You are on page 1of 7

Can Wood Become an Alternative Material for Tall

Building Construction?
Jamshid Mohammadi, M.ASCE1; and Lei Ling2
Downloaded from ascelibrary.org by University of California, San Diego on 06/30/17. Copyright ASCE. For personal use only; all rights reserved.

Abstract: With a growing interest in using timber as a structural material for tall buildings in response to trends in sustainable design, several
timber buildings between 7 and 10 stories have been built in Europe and Australia. However, in the United States, timber buildings are still
limited to 6 or fewer stories. The availability of glued laminated timber enables the fabrication of structural members with cross-section sizes
and shapes capable of supporting heavy loads. Although this may be appealing for use in tall buildings, from a structural engineering point of
view, there are also other timber properties that may be attractive to architectural designers as well. Among these properties, timber has favor-
able performance in fire (at least for solid sections). These properties may offer some advantages for wood to become a candidate for tall build-
ing construction as an alternative material. This paper presents an overview of available literature on the use of timber as a potential material
for use in tall buildings. Specific properties of wood in offering a renewable material with a favorable strength-to-density ratio and fire-
resistance rating are discussed. Recent and ongoing fire tests and the significance of adhesive effects in laminated timber members are
reviewed. Fire safety and structural design challenges in making wood an alternative material in tall building construction are also pre-
sented. DOI: 10.1061/(ASCE)SC.1943-5576.0000334. © 2017 American Society of Civil Engineers.

Introduction similar fire performance to heavy timber. The fire endurance capa-
bility of laminated and sawn timber, when used in design properly,
With the growing interest in green materials for construction and is believed to be the same because of the formation of a charring
reduction of carbon footprints in urban development, building mate- layer on the surface of the timber member that protects a structural
rials are becoming a major candidate for sustainability and zero- member for an acceptable length of time (Ling 2014). Solid wood
waste-usage practices (following factories and automobiles). can only be sawn into column or beam shapes; therefore, the cross-
Among various construction materials, timber has the potential for section dimensions are limited by the maximum available size of
use in green design and in meeting the objective of zero-waste- timber in nature. With the development of glued laminated prod-
usage practices. In building construction, timber has been primarily ucts, structural members can be fabricated in the form of panels
used in low-rise buildings and single family dwellings. However, with a variety of cross-section sizes and shapes, making them more
because of a favorable strength-to-density ratio, timber has attracted versatile for use as main structural components for tall buildings for
the attention of design engineers and architects as a potential mate- which there is a demand for large members.
rial that can also be considered in high-rise building applications. A As well-understood, tall buildings require relatively stringent
favorable strength-to-density ratio, versatility, and flexibility, as requirements for structural and architectural design. Specifically, in
needed in architectural design, may make timber a compatible ma- terms of fire endurance and emergency evacuation, tall buildings
terial with concrete and steel in certain tall building applications, demand more rigorous performance than low-rise or mid-rise build-
especially because the procedures for manufacturing concrete and ings. To accommodate a growing interest in using timber as a mate-
steel generate large amounts of greenhouse gases that leaves a rial of choice in high-rise buildings, adequate information on timber
much larger carbon footprint than does timber. performance in meeting structural and fire protection design
Timber has another architectural property: resistance to fire. requirements will need to be compiled. Furthermore, it is important
Contrary to popular belief, timber may actually perform relatively to determine whether there are any gaps in knowledge about fire-
well in resisting fire and providing a fire endurance rating that is de- endurance properties of timber and to identify what needs to be
sirable in architectural design. Used in some low-rise building done to fill any gap(s).
applications in the past, heavy timber has already demonstrated
favorable fire-resistance capabilities. Laminated wood members,
with capabilities for use in fabricating large members, have the Overview of Timber Used in Tall Buildings
same integrity as that of sawn wood. They are perceived to have
Definitions
1
Professor, Civil, Architectural and Environmental Engineering, Structural wood refers to the type used as the main material in the
Illinois Institute of Technology, Chicago, IL 60616 (corresponding framing of a structure. Depending on the type of application, vari-
author). E-mail: mohammadi@iit.edu ous terms are used within the industry to refer to the type and/or
2
Graduate Student, Computer Science, Illinois Institute of Technology, shape of the structural member used. In this section, an overview of
Chicago, IL 60616.
these terms is presented to specifically distinguish differences
Note. This manuscript was submitted on December 20, 2016; approved
on March 22, 2017; published online on June 14, 2017. Discussion period between what is termed as heavy timber construction and dimension
open until November 14, 2017; separate discussions must be submitted for lumber. Although at times the two terms are used interchangeably,
individual papers. This paper is part of the Practice Periodical on when it comes to structural design there are distinct differences in
Structural Design and Construction, © ASCE, ISSN 1084-0680. terms of (1) the way these types of lumber are made, (2) cross-

© ASCE 04017014-1 Pract. Period. Struct. Des. Constr.

Pract. Period. Struct. Des. Constr., 2017, 22(4): 04017014


sectional sizes, and (3) physical properties, especially fire endur- the same grain and glued together; the lamination and members
ance. Dimension lumber refers to structural members with standard joined together will provide the Glulam member with adequate
cross-sectional sizes of pieces of wood that are cut and roughly strength that is one directional. Hence, Glulam members are mainly
sawn to specific lengths, such as, 2.4, 3.0, or 3.8 m (8, 10, or 12 ft). used as joists and beams. Furthermore, the one-directional strength
For dimension lumber, the material thickness measures between of Glulam is very different from that of a CLT member (panel) in
51 mm (2 in.) and 102 mm (4 in.). For joists and beams, the nominal which the strength is two directional (Mohammad et al. 2012).
cross-sectional width is usually 51 mm (2 in.) with the actual width
38 mm (1.5 in.). The nominal cross-section depth runs at 152, What Constitutes a Tall Building in
203, and 254 mm (6, 8, and 10 in.) with the actual cross-section Timber Construction?
depth approximately19 mm (0.75 in.) less than the nominal
value. Structural members with cross-section sizes of 127 mm In wood construction, the term tall building has been used in the lit-
(5 in.)  127 mm (5 in.) or larger are classified as timber. Thus, erature in reference to buildings with 8 or more stories in height.
Downloaded from ascelibrary.org by University of California, San Diego on 06/30/17. Copyright ASCE. For personal use only; all rights reserved.

for posts and columns, a member with a nominal cross-section However, with more development in technologies in fabricating
of 102 mm (4 in.)  102 mm (4 in.), an actual cross-section of large built-up sections, it is conceivable that in the future more tim-
89 mm (3.5 in.)  89 mm (3.5 in.), is under the dimension lum- ber buildings with more stories will be built. Therefore, the 8-story
ber category. However, a column with a nominal size of 152 mm limit, as a designation for tall timber buildings, will probably be
(6 in.)  52 mm (6 in.), an actual cross-section size of 140 mm changed. On the basis of our review of literature, Table 1 presents
(5.5 in.)  140 mm (5.5 in.), is under the timber classification. the definitions for low-, mid-, and high-rise (or tall) buildings in
When heavy loads require using large cross sections, two or timber construction.
more of these members will need to be used together. Heavy
timber construction also refers to large structural members— Examples of Current Tall Buildings in Timber
with cross-section dimensions of 127 mm (5 in.)  127 mm (5 in.) or
Built in London in 2009, the first tall timber structure is perhaps
larger—roughly sawn as one piece. Examples of this type of struc-
the 9-story Stadthaus Building. Since it was erected, tall-timber-
tural members are typically found in older buildings. One-piece col-
building ideas began to grow. The LCT One Building, which is 8
umn sizes as large as 394 mm (15.5 in.)  394 mm (15.5 in.) in cross
stories tall, and the Forte Building, which is 10 stories tall (both
sections made of heavy timber are not uncommon. In terms of fire
constructed in Australia), are other examples of tall wood struc-
durability, as described later in this paper under the “Fire-Resistance
tures. The Stadthaus Building was designed by architect Waugh
of Timber Members,” the charring rate commonly used in design
Thistleton and the structural engineering firm of Techniker.
standards is the same for all types of structural wood, whether heavy
Using CLT panels, a cellular structure was formed to resist grav-
timber or dimension wood. However, massive heavy timber mem-
ity and lateral loads. With practically no columns used, a rela-
bers may offer a more favorable charring rate because of their large
tively clear opening space for the occupants was accommodated.
size and the formation of a charred outer layer. Furthermore, timber
The building has a one-hour fire-resistance rating from the potential
members with higher densities exhibit a relatively low charring rate.
charring effect of the timber and a one-half hour fire-resistance rat-
These characteristics may result in additional improvements in the
ing from plasterboards that meet the local fire code requirement
fire durability of heavy timber.
(TRADA 2009).
In addition to heavy timber and dimension wood, there are also
The LCT One Building, built in 2012 and designed by architect
two other types of timber structural members: (1) cross-laminated
Hermann Kaufmann, uses a vertical concrete core and precast con-
timber and (2) glued laminated timber. Although generally mis-
crete slab on timber beams and glulam columns. The fire resistance
taken for one another, these types of timber present differences
in design accommodates for approximately 38 mm (1.5 in.) per
in terms of both fabrication and strength. Cross-laminated timber
hour charring rate, which can be significant in terms of the time rat-
(CLT) panels were introduced in the early 1990s in Austria and
ing, especially for thick structural members. The hybrid concrete-
Germany. CLT panels are made by gluing timber planks together,
timber floor system concept was reported to have passed 2 h of fire
through cross-wise angles, to neighboring layers by applying high
test (Tahan 2013). The favorable fire rating was reported as having
compression. Because of the orthogonal arrangement of the timber
a positive effect on decisions to attempt design of tall buildings, up
layers, a CLT panel demonstrates good dimensional stability similar
to 30 stories, using timber or a hybrid of timber and other materials.
to a concrete slab and wall. According to the FPInnovations and
Skidmore, Owings & Merrill, LLP (SOM) presented an example
Binational Software Lumber Council (BSLC), a CLT panel has a
of a hybrid concrete-timber application in a tall-building-concept
favorable stiffness, and hence, it can be built in large dimensions—
design (SOM 2013). Called the prototypical building of Dewitt-
up to 18.3 m (60 ft) long and 0.5 m (20 in.) thick (FPInnovations
Chestnut Apartments, the concept design is 42 stories high and is an
and BSLC 2013). CLT panels are ideal for load bearing or shear
example of a tall building that can be constructed in urban areas
walls and floor panels in tall buildings. Since becoming available,
such as Chicago. The design was applied to an existing concrete
CLT products have changed the notion that timber can only be used
benchmark for comparison. The concrete benchmark building (the
in low-rise buildings as simple framing systems. Glued laminated
Dewitt-Chestnut Apartments), a 120-m (395-ft) tall building was
timber, commonly known as Glulam, is typically manufactured
designed by SOM and built in 1965. The concept design used tim-
from lumber with 51 mm of nominal thickness (2 in.), an actual
ber for floors, columns, and shear walls with timber members
thickness of 38 mm (1.5 in.). Examples of such sections are 51 mm
(2 in.)  152 (6 in.) and 51 mm (2 in.)  203 mm (8 in.). However,
Table 1. Definition of Timber Building Types by Story Number
some Glulam members have been manufactured with 25-mm (1-in.)
thick boards with an actual thickness of 19 mm (0.75 in.). Examples Definition Number of stories
of such boards include 25 mm (1 in.)  152 mm (8 in.) and 25 mm
Low-rise ≤4
(1 in.)  203 mm (8 in.). It is noteworthy that a Glulam member is
Mid-rise 5–7
different from a CLT member in the way the lamination of the layers
High-rise ≥8
is formed. In a Glulam structural member, timber layers are cut in

© ASCE 04017014-2 Pract. Period. Struct. Des. Constr.

Pract. Period. Struct. Des. Constr., 2017, 22(4): 04017014


connected to concrete joints through reinforcement. Spandrel towers can be conceivably considered in design as long as the
beams were used to resist torsion and transfer of bending moments design follows the procedures of a performance-based design phi-
to the columns. Other features of this concept design included using losophy, which requires fire safety considerations (NFPA 2011b).
(1) timber floor panels; (2) reinforced-concrete spandrel beams and Along this line, it is thus important to discuss the fire durability
timber columns to resist the gravity load; and (3) timber shear walls properties of timber and incorporate information on the time
(located at the central core extended to the perimeter of the build- delay rating of it in any performance-based design approach. In
ing) to work with the concrete link beams and form a vertical canti- terms of fire endurance, the property of wood in promoting fire
lever system to resist lateral loads. The building was designed with safety in a single-family dwelling has been confirmed by research
an all-concrete system from the foundation to Level 2 with the and experience. It has been shown that vulnerability to fire has lit-
entire structure made up of approximately 70% timber and 30% tle to do with the combustibility of the timber used in a building’s
concrete. Using a performance-based design philosophy, the SOM construction [Canadian Wood Council (CWC) (CWC 2000); Hall
project demonstrated that this prototypical building is, indeed, tech- (2013)]. Timber members can resist fire inherently as long as they
Downloaded from ascelibrary.org by University of California, San Diego on 06/30/17. Copyright ASCE. For personal use only; all rights reserved.

nically achievable. are designed to have a sufficient “thickness sacrifice” for pyroly-
Another example of a tall timber building was reported by Green sis process (charring). In particular, heavy or mass timber mem-
and Karsh (2012). The system uses a “strong column–weak beam” bers possess this charring capability and can be designed without
design concept with timber panels providing for vertical members, any fire-protective covers. The fire-protection capabilities of tim-
shear walls, and floor slabs. The beams were made of wide-flange ber members can be determined using either the direct exposure
ductile steel beams bolted to the mass timber panels to provide for or membrane-protection method (Ling 2014).
the ductility needed in the system for performance and resistance
against lateral loads. Four possible alternatives to building tall tim-
ber buildings were also discussed with different story levels
depending on whether only a core or a core with other components Direct Exposure Method
(including interior shear walls and perimeter moment frame sys- When wood is ignited, the pyrolysis process starts, and the char will
tems) were used. The potential number of stories can be 12, 20, or be formed on the surface of the timber to consume energy. During
30, depending on the type and combination of the resisting system this charring process, the outermost timber layer starts losing
adopted, with a cost comparable to that of a concrete building with a strength. In fact, as the temperature on the surface reaches 300°C
similar configuration (for 12- and 20-story buildings). (572°F), the member strength drops to 0, generating a complete
Perhaps one of the major reasons timber is not perceived as an char (Schaffer 1984). In the case of mass laminated timber, usually
alternative material for tall buildings is the common belief that (1) members have a relatively large size (compared to the typical fram-
timber members can only be reasonably designed to support small ing members used in houses). The outer layers generate a thick coat
loads, and (2) the material does not possess adequate ductility to of char to insulate the inner remaining wood to sustain loads contin-
withstand lateral loads. However, there has been at least one study ually. According to the American Forest & Paper Association
showing a totally different outcome, as described to conclude this (AF&PA), knowing the charring rate of timber, one can calculate
section. the remaining section of members after a certain length of time has
To demonstrate the capability of wood frame systems in earth- elapsed. The fire-resistance capability and rating can then be deter-
quake-load environments, the NEESWood Capstone Test project mined (AF&PA 2005).
was conducted in 2009 through the sponsorship of the National The calculation of exposure timber is used widely in national
Science Foundation (NSF). As reported by Van de Lindt et al. design specification (NDS) method (AF&PA 2005). In this method,
(2010), the building, seven stories tall with a 12 m (40 ft)  18.3 m a nominal charring rate is commonly suggested as b n equal to
(60 ft) area, was made up of a steel moment frame in the first story 38 mm/h (1.5 in./h) for sawn and glued laminated timber. The effec-
and wood frames in the second to seventh levels. The building was
tive charring rate (in./h) can be estimated using the equation b eff ¼
subjected to 180% of the 1994 Northridge earthquake. Although no
1:2 b n =t0:187 with t being time and using the nominal charring rate
major failure was observed in the structure, slight crack formations
in (in./h). Because the method being used in NDS to calculate the
in drywall appeared and a minimal withdrawal of fasteners (nails)
charring rate is a general equation for all species of wood, the esti-
were noticed after the test.
mation may vary depending on the type of wood that has been
exposed to fire. Experimental results have shown that the density of
Fire Resistance of Timber Members wood significantly affects the charring rate. Njankouo et al. (2004)
conducted a study on 10 species of wood (8 hardwoods, 2 soft-
According to common belief, timber is more susceptible to fire woods) and found that hardwoods with densities ranging from
because it can contribute to fire loads when ignited, but because 692 kg/m3 (43.2 lb/ft3) to 1,060 kg/m3 (66.2 lb/ft3) have a charring
they do not act as combustible materials, steel and concrete are con- rate from only 22 mm/h (0.85 in./h) to 35 mm/h (1.37 in./h), which
sidered to outperform timber during fire. However, research has is clearly lower than the nominal charring rate of 38 mm/h (1.5 in./h).
shown that this may not be necessary true. In the following section, According to the Standards Australia (SA) 1720.4 (SA 2006), char-
an overview of fire-resisting capabilities of timber is covered. ring rate can simply be obtained on the basis of wood density as
follows:
Combustible Nature of Timber  2
280
Following the International Code Council (ICC) and National Fire b ¼ 0:4 þ
r
Protection Association (NFPA), timber is presumably a combusti-
ble material, and as such, it is not recommended for use as a primary
structural material in tall buildings (ICC 2011; NFPA 2011b). where b = notational charring rate (millimeters/minute); and r =
However, because building codes require the same performance timber density at a moisture content of 12% (kilograms/cubic
and safety for all buildings regardless of material types, tall timber meter).

© ASCE 04017014-3 Pract. Period. Struct. Des. Constr.

Pract. Period. Struct. Des. Constr., 2017, 22(4): 04017014


Membrane Protection between the CLT panel and the wall board. The ceilings were pro-
tected by a 12-mm (1/2-in.) fireproof gypsum board with 27 mm
This method uses protective membranes, generally gypsum board,
(1 in.) mineral wool insulation. A total fire load of 790 MJ/m2
to cover the external surface of the structural members, leaving
(69,560 Btu/ft2) was applied to duplicate a real residential fire situa-
them to gain an increase in their fire-resistance ability. White
tion. This was a 1-h fire test, within which the flashover occurred
(2009) indicated that a single layer of 16 mm (5/8 in.) Type X gyp-
after approximately 40 min. After approximately 55 min, the fire in-
sum board has an average 30-min- long fire-resistance rating.
tensity declined, and the fire was extinguished manually after
According to the American Wood Council (AWC), for a double
60 min into the test. The test results showed that the solid mass tim-
layer of Type X gypsum board, at least 60 min can be added to the
ber structures have good performance in fire, with no temperature
fire-resistance rating of unprotected wood. By adding an appropri-
increases measured and no smoke observed on the upper story.
ately designed protective membrane to an unprotected structural
Furthermore, the charring depth was small, varying between 5 mm
element, a desired fire-resistance rating for the system (i.e., a combi-
(3/16 in.) and 10 mm (3/8 in.).
Downloaded from ascelibrary.org by University of California, San Diego on 06/30/17. Copyright ASCE. For personal use only; all rights reserved.

nation of the structural wood element and the protective membrane)


can be achieved (AWC 2010). On the basis of the bilinear model
adopted by European Committee for Standardization (CEN), EN Significance of Wood Properties on Fire Performance
1995-1-2 (CEN 2004), the initially protected surface can be used to of Timber Members
calculate the charring rate after the protective membrane or the char As it is known, the moisture content and density affect the fire per-
layer falls off. For unprotected members exposed to ISO-fire formance of solid sawn timber. While the moisture content
(one-dimensional charring rate), it has a value of about 38 mm/h improves the fire performance, it may reduce the capacity of the
(1.5 in./h) charring rate. When the protected member has a mem- member. For glued laminated timber, such as CLT panels, the adhe-
brane fall off, the charring rate nearly doubles the initial charring sive performance is also an important property to consider. There
rate. This is because by time the membrane falls off the temperature are two types of adhesives commonly used in CLT panels: (1) the
has already reached a high level such that any additional gain in fire melamine urea formaldehyde (MUF) based adhesive; and (2) the
endurance will be in addition to that the member already offers. polyurethane (PU) based adhesive. Both compounds provide satis-
Once the charring depth reaches 25 mm (1 in.), with a relatively factory performance in normal temperatures. However, with the
large thickness to protect the inner wood the charring rate drops presence of heat from fire, the performances of the two types are dif-
back to the initial charring rate at the start of the fire. ferent. Experimental results have shown that when MUF-based
adhesives are used, the CLT panels perform similar to homogene-
Recent Fire Tests ous timber panels (Frangi et al. 2009). The charring rate reported
The standard methods of fire tests for building construction and was at 0.6 mm/min (1.42 in./h) comparable with 0.65 mm/min
materials per ASTM E 119–11a were reported by Rizzo and (1.54 in./h) of solid sawn timber, which was adopted by EN 1995-
Menchetti (2012). The study reports on a fire test of a CLT and gyp- 1-2 (CEN 2004), and the nominal charring rate of 0.64 mm/min
sum board wall assembly. The study was conducted for the (1.5 in./h) recognized by the NDS (AF&PA 2005). When the PU-
American Wood Council to evaluate the fire performance of CLT based adhesive was used in the experiments, the charring rate
panels with protective membrane when exposed to standard fire. increased significantly; however, it was not constant during fire dura-
The CLT wall used in the study was made up of two 5-ply CLT pan- tion. At high temperatures, heat delamination is also possible.
els, each approximately 175-mm (6-7/8-in.) thick, with a single
layer of 16 mm (5/8 in.) thick Type X gypsum board attached on Treatment for Wood Members
each side of the CLT panel. The CLT wall was subjected to a maxi- The fire retardant treatment (FRT) is widely applied to decorative
mum 389 kN (87 kips) load during the fire test. The result of the test wood used in interior walls, hallways, or stairways. The FRT delays
showed that the test specimen could withstand fire exposure for a ignition, reduces the heat release rate and flammability, and slows
period of 3 h, 5 min, and 57 s before failure. During the test, the tem- the spread of flames when wood is exposed to fire. The FRT is gen-
perature on the unexposed surface of the CLT wall (above the gyp- erally believed to result in reducing the mechanical properties of
sum board) remained almost the same. wood and exposing metal fasteners to corrosion. Therefore, it is
Another CLT panel fire test was reported by the National rarely used in structural members made of heavy timber. However,
Research Council (NRC) of Canada. In this test, 8 full-scale experi- Tondi et al. (2012) indicated that the tannin-boron impregnation so-
ments (3 wall and 5 floor members) were exposed to standard ULC lution not only increased the mechanical properties of scots pine
S101 fire (Underwriters Laboratory of Canada, Standard Methods and European beech by 10–30%, but also reduces the flammability
of Fire Endurance Tests of Building Construction and Materials). and ember time while lengthening the ignition time. This is another
Some panels were unprotected and directly exposed to fire, while area where the state-of-the-knowledge on structural wood might
others were covered by Type X gypsum board (NRC 2012). benefit from additional tests and further investigations.
Aguanno (2013) also reported on fire tests on eight CLT floor pan-
els. The results were then used in a series of analytical models for
predicting the performance of CLT floor panels when exposed to Current Challenges in Using Wood in Tall Buildings
fire or external loads.
Frangi et al. (2008) reported on a full-scale fire test on a 3-story
Experimental Studies
CLT timber building in Tsukuba, Japan. With an area of approxi-
mately 7 m (23 ft)  7 m (23 ft) and a height of 10 m (32.8 ft), the The literature on the application of timber in tall buildings is rather
building consisted of 85-mm (3 3/8-in.) thick CLT panels as a struc- scarce and certainly not to the dimensions of data available for steel
tural wall component and 142-mm (5-5/8-in.) thick CLT panels as and concrete systems. Although a few timber towers have already
floor members. The protection for wall members was provided by a been built, with 8, 9, or 10 stories, with the exception of the
12-mm (1/2-in.) standard gypsum board and a 12-mm (1/2-in.) fire- NEESWood Capstone Test (Van de Lindt et al. 2010), very limited
proof gypsum board, with 27 mm (1 in.) of mineral wool insulation experimental data on full-scale models are available. On the subject

© ASCE 04017014-4 Pract. Period. Struct. Des. Constr.

Pract. Period. Struct. Des. Constr., 2017, 22(4): 04017014


of wood joints, the performance of CLT-panel connections is resist a 3- or 4-h fire-resistance rating with large axial loads. Thus,
available in some publications [e.g., in the CLT Handbook, to design a column with a large axial load, one alternative would be
FPInnovations and BSLC (2013)]. However, such tests are only to use built-up glued laminated timber sections. The current techni-
limited to simple connections (e.g., panel to panel) used in low-rise cal challenge to this practice is that if the laminated timber is applied
or mid-rise buildings. To promote the idea of using timber in tall in too many plank layers, the integrity, stability, or reliability of the
construction, perhaps more data on performances of wood frame member may be affected because of the demand for more adhesive
systems and their connections, especially in seismic-load environ- (Ling 2014).
ments, may be necessary.
Adhesive Effect
Fire-Resistance Issues
As mentioned herein, the PU-based adhesive has a high probability
According to the 2012 International Building Code by the to trigger delamination when a CLT panel is subjected to fire. To
Downloaded from ascelibrary.org by University of California, San Diego on 06/30/17. Copyright ASCE. For personal use only; all rights reserved.

International Code Council (ICC), buildings over 11 stories (with- maximize the fire performance of timber assemblies (such as those
out automatic sprinkler systems) need to have a 3-h fire-resistance made up of multiple relatively thick layers of laminated timber),
rating (ICC 2011). This requirement is 4 h when based on the some type of industrialization process needs to be developed for the
NFPA 5000 (NFPA 2011a). These ratings were for noncombustible application of the adhesive to reduce the potential for delamination,
materials. When a new fire-resistance rating for timber is devel- which significantly reduces fire resistance and shortened the rating
oped, one must take into account the fact that timber is mostly con- hour. Currently, the CLT Handbook (FPInnovations and BSLC
sidered to be a combustible material. Furthermore, the fire and fuel 2013) included the delamination effect in design methods.
load effects are also critical and need to be considered. However, the NDS (AF&PA 2005) did not include any such consid-
The fire-resistance requirement for timber members is limited to eration in design recommendations. If only the inherent charring
2 h (for small or medium cross-section members) in the design rate is considered, but not the potential delamination effect, the
method of the NDS. New adjustment factors for the allowable result obtained will not reflect the real fire-resistance hour.
design stress for section reduction (because of fire) will need to be
developed to meet the 3-h and 4-h fire rating required by the ICC
Significance of Charring
and NFPA, respectively. It is noteworthy that the currently available
fire-test results related to this issue are limited and only cover mem- In both the NDS specifications (AF&PA 2005) and the CLT
bers with small cross sections. Furthermore, these tests were usually Handbook (FPInnovations and BSLC 2013), the nominal charring
designed for approximately 1h of burning, and hence, they may not rate suggested was b n of 1.5 in./h; presumably, this value can be
offer much insight regarding the code-required 3-h endurance. In applied to all types of wood species. As expected, a heavier timber
addition, most such tests are under relatively small loads, but for a with higher density has a slower charring rate. Thus, it is imperative
timber column, for example, the applied load is much higher. Thus, to use a variable charring rate that is most representative of the type
adjustment factors must be considered to apply the results of avail- of timber. In a more appropriate approach, a list of charring rates
able tests to full-scale members or new tests on larger scales speci- corresponding to different wood species needs to be provided to
mens will need to be conducted. architects and design engineers in an effort to offer more flexibility
In another issue, it is noteworthy that most reported fire tests in the selection of a suitable fire-resisting value in tall building
were designed for limited time duration with manually extinguished applications.
fire. In the most appropriate process, any full-scale fire tests initiated
must continue without being manually extinguished. This continued
burn may provide valuable information on whether large-scale tim- Exterior Fire Exposure and Structurally
ber members can be self-extinguishable by the formation of the Significant Fire
char to isolate further combustion. If self-extinguishing can be real-
ized, a major finding regarding timber is achieved, and the current In NFPA 80 A (NFPA 2011b), special requirements for buildings
label as a combustible material (used in regulations) may be further with combustible exterior walls were prescribed. These require-
challenged (Ling 2014). ments were primarily based on timber frame systems. However,
Fire tests on timber joints have also been reported on a very insufficient data are available to demonstrate how an exterior timber
limited basis. In some tests, metal plates have been used as fasten- structural wall actually performs in fire. Perhaps more tests will be
ers. This is an unfavorable design arrangement for fire tests needed along this line to provide data on how and for how long exte-
because metal will lose strength quickly when subjected to high rior walls endure fire. According to NFPA 557 (NFPA 2012), struc-
temperatures. Because of the current application of CLT panels, turally significant fires are those that are intense enough (in size) to
concealed panel-to-panel connection systems became popular, impose threats to timber structural components of a building.
which are different from the traditional beam-to-column connec- Whether this definition can be applied to any type of element,
tions. Other types, such as timber-concrete joint connections have including those made of CLT panels, is subject to debate. As timber
also been proposed by SOM (2013) and in Green and Karsh becomes popular in taller constructions by using CLT panels of dif-
(2012); however, the performance of these other joints will need ferent cross-section configurations, certain clarifications or modifi-
to be verified in fire performance tests and/or through computer cations will perhaps need to be considered in the documentation.
simulation studies (Ling 2014).

Durability
Demand for Large Cross Sections for Columns
Currently, the largest sawn timber size available in NDS (AF&PA Compared with sawn timber, CLT panels may have a shorter serv-
2005) is 610 mm (24 in.)  610 mm (24 in.), while the largest glued ice life because of issues with the adhesive limited shelf life.
laminated timber is 1,524 mm (60 in.)  311 mm (12 1/4 in.). According to technical information from KLH UK (2013), the life
Obviously, these sizes are not suitable to be used as columns to span of a CLT panel (such as provided by KLH UK) is

© ASCE 04017014-5 Pract. Period. Struct. Des. Constr.

Pract. Period. Struct. Des. Constr., 2017, 22(4): 04017014


approximately 60 years. Considering the service life of most ordi- Table 2. Summary of Additional Data Needed
nary buildings, this life span seems to pose no problem with durabil-
Design and/or performance
ity. However, for a tall building, the life span is expected to be lon- issue Data needed
ger. In glued laminated timber members the deterioration of the
adhesive is expected to take place approximately 50 or 60 years af- Limited experimental studies Small-scale tests and full-size connection
ter the application—and this may not adequately respond to the tests
demand for a longer service life in tall buildings. Because the adhe- Low-rise buildings based on Fire-resistance ratings and calculations
sive performance is critical for the integrity of glued laminated tim- fire design and fire tests for mass timber; fire tests and fire
ber members, more durability tests will need to be conducted to modeling for high-rises
make sure the adhesive can maintain the expected performance for Demand for large sections New and enhanced technologies for
a long time period. for columns fabricating large cross sections with glued
Splitting occurs in sawn wood structural members, especially at laminated timber
Downloaded from ascelibrary.org by University of California, San Diego on 06/30/17. Copyright ASCE. For personal use only; all rights reserved.

connections. The significance of a split and whether or not it may Adhesive effect Adhesives used in laminated timber to
cause failure depends on the type of separation that can occur reduce potential delamination
between the wood fibers. Three types of splits are recognized and Charring rate of different Development of charring-rate list for
discussed herein. These are (1) checks or seasoning checks, which species different species of wood
are partial depth or thickness separations of the member primarily in Exterior-wall fire exposure Tests to further determine how exterior
the direction of the grain; (2) full depth splits, which are separations walls perform in actual fires
of the member in the direction of the grain; and (3) fractures, referred Durability Durability tests of laminated timber for
to as tension failures in ASTM D143 (ASTM 2004), which are sepa- long service lives and resistance to
rations perpendicular to the direction of the grain. If a split does not splitting
propagate to become a long split, the durability is not affected signif- Overcapacity issue Information on suitable structural
icantly. For some CLT panel connections, Mohammad et al. (2012) members with favorable strength-to-vol-
indicated an increase in splitting resistance caused by the cross-lami- ume ratio with enhanced fire endurance
nation, which develops a two-way action effect in the panel. This using proper protective coating
effect is attributed to favorable in-plane and out-of-plane strength
and stiffness properties in both directions and is believed to be analo- alternative material in tall building applications. Additional infor-
gous to the two-way action in a reinforced-concrete slab system. mation on areas where further data are needed can also be found in
However, further research will need to be performed to investigate the literature. The Timber Tower Research Project (SOM 2013) and
the cracking of large glued-laminated-wood members, or CLT pan- The Case for Tall Wood Buildings study (Green and Karsh 2012)
els, in particular in their connections, to provide more data on dura- are the leading publications on the tall-timber-tower concept. These
bility against potential for splitting and split propagation. studies also provide recommendations on the need for additional
studies. In a somewhat parallel study entitled Fire Safety
Overcapacity Issue Challenges of Tall Wood Buildings (Gerard et al. 2013), a literature
review and gap analysis were presented highlighting fire-safety
To provide 3- or 4-h fire resistance, a member will need to be built challenges in using wood for tall buildings. The analysis was very
with a relatively large cross section to sacrifice the thickness to fire, comprehensive and entailed information on the gap in the knowl-
so that it can maintain the ability to safely resist the applied load. edge and areas where additional studies and/or tests are needed to
This may cause an overcapacity issue in design. Furthermore, tim- overcome these challenges. Broad categories of issues related to fire
ber members with large cross-sectional areas may reduce the dynamics, economics, environment, and society were also covered.
strength-to-volume ratio, reduce usable areas, increase the total
weight of the building, and add to transportation energy consump-
tion and cost. For the purpose of reducing member size when it is
Conclusions
designed for fire, a high-performance protective membrane may
need to be developed to offer the fire resistance and mechanical
This paper presents an overview of available studies on timber
strength needed in tall building applications. Richardson and
structural and fire-endurance properties. The use of timber as an al-
Cornelissen (1987) indicated that coating can improve the fire-
ternative material for tall building construction is discussed. The pa-
resistance ratings of wood building members. In fact, it was
per also covers areas where additional data are needed to overcome
shown that the char formation can be reduced by as much as 70%
challenges in using timber in tall building applications. The follow-
(first 30 min exposed to a standard fire). However, currently coat-
ing are the main conclusions:
ing is not accepted by most codes as a method to improve the fire
1. Timber possesses certain unique properties and may have
performance of wood products. Additional studies will need to be
potential for use in tall buildings as the primary structural mate-
conducted to investigate how intumescent or fire-retardant coat-
rial. However, in tall building applications, there are many
ings improve fire performance of glued laminated timber during
challenges to make the material appealing to building owners,
lengthy exposure to fire, and whether they will improve the dura-
engineers, and architects. Among these is the perception of the
bility of large-timber built-up sections with a favorable strength-
to-volume ratio. public that timber is more suitable for use in single-family and
low-rise buildings. However, recent studies and limited tests
have shown that this belief is unwarranted.
Gap Analysis 2. Among various timber properties, the strength and fire durabil-
ity are those that will need to be further investigated to provide
Table 2 presents a summary of the areas where additional studies data that would support the use of timber in tall building
are needed to fill the gap in information in making wood an construction.

© ASCE 04017014-6 Pract. Period. Struct. Des. Constr.

Pract. Period. Struct. Des. Constr., 2017, 22(4): 04017014


3. The strength requirements for tall timber buildings have been Ling, L. (2014). “An investigation into fire endurance of wood as a struc-
demonstrated through innovative design by fabricating load- tural material in tall buildings.” Master thesis, Illinois Institute of
bearing walls along with columns using CLT panels. Technology, Chicago.
4. The fire-resistance capability of timber has been documented to Mohammad, M., Gagnon, S., Douglas, B. K., and Podesto, L. (2012).
“Introduction to cross laminated timber.” Wood Des. Focus, 22(2),
a certain extent. Specifically, this property is attributed to char-
3–12.
ring of the outside layer and protecting the wood inside. This NFPA (National Fire Protection Association). (2011a). NFPA 5000:
property is believed to provide a favorable fire rating for tim- Building construction and safety code, Quincy, MA.
ber. However, additional tests are needed to develop data for NFPA (National Fire Protection Association). (2011b). NFPA 80A:
high-rise buildings on the basis of fire ratings of various wood Recommended practice for protection of buildings from exterior fire
species and the performance of adhesives in CLT panels when exposures, Quincy, MA.
exposed to high temperatures. NFPA (National Fire Protection Association). (2012). NFPA 557: Standard
for determination of fire loads for use in structural fire protection
Downloaded from ascelibrary.org by University of California, San Diego on 06/30/17. Copyright ASCE. For personal use only; all rights reserved.

design, Quincy, MA.


References Njankouo, M. J., Dotreppe, J.-C., and Franssen, J.-M. (2004). Experimental
study of charring rate of tropical hardwoods, John Wiley & Sons, Ltd.,
AF&PA (American Forest & Paper Association, Inc.) (2005). National New York.
design specification (NDS) for wood construction, Washington, DC. NRC (National Research Council of Canada). (2012). “Full-scale fire
Aguanno, M. (2013). “Fire resistance tests on cross-laminated timber floor resistance tests on cross-laminated timber.” Constr. Innovation,
panels: An experimental and numerical analysis.” Master thesis, 17(4) hhttp://www.nrc-cnrc.gc.ca/ci-ic/article/v17n4-4i (Mar. 29,
Carleton Univ., Ottawa, ON, Canada. 2014).
ASTM. (2004). “Standard test methods for small clear specimens of timber. Richardson, L. R., and Cornelissen, A. A. (1987). “Fire-resistant coatings
D-143-94. Vol. 04.10 on Wood.” Annual book of ASTM standards, for roof/ceiling deck timbers.” Fire Mater., 11(4), 191–194.
West Conshohocken, PA. Rizzo, J. M., and Menchetti, J. R. (2012). “Test rep. for American Wood
AWC (American Wood Council). (2010). Design for code acceptance: Council: Standard methods of fire tests of building construction and
Component additive method (CAM) for calculating and demonstrating materials ASTM E119–11a.” Test Rep. No. WP-1950, NGC Testing
assembly fire endurance, Madison, WI. Services, Buffalo, NY.
CEN (European Committee for Standardization). (2004). “Eurocode 5: SA (Standards Association of Australia). (2006). “Timber structures—Part
Design of timber structures—Part 1-2: General–Structural fire design.” 4: Fire resistance for structural adequacy of timber members.” AS
EN 1995-1-2, Brussels, Belgium. 1720.4-2006, Sydney, NSW, Australia.
CWC (Canadian Wood Council). (2000). “Fire resistance and sound trans- Schaffer, E. L. (1984). “Structural fire design: Wood.” Research Paper FPL
mission in wood-frame residential buildings.” International Building 450, U.S. Dept. of Agriculture, Forest Service, Forest Products
Series No. 3, Ottawa, ON, Canada. Laboratory, Madison, WI.
FPInnovations and BSLC (FPInnovations and Binational Software Lumber SOM (Skidmore, Owens & Merrill, LLP). (2013). “Timber tower
Council). (2013). CLT handbook, U.S. Ed., Quebec. research project.” Final Rep., Skidmore, Owings & Merrill, LLP,
Frangi, A., Bochicchio, G., Ceccotti Cnr-Ivalsa, A., and Lauriola, P. M. (2008). Chicago.
“Natural full-scale fire test on a 3 storey XLam timber building.” Proc, 10th Tahan, N. (2013). “LCT ONE: Case study of an eight-story timber office
World Conf. on Timber Engineering, Curran, Red Hook, NY, 528–535. building.” hhttps://www.constructionspecifier.com/lct-one-a-case-study
Frangi, A., Fontana, M., Hugi, E., and Jübstl, R. (2009). “Experimental -of-an-eight-story-wood-office-building/i (Mar. 29, 2014).
analysis of cross-laminated timber panels in fire.” Fire Saf. J., 44(8), Tondi, G., Wieland, S., Wimmer, T., Thevenon, F. M., Pizzi, A., and
1078–1087. Petutschnigg, A. (2012). “Tannin-boron preservatives for wood build-
Gerard, R., Barber, D., and Wolski, A. (2013). Fire safety challenges of tall ings: Mechanical and fire properties.” Eur. J. Wood. Wood. Prod.,
wood buildings, Fire Protection Research Foundation, Quincy, MA. 70(5), 689–696.
Green, C. M., and Karsh, J. E. (2012). The case for tall wood buildings, mgb TRADA. (2009). Case study: Stadthaus, 24 Murray Grove, London,
Architecture þ Design, Equilibrium Consulting, Vancouver, BC, Canada. TRADA Technology Ltd., Buckinghamshire, U.K.
Hall, R. J. (2013). High-rise building fires, National Fire Protection Van de Lindt, J. W., Pei, S., Pryor, S. E., Shimizu, H., and Isoda, H. (2010).
Association, Quincy, MA. “Experimental seismic response of a full-scale six-story light-frame
ICC (International Code Council, Inc.) (2011). 2012 International building wood building.” J. Struct. Eng., 10.1061/(ASCE)ST.1943-541X
code, Country Club Hills, IL. .0000222, 1262–1272.
KLH UK. (2013). “Technical information-technical characteristics of KLH White, H. R. (2009). “Fire resistance of wood members with directly
cross-laminated timber panels.” hhttp://www.klhuk.com/product applied protection.” Wood design focus, Forest Products Society,
-/technical.aspxi (Mar. 29, 2014). Peachtree Corners, GA, 1–13.

© ASCE 04017014-7 Pract. Period. Struct. Des. Constr.

Pract. Period. Struct. Des. Constr., 2017, 22(4): 04017014

You might also like