frame timber construction industry in British
Columbia. The objective of this research is to investigate the effect
of I-joist reinforcement around web openings, specifically in terms
of capacity, stiffness, and failure mode.
Experimental Investigation
Materials
All wood I-joist specimens for the experiment were prepared at
the facility of AcuTruss Industries, Kelowna, British Columbia,
Canada. The specifications were chosen from the NASCOR
NJH12 I-joist series (Nascor 2010). Flanges were made of LVL
from CSI-certified spruce-pine-fir (SPR) No. 2 grade, kiln-dried
and heat treated (KD-HT) with a moisture content between 16
and 19%. Webs were made from OSB manufactured by Ainsworth
Engineering following the requirements of the Performance
Standard for Wood-Based Structural-Use Panels [PS2 (Voluntary
Product Standard 2010)] and CSA-O325 (CSA 2012). The OSB
was APA-rated sheathing grade with a span rating of 24/0 and
durability classification of Exposure 1. The height of the specimens
was 302 mm; the flange width and height were 63.5 and 38 mm,
respectively; and the thickness of the web was 9.5 mm. The
material for the subsequent retrofit was OSB from the I-joist
fabrication.
Specimen Description
I-joists with two different span lengths of 3.66 m (12 ft) and 6.10 m
(20 ft) were tested. The design capacities for the 3.66-m (12-ft) and
6.10-m (20-ft) I-joists according to the manufacturer’s guidelines
were 12 and 7.2 kN, respectively (Nascor 2010). They were cat-
egorized into five series (A–E) of specimens. Ten beams from each
series of a total of 100 specimens were tested. Series A was the
control test series without any openings.
The study was limited to circular openings as these are most
commonly encountered in practice. Since a previous study showed
that up to 33% web removal had no effect on I-joist capacity (Wang
and Cheng 1995), larger opening sizes were chosen. The present
study investigated I-joists with an approximate opening diameter
of between 44 and 94% of the web height, thereby covering a wider
variation of the opening diameter size. Finally, a minimum clear
distance of 150 mm was provided between supports and openings,
as the manufacturer recommended avoiding openings over supports
(NASCOR 2010). Series B and C had an opening of diameter (D)
equal to the height of the web (212.7 mm). The distance of the
opening from the edge (Le) in Series B was 305 mm for both
3.66 m (12-ft) and 6.10 m (20-ft) specimens, while in Series C
openings were located at 610 and 915 mm for 3.66-m (12-ft)
and 6.10-m (20-ft) specimens, respectively. In Series D and E,
the diameter of the opening was 152.4 and 101.6 mm, respectively,
and the opening was located 305 mm from the edge. The size and
location of the openings are illustrated in Fig. 1and Table 1.
Reinforcement of I-Joists
I-joists with the same dimensions and openings as Series B–E were
reinforced with OSB collars to investigate the effect of the
reinforcement on capacity and stiffness. The minimum reinforce-
ment length of the OSB collar (Lr) on each side of the opening was
equal to the diameter of the opening (D) based on the hypothesis
that shorter reinforcements do not lead to improvements of the
I-joist stiffness or capacity. Five reinforced series (F–J) of speci-
mens with the same span length of 3.66 m (12 ft) and 6.10 m
(20 ft) were tested after being reinforced around the opening with
an OSB collar. Collars were located on only one side of the web and
consisted of two layers, each layer composed of 9.5-mm (3=8-in:)
OSB. The first layer was arranged around the opening and glued
directly onto the web with its strong layer oriented perpendicular to
the joist span. The second layer was glued on top of the first collar
where the orientation was perpendicular to the first layer to prevent
crack propagation around the opening. In addition, in each layer,
the collar consisted of two pieces as shown in Fig. 2. The adhesive
was a two-component epoxy containing a resin (Cascophen
4001-5) and a catalyst (Cascoset 5830S.5). The reinforcement
length (Lr) of the OSB collar on each side of the opening was kept
equal to the diameter of the opening (Series F–I). Only for Series J,
the collar (reinforcement) length was doubled to evaluate the
capacity improvement because of the OSB collar length. The
details of the reinforced I-joists are given in Fig. 2and Table 1.A
total of 100 reinforced specimens were tested with 10 replicates in
each test series.
Methods
The specimens were tested as simply supported beams in four-point
bending according to ASTM D5055 (ASTM 2013). The loads were
applied by a hydraulic actuator with a loading rate of 4mm=min to
comply with ASTM D5055 (ASTM 2013). A schematic of the test
setup and a photo of an actual test specimen are displayed in Fig. 3.
Hollow structural section (HSS) rectangular tubes were placed ver-
tically on both sides of the flanges along the length of I-joists at a
spacing of 305 mm to ensure concentric loading and to prevent
lateral buckling (as shown in Fig. 3). The joist deflections were
measured by placing an extensometer at midspan. The stiffness
was calculated for the range of 10–40% of capacity according
to EN 26891 (CEN 1991). Three cameras, focused at the midspan,
(a) (b)
Fig. 1. (a) Schematic of reinforced I-joist with lateral support along the length; (b) cross section of I-joist (dimensions are in millimeters)
© ASCE 04017022-3 J. Struct. Eng.
J. Struct. Eng., -1--1
Downloaded from ascelibrary.org by University Of British Columbia on 02/14/17. Copyright ASCE. For personal use only; all rights reserved.
at the location of the opening and at the loading point, were in-
stalled to monitor the crack pattern and failure of the specimens.
Analysis of variance (ANOVA) was performed. ANOVA can be
used for comparing means when there are more than two levels of a
single factor and to determine whether observed differences in
means are statistically significant. A P-value is calculated and com-
pared to α, the significance level, also called probability of Type I
error. If the P-value is smaller than α, then the null hypothesis of no
differences between means is very unlikely and is rejected. Detailed
explanations can be found in textbooks (Montgomery 2008). A
probability of Type I error, α¼0.05, was considered; the response
variables were the capacity and stiffness of I-joists. To evaluate
which individual test series are different from others, Fisher’s least
significant difference (LSD) test was used to identify the groups
that are different (Williams and Abdi 2010).
Results
The load-deflection curves of all specimens were linear up to
failure. Since all individual specimens exhibited similar load-
deformation behavior, only the average load-deflection curves are
plotted in Fig. 4. The average capacities and stiffnesses for all test
series as well as the corresponding coefficients of variation (COVs)
Table 1. Summary of Test Series and Test Results
I-joist
Series
identifier
I-joist length
[L(mm)]
Opening depth
[D(mm)]
Opening location
[Lea(mm)]
Reinforcement
length [Lr(mm)]
Stiffness ðkÞ
Capacity
ðFexpÞ
N=mm COV kN COV
Control 12-A 3,650 —— —1,220 17 40.5 30
20-A 6,100 —— —310 10 28.9 14
I-joists with
opening
12-B 3,650 213 305 —1,000 10 18.3 12
12-C 3,650 213 610 —1,035 10 17.1 10
12-D 3,650 150 305 —1,080 14 27.5 10
12-E 3,650 100 305 —1,165 16 36.5 19
20-B 6,100 213 305 —364 12 20.7 13
20-C 6,100 213 914 —315 14 20.1 13
20-D 6,100 150 305 —345 11 27.9 15
20-E 6,100 100 305 —305 9 26.3 25
Reinforced
I-joist
12-F 3,650 213 305 D + D + D 1,045 13 21.9 13
12-G 3,650 213 610 D + D + D 1,085 8 20.5 10
12-H 3,650 150 305 D + D + D 1,115 9 35.2 14
12-I 3,650 100 305 D + D + D 1,180 14 40.7 24
12-J 3,650 100 305 2D + D + 2D 1,240 15 45.9 25
20-F 6,100 213 305 D + D + D 310 12 21.1 13
20-G 6,100 213 914 D + D + D 310 8 22.2 11
20-H 6,100 150 305 D + D + D 305 9 30.5 17
20-I 6,100 100 305 D + D + D 320 15 29.2 10
20-J 6,100 100 305 2D + D + 2D 315 11 26.7 15
aOpening location was measured from support.
(a)
(c)
(b)
Fig. 2. Reinforcement technique using OSB collar: (a) schematic;
(b) plan; (c) test specimen
Fig. 3. Test setup: (a) schematic; (b) photo
© ASCE 04017022-4 J. Struct. Eng.
J. Struct. Eng., -1--1
Downloaded from ascelibrary.org by University Of British Columbia on 02/14/17. Copyright ASCE. For personal use only; all rights reserved.
are summarized in Table 1. Typical failure patterns in different
I-joists, i.e., control, unreinforced, and reinforced series, are plotted
in Figs. 5and 6.
Series A (Control I-Joist)
Series A represents the control beams without any opening. The
3.66-m (12-ft) I-joists failed in either shear at support or flexure
at midspan. The failure was initiated mostly by the presence of
knots in the flanges or as a result of the debonding of the OSB
webs. Three specimens (12-A3, 12-A4, and 12-A6) failed in shear
as shown in Fig. 5(a), and the rest of the specimens failed in flexure
at midspan as shown in Fig. 5(b). In the case of 6.10-m (20-ft)
I-joists, all the specimens failed in flexure at midspan and there
was no shear failure in any of the 6.10-m (20-ft) specimens.
The average capacity of the 3.66-m (12-ft) and 6.10-m (20-ft) con-
trol test series was 32 and 24 kN, respectively, with COVs of 30 and
25%. The experimental capacity for the 3.66-m (12-ft) and 6.10-m
(20-ft) I-joists was found to be 3.4 and 4.0 times higher, respec-
tively, than the specified design capacity from the manufacturer.
Similarly, the average stiffness of 3.66-m (12-ft) and 6.10-m
(20-ft) control I-joists was 1,220 and 310 N=mm, respectively,
which was 2.4 and 2.8 times higher than the specified design stiff-
ness from the manufacturer [design stiffness of 511 and 110 N=mm
for 3.66-m (12-ft) and 6.10-m (20-ft) I-joists, respectively].
0
10
20
30
40
50
0 1020304050
Load (kN)
Deflection (mm)
12-A
12-B
12-C
12-D
12-E
0
10
20
30
40
50
0 1020304050
Load (kN)
Deflection (mm)
12-F
12-G
12-H
12-I
12-J
0
10
20
30
40
50
0 20406080100
Load (kN)
Deflection (mm)
20-A
20-B
20-C
20-D
20-E
0
10
20
30
40
50
0 20 40 60 80 100
Load (kN)
Deflection (mm)
20-F
20-G
20-H
20-I
20-J
(a) (b)
(c) (d)
Fig. 4. Average load-deflection curves of I-joists: (a) 12-ft unreinforced; (b) 20-ft unreinforced; (c) 12-ft reinforced; (d) 20-ft reinforced
Series B:
Series A:
(a)
(c) (d)
(e) (f) (g)
(b)
Series C:
Series D: Series E:
Fig. 5. Typical failures for I-joists, Series A–E
© ASCE 04017022-5 J. Struct. Eng.
J. Struct. Eng., -1--1
Downloaded from ascelibrary.org by University Of British Columbia on 02/14/17. Copyright ASCE. For personal use only; all rights reserved.
According to ASTM D5055 (ASTM 2013), wood I-joists
require the allowable capacity to be “the lower 5% tolerance limit
with 75% confidence divided by 2.1.”The factor of safety from the
present test results was found to comply with this requirement.
I-Joist Series with Opening (Series B–E)
The presence of an opening changed the failure mode and capacity
of the I-joists. Typical failure patterns from Series B–E are shown
in Figs. 5(c–g). The comparisons among I-joist capacity and stiff-
ness are shown in Figs. 7and 8, respectively.
Series B specimens featured an opening equal to the height of
the web and located 305 mm from the leading edge. All 3.66-m
(12-ft) and 6.10-m (20-ft) Series B I-joists failed in shear at the
opening [Fig. 5(c)], the exceptions being Specimen 20-B9, which
failed in shear right next to the opening, and Specimen 20-B10,
which experienced flexural failure at midspan. In both specimens,
failure was initiated at a knot. Compared to the control series, the
average capacity of 3.66-m (12-ft) Series B I-joists was reduced by
53% and for the 6.10-m (20-ft) I-joists it was reduced by 21%, as
shown in Fig. 7(a). While the average stiffness of the 3.66-m (12-ft)
I-joists in Series B was reduced by 18%, the stiffness of the 6.10-m
(20-ft) I-joists was not affected by the presence of openings (as
presented in Table 1and Fig. 8).
Series C featured the same opening size as Series B, but located
at 610 and 915 mm from the support for 3.66-m (12-ft) and 6.10-m
(20-ft) I-joists, respectively. All specimens of Series C failed in
shear at the location of the opening [Fig. 5(d)]. The average capac-
ity of Series C for the 3.66-m (12-ft) I-joists was reduced by 54%,
and by 20% for the 6.10-m (20-ft) I-joists when compared to Series
A; see Fig. 7. Though the average stiffness of the 3.66-m (12-ft)
I-joists in Series C was reduced by 15%, the stiffness of the
6.10-m (20-ft) I-joists was not affected by the presence of openings
(as shown in Table 1and Fig. 8).
All specimens of Series D (openings 66% the height of the web
and located 305 mm from the support) failed in shear, with failure
starting diagonally at the opening by cracking of OSB followed by
web-flange joint debonding and finally diagonal splitting of the
flange [Fig. 5(e)]. While the average capacity of the 3.66-m
(12-ft) I-joists was 30% lower than Series A, the average capacity
of 6.10-m (20-ft) I-joists was found to be the same. Half of the
3.66-m (12-ft) specimens of Series E (openings about 50% of
the web height, also located at 305 mm from the support) failed in
shear diagonally along the opening, similar to Series D [Fig. 5(f)].
The rest of the 3.66-m (12-ft) span I-joists specimens failed in
flexure at midspan, with the average capacity of 31 kN for
3.66-m (12-ft) I-joists very close to that of Series A. The average
stiffness of the 3.66-m (12-ft) I-joists was reduced by 5%, whereas
the average stiffness of 6.10-m (20-ft) I-joists was close to the
control series. The failure patterns of the 6.10-m (20-ft) I-joists
were similar to the control Series A, with all specimens failing
in flexure close to midspan, and capacity was similar to the control
beams.
Reinforced I-Joist Series (Series F–J)
The opening ineach I-joist of Series F–I were reinforced by attaching
an OSB collar (Lr=D,whereLris the reinforcement length and Dis
the diameter of the opening) around the opening. This collar stopped
shear cracks from propagating, which helped prevent shear failure in
some specimens. While test specimens without any reinforcement
failed abruptly in shear at opening, the OSB collar either prevented
shear failure or prevented shear failure from propagating when it
initiated at other weak locations such as knots or the web-flange
intersection. The majority of specimens still failed in shear
diagonally, which was followed by debonding of web-flange joint
[Figs. 6(a and b)]. The OSB collar debonded at the end of the failure
process. In two Series F specimens (20-F2 and 20-F10), the OSB
collar prevented the shear failure and instead induced a flexural fail-
ure. This is because of the contribution of the OSB collar, which made
the opening location stronger in shear compared to the flexure capac-
ity of the I-joist. In Series G, all 3.66-m (12-ft) I-joists and all but two
6.10-m (20-ft) specimens failed in shear diagonally at opening
[Fig. 6(c)]. The exceptions failed in flexure at midspan. Compared
to unreinforced Series B, the capacity of 12-ft specimens from both
F and G series was found to be 19% higher, as shown Fig. 7.
Most 3.66-m (12-ft) and 6.10-m (20-ft) specimens from Series
H failed in flexure, similar to control I-joists Series A [Fig. 6(e)].
Additionally, compared to unreinforced Series D, their capacity in-
creased by 27 and 13% for 3.66-m (12-ft) and 6.10-m (20-ft)
I-joists, respectively. Likewise, in Series I the reinforcement collar
Series F:
(a) (b)
Series G:
(c) (d)
Series H:
(e) (f)
Series I:
(g) (h)
Series J:
(i) (j)
Fig. 6. Typical failures for reinforced I-joists, Series F–J
© ASCE 04017022-6 J. Struct. Eng.
J. Struct. Eng., -1--1
Downloaded from ascelibrary.org by University Of British Columbia on 02/14/17. Copyright ASCE. For personal use only; all rights reserved.
efficiently prevented failure at the opening and the failure of both
the 3.66-m (12-ft) and 6.10-m (20-ft) I-joists was in flexure.
Compared to unreinforced Series E, average capacity improved
by 4 and 13%, respectively, for 3.66-m (12-ft) and 6.10-m (20-ft)
reinforced I-joists, as shown in Fig. 7.
In Series J, the length of the OSB collar on each side of the
opening was doubled (Lr=D) compared to Series I. Most of
the 3.66-m (12-ft) and all of the 6.10-m (20-ft) reinforced I-joists
failed similarly to the control specimens either in flexure at mid-
span or in shear at the loading point [Figs. 6(i and j)]. The capacity
12A
(Control)
12B,
-53%
12C,
-54%
12D,
-30%
12E,
-3%%
12F,
+19%
12G,
+19%
12H,
+27%
+12I,
4%
12J,
+17%
0
5
10
15
20
25
30
35
40
Nk ,yticapaC
Design
Capacity=
12kN
Negative sign means capacity decrease in % compared to control beams
and positive sign means improvement after retrofit.
Control I-joist Unreinforced I-joist
Reinforced I-joist
20A
(Control) 20B,
-21%
20C,
-20%
20D,
-7% 20E,
-6%
20F,
+4%
20G,
+6%
20H,
+6%
20I,
+20% 20J,
+8%
0
5
10
15
20
25
30
35
40
Capacity, kN
Negative sign means capacity decrease in % compared to control beams and
positive sign means improvement after retrofit.
Design
Capacity=
7.2 kN
Control I-joist Unreinforced I-joist
Reinforced I-joist
(a)
(b)
Fig. 7. Comparison of capacity: (a) 12-ft I-joists; (b) 20-ft I-joists
12A
(Control) 12B,
-14%
12C,
-11%
12D,
-9%
12E,
-3%%
+12F,
+5%
12G,
+5%
12H,
+3%
12I,
+1%
12J,
+6%
0
300
600
900
1200
1500
1800
Design
Stiffness=
511N/mm
Negative sign means capacity decrease in % compared to control beams
and positive sign means improvement after retrofit.
Control I-joist Unreinforced I-joist Reinforced I-joist
20A
(Control)
20B,
+18% 20C,
+2%
20D,
+11% 20E,
-2%%
20F,
-14%
20G,
-2%
20H,
-12%
20I,
+5%
20J,
+3%
0
100
200
300
400
500
600
Stiffness, N/mm Stiffness, N/mm
Design
Stiffness=
110N/mm
Negative sign means capacity decrease in % compared to control beams
and positive sign means improvement after retrofit.
Control I-joist Unreinforced I-joist Reinforced I-joist
(b)
(a)
Fig. 8. Comparison of stiffness: (a) 12-ft I-joists; (b) 20-ft I-joists
© ASCE 04017022-7 J. Struct. Eng.
J. Struct. Eng., -1--1
Downloaded from ascelibrary.org by University Of British Columbia on 02/14/17. Copyright ASCE. For personal use only; all rights reserved.
of the 3.66-m (12-ft) and 6.10-m (20-ft) I-joists increased by 17 and
8%, respectively, compared to unreinforced Series E, as shown
in Fig. 7.
The average stiffness of the reinforced 3.66-m (12-ft) I-joists
was lower than that of the control series, except in Series J, where
it remained approximately the same. Compared to the unreinforced
test series with openings, however, the average stiffness increased
up to 6%. In the 6.10-m (20-ft) I-joist series, the reinforcement did
not alter stiffness, as illustrated in Fig. 8.
Statistical Analyses
Tables 2and 3summarize the ANOVA results from 3.66-m (12-ft)
and 6.10-m (20-ft) I-joists, respectively. The P-value for the re-
sponse variables (capacity and stiffness) were found to be less than
α¼0.05 for both the 3.66-m (12-ft) and 6.10-m (20-ft) test series.
Therefore, the null hypothesis of the averages being equal was
rejected and it was statistically confirmed that web openings and
reinforcements create significant changes in I-joist capacity and
stiffness.
Fisher’s least significant difference (LSD) was calculated to
identify the statistical significance of the observed differences.
Openings in joist Series B–D had a significant mean difference
in capacity and stiffness compared to control Series A.
The LSD value for Series E indicated that an opening with a diam-
eter of half of the web height does not affect the I-joist capacity.
For 6.10-m (20-ft) joists, it was found that the mean strength of
Series B and C and the mean stiffness for Series B and D are
different from the control series. The other opening layouts had
no statistically significant effect on capacity and stiffness.
The capacity of 3.66-m (12-ft) joists of Series H and J had a
significant improvement compared to their unreinforced Series
of D and E, respectively. For the 6.10-m (20-ft) series, however,
only I-joist Series I exhibited an improvement in capacity. While
stiffness of the 3.66-m (12-ft) specimens was not affected by rein-
forcements (Table 2), some 6.10-m (20-ft) reinforced series (F, H,
and J) showed significant improvement in stiffness (Table 3).
The present study was limited to OSB collar reinforcement at
one side of the I-joist only. In practice, such reinforcement might
potentially lead to out-of-plane bending effects. In the tests, such a
failure mechanism was prevented using lateral supports, i.e., HSS
tubes as shown in Fig. 3. However, in a typical light-frame wood
construction, floor I-joists are laterally supported along their length
by the floor. A future study should investigate the effect of out-of-
plane bending on single-sided reinforced I-joists.
Proposed Analytical Model to Estimate I-Joist Capacity
A regression analysis was performed using the current test results to
develop models to predict the capacity of unreinforced and rein-
forced I-joists with web openings. I-joist span length-to-height ratio
(L=h) and opening size-to-web height ratio (D=hw) affect the
capacity of I-joists (Afzal et al. 2006) and were considered in
the regression model. The interactions among the parameters are
also considered in the models. The proposed equations for unrein-
forced and reinforced I-joists with web opening are
Punreinforced ¼64.8−1.5ðL=hÞ−54.3D
hwþ1.9ðL=hÞD
hw
ð3Þ
Preinforced ¼105.6−3.5ðL=hÞ−90.8D
hwþ3.8ðL=hÞD
hw
ð4Þ
where L= I-joist span length (m); h= height of I-joists; D= size of
opening; and hw= height of web.
The capacities of unreinforced and reinforced I-joists with open-
ings from the tests as presented herein, as well as previous tests
performed by Afzal et al. (2006) and Polocoser et al. (2013), were
calculated using various analytical models: Eq. (1) proposed by
Zhu et al. (2005); Eq. (2) proposed by Pirzada et al. (2008);
and the new models proposed in this research, i.e., Eqs. (3) and
(4). Afzal’s tests (Afzal et al. 2006) on wood I-joists with circular
openings were considered for the unreinforced I-joist model com-
parison of Eq. (3), where I-joists were 302 and 406 mm deep and
the opening size varied from 20 to 100% of web height. On the
other hand, Polocoser’s tests (Polocoser et al. 2013) on reinforced
wood I-joists around the openings using OSB collars were used for
the reinforced I-joist model comparison of Eq. (4). The reinforced
Table 2. ANOVA Results for 12-ft I-Joists
Parameter P-value Test series
Mean
difference LSD Comment
Capacity 0.00 <α12A-12B 16.2 4.8 Significant
12A-12C 16.9 Significant
12A-12D 9.2 Significant
12A-12E 0.8 Not significant
12B-12F 2.6 Not significant
12C-12G 2.8 Not significant
12D-12H 6.0 Significant
12E-12I 1.0 Not significant
12E-12J 5.3 Significant
Stiffness 0.00 <α12A-12B 217.2 129.7 Significant
12A-12C 182.4 Significant
12A-12D 136.8 Significant
12A-12E 49.7 Not significant
12B-12F 47.6 Not significant
12C-12G 53.6 Not significant
12D-12H 36.8 Not significant
12E-12I 15.2 Not significant
12E-12J 72.9 Not significant
Table 3. ANOVA Results for 20-ft I-Joists
Parameter P-value Test series
Mean
difference LSD Comment
Strength 0.00 <α20A-20B 4.7 3.2 Significant
20A-30C 5.2 Significant
20A-20D 1.6 Not significant
20A-20E 1.0 Not significant
20B-20F 0.6 Not significant
20C-20G 1.8 Not significant
20D-20H 1.5 Not significant
20E-20I 4.0 Significant
20E-20J 1.5 Not significant
Stiffness 0.00 <α20A-20B 71.7 32.9 Significant
20A-30C 25.6 Not significant
20A-20D 52.3 Significant
20A-20E 12.4 Not significant
20B-20F 54.3 Significant
20C-20G 8.7 Not significant
20D-20H 37.6 Significant
20E-20I 14.5 Not significant
20E-20J 7.9 Not significant
© ASCE 04017022-8 J. Struct. Eng.
J. Struct. Eng., -1--1
Downloaded from ascelibrary.org by University Of British Columbia on 02/14/17. Copyright ASCE. For personal use only; all rights reserved.
I-joists were 356 and 406 mm deep with spans of 2.4 and 4.9 m, and
openings of up to 63% of the joist depth.
The predictions using these three models compared against the
test results are illustrated in Fig. 9and summarized in Tables 4and
5. The percentage errors of the models as compared to test results
are also reported in Tables 4and 5. While Zhu’s model significantly
underpredicts capacity by up to 73%, Pirzada’s model seems
appropriate for small web opening sizes up to web depth ratio
(D=hw<0.5) but also significantly underpredicts for larger open-
ings D=hw>0.5. In the case of reinforced I-joists, it was observed
that both Zhu’s and Pirzada’s models significantly underpredict
capacity by up to 67 and 73%, respectively. One reason is that both
models ignore the effect of the span-to-height ratio of I-joists,
which is a key parameter. In addition, the previous models were
derived from fewer test specimens and were not verified against
other test results. The main reason, however, is that both models
were derived for nonreinforced I-joists with openings.
The accuracies of the proposed model and the previous models
were compared using four descriptive statistical parameters:
(1) average performance factor (PF), the average ratio of experi-
mental capacity to calculated capacity (Fexp =Fcal); (2) χfactor,
the inverse of the slope of a linear least-square regression of the
calculated capacity (Fcal) versus the experimental capacity
(Fexp); (3) standard deviation (SD); and (4) sample variance
(VAR). The analysis results showed that the previous models pro-
posed by Zhu et al. (2005) and Pirzada et al. (2008) are underpre-
dicting the experimental capacity, with all statistical measures
being rather high compared to the new model, as shown in Table 6.
The proposed models [Eqs. (3) and (4)] show good accuracy, with
average PF and χvalues being close to 1.0. Moreover, SD values of
0
10
20
30
40
50
60
70
0 10203040506070
Nk ,yticapaC
d
etc
id
e
rP
Experimental Capacity, kN
Zhu's Model (χ= 1.47)
Pirzada's Model (χ= 1.36)
Proposed Model (χ= 1.0)
0
10
20
30
40
50
60
70
0 10203040506070
Predicted Capacity, kN
Experimental Capacity, kN
Zhu's Model (χ= 2.01)
Pirzada's Model (χ= 1.94)
Proposed Model (χ= 1.02)
(a) (b)
Fig. 9. Predicted versus experimental capacity of different models for (a) I-joists with opening; (b) reinforced I-joists
Table 4. Comparison between Analytical Models for I-Joists with Openings
Series
identifier
I-joist span
length-to-height ratio (L=h)
Opening size to web
height ratio (D=hw)
Capacity
Fexp FzhuaFpirzada bFproposedc
kN kN Δ(%) kN Δ(%) kN Δ(%)
12-B 12.1 0.94 18.3 12.0 34 6.7 63 17.7 3
12-C 12.1 0.94 17.1 12.0 30 6.7 61 17.7 4
12-D 12.1 0.66 27.5 18.9 31 19.5 29 26.4 4
12-E 12.1 0.44 36.5 24.7 32 29.7 19 33.3 9
20-B 20.2 0.94 20.7 12.0 42 6.7 68 20.4 1
20-C 20.2 0.94 20.1 12.0 40 6.7 67 20.4 2
20-D 20.2 0.66 27.9 18.9 32 19.5 30 24.8 11
20-E 20.2 0.44 26.3 24.7 6 29.7 13 28.4 8
She-16-66d4.3 0.2 53.8 31.2 42 40.9 24 49.3 8
She-16-132d4.3 0.4 42.8 26.0 39 31.7 26 40.1 6
She-16-198d4.3 0.6 31.0 20.9 33 22.5 27 30.9 0
She-16-264d4.3 0.8 22.0 15.7 29 13.3 39 21.7 2
She-16-330d4.3 1 15.5 10.5 32 4.2 73 12.5 20
She-16-264-18d8.8 0.8 20.5 15.7 23 13.3 35 22.0 7
She-12-56d5.2 0.25 42.6 29.9 30 38.6 9 46.2 8
She-l2-113d5.2 0.5 30.2 23.5 22 27.1 10 35.1 16
She-12-170d5.2 0.75 21.3 17.0 20 15.6 27 23.9 12
She-l2-226d5.2 1 14.4 10.5 27 4.2 71 12.8 11
She-12-170-12d9.2 0.75 21.3 17.0 20 15.6 27 23.8 12
She-12-170-24d13.2 0.75 21.3 17.0 20 15.6 27 23.7 11
aZhu et al. (2005).
bPirzada et al. (2008).
cProposed model.
dTests as reported by Afzal et al. (2006).
© ASCE 04017022-9 J. Struct. Eng.
J. Struct. Eng., -1--1
Downloaded from ascelibrary.org by University Of British Columbia on 02/14/17. Copyright ASCE. For personal use only; all rights reserved.
the proposed equations (unreinforced and reinforced I-joists)
were only one-tenth and one-sixteenth, respectively, compared to
Pirzada’s equation.
Conclusions
The experimental investigation on the performance of 3.66-m
(12-ft) and 6.10-m (20-ft) composite wood I-joists with web open-
ings (unreinforced and reinforced) allows the following conclu-
sions to be drawn:
1. The experimental capacity and stiffness of the regular I-joists
without openings were at least three times higher than the
specified design values, which satisfies the requirements accord-
ing to ASCE 7-10.1 (ASCE 2010) and ASTM D5055
(ASTM 2013).
2. A stiffness reduction of up to 14% in 3.66-m (12-ft) and 6.10-m
(20-ft) series was observed compared to the control series. There
is no clear trend observed for the stiffness reduction of an I-joist
with an increase in web opening size. The change in the position
of the opening in shear span also had no effect on the joist
stiffness.
3. An opening had more effect on smaller span beams. Most of the
3.66-m (12-ft) test specimens with openings showed premature
shear failures at the location of the opening. The capacity of
3.66-m (12-ft) I-joists was reduced up to 54%, while for
6.10-m (20-ft) I-joists, the capacity was reduced only up to
21%.
4. In the case of 6.10-m (20-ft) I-joists, presence of openings about
half of the web height did not have any effect. The statistical
analyses confirmed these findings.
5. The reinforcement of I-joists using OSB collars was effective to
prevent shear failure close to a web opening and increased the
stiffness of the I-joists up to 6 and 8% for 3.66-m (12-ft) and
6.10-m (20-ft) I-joists, respectively, compared to unreinforced
series.
6. The reinforcement of I-joists with OSB collars significantly
increased the capacity [27 and 20%, respectively, for 3.66-m
(12-ft) and 6.10-m (20-ft) I-joists] compared to unreinforced
series. Furthermore, the capacities of the reinforced I-joists
(Series I and J) were found to be almost equal to the control
series capacity.
7. The newly proposed model to predict I-joist capacity with open-
ings was superior compared to existing models from the litera-
ture, while the new model to predict the capacity of an I-joist
with reinforced openings was also sufficiently accurate in pre-
dicting capacities from previous tests.
8. Future studies should include wood I-joists with multiple web
openings and should examine the effect of OSB collar reinfor-
cement on both sides of the web opening.
Acknowledgments
The research was supported by the Natural Sciences and Engineer-
ing Research Council (NSERC) of Canada through an Engage
Grant and Industrial Postgraduate Scholarship (IPS) in collabora-
tion with AcuTruss Industries of Kelowna, British Columbia,
Table 5. Comparison between Analytical Models for Reinforced I-Joists
Series
identifier
I-joist span
length-to-height ratio (L=h)
Ratio of opening size
to web height (D=hw)
Capacity
Fexp FzhuaFpirzada bFproposedc
kN kN Δ(%) kN Δ(%) kN Δ(%)
12-F 12.1 0.94 21.9 12.0 45 6.8 69 21.7 1
12-G 12.1 0.94 20.5 12.0 42 6.8 67 21.7 6
12-H 12.1 0.66 35.2 19.2 45 19.6 44 33.0 6
12-I 12.1 0.44 40.7 24.9 39 29.7 27 42.0 3
12-J 12.1 0.44 45.9 24.9 46 29.7 35 45.9 0
20-F 20.2 0.94 21.1 12.0 43 6.8 68 22.4 6
20-G 20.2 0.94 22.2 12.0 46 6.8 69 22.4 1
20-H 20.2 0.66 30.5 19.2 37 19.6 36 27.2 11
20-I 20.2 0.44 29.2 24.9 15 29.7 2 31.0 6
20-J 20.2 0.44 26.7 24.9 7 29.7 11 26.7 0
Bd13.7 0.71 38.8 18.0 54 17.5 55 36.0 7
Cd13.7 0.71 39.8 18.0 55 17.5 56 36.0 10
Dd12.0 0.59 33.1 21.0 36 22.8 31 41.3 25
Ed12.0 0.68 40.6 18.7 54 18.8 54 38.6 5
Fd12.0 0.60 55.6 20.8 63 22.3 60 41.3 26
Hd6.9 0.71 40.8 18.0 56 17.5 57 44.9 10
Id6.9 0.71 38.3 18.0 53 17.5 54 44.9 17
Jd6.0 0.59 52.0 21.0 60 22.8 56 50.7 2
Kd6.0 0.60 52.1 20.8 60 22.3 57 50.7 3
Ld6.0 0.60 48.0 20.8 57 22.3 54 50.7 6
aZhu et al. (2005).
bPirzada et al. (2008).
cProposed model.
dTests as reported by Polocoser et al. (2013).
Table 6. Performance of Analytical Models for I-Joists with Openings
I-joist Models Average PF χSD VAR
Opening Zhu et al. (2005) 1.44 1.47 0.17 0.03
Pirzada et al. (2008) 1.85 1.36 0.88 0.77
Proposed 1.01 1.00 0.09 0.01
Reinforced Zhu et al. (2005) 1.94 2.01 0.43 0.18
Pirzada et al. (2008) 2.13 1.94 0.69 0.48
Proposed 1.01 1.02 0.11 0.01
© ASCE 04017022-10 J. Struct. Eng.
J. Struct. Eng., -1--1
Downloaded from ascelibrary.org by University Of British Columbia on 02/14/17. Copyright ASCE. For personal use only; all rights reserved.
Canada. Thanks are also extended to Kader Newaj Siddiquee, for
his priceless help during the experimental tests.
References
Afzal, M. T., Lai, S., Chui, Y. H., and Pirzada, G. (2006). “Experimental
evaluation of wood I-joists with web holes.”Forest Prod. J., 56(10), 26.
American Forest & Paper Association. (2001). “Allowable stress design
(ASD) manual for engineered wood construction.”Washington, DC.
ASCE. (2010). “Minimum design loads for buildings and other structures.”
ASCE 7–10, Reston, VA.
ASTM. (2013). “Standard specification for establishing and monitoring
structural capacities of prefabricated wood I-joists.”ASTM-D5055,
West Conshohocken, PA.
Borri, A., Corradi, M., and Grazini, A. (2005). “A method for flexural
reinforcement of old wood beams with CFRP materials.”Compos.
Eng., 36(2), 143–153.
CEN (European Committee for Standardization). (1991). “Timber struc-
tures: Joints made with mechanical fasteners—General principles for
the determination of strength and deformation characteristics.”EN
26891, Brussels, Belgium.
CSA (Canadian Standards Association). (2012). “Construction sheathing.”
CSA O325-07, Rexdale, ON, Canada.
CSA (Canadian Standards Association). (2014). “Engineering design in
wood.”CSA O86-14, Mississauga, ON, Canada.
Dietsch, P., and Tannert, T. (2015). “Assessing the integrity of glued-
laminated timber elements.”Constr. Build. Mater., 101, 1259–1270.
Fergus, D. A. (1979). “Effect of web voids and stiffeners on structural per-
formance of composite I-beam.”Purdue Univ., West Lafayette, IN.
Franke, S., Franke, B., Harte, A. M. (2015). “Failure modes and reinforce-
ment techniques for timber beams: State of the art.”Constr. Build.
Mater., 97, 2–13.
Guan, Z. W., and Zhu, E. C. (2004). “Non-linear FE modeling of crack
behavior of openings in OSB webbed I-beams.”J. Struct. Eng.,
10.1061/(ASCE)0733-9445(2004)130:10(1562), 1562–1569.
Guan, Z. W., and Zhu, E. C. (2009). “Finite element modelling of aniso-
tropic elasto-plastic timber composite beams with openings.”Eng.
Struct., 31(2), 394–403.
Hallström, S. (1996). “Glass fibre reinforced holes in laminated timber
beams.”Wood Sci. Technol., 30(5), 323–337.
Hilson, B. O., and Rodd, P. D. (1984). “The effect of web holes on the
behaviour and ultimate shearing strength of timber I-beams with
hardboard webs.”Struct. Eng., 62B(4), 69–78.
Islam, A., Nwokoli, S. U., and Debebe, T. (2011). “Bearing capacity of
I-joists.”Master’s thesis, Faculty of Science and Engineering, School
of Engineering, Linnaeus Univ., Vaxjo, Sweden.
Islam, M. S., Shahnewaz, M., and Alam, M. S. (2015). “Structural capacity
of timber I-joist with flange notch: Experimental evaluation.”Constr.
Build. Mater., 79, 290–300.
Leichti, R. J., Falk, R. H., and Laufenberg, T. (1990a). “Prefabricated wood
I-joists: An industry overview.”Forest Prod. J., 40(3), 15–20.
Leichti, R. J., Falk, R. H., and Laufenberg, T. (1990b). “Prefabricated wood
composite I-beams: A literature review.”Wood Fiber Sci., 22(1), 62–79.
Leichti, R. J., and Tang, R. C. (1983). “Analysis of wood composite
I-beams with glued flange-web joints.”Proc., Spring Conf. on
Society for Experimental Stress Analysis, Brookfield Center,
Brookfield, MA, 45–50.
Leichti, R. J., and Tang, R. C. (1986). “Creep analysis of wood composite
I-beams.”Proc., Southeastern Conf. on Theoretical and Applied
Mechanics XIII, Univ. of Southern Carolina, Columbia, SC, 484–490.
Li, Y. F., Xie, Y. M., and Tsai, M. J. (2009). “Enhancement of the flexural
performance of retrofitted wood beams using CFRP composite sheets.”
Constr. Build. Mater., 23(1), 411–422.
Maley, J. D. (1987). “Wood I-joists: A closer look.”Proc., Structures
Congress, Building Structures, ASCE, New York, 221–35.
Montgomery, D. C. (2008). Design and analysis of experiments, 7th Ed.,
Wiley, Toronto.
Morris, V., Gustafsson, P. J., and Serrano, E. (1995). “The shear strength of
light-weight beams with and without a hole: A preliminary study.”
Proc., Wood Mechanics: Workshop on Mechanical Properties of Panel
Product, Building Research Establishment, Watford, U.K.
Morrissey, G. C., Dinehart, D. W., and Dunn, W. G. (2009). “Wood I-joists
with excessive web openings: An experimental and analytical investi-
gation.”J. Struct. Eng.,10.1061/(ASCE)ST.1943-541X.0000013,
655–665.
Nascor. (2010). “Specifier guide NJ, NJH, NJU series I-joists.”Ottawa.
NDS (National Design Specification). (2015). “National design specifica-
tion for wood construction.”American Forest & Paper Association
(AF&PA), American Wood Council, Washington, DC.
Nowak, T. P., Jasienko, J., and Czepizak, D. (2013). “Experimental tests
and numerical analysis of historic bent timber elements reinforced with
CFRP strips.”Constr. Build. Mater., 40, 197–206.
Pirzada, G. B., Chui, Y. H., and Lai, S. (2008). “Predicting strength of
wood I-joist with a circular web hole.”J. Struct. Eng.,10.1061/(ASCE)
0733-9445(2008)134:7(1229), 1229–1234.
Polocoser, T., Miller, T. H., and Gupta, R. (2013). “Evaluation of
remediation techniques for circular holes in the webs of wood
I-joists.”J. Mater. Civ. Eng.,10.1061/(ASCE)MT.1943-5533
.0000737, 1898–1909.
Raftery, G. M., and Harte, A. M. (2011). “Low-grade glued laminated tim-
ber reinforced with FRP plate.”Compos. Eng., 42(4), 724–735.
Samson, M. (1983). “Influence of flange quality on the load-carrying
capacity of composite webbed I-beams in flexure.”Forest Prod. J.,
33(1), 38–42.
Vallée, T., Tannert, T., and Fecht, S. (2017). “Adhesively bonded connec-
tions in the context of timber engineering: A review.”J. Adhesion,
93(4), 257–287.
Voluntary Product Standard. (2010). “Performance standard for wood-
based structural-use panels.”DOC PS2-10, National Institute of
Standards and Technology, Washington, DC.
Wang, A., and Cheng, J. J. R. (1995). “Shear behaviour of OSB wood
composite I-beams with web opening.”Canadian Forest Service, Dept.
of Civil Engineering, Edmonton, AB, Canada.
WIJMA (Wood I-Joist Manufacturers Association). (1999). “Establishing
shear capacities for prefabricated wood I-joists with hole.”American
Forest & Paper Association, Washington, DC.
Williams, L. J., and Abdi, H. (2010). “Fisher’s least significant difference
(LSD) test.”Encyclopedia of research design, Sage Publications, Thou-
sand Oaks, CA, 1–5.
Zhu, E. C., Guan, Z. W., Pope, D. J., and Rodd, P. D. (2007). “Effects of
openings on oriented strand board webbed wood I-joists.”J. Struct.
Eng.,10.1061/(ASCE)0733-9445(2007)133:1(145), 145–149.
Zhu, E. C., Guan, Z. W., Rodd, P. D., and Pope, D. J. (2005). “Finite
element modelling of OSB webbed timber I-beams with interactions
between openings.”Adv. Eng. Software, 36(11-12), 797–805.
© ASCE 04017022-11 J. Struct. Eng.
J. Struct. Eng., -1--1
Downloaded from ascelibrary.org by University Of British Columbia on 02/14/17. Copyright ASCE. For personal use only; all rights reserved.
