Use of Timber On Multi-Storey Building
Use of Timber On Multi-Storey Building
13
13
Use of Timber in Tall
Multi-Storey Buildings
All rights reserved. No part of this book may be reproduced in any form or by any means, electronic
or mechanical, including photocopying, recording, or by any information storage and retrieval
system, without permission in writing from the publisher.
ISBN 978-3-85748-133-8
Publisher:
IABSE
c/o ETH Zürich
CH-8093 Zürich, Switzerland
1 Introduction 1
1.1 Historical use of timber for construction 1
1.2 Modern renaissance of timber as a construction material 5
1.3 Other chapters 6
13 References 171
Much has been written in the last few decades about the relative merits of alternative materials
for building construction. As part of such efforts, this Structural Engineering Document (SED)
provides guidance to engineers on how to properly design multi-storey buildings that incor-
porate timber and timber-based products as superstructure elements. The scope encompasses
traditional systems for buildings up to 10 storeys made from conventional timber products and
innovative systems that employ modern timber-based composites, as well as emerging possibili-
ties for using timber elements in very tall buildings.
Poor building performance is usually accompanied by a failure to integrate design across all
aspects of a project; or a failure to link design concepts with the realities of local construction
and maintenance practices. For example, if timber elements are not properly protected from
wetting (i.e. more than occasionally wetted at rates that exceed ambient drying rates), they are
unlikely to be durable. However, if they are protected adequately, timber elements are likely
to retain their initial properties for centuries. This document emphasises attainment of Total
Performance Goals on a cradle to grave basis, taking account of structural and non-structural
considerations. In the contemporary parlance, structural design decisions must support attain-
ment of Total Performance Goals from cradle to grave. Even though the lifespan of most build-
ings are indeterminate at the time of their conception, their design and construction must address
issues like capability of the fabric to retain integrity up to and beyond the likely lifespan and
eventual dismantling.
The intended audience for this SED is structural engineering practitioners, construction profes-
sionals, academic researchers, code drafting bodies, and students. However it is hoped that
there will be ancillary audiences amongst architects, property developers, town planners, and
governmental policy makers.
Ian Smith
Andrea Frangi
1
Chapter
1
Introduction
Summary: Since the dawn of civilization, timber has been a primary material for achieving
great structural engineering feats. Yet during the late 19th century and most of the 20th century
it lost currency as a preferred material for construction of large and tall multi-storey build-
ing superstructures. This Structural Engineering Document (SED) addresses a reawakening of
interest in timber and timber-based products as primary construction materials for relatively
tall, multi-storey buildings. Emphasis throughout is on the holistic addressing of various issues
related to performance-based design of completed systems, reflecting that major gaps in know-
how relate to design concepts rather than technical information about timber as a material.
Special consideration is given to structural form, fire vulnerability, and durability aspects for
attaining desired building performance over lifespans that can be centuries long. This chapter
discusses the historical use of timber as a high-performance construction material and lays
the groundwork for detailed discussion of modern practices and possibilities in other chapters.
From antiquity onwards, urban utilization of construction materials has been shaped by their
fire performance when assembled into buildings. City-wide or district-wide conflagrations were
the impetus for prescriptive building regulations that date back to the Roman Empire [2]. More
modern catastrophes like The Great Fire of London in 1666 and The Boston Fire in 1872 have
reinforced fear of urban fires, and many specific building code restrictions created between
17th and 19th centuries are recognizably alive today in some jurisdictions (e.g. not allowing
timber buildings to have more than four storeys above ground). Historical building regulations
2 CHAPTER 1. INTRODUCTION
and material prohibitions were enactments to mitigate the extent of destruction that could be
wrought by invading enemies, as well as safeguards against disproportional consequences of
accidental events. Modern knowledge in fire-proofing (as well as modern firefighting techniques)
precludes the likelihood of the uncontrollable fires of previous millennia. However, until the lat-
ter part of the 20th century societies have been slow to adapt to changes in fire technologies and
threat levels, but now that they have adapted, a new era in timber construction has commenced.
During the last 50 years, fire performance and safety of buildings has become the subject of
systematic study, and fire engineering has developed as a science that can be applied robustly
in practice. Contemporaneously, fire detection and suppression methods have been developed
to an extent where it is no longer necessary to simply contain localized fires within non-com-
bustible compartments long enough for building evacuation or arrival of the firefighters. Tech-
nical measures available today can detect and extinguish fires quickly using a wide range of
devices. In practical terms, modern fire engineering and available technical measures result in
mitigation of fire damage to the extent that large buildings can be easily repaired after fires,
and minimise the possibility of fire spreading from one building to another. Building codes and
governmental authorities are starting to retreat from blanket prohibitions of timber materials
for certain uses. Additionally, prescriptive building regulations have been replaced by explicit
Building Performance Outcome (BPO) requirements (e.g. in Australia, Canada, New Zealand,
and Switzerland). BPO codes recognise that satisfactory fire performance can be achieved in
circumstance-specific ways, and that previous “one size fits all needs” approaches are not nec-
essarily ideal. Such codes embody principles, objectives, and quantitative performance-related
technical limitations that are independent of the construction materials employed. In Canada for
instance, there is now no material-specific limitation on building heights as long as a building’s
design meets required objectives in accordance with applicable regulations. Although the BPO
approach extends beyond fire aspects of building design, it has so far been the most effective
mechanism to overcome outdated and overly restrictive codes and practices related to timber.
Despite nature designing trees to decompose once they are dead, structures like the Yingxian
Wooden Pagoda in China and the Gümmenen Bridge in Switzerland demonstrate that decom-
position can be delayed indefinitely, if structures are appropriately designed, constructed, and
maintained (Fig. 1.1). One of the most prestigious examples of this is the pagoda of the Horyu-
ji Buddhist temple in Japan, arguably the world’s oldest timber building. One of its Japanese
cypress posts is thought to have been felled in 594 AD, thus that structure might actually be
around 1400 years old. Its longevity reflects that if timbers are cut and dried soon after trees are
felled and then kept dry in service, they are unlikely to experience extensive damage from bio-
logical attacks by decay fungi or other destructive agents (Chapter 4). Ancient timber buildings
owe their longevity to people who constructed them understanding the characteristics of tim-
ber and employing system level design concepts that protected the primary structural elements
from durability threats. Similarly, if longevity is a goal of modern design then timber should be
shielded from direct or indirect wetting, especially if the timber is a portion of a critical struc-
tural load path. Simple measures like large roof overhangs, draining water away from founda-
tions, and providing barriers to capillary wetting are highly effective [2,3].
In relatively modern times it has been a practice in some countries to treat exposed timbers
using creosote or synthetic chemicals. However, although durability is improved by introducing
these toxic materials, the serviceable lifetimes of wetted timbers remain finite. Furthermore, the
1.1 HISTORICAL USE OF TIMBER FOR CONSTRUCTION 3
(a)
(b) (c)
Fig. 1.1: Durable historical timber structures: (a) Yingxian Wooden Pagoda (courtesy of
Dr. Lin’an Wang, China National Institute of Cultural Property) From Lam & He SEI 2, 2009;
(b & c) Gümmenen bridge over the river Saane close to Mühlenberg, Switzerland (length about
100 m, built in 1555 and still used)
leaching of these chemicals from timbers can poison groundwater. Consequently this approach
is being increasingly banned. The philosophy espoused in this SED is that engineers should not
use timber for structural purposes if it cannot be kept dry. The meaning of dry can be ambiguous.
As used here the meaning corresponds to Service Class 1 of Eurocode 5 or similar definitions,
i.e. “... characterized by a moisture content in the material corresponding to a temperature of
20°C and the relative humidity of the surrounding air only exceeding 65 percent for a few weeks
per year” [4]. As discussed in detail in Chapter 4, there can be valid durability design reasons
for inoculating structural timbers with low dosages of synthetic preservatives, with combating
termite attacks being the most common one.
In the 1940s Howard Hughes and his engineers designed, constructed, and flew the H-4 Hercu-
les single hull flying boat constructed from laminated birch. The erroneously named “Spruce
Goose” was designed to carry 750 troops, is 24.18 m high, 66.65 m long, and has a wingspan of
97.54 m [5], making it approximately the same size as the Airbus A380 (the world’s largest air-
liner at the beginning of the 21st century). That the H-4 Hercules could fly and resist enormous
internal force flows generated during takeoff and landing is irrefutable. Furthermore, aeroplanes
were not the first large high-performance timber structures. More than 5000 years before around
3500 BC enormous timber sailing ships were common sights in Egypt and Mesopotamia [2].
The most famous was the three-mast Syrakosia of Alexandria, thought to have been about 70 m
4 CHAPTER 1. INTRODUCTION
long, which was commissioned by the tyrant Hieron and superintended and launched by the
mathematician Archimedes [6].
There are also many examples of historic wood structures that towered over the landscape. Built
in 1056, China’s 67.13 m tall Sakyamuni Pagoda was the world’s tallest building for several
centuries and has withstood many earthquakes [7]. Currently the tallest timber structure is the
118 m high Gliwice Radio Tower in Poland that was constructed in 1935 (Fig. 1.2). Table 1.1
summarises the tallest known historical and large man-made timber structures. For further infor-
mation, Langenbach [8] provides additional details of many of the listed structures, and Foliente
[9] provides an engineering overview of the historical development of human use of timber as
a construction material.
In summary, timber is a material that can be used highly effectively to create large and tall
structural systems capable of withstanding intense external actions. However, as with any other
structural material, good performance is intrinsic to the skill of the designers and builders and
not only an attribute of the material itself. Timber buildings that are well designed and properly
maintained, on the basis of educated know-how, typically exhibit exemplary performance under
normal and abnormal circumstances associated with human usage and natural events.
1.2 MODERN RENAISSANCE OF TIMBER AS A CONSTRUCTION MATERIAL 5
Despite broad loss of interest during the 20th century in using timber as a fully fledged structural
material, the know-how necessary to use it did not die. This was largely because of the dedi-
cated efforts of a cadre of aficionados who collectively maintained existing knowledge and further
developed it in readiness for a reawakening of broader interest in timber as a construction material
for large and tall building superstructures. Instrumental in this context have been expert panels like
Working Commission W18 – Timber Structures who operate under the umbrella of the Interna-
tional Council for Research and Innovation in Building and Construction; and the IABSE Working
Commission 2 – Steel, Timber and Composite Structures. Their efforts were not in vain, and dur-
ing the last three decades there has been renewed technical interest in timber by researchers and
practitioners around the world. Consequently, new ideas have emerged for design and construction
of timber structures in ways consistent with modern needs and practices.
Recent renewed interest in timber coincides with an almost total reversal of the factors that led
to suppression of interest in it approximately one century earlier. As already indicated, modern
BPO-based codes, circa 1990 onwards, permit timber usage in relatively tall multi-storey build-
ings provided it can be demonstrated that fire performance objectives will be met by design
solutions. In addition, the availability of new high-performance timber-based composites (e.g.
Laminated Strand Lumber), innovative building element manufacturing technologies (e.g. Com-
puter Numerical Controlled cutting machines), novel architectural styles and vernaculars; and
the search for “greener” construction methods have fuelled a renewal of interest in timber. In
response, timber design codes in various parts of the world have begun to be reoriented towards
provision of generalized structural engineering information, instead of only focussing on infor-
mation appropriate for low-rise building applications [4,10].
Currently educational and professional training mechanisms related to modern applications of timber
in construction are not at a level comparable with what exists in relation to materials like structural
steel and concrete, but significant initiatives addressing this are underway in Austria, Canada, China,
France, Germany, Italy, New Zealand, Sweden, Switzerland, United Kingdom, and elsewhere.
Chapter
2
Structural Design Issues
Summary: This chapter relates to the application of engineering methods where provisions of
written loading and material design codes are used as explicit justification of the acceptability
of structural system designs. In this context, it is assumed that all elements within engineered
load paths will be sized sufficiently that it is highly unlikely that they will sustain damage during
their functional lives, irrespective of events. Even then, it is necessary to design large and com-
plex structural systems in ways that will accommodate any localized damage that does occur
and limit it to an extent admissible for structural systems only constructed from elements that
fail by ductile processes. Such provisions are not expected to adversely impact costs because
the sizing of structural elements in relatively tall buildings is controlled by the magnitude of
lateral drift under wind and seismic loading scenarios. Recommendations are made which when
applied together facilitate proper design of any type of large and tall timber superstructure. The
recommendations concern safety, serviceability, and avoidance of unacceptable damage in the
event of overload.
2.1 Introduction
Just like other engineering materials, timber has specific characteristics that need to be con-
sidered to maximize strength and stiffness capabilities of structural elements, substructures,
and structural systems. This chapter discusses key issues related to the design of mechanically
efficient building superstructures.
The most important engineering characteristic of some, but not all, timber structural elements
is that mechanical properties are directionally dependent. Glued laminated timber (glulam) ele-
ments are illustrative of this as the mechanical properties depend upon the direction relative to
the axes of regular prismatic members. Table 2.1 shows typical direction of loading design prop-
erties of glulam members, based on values in the Canadian timber design code. Quoted specified
strength properties in the table apply to glulam made from softwood species, and apply to what
is termed Standard Term Loading conditions, which means that those properties are directly
applicable to most loading combinations associated with self-weight of building superstruc-
tures, plus roof snow loading and/or normal building floor occupancy loads [10]. The Canadian
specified strength values are 5-percentile strengths, multiplied by 0.8 to adjust from capaci-
ties observed in short-term laboratory tests to those appropriate to Standard Term Loading.
Relative to discussion in this Structural Engineering Document (SED), loading combinations
10 CHAPTER 2. STRUCTURAL DESIGN ISSUES
most likely to govern design of structural glulam elements in superstructure systems would be
appropriately sized using such properties. As tabulated values show, glulam members are much
stiffer and stronger when loaded parallel to their axes in tension or compression or bending, than
when loaded in directions transverse to their axes. This reflects the orientation in which nature
manufactured tree stems to be efficient in resisting effects of self-weight and environmental
loads. Care needs to be taken to avoid construction details that create high stresses transverse to
grain or in shear. As Table 2.1 indicates, manipulation of the wood species, quality of lamellas,
and layering arrangements of lamellas all enable the creation of a broad range of engineering
property combinations. As such, particular element capabilities can be accurately matched with
internal force demands on elements. Thus, although the starting point of glulam manufacture is
a tree stem with variable characteristics, the end product is an efficient structural product. The
same applies to other engineered timber composite products, like cross laminated timber (CLT),
as discussed in Chapters 6 and 7.
Contemporary timber design codes and most proprietary timber-product information are based
on representing strength and stiffness properties as apparent properties corresponding to lin-
ear elastic material response. This makes the calculation of member properties quite simple.
Usually, the greatest complexity is in the design of the connections and that is the aspect that
separates the skilled from novice timber designers. For that reason, there is great emphasis in
Chapters 5–9 on connection selection and design. Within this SED, the term “structural ele-
ments” refers to linear members, plates, connections, and any other parts within engineered load
paths of a superstructure.
or by capillary action where timber bears on wet materials. These are neither typical nor
appropriate uses of timber elements within the superstructure of a large building. This topic
is discussed further in Chapter 4 with respect to long-term wetting and subsequent durability
problems.
Some timber design codes include what appear to be stringent requirements that discount
strength properties of timber elements when their volumes increase beyond reference dimen-
sions (i.e. sizes associated with indexing of tabulated design values). However, such adjustments
are largely the result of the already mentioned employment of apparent elastic strength proper-
ties, which results in inclusion of post-elastic behaviour. Dependencies of apparent strength
properties of elements on their sizes are not highly relevant to the discussion herein because
adjustments generally relate to unsuitable products such as small dimension lumber.
Pathway 2 – Engineered Design is the normal approach by which engineers design structural
systems and is based on the application of provisions in written design codes that specify load-
ings and how structural elements manufactured using various materials [e.g. timber, structural
steel, reinforced concrete (RC)] should be sized. In general, this pathway applies to the use of
generic construction materials and elements manufactured to specific product specifications.
This is the default approach discussed in this chapter, and the SED in general.
Pathway 3 – Alternative Solutions is the route by which engineers can apply technical infor-
mation that is beyond the scope of written design codes to demonstrate that designed solu-
tions are at least technically equivalent to Acceptable Solutions based on pathways 1 or 2. The
Alternative Solutions pathway is discussed most extensively in this SED in the context of fire
design (Chapters 3 and 8–10), and in respect of structural design of systems using timber-based
engineered composites like CLT (Chapters 5–10).
Discussion in this SED is based on the assumption that the structural design will be undertaken
via the Load and Resistance Factor Design (LRFD) approach, also known as Partial Coefficients
Design (PCD) approach. Essential characteristics of LRFD/PCD are as follows:
(i) Consideration of the effects possible loading combinations have on stability and deforma-
tion of the structural systems.
(ii) Explicit separate consideration of various potential limiting states related to stability and
functionality of the structural systems.
12 CHAPTER 2. STRUCTURAL DESIGN ISSUES
(iii) Accounting for systemic and statistical variability in properties of elements that are parts
of the structural systems (i.e. variability associated with or resulting from material manu-
facturing, placement, and service environment).
(iv) Consistency in requirements for the design of structural elements made from various ma-
terials (e.g. timber, structural steel, structural glass, RC, reinforced masonry).
(v) Consideration of the deformation compatibility of various elements in the structural sys-
tems (i.e. employing structural analysis techniques that account rationally for boundary
and other applicable deformation constraints and for the interaction of elements in sys-
tems).
Notably, there is no explicit discussion in this SED about foundation substructures beyond a brief
reporting on systems employed in example projects and case studies of Chapters 8–11. Never-
theless, it is intended that information given will be used in combination with well-established
principles in foundation design applied by engineers experienced in such matters. Additionally,
it is presumed that engineers designing building superstructures and their foundations will be
fully cognizant of the principles and details of complete building design.
The fundamental principle embedded in Eq. (2.1) is that the maximum effect of all possible
loading combinations must be resisted in respect of effects related to loss of stability or unac-
ceptable deformation. This principle is applicable with respect to the behaviours of the complete
superstructures, substructures, and individual structural elements.
Loading codes account for the geographic locations of buildings and define all necessary
information related to magnitudes and combinations of loads to be considered by engineers
designing superstructure systems. In most instances, the partial load factors are independent
of the type(s) of material used to construct building superstructures. However, there are
some important exceptions, with the most typical being cases where engineers are allowed
to perform equivalent static force design of relatively simple buildings, as in the example
project of Chapter 8, and cases where temporally varying processes associated with mate-
rial responses (e.g. shrinkage and creep) can generate significant stresses in the structural
systems. Material specific codes define characteristic material properties necessary for sizing
structural elements and making deformation predictions and specify partial material coef-
ficients. Those codes include guidance on necessary adjustments of characteristic material
properties to transform them from reference values to values appropriate to particular service
situations.
Technically, resistance factors in LRFD (commonly termed phi f values) and material partial
factors in PCD (commonly termed gM values) are the inverse of one another:
f = 1/gM (2.2)
2.2 DESIGN PRACTICES AND ASSUMPTIONS 13
In the context of this document, it is presumed that engineers will employ LRFD/PCD design
practices deemed appropriate to the design of the structural systems in particular geographic
locations.
Under pathway 2 to Acceptable Solutions, there is an implicit requirement that all elements
in a structural system be able to distort sufficiently to achieve the levels of deformation asso-
ciated with their characteristic design strength properties. However, the only reliable way to
ensure that this presumption is not violated is to design and construct structural systems such
that they are statically determinate. In all cases where systems are statically indeterminate,
adoption of codified design strengths of elements (i.e. element design capacities) presupposes
that substructures can accommodate deformations required to achieve full strengths of criti-
cal parts. Similarly, adoption of codi-
fied strength of elements presupposes
Load
Relevant to this SED, it is therefore acknowledged that in practice there is no exactness in the pro-
tocol by which engineers combine information in loading and material codes to design structural
systems. Except when loaded in compression, structural timber elements fail by pseudo-brittle
processes (i.e. exhibit limited post-elastic response prior to loss of capacity) [11]. Therefore,
one should not presume that large and complex structural systems that employ structural timber
elements will be capable of accommodating localized damage of an extent that may be judged
admissible for structural systems only constructed from elements that fail by ductile processes.
Engineers perform analyses of structural systems that transform code load information into the
effects of loads on structural elements. Irrespective of what design analysis methods and tools
are employed, materials’ information contained in codes alone is not sufficient because some
important stiffness property information is missing. In particular the information about stiffness
of connections is minimal or non-existent. Most connections joining structural timber elements
to each other and to other structural materials are comprised of mechanical fasters or carpentry
joints [12]. Connection response is not simple to estimate with exactitude. Therefore, either it
is assumed that timber end connections are pinned or experimental research information is used
to characterize their responses.
• When building superstructures are low-rise and not slender in shape, primary influences
on structural element sizing are satisfaction of strength demands associated with resisting
gravity forces on walls, columns, slabs, and girders. In the case of lightweight elevated
floors, avoiding vibration serviceability problems will often control elements sizes.
• For medium-rise and moderately slender buildings, superstructure element sizing is often
related to provision of adequate strength against the combined effects of gravity and lateral
loads (e.g. wind and seismic).
• When building superstructures are tall and slender, the dominant influence on structural ele-
ment sizing is the need to control lateral motions associated with wind and seismic loads.
2.2 DESIGN PRACTICES AND ASSUMPTIONS 15
1.0
Sawn timber
0.9
Relative specific properties Steel
0.8 Plain conc.
(per unit mass basis)
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
Tension Compression MOE
Fig. 2.2: Approximate relative specific mechanical properties of common construction materials
As discussed further for example projects in Chapters 8–11, experience to date is that sizing
of elements in relatively tall superstructure systems employing timber elements can be gov-
erned by strength considerations. However, more usually control of sway is the governing
considerations with emphasis on the provision of horizontal diaphragms having sufficient
rigidity to maintain the building shape under design-level wind or seismic loads. This reflects
the relative high strength to mass and stiffness to mass ratios for timber (Fig. 2.2).
Timber structural systems are often claimed to be highly compliant and able to absorb
energy well in elastic and post-elastic regimes during extreme loading events [13]. This
is true of some highly statically indeterminate traditional and modern structural systems
employing timber which can accommodate very high levels of local deformation and absorb
large quantities of kinetic energy in their elements without more than localized damage.
Figure 2.3 illustrates such systems. In the case of the traditional timber-laced and timber
infill-frame buildings shown, they accommodate very high distortion, because all of the
building’s substructures are flexible. They dissipate significant energy through frictional
contact with the constrained masonry pieces. Their survival from major earthquakes has
proven that traditional timber-laced and timber infill-frame buildings have remarkable capa-
bilities to recover elastically once unloaded [14]. Modern CLT buildings can also accom-
modate large deformations and absorb energy because the intra-wall segments connectors
and those between the walls and horizontal floor diaphragms are metal and thus flexible and
can distort plastically [15]. Additionally, special multi-storey timber construction systems
employing post-tensioning as a means of absorbing potentially damaging kinetic energy
flows through connections during earthquakes are being developed in New Zealand [16].
However, despite it clearly being possible to avoid damage to structural timber elements by
utilization of special techniques, it is not envisioned that engineers will often want to avail
themselves of exotic design options.
(a) (b)
(c)
Fig. 2.3: Old and modern highly statically indeterminate superstructure systems known to
accommodate very high levels of local deformation in elements: (a) timber-laced bearing wall
house (courtesy of R. Langenbach); (b) timber infill-frame house (courtesy of R. Langenbach);
(c) modern massive timber CLT assemblies (courtesy of A. Ceccotti)
Recommendation 1: Designers should aim to achieve strict compliance with loading codes and
material specific design codes and provide clients with reasonable surety that superstructure
systems will not be damaged by loading scenarios anticipated by design codes.
Recommendation 2: For buildings that do not contain special measures to absorb kinetic
energy during design-level earthquakes or other unusual design events, the Engineered Design
approach is an appropriate method for demonstrating acceptability of superstructure system
designs.
Recommendation 4: For buildings that contain special measures to absorb kinetic energy dur-
ing design-level earthquakes or other unusual design events the Alternative Solutions approach
is an appropriate method for demonstrating acceptability of superstructure system designs.
2.2 DESIGN PRACTICES AND ASSUMPTIONS 17
Is NO
j=J
?
YES
Resize elements Resize Elements:
OUTPUT iteration k:
(selectively or globally to Ri,k+1 = (Ri,k)2 / IFRMax_i,k (i = 1, I)
Fi,j,k values
achieve target displacement Note: element stiffness resized as a result
system deflections k
responses)
Is
NO
GIFRMax_k = 1.0 ± tol.
?
YES
Are
NO YES
k acceptable END
?
Figure 2.4 shows a suitable design methodology for implementing “normal” design in a way
consistent with recommendations 1–3. To follow this approach designers need to classify struc-
tural elements by failure mechanism: brittle (Type B) or ductile (Type D). Table 2.2 suggests a
simplified method of classifying elements. Alternatively numerical analysis or experiment test
information can be used for this classification.
Notably, the suggestion to employ distinctions between brittle and ductile elements within engi-
neered load paths of superstructures is not an explicit suggestion that engineers adopt Capacity
Based Design (CBD) practices that are permitted options under some design codes [17,18].
Adopting CBD practices can provide an added layer of surety against unstable damage propa-
gation. However, they are not sufficiently reliable to be used as the first layer of protection that
load paths containing timber members will behave as desired, which is why retention of an
elastic response is advised herein.
It is suggested that those engineers interested in following the Alternative Solutions approach
to proving the acceptability of proposed designs refer to the comments in Subsection 2.5.2 that
18 CHAPTER 2. STRUCTURAL DESIGN ISSUES
address robustness in the context of need to block the possibility of damaging kinetic energy
flows which can lead to disproportionate damage of large structural systems.
Structural analysis of timber construction is essentially no different from design of other sys-
tems, if the structural skeleton is well defined, even though timber frameworks will typically
contain less moment transferring connections and more bracing diagonals and/or supplementary
stabilizing components (as discussed further in Chapter 5).
As noted in Subsection 2.2.2.1, there exist approximate relationships between slenderness and
which limiting state will govern lateral load responses of building superstructures. The simplest
superstructures to design and construct are those having squat shapes (i.e. a modest ratio of
2.3 EFFECT OF SUPERSTRUCTURE SHAPE AND HEIGHT 19
height to footprint dimensions) and containing many internal divisions. This is true irrespective
of the absolute height. In such circumstances, satisfying ultimate limiting state design criteria
is mostly a case of being able to provide sufficient resistance to gravity forces and sufficient
shear walls or bracing to resist effects of horizontal shear forces during extreme wind or seismic
events. From a strength point of view, principally what limits the height of squat buildings is the
point at which columns or walls occupy an unacceptably large proportion of the floor plans at
the lowest levels in the system.
Figure 2.5 illustrates internal force flows that can be expected to be of primary importance for
low-rise squat and relatively tall and slender buildings, based on the illustrative case of plat-
form timber construction and seismic ground shaking. Essentially, in low-rise buildings the
key internal actions within superstructures are the horizontal forces between layers through
Inertial
forces
Reaction
forces
Fig. 2.5: Effect of building height and shape on internal force flows: seismic actions
20 CHAPTER 2. STRUCTURAL DESIGN ISSUES
the system’s depth. In taller and more slender buildings, the important internal actions within
superstructures are the horizontal shear forces and axial forces because of gravity and the bend-
ing forces. Even without specific variables or calculations, it is readily apparent that for slen-
der buildings controlling horizontal drift will be problematic. Resisting the total applied shear
forces and bending moments near to the base is also likely to be a concern. Controlling of drift
levels for slender, all-timber buildings beyond 8 or 10 storeys is likely to be impossible under
normal economic constraints. Thus, the most viable option for using timber in relatively tall
and slender superstructures is hybrid construction that combines timber substructures with other
substructure types [19,20].
As exemplified by the behaviour of slender steel and RC superstructure frameworks that work
in conjunction with RC building cores and massive timber floor and roof diaphragms (Chapter
6), torsion must be explicitly considered for superstructure deformation under wind and seismic
design events. This includes the need to consider demands on substructures and structural ele-
ments that can result from irregularities in the architectural or structural shapes.
In the modern context, the essential principles of ensuring efficient structural actions of slender
structures are the same in respect of the importance of the diaphragms. Building shapes may be
more complex and the construction less massive, but the use of horizontal diaphragms as sub-
structures to maintain the general shape of systems remains critical to efficient design of the ver-
tical substructure members. Multi-storey building superstructures of types discussed in this SED
employ large timber horizontal diaphragm elements as critical parts of fully three-dimensional,
self-bracing arrangements. As discussed in detail in Chapter 6, it is possible to construct massive
CLT diaphragms that are as rigid and as strong as RC diaphragms at near equal thickness. The
ability to do this is important to the utilization of timber in very tall superstructures, because the
rigidity of those substructures is sufficient to ensure that framework substructures will deflect
laterally in unison with RC or other massive shear walls. Making parts of superstructures so that
they move in unison dynamically is important, because it minimizes the potential that damaging
forces will flow through connections (e.g. between framework and shear wall substructures)
during seismic or other events that shake the buildings.
2.5 ACCEPTABLE RISK LEVELS AND AVOIDANCE OF DISPROPORTIONATE DAMAGE 21
Apart from philosophical and subjective issues of relative security of tall versus low-rise build-
ings, there are many compelling fracture mechanics-based reasons why the design of large
building systems should be based on elastic analyses. Irrespective of whether damage occurs in
a small object like a fracture toughness test specimen or in a large object like a building super-
structure, damage will continue to propagate if the conditions in Eq. (2.3) occur [22,23]:
d We > d U + Wf + d V + d Wp + d Wd + d L (2.3)
where d We is the increase in the work to the system since the last increment of damage, d U is
the increase in strain energy in the complete system because of the last increment of damage,
Wf is the work required to create the next increment in damage, d V is the increase in kinetic
energy in the system because of the last increment of damage, d Wp is the plastic work to distort
“other” components of the system, d Wd is the work to overcome damping in the system, and
d L is other energy loses in the system, because of phenomena like creep and generation of heat.
If only the first three terms in Eq. (2.3) are considered, which corresponds to standard static
analysis, it will be concluded that the capacities of structural components and complete struc-
tural systems are unique. However, when the equation is considered in its entirety, it becomes
clear that capacities of components are always conditional quantities. Therefore, the capacities
of two identical components situated within two different structural systems (or two different
locations within the same system) will fail at different levels of internal force [22]. In practical
terms, the implication of Eq. (2.3) is that as systems get larger, they tend to become more brittle,
because d U becomes larger with respect to Wf. For buildings constructed with non-timber struc-
tural materials, Eq. (2.3) has been applied in conjunction with search tree algorithms to predict
which amongst a large range of the possible failure mechanisms and sequences is likely to
govern the failure of a large structure under various loading scenarios [23]. Such an application
is illustrated in Fig. 2.6. The approach could be carried out to predict the minimum failure load-
ing for the system or to identify system modification strategies that mitigate against occurrence
Damage evolves
Dynamic
roots Multiply by # type events (e.g.)
• Seismic loading
• Gas explosion = # Stems
Undamaged structure Damage event • Bomb
of disproportional collapse under various damage scenarios (e.g. overloading and explosions).
Although the details of any outcome would depend on the structural form and construction
details for a building, it is often the d V term that is responsible for occurrence of dispropor-
tional damage. Thus, fracturing is a phenomenon for consideration that emphasizes the need to
maximize material damping in the deformation range before there are unrecoverable localized
distortions in elements.
As a sidebar to the present discussion, it is worth mentioning that Eq. (2.3) leads to the unerring
conclusion that exact designs for strength are only possible if the structural analyses of systems
are based on energy methods. Of course, the implication is that failure criteria for components
should be related to how much energy it takes to destroy them, but it can be anticipated that most
structural designers will not wish to embrace this idea because it requires abandoning a long
history of designing structural systems according to apparent strength criteria, in which they
mostly have great faith. For this reason, the discussion in this subsection should be regarded
by design practitioners as simply conceptual reasoning supporting recommendations in this
chapter.
Fig. 2.7: Modern podium construction with light-frame timber construction above two RC
plinth storeys (courtesy of FPInnovations)
24 CHAPTER 2. STRUCTURAL DESIGN ISSUES
earthquakes. Reportedly, the practice normally being adopted is to design RC plinth and timber
upper storeys as fully RC superstructures and fully timber superstructures, respectively.
The construction of podium or other superstructure systems that couples substructures with
different uncoupled dynamic response characteristics has the potential to create articulation in
the mode shapes and to result in unexpected force flows at points of connectivity/articulation.
Practices such as analysing substructures as though they are separate structural systems are
unreliable and can lead to inconsistencies in values assigned to ductility-related force modi-
fication factor (Rd) and the overstrength-related force modification factor (Ro). The product 1/
RdRo is used to adjust equivalent static design forces applicable to systems with brittle–elastic
responses to equivalent design forces applicable to the expected “actual” response of the system.
Technically RdRo is equivalent to what in Europe is called the Behaviour Factor q, defined as the
“factor used for design purposes to reduce the seismic actions in a linear static or modal analy-
sis in order to account for the non-linear response of a structure, associated with the material,
the structural system and the design procedures” [18]. In principle, the issue related to seismic
design of podium construction is the same as issues addressed in design of the hybrid timber and
RC superstructure of Example Project 1, as presented in Chapter 8.
The issue is addressed in this chapter because it is important to be clear about what this SED
recommends in cases where articulated dynamic responses of substructures must be evaluated.
The approaches recommended are:
Recommendation 6: In cases that fall outside the scope of recommendation 5 the effects of seis-
mic forces on superstructure systems should be evaluated on the basis of a full dynamic analysis.
Design engineers need to have available as wide a range of choices as possible. The discussion
and commentary in this chapter is aimed at facilitating that with respect to traditional timber
products (e.g. sawn timber and glulam) and modern timber-based, engineered composites as
high-performance structural materials. Chapters 5–7 place what is said in this chapter in context
relative to various types of structural systems for buildings.
25
Chapter
3
Fire Design Concepts
Summary: Combustible building materials like timber burn on their surface, release energy and
thereby contribute to fire propagation and the development of smoke during building fires. Largely
because of these characteristics, building codes have historically prescriptively limited the use
of timber as a building material. In particular, codes have restricted the number of storeys and
heights of timber building superstructures. However, timber elements are very predictable in terms
of how they burn, and if properly designed, they will perform excellently during building fires.
Effective fire design and firefighting is essential for very large, low-rise timber buildings, as well
as for tall buildings constructed of other materials. There is no reason that this knowledge cannot
be applied to high-rise timber construction. Notably, research has already verified the creation of
appropriate construction details and produced models of how tall timber buildings should perform
in fires. Modern performance-based codes recognize this, thereby making it possible to rationally
account for how buildings containing timber will behave as constructed systems in the event of
fires. Such codes recognize that timber elements can be designed to perform well by either being
large enough that they only char on their surfaces during fires, or being shielded from fire by
being encapsulated in layers of non-combustible materials. Fire compartmentalization is crucial
to good fire performance of any large building and that is emphasized in the discussion herein,
as a primary mechanism for ensuring that tall multi-storey building superstructures containing
timber will perform as desired. Specifically included herein are discussions that outline the histori-
cal development and current state of international building fire design regulations. These explain
how the performances of individual timber parts of buildings can be predicted; discuss building
system design concepts; and illustrate how tall buildings with desirable fire performance can be
constructed. As the example of the 120 m tall Dock Tower project concept illustrates, tall buildings
incorporating timber as a primary construction material can be expected to have fire performance
at least equal to, and potentially better, than typical traditional non-combustible tall buildings of
today. As with all other aspects of design and construction, finding acceptable solutions depends
on equipping engineers with knowledge and tools with which to exploit the potential of fire engi-
neering. Discussion in this chapter presents the relevant know-how.
3.1. Introduction
Many building codes have historically limited use of timber as a building material because of its
combustible nature. Such limitations manifest themselves as prescriptive limits on the number
26 CHAPTER 3. FIRE DESIGN CONCEPTS
of storeys and heights of timber buildings [24]. For example until 2005, new timber structures
in Switzerland were mostly limited to low-rise buildings of no more than two storeys. As dis-
cussed in Chapter 1, the basis of this was societal memory of past disastrous urban fires that had
occurred in various places over centuries. Fire safety is always a vital factor for humans to feel
comfortable and an important criterion for the choice of material for buildings, especially in
residential buildings where people sleep and can be slow to react to fire events.
Since the 1990s, many research projects around the world have focused on the fire behaviour of
timber structures [25–27], with some specifically aimed at supplying basic data and information
on the safe use of timber in multi-storey buildings. This led to novel fire design concepts and
models, based on extensive element and full scale testing [28,29] and a large statistical database
on fires in timber and concrete/brick buildings [30]. This improved knowledge, combined with
technical measures, especially sprinkler and smoke detection systems and well-equipped fire
brigades, allow for the safe use of timber in a wide field of applications. Again taking the Swiss
example, research has supported changes to the building code regulations that allow construction
of mid-rise residential buildings with up to six storeys. Such code alterations have been made in
other countries, similarly based on improved knowledge of how completed buildings perform
during fires. Consequently, there is now widespread international convergence of codified fire
performance-related regulations founded on requirements that buildings once constructed will
meet performance-based design objectives (Sections 3.3 and 3.5).
While, there are only a few examples of applying modern fire design practices to genuinely
very tall timber buildings, the practices are well proven in other contexts. For example, it is well
known how to handle design and fight fires in very large low-rise timber buildings and how to
do the same for tall buildings constructed of other materials. Research has verified the creation
of appropriate construction details and produced models of how tall timber buildings should
perform in fires. This reflects that fire safety engineering has developed rapidly as a discipline
that integrates all aspects of fire safety (structural, technical, and organizational) and combines
knowledge into the design of buildings. However, it is perhaps unrealistic to suppose that build-
ing regulatory authorities will rapidly permit unlimited use of timber in tall buildings worldwide
given the historical aversion to that. This chapter presents a generic fire safety concept for future
tall buildings in which a large proportion of the volume and mass of the structure (wall and
floors) is constructed using timber or wood-based composite products.
• Fire load, which is the amount, type, and arrangement of combustible materials.
• Ventilation conditions in the room of a fire’s origin.
• Thermal properties of the enclosures, that is, walls, ceiling, and floor.
• Firefighting actions.
Parametrical fire curves are calculated with simplified formulas developed for limited bound-
ary conditions [35]. For a more comprehensive and more detailed analysis, computer simula-
tions may be used like multiroom zone models or computational fluid dynamics models. Using
probabilistic approaches to fire simulation, it is possible to incorporate the input parameters as
variables and, thereby, account for uncertainties relevant to an analysis [36]. However, because
of the complexity and expense, such an approach is restricted to special situations and buildings
where simple calculations cannot be applied with confidence.
As achieving absolute safety is impossible, the level of acceptable risk is in general quantified
by building regulatory authorities, or by building owners and insurance companies in the case of
financial losses. Objectives can be reached with different fire safety concepts taking into account
the type of structure and occupancy. The most efficient ones should be selected through compar-
ison of options based on criteria like the total cost of fire safety measures, flexibility of applica-
tion, and architectural constraints on acceptability of alternative fire safety concepts. Depending
on the main focus of the measures the following generic concepts are distinguishable:
– Structural concept: Buildings are divided into compartments with fire-resistant floors and
walls that limit the spread of fires beyond where they started, with compartment sizing de-
pending on building type and usage. Floors are usually the primary fire separating elements
in multi-storey buildings. It is accepted that contents inside the building will be totally
destroyed and that building compartments will be damaged. This favours construction of
buildings with cellular structural forms. However, small rooms may limit use and flexibility,
especially for occupancies like shopping centres and industrial facilities.
– Surveillance concept: Automatic fire detection systems can locate fires and alert nearby fire
brigades, thereby allowing adoption of less stringent fire protection measures. Sometimes
this even leads to acceptability of unprotected steel or timber, especially in low-rise build-
ings with small fire loads and slow fire development (e.g. offices, schools, and gymnasi-
ums). However, in such instances fire detection systems must be properly maintained and
regularly checked for functionality.
– Extinguishing concept: A sprinkler system can extinguish fires at an early stage or keep
their intensity low, until firefighters arrive. In some jurisdictions, deployment of a sprinkler
system is considered to greatly reduce the necessity of the passive fire resistances of struc-
tures and fire compartment separations. At least yearly checks are important to achieving
high reliability of such sprinkler systems.
Prescriptive, also known as “deemed to comply” fire regulations give detailed rules leading to
standard concepts. Standard concepts are considered to give an acceptable level of fire safety.
If prescriptive fire safety concepts are employed within the scope of building regulations, no
quantification of the fire risk is required. The scope of such approaches is usually limited to
relatively small low-rise buildings of types known to perform satisfactorily in fires. For large
and uncommon buildings, detailed probabilistic and engineering-based analysis is necessary
and appropriate to design fire safety measures [29,39].
dchar,n
dchar,0
Key:
1. Border of residual cross-section (real shape)
2. Border of equivalent rectangular residual cross-section
2
Fig. 3.2: Charring depth dchar,0 for one-dimensional charring and notional charring depth dchar,n [40]
Charring depth dchar,0 or dchar,n (mm)
40
Key:
1. Relationship for initially unprotected
members for charring rate 0 and n
30 2b
2. Relationship for initially protected members
where charring tch starts at the same time as
20 1 dchar,0 =25 mm the failure of protection tf:
or 2a. After protection has fallen off charring
dchar,n = 25 mm increased at double rate 0 and n
10 2b. After char depth exceeds 25mm
2a charring rate reduces to 0 and n
0
tf ta
Time t
Fig. 3.3: General description of charring for initially protected timber surfaces according to
EN 1995-1-2 when start of charring tch occurs at the same time as the failure (i.e. fall off) of the
fire protective cladding tf (line 2). Line 1 is for initially unprotected timber surfaces
the value observed for one-dimensional heat transfer under ISO-fire exposure of a semi-infinite
timber slab. EN 1995-1-2 [41] suggests the value b0 = 0.65 mm per minute for softwood, as con-
firmed by experimental studies [42,43]. To take into account the effects of corner rounding and
fissures and to simplify the calculation of cross-sectional properties (area, section modulus, and
second moment of area) by assuming an equivalent rectangular residual cross section, design
codes generally define charring rates greater than the one-dimensional charring rate. These are
called notional charring rates bn [41]. Figure 3.2 shows the definition of the charring depth dchar,0
for a one-dimensional charring versus a notional charring depth dchar,n.
For protected timber surfaces, specific charring rates should be applied during different phases
of fire exposure [44]. Figure 3.3 gives the simplified model adopted by EN 1995-1-2 wherein
the start of timber charring tch occurs at the same time as the failure tf of the covering cladding.
Phase 2a describes the increased charring of timber after cladding has fallen off. The approach
assumes that timber charring takes place at double the rate for an initially unprotected surface.
The main physical reason for the increased charring rate is that at the time of failure of the
cladding the fire temperature will already be high and that there will be no protective char layer
to reduce the rate at which the timber continues to get heated [45]. The protection provided by
30 CHAPTER 3. FIRE DESIGN CONCEPTS
Key:
1. Initial surface of member
1
2. Border of residual cross-section
2 3. Border of effective cross-section
3
dchar,n dchar,n notional charring depth
d0 zero strength layer: d0 = 7mm
k0 d0 k0 k0 = 1.0 for t ≥ 20 minutes;
def k0 = t/20 for t < 20 minutes
t time of fire exposure
def effective charring depth
Fig. 3.4: Definition of residual cross section and effective cross section [40]
the char layer is assumed to grow progressively until its thickness has reached 25 mm. After
that, the charring rate decreases to the value for initially unprotected surfaces. For simplicity, a
25 mm criterion is adopted for calculations based on either one-dimensional or notional charring
rates. The simplified model can be used for protective claddings made of wood-based panels or
wood panelling and normal gypsum plasterboards [40].
Fire reduces the cross section, stiffness and strength of heated timber close to a burning surface,
with stiffness and strength decreasing significantly when temperature increases [46,47]. At a tem-
perature of about 200°C wood begins to undergo rapid thermal decomposition. The pyrolysis zone
has temperatures between 200°C and 300°C, with the char front being at about 300°C [40,42].
Because of the efficient insulating behaviour of the char layer and the timber, typical temperature
profiles exhibit steep temperature gradients through the burning timber members. Temperature-
dependent reductions in strength and stiffness near the charred layer can be considered in different
ways. EN 1995-1-2 for example gives two methods: the “Reduced cross-section method” and
the “Reduced properties method” [40]. The reduced cross-section method is the simplest and
accounts for strength and stiffness reductions near the charred layer by adding an additional depth
k0.d0 (called zero strength layer) to the charred layer depth dchar,n (Fig. 3.4). This zero strength layer
is assumed to build up linearly with time during the first 20 minutes of fire exposure. This method
permits designers to use strength and stiffness properties for ambient temperature for the resulting
effective cross section. Thus, the temperature-dependent reduction factor is taken as kmod,fi = 1.0
for the effective cross section. The reduced properties method, by contrast, downwardly adjusts
mechanical properties of materials and uses those in combination with realistic estimates of resid-
ual cross-section dimensions, making the method relatively complex [40].
On the basis of extensive experimental and numerical analysis novel fire resistance models were
developed for load-bearing structural elements. These include timber-concrete composite slabs,
cross-laminated timber panels, timber slabs made of hollow core elements, and multiple steel-
to-timber dowelled connections with slotted-in steel plates. Details of the design models can be
found in reference documents [48–51].
(E) and insulation (I). The required period of time is normally expressed in terms of fire resis-
tance using the standard fire exposure and is specified in building regulations. While fire tests
are still used extensively for the verification of the separating function of timber substructures
like light-frame wall-and-floor assemblies, design models are becoming increasingly common.
For the ISO-fire exposure criterion, I (insulation requirement) may be assumed to be satisfied, if
the average temperature rise over the whole of the non-exposed surface is limited to 140°C, and
the maximum temperature rise at any point of that surface does not exceed 180°C. This prevents
ignition of objects in neighbouring compartments. The criterion E (integrity requirement) may
be assumed to be satisfied, if no flames or hot gases on the non-exposed side of an assembly
can be observed. As I is clearly defined, verification can be made by heat transfer calculations
instead of testing. The E, on the other hand, is mostly determined by observations, because
calculations are too complex and involve prediction of when cracks will form, hot gas dynam-
ics, and other factors commonly beyond analytical capabilities. Premature integrity failure may
occur because of sudden failure of claddings or opening of gaps, which often is dependent on
how the material layers are fastened together. Extensive experience with full-scale testing of
wall-and-floor assemblies supports rules about the construction detailing of some wall-and-floor
assemblies included in EN 1995-1-2, for example. The integrity criterion may be assumed to
be satisfied, if the insulation criterion has been satisfied and panel materials remain fastened to
timber frameworks on non-exposed sides of assemblies.
In timber buildings, walls and floors are mostly built up by adding different layers to form an
assembly. For the verification of the separating function of timber assemblies, additive methods
are common for determining combined behaviour of components. With additive models, the
fire resistance of layered constructions is obtained by simply summing the contribution of the
individual layers. Calculation models for verifying the contributions of the different layers to
the separating function of light timber frame wall-and-floor assemblies are used in the UK [52],
Canada [17] and Sweden [53]. The current design method according to EN 1995-1-2 (Annex
E) is based on the Swedish component additive method. As an enhancement of the Canadian
method, the Swedish approach considers the influence of adjacent materials on the fire perfor-
mance of each layer and, therefore, more realistically describes the fire performance. However,
it should be cautioned that the method is based on data from only a limited number of fire tests
on wall assemblies and, therefore, only covers a limited range of timber structures. König et al.
[54] give a comprehensive review of the related international practices.
A comprehensive research project on the separating function of light timber frame wall-and-
floor assemblies with cladding made of gypsum plasterboards and wood-based panels was car-
ried out at ETH Zurich, in collaboration with Empa Dubendorf [55,56]. The objective was the
development of an improved design model for the verification of the separating function (E and
I criteria) of light timber frame wall-and-floor assemblies. A large number of small-scale fire
tests permitted the analysis of the influences of material type, thickness, position, and number
of the layers on the thermal behaviour of protective cladding made of gypsum plasterboards
and wood-based panels. Results of the fire tests allowed the verification and calibration of ther-
mal properties used in thermal finite element (FE) analysis. An extensive FE parametric study
enabled calibration of coefficients of a model used for the verification of the separating function
of light timber frame wall-and-floor assemblies. The resulting model is capable of considering
timber assemblies with an unlimited number of layers of gypsum plasterboards, wood panels,
or combinations thereof, with wall cavities that may be either empty or filled with rock or glass
32 CHAPTER 3. FIRE DESIGN CONCEPTS
1 4 5 6
Key
1. Timber frame member
2. Panel
3. Void cavity
4. Cavity insulation
5. Panel joint
6. Position of services
a–e Heat transfer paths
6 5 2 3
c a e b d
Fig. 3.5: Possible heat transfer paths through separating multiple layer construction
Layer i
Layer 1
Fig. 3.6: Timber frame wall-and-floor assemblies: numbering and function of the different
layers
fibre insulation. The model is based on the additive component method given in EN 1995-1-2.
Thus, the fire resistance tins of the timber assembly is taken as the sum of the contributions from
the different layers for the worst possible path of heat transfer (Fig. 3.5), and according to their
function and interaction (Fig. 3.6):
i = n −1
tins = ∑t
i =1
prot, i + tins,n (3.1)
i = n −1
with ∑ tprot,i = Sum of the protection values tprot,i of the layers (in the direction of the heat flux)
i =1
preceding the last layer of the assembly on the fire-unexposed side [minute]
tins,n = Insulation value tins,n of the last layer of the assembly on the non-exposed side [minute]
Protection and insulation values of the layers can be calculated according to Eqs. (3.2) and (3.3),
taken into account the values of the layers:
tprot,i = (tprot,0,i · kpos,exp,i · kpos,unexp,i + Δti) · kj,i (3.2)
with tprot,0,i = basic protection value [minute] of layer i, see Fig. 3.6
3.5. DESIGN MODEL FOR THE VERIFICATION OF THE SEPARATING FUNCTION 33
tins,0,n = basic insulation value [minute] of the last layer n of the assembly on the non-unex-
posed side, see Fig. 3.6
Δti, Δtn = correction time [min] for layers protected by gypsum plasterboards of type F or
type × as well as gypsum fibreboards (GF)
kpos,exp,i, kpos,exp,n = position coefficient that takes into account the influence of layers preced-
ing the layer considered
kpos,unexp,i = position coefficient that takes into account the influence of layers backing the
layer considered
kj,i, kj,n = joint coefficient
The basic insulation value tins,0 corresponds to the fire resistance of a single layer without the
influence of adjacent materials and joints (i.e. the average temperature rise over the whole of
the non-exposed surface is limited to 140°C, and the maximum temperature rise at any point
of that surface does not exceed 180°C). The basic insulation value can be assessed by tests or
FE thermal analysis. For FE thermal analysis, only the temperature criterion of 140°C is used.
The temperature of the layer at the beginning of the analysis on the exposed side and the non-
exposed side is assumed to be 20°C. The definition of the basic insulation value tins,0 is illustrated
in Fig. 3.7.
Wall-and-floor assemblies with only one layer are rarely used in buildings. Most assemblies
consist of two or more layers. The contribution to the separating function of the construction of
each layer, except the last layer of the assembly on the exposed side, is mainly protection of the
layer(s) below (see Fig. 3.6). Therefore, it is appropriate to introduce a basic protection value
tprot,0 defined as the time until failure of the protective function. Analogous to the calculation for
fire protective claddings of load-bearing timber constructions, according to EN 13501-2 [57],
the definition of the basic protection value tprot,0 is as illustrated in Fig. 3.8. The testing method
for fire protective claddings is performed with a particleboard with a thickness of 19 mm back-
ing the layer studied. The contribution of the cladding to the fire protection of the particleboard
may be assumed to be satisfied, if the average temperature rise over the whole exposed surface
of the particleboard is limited to 250°C and the maximum temperature rise at any point on
that surface does not exceed 270°C. For FE thermal analysis, only the temperature criterion of
250°C is used. The temperature of the layers on the exposed and non-exposed sides at the begin-
ning of an analysis is assumed to be 20°C (Fig. 3.8). Analytical equations for the calculation of
the basic insulation value tins,0 and the basic protection value tprot,0 for different materials have
been calculated by FE simulations and verified by fire tests.
The position coefficient considers the position of a layer within an assembly, in the direction
of the heat flux, because the layers preceding and backing a layer under consideration have an
influence on its fire behaviour. The physical meaning of the position coefficient is illustrated in
Temperature
distribution
Fig. 3.7: Definition of the basic insulation value tins,0 using FE thermal analysis
34 CHAPTER 3. FIRE DESIGN CONCEPTS
Temperature distribution
19 mm particleboard
20°C 20°C + 250°C = 270°C
tprot,0
Considered layer (average temperature rise)
Basic insulation value
20°C according to
EN 13501-2, 2003
Fig. 3.8: Definition of the basic protection value tprot,0 using FE thermal analysis
Temperature distribution
Temperature distribution
h 3. layer h 3. layer h 3. layer
20°C ≥20°C
(2. layer falls off) 270°C
h 2. layer h 2. layer
20°C
(1. layer falls off) 270°C
h 1. layer
20°C
(a) Start of fire: (b) Second layer exposed to fire: (c) Third layer exposed to fire:
t=0 t = tprot,1 t = tprot,1 + tprot,2
Fig. 3.9: Temperature distribution at different times of timber assembly with three layers
Fig. 3.9 for a three-layer assembly. For simplicity, it is assumed that each layer has the same
thickness and density and that the influence of joints between layers is neglected. In this case,
the basic protection value for each layer is the same (tprot,0,1 = tprot,0,2 = tprot,0,3). The first layer is
directly exposed to fire and backed by the second layer. The temperature of all layers at the
beginning of the fire on the exposed side and the non-exposed side is assumed to be 20°C
(Fig. 3.9a). The contribution of the first layer to the total fire resistance is defined as tprot,1.
The position coefficient kpos,1 of the first layer can be described as the ratio tprot,1 to tprot,0,1 and
depends on the layer backing the first layer. The second layer is protected by the first layer. It
is conservatively assumed that, after failure of the protection provided by the first layer that the
second layer is directly exposed to fire, with failure occurring when temperature at the interface
between the first and second layers reaches 270°C time t = tprot,1. The main differences in com-
parison with the initially unprotected first layer is that the temperature in the fire compartment
is already at a high level and that no protective char layer exists [45]. Further, the temperature of
the second layer on the fire-exposed side is 270°C, as defined previously, while the temperature
on the non-exposed side is equal or greater than 20°C depending on the thickness of the second
layer and the material preceding and backing the layer (Fig. 3.9b). For these reasons, the con-
tribution of the second layer to the total fire resistance is lower than the contribution of the first
layer, that is, tprot,2 < tprot,1. The position coefficient kpos,2 of the second layer can be described as
the ratio tprot,2 to tprot,0,2 and is < 1.0. For the same reasons, EN 1995-1-2 assumes that after failure
of a cladding, charring occurs at the increased rate of initially unprotected surfaces [41]. The
third layer is protected by the second layer. After failure of the protection by the second layer at
270°C time t = tprot,1 + tprot,2, the third layer is directly exposed to fire.
3.5. DESIGN MODEL FOR THE VERIFICATION OF THE SEPARATING FUNCTION 35
The third layer is the last layer in the assembly and takes an insulating function (Fig. 3.6). There-
fore, for that layer, the temperature criterion of 140°C is applicable, and an insulation value of
tins,3 must be calculated. Because of the further increased temperature in the fire compartment
and the missing protective char layer, as well as the temperature rise criterion of 140°C, instead
of 250°C, the contribution of the third layer to the total fire resistance is lower than the contri-
butions of the first and second layers, that is, tins,3 < tprot,2 < tprot,1. The position coefficient kpos,3
of the third layer can be described as the ratio tins,3 to tins,0,3 and is < 1.0. The fire resistance of
the entire timber assembly is the sum of the contributions from the different layers, that is,
tins = tprot,1 + tprot,2 + tins,3.
The influence of the layers preceding and backing the layer considered is analysed separately.
In that case, the position coefficient kpos,exp accounts for the influence of the layers preceding
a layer considered, whereas the influence of the
layer backing the layer studied is considered 1.0
by kpos,unexp. The position coefficients calculated
using FE simulations showed that the position
Position coefficient kpos,exp
0.8
coefficient kpos,exp is mainly influenced by the fire
temperature at the time when a layer being con-
0.6
sidered is exposed directly to the fire, as opposed
to the thickness of the layer itself. The influence
of the preheating was shown to be small. The 0.4
design model for the verification of the separat- 25 mm
18 mm
ing function of timber assemblies was developed 0.2 15 mm
for the ISO fire exposure and is a function of the 12.5 mm
10 mm
sum of the protection values of the preceding 0
layers (i.e.∑ tprot,i − 1). The thickness of the layer 0 10 20 30 40 50 60
considered is expressed as the basic protection Protection time Σt prot,i-1 (min)
value tprot,0,i or basic insulation value tins,0,n. As an
example, Fig. 3.10 shows the position coefficient Fig. 3.10: Position coefficient kpos,exp of
kpos,exp of gypsum boards with different thick- gypsum boards with different thickness
nesses, as a function of the sum of the protection as a function of the sum of the protection
values of the layers preceding the gypsum board values of the layers preceding the gyp-
(i.e.∑ prot,i − 1). The position coefficient kpos,exp
t sum board (i.e. ∑ tprot,i − 1)
decreases with increasing ∑ tprot,i − 1, because the
gypsum board is protected longer and at higher exposure temperatures. The position coefficient
kpos,exp decreases with reducing thicknesses of the gypsum board. Similar effects are observed
for wood-based panels and batt cavity insulation, making it possible to determine the position
coefficient kpos,exp as a function of the sum of the protection values of the layers preceding a layer
(∑ tprot,i − 1) and the basic values of a layer considered (tprot,0,i and tins,0,n). Calculation of position
coefficient kpos,exp is therefore simple.
The influence of a layer backing the layer of interest is defined by the position coefficient kpos,unexp.
Results of fire tests, Fig. 3.11, supported by FE simulations show that the influence of a backing
layer is small, if the backing layer is made of gypsum or timber. Also, batt insulation backing
layers cause other layers to heat up more rapidly reducing the protection time of the layer, and
therefore the position coefficient kpos,unexp should be evaluated for the case of batt insulation back-
ing. For timber or gypsum backing assuming kpos,unexp,i = 1.0 is conservative.
36 CHAPTER 3. FIRE DESIGN CONCEPTS
Timber
Insulation Cavity Air
GF GF GF GF
Fig. 3.11: Time taken to reach the temperature rise of 250°C (average) and 270°C (at any
point) on the fire-unexposed side of 15 mm thick GF tested as a single board or backed with
different materials
The methods described in this chapter extend the range of application of EN 1995-1-2 approaches
significantly, with further details given elsewhere [55,56].
For light-frame multi-storey structures, all floors and nearly all walls are load-bearing elements.
Therefore, timber framing materials are not allowed to be consumed at all on the fire-side of an
assembly until the R value is achieved. Consequently, the presence of non-combustible material
on interior surfaces is a universal requirement for buildings with more than four storeys.
Figure 3.12 shows examples of cross sections with fire resistance REI60/EI30(nbb). In
Fig. 3.12a, the claddings on either side of the timber framing are EI60. This means that the
interior of the timber structure need not be designed for a specific fire resistance rating, as
attainment of the required performance is guaranteed by the claddings alone. The solution in
Table 3.1: Fire requirements for load-bearing and separating elements in residential, office,
and school buildings designed based on the structural concept depending on the number of
storeys [31]
3.6. FIRE DESIGN CONCEPT FOR TALL TIMBER BUILDINGS 37
R60/EI30(nbb)
EI60/EI30(nbb) (a)
REI60/EI30(nbb)
R60/EI30(nbb)
EI60/EI30(nbb) (b)
REI60/EI30(nbb)
Cladding EI30(nbb)
Cladding EI30(nbb)
R60/EI30(nbb)
EI60/EI30(nbb) (c)
REI60/EI30(nbb)
Cladding EI30(nbb)
Fig. 3.12a will therefore, ensure a fire resistance longer than 60 minutes. In Figs. 3.12b and
3.12c, the claddings on either sides of the timber framing are EI30, in which case the fire resis-
tance of the timber structure shall be at least R30. Notably, the presence of insulation with a
melting point ≥ 1000°C inside wall cavities (Fig. 3.12b) can improve the fire resistance, but only
if the insulation remains in place after the failure of the cladding on the side exposed to fire.
Besides requirements related to use of combustible material and fire resistance of building ele-
ments, the fire regulations include mandatory rules for the design of escape routes consisting of
corridors and staircases, emergency exits, and necessary organizational and technical measures
such as smoke detectors and alarm systems, sprinkler systems, and smoke exhaust systems. All
such additional measures are required, irrespective of the type of construction materials used.
Basic safety objectives for prescriptive requirements are summarized in Table 3.2 for mid-rise
(five or six storeys) residential buildings, with those objectives reflecting the assumption that
occupants can leave buildings or be evacuated by fire brigades in case of fire. The limited nature
of the number of storeys and building heights is also crucial in respect of firefighting actions,
with it being assumed that a fire will usually be fought from the outside of a building.
The Swiss regulations recognize that the fire safety objectives adopted for mid-rise residen-
tial buildings can be achieved via prescriptive requirements, even when combustible structural
materials are used. The same applies in most other countries where use of timber is common.
3.6.1 Main differences between mid-rise and tall buildings with regard
to fire safety
Because of the height of tall buildings, occupants located on upper levels need considerable
time to leave the building, in the event of fire. Additionally, firefighters must be able to fight
fires from within buildings, which means that it can also take considerable time for them to
reach the fires. Additionally, some escape routes may become blocked and evacuation of build-
ing occupants may be by circuitous routes, or in extreme cases evacuation rescues may not be
possible at all. For these reasons, fire safety concepts for tall buildings are based on the scenario
that occupants located in upper parts of buildings cannot leave during fires. Additionally, it is
assumed that fires cannot be extinguished and will continue until all combustible material in
any affected fire compartment(s) has burned (i.e. full burn-out). On the basis of this scenario tall
buildings must comply with fire requirements that are more rigorous than for mid-rise buildings
(Table 3.2). Tall composite buildings of types discussed in Chapters 7, 9 and 11 must be treated
for fire in the same way as conventional tall buildings. In such cases requirements for building
elements are as follows:
• Separating building elements shall be designed in ways that sustain a full burn-out, thereby
preventing uncontrolled spread of fire to other parts of buildings throughout the duration of
a fire.
• Load-bearing building elements shall be designed in ways that prevent their structural col-
lapse during full burn-out without intervention of the fire brigade.
timber buildings of five and six storeys) can be considered as a partial building encapsulation
method (i.e. timber elements have fire protective non-combustible claddings with a fire resis-
tance of 30 minutes, Fig. 3.12). Here the classification of “partial” reflects that in many cases an
R-value of 30 minutes is insufficient to prevent the start of charring of timber structures during
a complete burn-out. Another way of guaranteeing effective compartmentalization is the use of
composite elements. For example, timber-concrete composite slabs can be designed in ways
that enable them to efficiently carry design loads applicable to normal and fire design situations,
while at the same time being effective fire barriers. Use of technical measures like sprinklers
is possible, as a way of reaching design objectives by preventing fires from developing. How-
ever, many building regulatory authorities are reluctant to accept sprinkler concepts alone for
fire protection of large and tall buildings. In some instances, approaches acceptable to regula-
tory authorities combine the passive ability of buildings to compartmentalize fires along with
sprinkler systems. Such solutions are commonly known as “Special Solutions” or “Alternative
Solutions” and require specialist fire engineering as their basis.
Multi-use
room Water mist
system
Staircase with
natural ventilation
Shaft
Elevator
Staircase
Corridor
Water mist
system
Multi-use
room
Projecting
concrete slab
staircases encapsulated in RC combined with projecting RC fire floors every three storeys. The
projecting concrete slabs have the function of preventing fire propagation across the building
facade. Residential apartments form secondary fire compartments with timber-concrete com-
posite floor slabs and timber walls designed according to the requirement of burn-out. The
building has five staircases in total, each placed as far away from each other as possible. In
addition, every apartment has direct access to at least two escape routes. Furthermore, two of the
exterior staircases are open to the outside environment, whereas the other two are pressurized to
avoid the entry of smoke. In order to control and extinguish a fire in an early stage, all building
rooms are equipped with a water mist sprinkler system. Activation of the water mist system is
temperature-actuated or controlled by a fire alarm system. In such a system, the nozzles break
the water down into very small drops creating a cooling and smothering water fog, which does
not allow a fire to persist. An additional measure is that the building has two high-pressure water
mist fire hydrants on each floor of the central core. Organizational measures were also designed
to enable safe evacuation and firefighting. The project shows the feasibility of using structural
measures in combination with the technical ones (water mist system, alarm signal, etc.) and
organizational measures to enable construction of tall timber building that have fire safety lev-
els higher than the standard measures for high-rise buildings. Essential to this is that measures
taken implement an approach of providing redundancy in each aspect of the fire design.
As part of that testing, four room modules (H1, H2 and G1, G2) were prefabricated off-site
with light-frame timber construction. Each module was 6.6 m long, 3.1 m wide, and 2.8 m high
and had a window (opening 1.5 m × 1.7 m) with a standard double layer of insulation glass.
The modules were identical apart from the wall and ceiling linings. For modules H1 and H2,
combustible wood-based panels (oriented strand board) were used as wall and ceiling linings. In
contrast, modules G1 and G2 had one to three layers of non-combustible gypsum plasterboard
wall and ceiling linings. The combustible floor was made of a light timber frame construction/
timber hollow core elements and was covered by a thin layer of linoleum. All modules were
equipped with an automatic fire detection system having four different sensors, as well as two
sprinkler systems. During the test fires, the contents (mobile fire load), as well as the combustible
construction materials contributed to the total fire load. Each module contained a mattress made
3.8. EXPERIMENTAL STUDIES 41
of polyurethane foam material and 11 wooden pallets, as additional mobile fire loads. The total
fire load density (calculated over the floor area) for the modules with non-combustible wall and
ceiling linings varied between 363 and 366 MJ/m2, and for the modules with combustible wall
and ceiling linings the total fire load density was approximately 855 MJ/m2.
After ignition, fire grew very rapidly in all tests, and the temperatures rose to flashover condi-
tion within only a few minutes. For the modules with combustible wall and ceiling linings,
flashover occurred after about 4 minutes. For modules with non-combustible wall and ceiling
Fig. 3.14: Damages on the mattress (left) and on the wall and the floor (right)
42 CHAPTER 3. FIRE DESIGN CONCEPTS
linings, flashover occurred after 6 to 7 minutes. In full-scale fire tests by the Technical Research
Centre of Finland where different timber compartments with and without gypsum plasterboard
protection of timber substructures were tested, flashover occurred between 4 and 6 minutes [62],
thereby confirming the results of the series BÜ tests. For the module with combustible wall and
ceiling linings, the fire spread through the window to the exterior. Following this, flashover
was much more severe than for the modules with non-combustible wall and ceiling linings
(Fig. 3.15). This was confirmed by observations of the facade with an infrared camera.
In the BÜ nbb demonstration test, the interior glass layer of the window of the upper module
failed after about 40 minutes. In the BÜ bb test with combustible linings, the heat flux from
the flames emerging from windows higher on the facade was much greater. Consequently, fire
spread could not be significantly delayed by merely placing a 250 mm protruding ledge made
from 1 mm thick steel sheet placed on the facade between the lower and the upper modules
(Fig. 3.15). In the bb test, the interior glass layer of the window of the upper module had already
failed within 7 minutes of ignition. Owing to excessive flaming in the BÜ bb test with combus-
tible linings fire could not be stopped for about 20 minutes.
3.8. EXPERIMENTAL STUDIES 43
Fig. 3.15: Fire development 7 minutes after fire ignition; left: lower module with combustible
linings, right: lower module with non-combustible linings
The BÜ nbb test with non-combustible linings made of one layer gypsum plasterboard, was
stopped after about 45 minutes and only after the interior glass layer of the window of the upper
module had failed. By that time, all mobile combustible material was burnt out. However, the
fire continued because of charring of the wood-based fibreboards placed behind the gypsum
plasterboard, which started to fall off after about 30 minutes. The BÜ nbb demo test with non-
combustible linings made of two or three layers of gypsum plasterboards was stopped after
about 60 minutes when no more flame emerged from the window opening and the fire inside
the lower module was in the later stages of decay. Because of the fire protection provided by the
gypsum plasterboard, the light timber framing in the walls and ceiling was not damaged, even
when complete burn-out was achieved.
Figure 3.16 shows the room temperatures measured on the ceilings of modules at the front
(near the window) and rear. Temperatures at the rear of the module were lower than at the front
because of the limited amount of oxygen away from the open window. Also, no significant
differences were observed in the early stage and peak temperatures for the modules with and
without combustible wall and ceiling linings. This confirms that pyrolysis gases released by
combustible wall and ceiling linings were only partially burnt. As unburned pyrolysis gases
flowed out of the window, they caused intense combustion outside where oxygen was freely
available. Other tests have yielded similar results [62].
Full-scale tests confirmed that by using purely passive structural measures, fire may be con-
tained to a single room within a timber structure. The BÜ nbb demo test in which timber fram-
ing was protected by three layers of gypsum plasterboards on the ceiling and two layers of
gypsum plasterboards on the walls is explicit proof that in multi-storey buildings, complete
burn-out of a lower room/module can occur without significant damage to the timber structure
44 CHAPTER 3. FIRE DESIGN CONCEPTS
1400
1000
800
600
Test BÜ nbb front
400 Test BÜ nbb rear
Test BÜ bb front
200 Test BÜ bb rear
0
0 10 20 30 40 50
Time (min)
Fig. 3.16: Room temperatures measured on the ceiling in the front as well as in the rear of the
lower modules for the fire test BÜ nbb with non-combustible linings as well as the fire test BÜ
bb with combustible linings (the temperature of the test BÜ bb front are not complete because
of a loss of electrical power occurred during the fire test after about 10 minutes)
or fire propagation to an upper room. Even smoke damage to upper rooms was limited until the
window broke in the room with the fire.
As with all other aspects of design and construction, finding acceptable solutions depends on
equipping engineers with knowledge and tools with which to exploit the potential of fire engi-
neering. This discussion is aimed at alerting them that the necessary knowledge and tools exist.
45
Chapter
4
Durability Design Concepts
Summary: Durability problems in buildings are always preventable through proper design,
construction, and maintenance practices. Timber construction elements in buildings can have
indefinite lifespans if they are properly protected against hazards to the durability. Primary
threats are biotic decay or insect damage to timber parts and corrosion of metal parts. The exact
natures of the threats can vary across and within geographic regions, as a function of climate
and human interventions within ecosystems. Critical elements within structural systems should
never be put at risk of attack from mechanisms that degrade timber or corrode metal parts,
because primary structural systems of buildings must always remain fully functional. However,
it may be an acceptable design decision to expose replaceable, non-critical structural and non-
structural timber elements, as part of life cycle cost optimization or architectural decisions.
Designing building elements so that they have appropriate durability is crucial to the implemen-
tation of construction systems that contain timber. The durability design of multi-storey build-
ings discussed in this Structural Engineering Document (SED) is as important as their having
proper structural and fire design, and must be similarly addressed at both the system and the ele-
ment levels. Attack mechanisms, risk mitigation methods, and the estimation of lifespan expecta-
tions for non-critical structural and non-structural timber elements are discussed in this chapter.
4.1 Introduction
This chapter addresses durability design of timber components in building superstructures from
the perspective of risk of material degradation over certain lifespans. This is required as a tool
for making initial material selection, planning building maintenance, designing replacement
programmes, and optimizing life cycle costs. Even though material selection decisions are ulti-
mately made at the building element/component level, degradation hazards are determined pri-
marily at the building system level. Thus, durability design depends on establishing building
element performance requirements based on their functional purposes. Herein, the discussion
of durability requirements is based on linking assessment of risk to element utilization classes:
• non-structural elements,
• non-critical structural elements, and
• critical structural elements.
46 CHAPTER 4. DURABILITY DESIGN CONCEPTS
From an engineering perspective, the supporting logic to the utilization classes is that buildings
can be designed such that non-structural and non-critical structural components are repairable
or replaceable. In such cases, designers should be allowed to make design decisions that trade
repair and replacement costs against the expected durability risk. However, critical structural
elements should not deteriorate and must last until a building is decommissioned. Crucial to this
is the recognition that, in the context of durability, the applicable performance requirement is
integrity of the primary structural system throughout a building’s actual lifespan. Specifically
this means that critical structural parts of buildings should not decay, corrode, or be other-
wise structurally compromised during normal service with respect to what the structural codes
assume as their available capacities.
Within this concept, the question arises as to what constitutes a critical structural element. There
will rarely be precise answers to this. However, a suitable definition may be one whose failure
could lead to the loss of life, blockage of emergency escape and access routes, or loss of critical
equipment (e.g. emergency electrical generators). Such a definition can be translated for imple-
mentation into requirements such as “a critical structural element is one whose failure would
cause the collapse of an area of building wall, floor or roof of more than 50 m2, or would sever
access to any portion of a building having a floor area greater than 50 m2”. Herein adoption of
50 m2 is purely illustrative and may differ significantly from applicable requirements in specific
regulatory jurisdictions.
The remainder of this chapter discusses quantitative approaches by which engineers can esti-
mate the likely service life of a timber or wood-based composite and affiliated fasteners and
hardware that connect such material to other structural parts. Although the presented approach is
applicable to only common, naturally occurring deterioration hazards, the conceptual principles
can be extended to rarer or human-induced hazards. With such tools, engineers can rationally
select materials for non-structural, non-critical structural, and critical structural parts of particu-
lar buildings.
4.2 State-of-the-art
Much scientific and design information exists on durability of generic and proprietary timber
and wood-based construction materials, including thermo-mechanically and chemically modi-
fied wood, and products created by fractionating and reconstituting wood in combination with
natural and synthetic adhesives. In Australia, for example, where a vast range of timber species
are used in construction, timber species and preservation treatments are grouped into classes
according to durability expectations [63]. However, such lists reflect how materials perform
under standardized assessment conditions, rather than providing direct indications of deteriora-
tion risks that can exist in specific construction situations. Consequently there is still a need for
more direct performance criteria for durability of building elements, as addressed by interna-
tional and national standards and best practice guidelines [64–66]. Standards and guidelines that
exist set a framework for establishing target performance durability levels for building elements
but do not give specific guidance relevant to design practice. The discussion below aims to fill
that void.
The ability of engineers to apply concepts to practice depends on access to robust and con-
sistent approaches for making building element service life predictions under defined service
4.2 STATE-OF-THE-ART 47
(a) (b)
DARWIN
D DARWIN D
Broome
Cairns
Broome
B Cairns
Townsville Townsville
Alice Springs
A B C Bundaberg A
Alice Springs
Bundaberg
Roma Roma
BRISBANE BRISBANE
Kalgoorlie Kalgoorlie
Coffs Harbour Coffs Harbour
PERTH PERTH
Dubbo Dubbo
ADELAIDE ADELAIDE
SYDNEY
CANBERRA
C SYDNEY
CANBERRA
ZONE A ZONE C MELBOURNE ZONE A ZONE C MELBOURNE
HOBART HOBART
(c) (d)
DARWIN DARWIN
F
Cairns Cairns
Broome Broome
D Townsville Townsville
C Roma
Roma Roma
BRISBANE BRISBANE
Kalgoorlie
Kalgoorlie Kalgoorlie
Coffs Harbour Coffs Harbour
PERTH
ADELAIDE
B Dubbo
Dubbo
PERTH
C ADELAIDE
Dubbo
E
SYDNEY
CANBERRA
D SYDNEY
CANBERRA
ZONE A ZONE C MELBOURNE
MELBOURNE
ZONE B ZONE D B
A A
HOBART HOBART
(e) (f)
A DARWIN
C DARWIN
Cairns Cairns
Broome Broome
B Townsville Townsville
Roma Roma
BRISBANE BRISBANE
Kalgoorlie Kalgoorlie
Coffs Harbour Coffs Harbour
PERTH PERTH
Dubbo Dubbo
E ADELAIDE ADELAIDE
SYDNEY
CANBERRA
C SYDNEY
CANBERRA
ZONE B
HOBART HOBART
Fig. 4.1: Hazard zones for Australian durability attack mechanisms [68]: (a) in-ground decay;
(b) above-ground decay; (c) termite attack; (d) marine borer attack; (e) atmospheric corrosion;
(f) embedded corrosion (courtesy of CSIRO)
48 CHAPTER 4. DURABILITY DESIGN CONCEPTS
situations. More specifically, the requirement is for models that enable calculation of the
strength and stiffness that structural elements are expected to have after a certain time in ser-
vice, or how long non-structural elements will remain functional for intended purposes under
defined threats and particular applications. Such models have been developed for timber con-
struction in Australia [67,68]. Those models were developed to predict decay from fungi,
marine borers and termites, and corrosion of steel parts. Using the resulting attack models, a
draft engineering design code has been developed [69–71]. The draft Australian design code
relates structural load capacity to target design life and gives design procedures intended to
generate design solutions with comparable levels of structural reliability under all hazard situ-
ations. Figure 4.1 shows the hazard zones of Australia associated with various attack mecha-
nisms. As structural engineers will recognize, the approach is in principle similar to mapping
of expected design wind speeds or seismic ground accelerations. Obviously, the principles
developed in Australia could be applicable to other parts of the world or to non-timber materi-
als, but so far that has not been done. However, by analogy engineers should be able to make
extrapolated judgements applicable elsewhere.
Design lives of non-structural and non-critical structural elements reflect how often they are
intended to be repaired or replaced, and there is no reason to match lifespans of such elements to
one another or to design lives for critical structural elements. The distinction made here speaks to
providing designers with options for life cycle cost optimization and not to circumstances under
which they might avail themselves of such an option. As already indicated, critical structural
elements must have design lifetimes associated with actual expectations of building lifespans
based on reasonable expectations of lifelong maintenance and evolution in building occupan-
cies. There is of course uncertainty in durability design, especially whether proper maintenance
will be done, but the situation here is in principle no different than for structural and fire design
which presumes that buildings will not be altered in ways that compromise the original design
intent.
4.3.2 Decay
At temperatures between 0°C and 60°C decay fungi attacks can progress rapidly when the
timber has a moisture content above the fibre saturation point (≥ 30%) [72]. Therefore, such
conditions should never be deliberately created internally or externally. The existence of mois-
ture conditions in timber that sustain decay fungi are always the result of bad initial design,
poor workmanship, inappropriate building renovation methods, unintended uses of buildings,
and/or improper building maintenance. Excessive moisture in timber is often associated with
4.3 ATTACK MECHANISMS 49
Decay pocket
Gap Decay
pocket
Decay
Impermeable skin pocket
(d) (e)
Wood sucks up water Masonry
from soil via footing wet by rain
Dry air
Sandstone
foundation Moist
soil Wood absorbs moisture
from masonry
Fig. 4.2: Avoidable mechanisms for moisture accumulation in traditional timber low-rise con-
struction: (a) direct wetting by rain; (b) water ingress via surface check; (c) water ingress via
construction joint; (d) moisture wicking via stone foundation; (e) moisture wicking via masonry
contact (courtesy of CSIRO)
capillary action (also known as wicking) and/or water entrapment with water coming from the
ground, rain, or leaks within buildings. Thus, decay prevention must focus on practices that keep
timber dry.
Figure 4.2 illustrates mechanisms by which moisture typically enters structural and non-struc-
tural elements of traditional timber structures. These examples illustrate that in all cases the
problems result from poor or shoddy practices. Site investigations of buildings always reveal
that best practices were not adopted during at least one stage in either design or maintenance.
Under proper practices, wetting of timbers should never occur except under rare or accidental
events such as damage to glazing, spillage, or leaks.
This SED does not recommend impregnation of timber with chemical preservatives as a primary
means of mitigating risk of decay in critical structural elements, because such treatments have
not been proven reliable over the long term. Instead, the recommended practice is to employ
durable design strategies at system and element levels that eliminate the hazards. In practice,
this recommendation translates into designing the building superstructure (and foundation,
where applicable) systems so that water is drained away from timber members and to ensure
that intentional spaces are ventilated. Elements should be designed to have geometries that
do not entrap moisture. This does not mean that exposing non-structural or non-critical struc-
tural elements to risk of decay is not an acceptable design decision. However, any replaceable
50 CHAPTER 4. DURABILITY DESIGN CONCEPTS
elements that are allowed to decay should not be directly attached to critical structural elements,
should be separated from them by impermeable barriers (e.g. vapour membranes), and should
be intentionally selected as sacrificial.
Models exist for estimating service lifespans of elements based on measures such as the time
taken for non-structural elements to develop a certain depth of surface decay under a particular
external service use, or the time for a certain depth of surface decay of non-critical structural
elements in contact with the ground [73]. These should be used where appropriate.
4.3.3 Termites
Termite attacks are possible in all ecosystems where the mean annual temperature is greater
than 10°C, with speed of travel and ferocity of attacks tending to accelerate with any increase
in temperature. Boundaries of termite infestations can advance at around 1 or 2 km per year.
Therefore, maps showing areas of infestation are only rough indicators of whether a termite
hazard might exist within the design life of a specific building (Fig. 4.1c). Accurate estimates of
possible termite arrival dates can be made using termite attack models and up-to-date infestation
data as discussed by Leicester et al. [74].
Although unlikely, termite attack may exist where the mean external temperature is below 10°C.
Termites have spread northward in both North America and Europe, because they are sometimes
transported along with traded commodities (e.g. in packing cases). For example, viable colonies
exist in the core areas of several Canadian cities because metro systems and other underground
spaces such as pedestrian walkways provide year-round conduits for termite movement under
sufficiently warm conditions [75]. A rarer possibility also exists that termites get blown by wind
onto the roofs of high-rise buildings during their alatory phase. They then establish colonies
where rain water has penetrated building envelopes, as has been observed in Hawaii and else-
where [76].
4.3.4 Corrosion
Corrosion attacks unprotected metal fasteners inserted into timber in two ways (Fig. 4.3). In the
first, airborne salt attacks exposed parts of fasteners. In the second, the acidity of any impreg-
nated chemicals for biological or fire protection and natural acidity of timber itself become
agents of attack on the embedded parts of fasteners. In the case of joints with bolts fitted in
tolerance holes, the corrosion rate of shanks is enhanced by any salt and water that accumulate
within oversized holes (Fig. 4.3c). This type of attack may be mitigated by protection mea-
sures such as covering a shank with a plastic sleeve, or using tight-fitting dowels instead of
loose-fitting bolts. Notably, using tight-fitting dowels, instead of bolts in tolerance holes also
has structural benefits in terms of greater stiffness and strength. Practical models are available
for estimation of rates at which exposed and embedded parts of steel fasteners in timber will
corrode [69,77,78].
The corrosion rate of embedded fasteners tends to be quite low, unless the timber has been
treated with a mineral-based chemical, such as alkaline copper quaternary (ACQ), or if the
timber has high moisture content [79]. For the case of atmospheric corrosion, the rate for mild
steel can be very high for buildings located within 1 km of saltwater (Fig. 4.4). Corrosion rates
4.4 DESIGN STRATEGIES 51
(a)
Atmospheric
corrosion Embedded
corrosion
(b)
(N2)
Corrosion on
bolt shank
(N1) cbolt
Oversized hole
NAIL
Fig. 4.3: Types of fastener corrosion: (a) nail corrosion; (b) bolt shank corrosion (courtesy of
CSIRO)
Situations exist where the prohibition of timber and wood-based composites as critical structural
elements of large and tall buildings is appropriate. This occurs, for example, when the risk of
fungal or insect attack cannot be acceptably mitigated via technical measures. Otherwise, risk
suppression against temporary consequences of rare events (e.g. cyclonic and seismic event
damage, floods) that partially destroy normal protection of timber critical structural elements
can sometimes be attained by impregnating such members with low dosages of preservative
treatment chemicals. Such products are always proprietary in nature and manufacturers provide
data on water resistance and durability capabilities of each type and brand; in the case of some
modern engineered wood-based composites, the materials have sufficient inherent capability to
repel water and preserve themselves. Irrespective of other measures, when buildings are con-
structed in locations where risk of termite attack exists low-level preservative treatments should
be applied to critical structural elements. The H2 preservative loading described in the Austra-
lian Standard AS 1604.1 [63], or similar, will typically be a suitable treatment. Use of timber
classified as naturally durable is insufficient, because of the potential presence of sapwood,
which is never reliably durable.
4.5 CALCULATION OF ENGINEERED SERVICE LIFE 53
The realizable benefits from taking great care during design and construction of the structurally
critical parts of tall timber structures are easily demonstrated by the longevity of monumen-
tal timber structures (Fig. 1.1) that have remained serviceable for many centuries with only
repair to their non-structurally essential fabrics. A constant factor in longevity is that apart from
occasional temporary wetting timbers in structures have remained dry since installation [2].
As discussed in Sections 4.3.4 and 4.4.2, embedded corrosion of steel fasteners can be rapid if
timber is inadvertently wetted. Therefore, given the potential consequences it is recommended
that stainless steel fasteners, and if appropriate other stainless steel parts, be used in conjunction
with timber and wood-based composite elements that are critical to the structural integrity of
buildings.
In that document, more than 100 Australian timber species are considered based on groups clas-
sified according to their resistance to in-ground decay, above-ground decay, marine borer attack,
and wood acidity. Chemical treatment may also be considered based on standard classifications
given in Australian Standard 1604.1 [63]. Local climate effects are accounted for based on the
hazard zones presented in Fig. 4.1 and consideration of other locality-specific parameters. For
example, in assessing the degradation of metal fasteners because of atmospheric corrosion, the
design hazard also takes into account the distance from the ocean coast, the coastal exposure,
the type of terrain in which the building is embedded, and the proximity of any industrial plants
producing potentially corrosive emissions.
Not all provisions of the draft Australian service life design code will be appropriate for initial
material choices and selection of maintenance and repair strategies for non-structural and non-
critical structural timber and wood-based composite elements. However, those provisions are
the most appropriate existing framework for bringing such decisions into the quantitative arena
of engineering practice.
Consider a 400 mm diameter pole of untreated hardwood from durability Class 2 (second best of
four classes) from which the sapwood has been removed and which is located in a region of Haz-
ard Class B (Fig. 4.1a). For this case, the attack is assumed to comprise a simple perimeter and
core decay pattern, as shown in Fig. 4.6a. Also shown in that figure are the computed mean values
of perimeter and core decay. The mean residual strength ratio of the pole after any chosen service
(a) (b)
60 1.5
Perimeter
Core
Relative load capacity
Decay depth (mm)
20 0.5
0 0
0 10 20 30 40 50 0 10 20 30 40 50
Time (years) Time (years)
Fig. 4.6: Decay and bending load capacity of 400 mm diameter hardwood pole, Class 2, located
in Hazard Zone B: (a) mean decay depths (b) relative load capacity (courtesy of CSIRO)
4.5 CALCULATION OF ENGINEERED SERVICE LIFE 55
Table 4.1 Design service life for 400 mm diameter hardwood pole (courtesy of CSIRO)
life is then calculated as the ratio of the elastic section modulus of the residual cross section to the
initial elastic section modulus, resulting in the relationship shown in Fig. 4.6b. Computation of
the design bending strength capacity proceeds in a similar manner, except that the mean depth of
decay d is replaced by an effective depth deff that takes into account uncertainties associated with
the prediction of the loss of cross section. This effective decay depth is given by Eq. 4.1:
and used properly timber elements can have design life expectancies consistent with needs of
building owners and occupiers. The technologies are in many cases well known and have been
applied successfully over many centuries, as testified by surviving buildings from various ages
and eras. Additional contemporary measures add to the available tools but do not alter the valid-
ity underlying applicable philosophy that building elements must be shielded directly from
durability risks such as termite attack, corrosion, and fungi.
Emerging modern durability design practices, such as those now existing in New Zealand and
Australia, for timber construction are formalizations of good practices. Although quantitative
engineering calculations are at their core, it is important to be cognizant during design that
desired service lives of non-critical structural and non-structural elements may not be achieved
through passive resistance alone to attacks by various agents. Provision of passive resistance
to some types of attacks on individual elements must be inextricably married to the selection
of an appropriate whole-building design concept. Whole-building concept practices need to be
ones that do not expose critical structural elements to aggressive service environments. In cases
where termite attack is possible, active whole-building protection is required for all parts of the
building fabric, rather than passive resistance termite attack. As readers who have consulted
Chapter 3 will realize, strong similarities exist between fire and durability design in the sense
that both are fundamentally about risk mitigation by protecting vulnerable elements from unde-
sired exposure, or where that is not possible making sure that sufficient sacrificial material exists
such that critical elements will retain functionality during and after any “attack”.
Acknowledgements
This chapter incorporates background information, figures, and diagrams provided by Dr. Robert
H. Leicester and Dr. Greg C. Foliente of the Commonwealth Scientific and Industrial Research
Organization (CSIRO). This does not mean that what is discussed herein matches practices
adopted in Australia or mirrors opinions of those contributors.
57
Chapter
5
Timber Frameworks with Rigid
Diaphragms: Special Considerations
Summary: This chapter discusses the structural performance issues associated with design and
construction of multi-storey building superstructures having frameworks constructed from rela-
tively large linear timber elements and “rigid” floor and roof diaphragm slabs. Emphasis is on
situations where superstructures have six or more storeys, with the primary function of the tim-
ber frameworks being to directly resist the effects of gravity loads. The rigid diaphragms trans-
fer forces associated with wind or seismic loadings to supplementary shear walls/building cores
made from material such as reinforced concrete (RC). Irrespective of other factors, composite
action always exists between skeletons and “other parts” of buildings. All parts must be aligned
together to make the total system work structurally and to prevent non-structural parts from
becoming detached under various loading scenarios. Therefore, although timber may form the
bulk of framework systems, it is in fact the connections that mostly define the structural perform-
ance characteristics of what are commonly called timber-framed buildings. Much discussion
herein concerns selection and design of connection methods. From the structural engineering
perspective, connections are arguably more important when using timber than when using other
framing materials to sympathetically meld selection of the system level concept with the capa-
bilities of the building elements. Hence this chapter concentrates on utilizing timber elements
in ways that maximize the overall structural efficiency of systems in normal and overloaded
situations. This recognizes that although timber design codes and other technical support tools
provide appropriate element sizing information, the skill of individual engineers is what results
in elegant solutions that transcend mere functionality.
5.1 Introduction
What is presented in this chapter is consistent with the structural design principle outlined in
Chapter 2. This means, for example, that it is intended that high-rise superstructure systems
having timber frameworks that act compositely with rigid diaphragms (of timber or other mate-
rial) should be designed and constructed in ways that ensure that primary elements within them
will not exceed their elastic response ranges under design level events. Although important
lessons have been learned from the behaviours of low-rise timber frameworks and high-rise
frameworks constructed from other materials, it is crucial to recognize that unique circum-
stances apply to design of high-rise timber frameworks. In particular, special attention needs to
be paid to selection of structural connection methods and marrying that to choice of structural
58 CHAPTER 5. TIMBER FRAMEWORKS WITH RIGID DIAPHRAGMS: SPECIAL CONSIDERATIONS
arrangements. Focus herein also places emphasis on the proper use of information contained
within contemporary timber design codes.
Traditional timber framed buildings are relatively flexible with respect to global response. Also
flexibility exists at subsystem levels. Furthermore, such systems have the ability to adjust the
geometry of their substructures during seismic events [14,87]. Relative to other structural sys-
tems there can be high levels of embodied structural damping, reflecting that the building fabric
is often not monolithic and can consume energy via frictional and other contacts. However, the
ability to achieve this is not always realized, because it requires careful and deliberate attention
to construction detailing. Lateral bracing of low-rise frameworks is normally achievable without
the need to make frameworks act compositely with shear walls. This is in contrast to tall timber
structures. However, various factors related to the intensities of design loads usually combine so
that structural systems can normally be very simple, with well-defined flow paths for resisting
gravity and lateral forces (Fig. 5.2).
In modern times, glued laminated timber (glulam) elements have been the most common type
of framing element, reflecting the lack of reliable supplies of new, large, sawn timbers in most
parts of the world. Length restrictions for glulam members depend mostly on transportation
methods and on-site lifting capabilities, which means that limits on possible dimensions match
restrictions for prefabricated steel and RC elements. As illustrated in Fig. 5.3, employing con-
tinuous girders and continuous columns in various configurations is technically possible and
often done. Importantly, continuity of primary framing members is employed to reduce moment
peaks and deflections within the girders and column members, but for reasons explained below
not to transfer moment forces between beams and columns.
60 CHAPTER 5. TIMBER FRAMEWORKS WITH RIGID DIAPHRAGMS: SPECIAL CONSIDERATIONS
Grider to column pin connection Column to foundation pin connection Wall bracing
(a) (b)
Fig. 5.4: Logic of how timber/glulam framework elements should interface with connections:
(a) bearing transfers of gravity forces; (b) axial transfers of tensile forces resulting from wind
and seismic forces
In the design and construction of timber heavy-frame systems, special attention must be given to
the connections between framework elements and for those between the superstructure frame-
work and the foundations. Fundamentally, choices for interconnecting system parts are car-
pentry joints, connections made using metal parts and metal fasteners (with and without local
reinforcement of timber members), hybrid carpentry-metal part connections, glued joints, and
hybrid glue-metal part connections. Figure 5.4 shows preferred types for timber frameworks.
Decisions about connection methods are always intrinsically linked to the overall structural
design strategy. Three-dimensional timber frameworks are not usually intended to work in an
analogous manner to most three-dimensional RC and structural steel frameworks. As shown in
Figs. 5.2 and 5.3, arrangements of timber columns and girders are not usually self-stabilizing
(i.e. they need to be supported by temporary bracing during erection and then permanently
braced). An arrangement of floor and roof diaphragms in conjunction with diagonal bracing in
wall planes is the normal approach to stabilizing systems. Wall bracing is normally widely dis-
persed across bay locations in timber superstructure frameworks. Bays that are not directly sta-
bilized must be anchored to those that are. Modern multi-storey buildings of any height nearly
always have elevators and stair shaft walls and/or fire walls with significant shear wall capabili-
ties. In contrast, timber low-rise superstructures are normally not designed to work compos-
itely with shear walls (i.e. the two types of substructures are not structurally interconnected).
5.3 MODERN RENAISSANCE TALL TIMBER FRAME SYSTEMS (CIRCA 20–80 M TALL) 61
Low-rise timber buildings that are large in plan are often designed so that each fire compartment
is an independent structure, with each fire compartment deriving its three-dimensional stabil-
ity from geometry and an independent bracing system. Such practices yield buildings that are
inherently robust and not prone to disproportionate spread of damage whatever the cause of the
damage. However, when large building volumes derive from them having significant height, the
same strategy cannot be followed.
Present timber design codes are written predicated on the design of traditional low-rise build-
ings. Therefore as usage of timber transitions to also include construction of high-rise build-
ings, it becomes prudent to consider whether codes still adequately reflect needs of designers.
When this Structural Engineering Document (SED) was written Canadian code committees, for
example, had commenced the task of reviewing relevance of provisions in their national build-
ing and timber design codes.
1. Timber frameworks must employ individual elements in a manner that sensibly approxi-
mates ways in which timber is designed by nature to perform efficiently.
2. Tall timber frameworks function most efficiently when working compositely with other
substructures to jointly resist effects of lateral loads.
The need to respect these concepts was recognized by architects and structural engineers in the
Victorian and Edwardian eras, and many of their structures are still in use today [14,87,88].
What is discussed further herein is based on the concept of using timber frameworks to resist
the primary effects of gravity loads, whereas systems constructed from RC (or other relatively
massive material) resist the primary effects of lateral loads.
Tree stems (and therefore sawn timbers and substitutes such as glulam) are inherently strong and
able to resist very large compressive, tensile, and bending forces that stress the material parallel
to the grain (i.e. parallel to axes of tree stems). Yet such products are quite weak in situations
that stress the material perpendicular to the grain. Utilizing large timber members efficiently
revolve on using them as simple beams, simple columns, simple ties, beam-columns, and beam-
ties. Maximizing height capabilities of timber gravity load-resisting frameworks depends on
maximizing capacities of vertical members in compression (because they are weaker in com-
pression than tension). That requires loading vertical members as closely to concentrically as
possible. Horizontal members in gravity load-resisting frameworks must resist bending and
associated shear forces. For overall efficiency of the total system, horizontal members will also
have to (at least partially) participate in the transfer of forces associated with lateral loads on
62 CHAPTER 5. TIMBER FRAMEWORKS WITH RIGID DIAPHRAGMS: SPECIAL CONSIDERATIONS
the buildings to the lateral load-resisting system. Normally horizontal members act as beam-
ties or beam-columns, but not because of any intent to create moment framework action as in
RC or steel frameworks. Of course, this does not mean that vertical, or in some instances also
horizontal, members cannot be continuous or semi-continuous, as illustrated in Fig. 5.3. How-
ever, framework connections need not be complex, and well-established traditional and modern
mechanical connection technologies can be employed.
Anchoring of timber frameworks to building cores, and when appropriate direct anchoring of
massive timber floor and roof diaphragms to RC or other structural building cores, is a critical
consideration. As with all connections involving timber, careful attention should be paid to
5.3 MODERN RENAISSANCE TALL TIMBER FRAME SYSTEMS (CIRCA 20–80 M TALL) 63
(a) (b)
(c) (d)
Fig. 5.6: Illustration of a suitable overall structural scheme for a modern renaissance tall
timber frame system: (a) foundation and RC building core; (b) timber framework; (c) horizontal
diaphragms; (d) assembled primary structural system
avoid connections that transfer loads other than thrusts and shear forces. Depending on con-
struction practices and sequencing, attention may also need to be directed towards accom-
modation of differential movements between timber and other materials. Shrinkage related to
moisture movements and drying in timber and curing of concrete can be significant in some
instances, as can temperature-related movements. Thus, these will need to be accommodated.
As in other structural engineering situations the “golden rules” are to design connections so that
they accommodate, rather than resist differential movements related to contraction or dilation
of components and to avoid solutions that grip members too close to their ends or edges. This
should, however, not be taken to imply that the difficulties associated with effectively tying
parts together are any greater with timber than any other materials.
Consistent with general principles of structural engineering, large and tall buildings employ-
ing timber as a primary structural material should be designed to handle effects of even the
most extreme loads in the elastic range (Recommendation 3). Doing that is the first line of
defence in achieving solutions that are robust and reliable. This need not entail loss of mate-
rial or economic efficiency. The consequence is in fact mostly that architects and engineers
will have to apply their skills in new ways (i.e. differently from how they conceptualize the
design of low-rise timber superstructures). Designing systems to respond elastically should
not be the only defence against damaging events. Avoiding the possibility of disproportion-
ate damage or catastrophic collapse of systems such as those illustrated in Fig. 5.6 requires
taking deliberate measures. Column failure presents the greatest danger of widespread dam-
age. There exist a number of viable ways of ensuring systems have the robustness to contain
64 CHAPTER 5. TIMBER FRAMEWORKS WITH RIGID DIAPHRAGMS: SPECIAL CONSIDERATIONS
damage, if for instance a column was removed because of a vehicular impact or because of
an explosion. Use of continuous horizontal beam/girder elements in the gravity load system
and continuity in floor and rood diaphragms can provide the possibility of alternative load
paths, if such damage were to occur. Practical implementation of this approach can be based
on ensuring that potentially critical elements would not exceed their ultimate capacities were
an adjacent parallel component(s) removed. Such calculations can be based on non-partial,
factored design loads for structural systems to avoid design solutions being unreasonably con-
servative. The determination of how many columns could be removed and survival be main-
tained should depend on considerations such as the class of building occupancy and physical
proximity to hazard sources. Another illustrative robustness and damage containment strategy
is to provide supplementary lateral bracing systems (located within timber frameworks) that
are located in regions of the building judged most potentially critical under disaster sce-
narios. Secondary triangulated bracing that is permanent or would be activated when required
could also be installed (e.g. at lower storey corners). The approaches suggested in this chapter
essentially replicate what has been followed for many centuries [14,87]. However, in situa-
tions where timber skeletons braced in the vertical plane act compositely with other substruc-
tures to resist lateral wind or seismic loads, full dynamic analyses of the entire hybrid system
is the preferred approach.
(a) (b)
Fig. 5.7: Installation of glulam superstructure framework segments (further details in Chapter
8): (a) two-storey segment; (b) interconnection of segments
Traditional connection methods that have proven successful all adhere to these simple prin-
ciples (Fig. 5.4). To be effective, modern connection methods also need to do so. In large
superstructure systems one must consider how elements will be manufactured and erected.
As illustrated by example projects in Chapters 8 and 10, timber/glulam framework elements
of large multi-storey superstructures are nearly always manufactured off-site in factory con-
ditions. This includes the exact, final element shaping and the cutting and drilling associ-
ated with connections. Similarly, metal connections are nearly always manufactured off-site.
Therefore, for building superstructures of types discussed in this chapter, the elements are
received as parts intended to fit together exactly. The situation is in no way akin to traditional
timber construction practices, where elements were typically manufactured from raw materi-
als such as pieces of lumber. In modern practice, large timber frameworks are often partially
preassembled before being lifted into place, as illustrated in Fig. 5.7. In principle, on-site
assembly practices are similar to those that apply to structural steel and precast RC element
frameworks.
Joints made using generic mechanical fasteners are pervasive in modern timber construction.
The term mechanical fastener encompasses all devices that are inserted into two abutting mem-
bers in ways that result in transferred forces flowing across joint interfaces via those devices.
Most generic mechanical fasteners have been mass-produced since the 19th or early 20th centu-
ries, with wire nails, steel bolts, and plain steel dowels being the most common. The popularity
of joints using generic fasteners reflects several factors including that the fasteners are typically
inexpensive and easy to install, and that joint strength can be reliably predicted using easily
accessed information (i.e. in codes and handbooks). In the context of large frameworks, joints
66 CHAPTER 5. TIMBER FRAMEWORKS WITH RIGID DIAPHRAGMS: SPECIAL CONSIDERATIONS
(a)
Wood
Forces on
gripping
element
Resistance
from wood
(bearing + skin
friction)
Gripping
element
(b)
(i) (ii)
Wood
Steel dowel
(embedded)
Fig. 5.8: Joints with laterally loaded fasteners: (a) principle of gripping action; (b) classifica-
tion of mechanical joints with metal parts: (i) exposed metal, (ii) concealed metal
in connections are typically made using laterally loaded steel dowels and bolts. As indicated in
Fig. 5.8, joints made with generic dowel fasteners often interconnect timber elements with metal
link elements that bridge connections, with the metal parts being at the hub. The slenderness of
metal dowel fasteners normally determines the flexibility and degree of ductility of joints and
connections employing them. Reliance on skin friction around dowel fasteners for primary force
resistance is something structural engineers usually try to avoid because of its relative lack of
reliability after timber members have dried or dried and then been rewetted. Therefore, with the
exception of bolts, plain dowels, and some types of screws, emphasis tends to be on using dowel
fasteners in situations where force transfer does not involve the possibility of pulling them out
of or through joined materials. In the context of structural frameworks formed from large tim-
bers, all mentioned types of generic fasteners have been employed. However, experience sug-
gests that using plain metal dowels and metal bolts are most appropriate and economical when
large forces need to be transferred.
There exists extensive literature on design and construction of joints employing generic dowel
fasteners. Nearly all general textbooks and design aids on timber engineering devote substan-
tial attention to the appropriate use of various fasteners and provide sample design solutions
[92,93]. This stated, designers who intend system responses to remain in the elastic range should
verify the response state associated with design capacities for recommended dowel fastener
connections [93].
5.4 EFFECTIVE CONNECTION METHODS 67
Glued-in rods
Steel connecting assemblies
Glulam members
Fig. 5.9: Illustration of glued-in rod connections suitable for multi-storey framework applications
Many types of proprietary joint/connection hardware are highly suitable for making connec-
tions in timber framework elements, for attaching diaphragms to frameworks, and for attaching
frameworks to shear walls and foundations. In all cases, necessary design information about
joint/connection stiffness, strength, and failure mechanisms is outside the scope of design codes
and must be obtained from the specific manufacturer.
Glued-in, rod joint systems involve the embedding of metal or other types of reinforcing bars
into structural elements. This creates local continuity akin to connections in RC frameworks, or
diffusely transfers forces into or from ends of members to other parts of connections (e.g. roll-
ers, pins, and swivels). Reinforcing rods are inserted into predrilled oversized holes in glulam or
other large timber members with gap-filling adhesives or grouts employed to bond the rods. For
economy, rods are normally of the types employed in RC construction processes. Considerable
research has been devoted to developing methods in Scandinavia, Russia, New Zealand, and
Canada. Most of the research has focused on the behaviour of joints/connections under static
and seismic loads for situations where glulam columns and girders intersect or where columns
connect to foundations. As expected, certain connection configurations, details, and fabrica-
tion methods are superior to others. Generally, it is preferable that reinforcing bars be installed
into timber members off-site under controlled factory conditions and then fastened to hub-like
connection devices by on-site bolting to create arrangements such as that illustrated in Fig. 5.9.
The illustrated arrangement would achieve a self-bracing frame action and steel connecting
assemblies could be designed to yield in desired ways, if the system was overloaded. Using
glued-in rod approaches can create force flows consistent with almost any desired structural
system behaviour [91].
A discussion of standard timber jointing methods would not be complete without mentioning
glued joints. Rigid adhesives are employed under factory conditions to create intra-component
joints in glulam members, multilayered timber-based composites such as laminated veneer lum-
ber and CLT, box-girders, and stressed-skin panels. On-site gluing of large timbers has been
practiced on a limited basis to create long members for low-rise construction and conceptu-
ally such technology could be applied to construction of relatively tall multi-storey framed
superstructures. However, the technology has not been proven suitable for construction of super-
structures of types that are the focus of this chapter, and hence, is not recommended at this point
of time.
68 CHAPTER 5. TIMBER FRAMEWORKS WITH RIGID DIAPHRAGMS: SPECIAL CONSIDERATIONS
(a) (b)
Surface or embedded
energy dissipation elements
(absorb energy via plastic
work during load cycling)
Glulam beam
(c)
(i) M (ii) M (iii) Elastic
M
Plastic
θ θ θ
Elastic
Plastic
Fig. 5.10: Logic of self-centring frame connections, based on Buchanan et al [16]: (a) compo-
nents of a frame connection; (b) deformed connection during seismic action; (c) logic of flag-
shaped hysteresis loops (moment M versus rotation q: (i) effect of post-tensioning, (ii) effect of
energy dissipation elements, (iii) combined response: flag-shaped hysteresis loop
On a final note, researchers from New Zealand and Italy have been performing experiments and
undertaking numerical modelling to define self-centring connection technologies suitable for
timber heavy-framed building superstructures. Focus is on creating frameworks that efficiently
resist the effects of seismic actions [16], based on adaptation of what have become fairly common
practices for post-tensioned prefabricated RC frameworks [94]. The principle is that by suitably
arranging post-tensioning tendons and sacrificial metal energy absorbing elements (that deform
plastically, if overloaded) it is possible to create a self position-restoring system response with-
out sustaining more than minimal damage (Fig. 5.10). Essentially the post-tensioning forces that
do not exceed the elastic range of the tendons pull the system back into its original position when
external forces are removed. The energy dissipation elements can undergo plastic deformation
while absorbing considerable amounts of system-level inertial work. If properly designed, the
dissipation elements can be easily replaced during post-disaster repair work. Buchanan et al.
[16] have also developed post-tensioning connection systems suitable for minimizing damage
in timber buildings containing massive wall panels. They claim that their connection methods
are suitable for constructing building superstructures with 10 or more storeys in high seismic
risk locations. No explicit proof exists yet of satisfactory field performance of post-tensioning
5.5 ADDITIONAL COMMENTS 69
connection systems, but it can reasonably be presumed that they will work well, because what
is done replicates in principle the approaches to seismic design employed in traditional bearing
wall and timber infill-frame buildings (Figs. 2.3a and 2.3b). The modern approach is to achieve
via post-tensioning and energy absorbing elements what was formerly achieved by overburden
pressures, frictional damping, and carpentry details.
Acknowledgements
This chapter draws on input from Drs. Mohammad A.H. Mohammad, Marjan Popovski, and
Chun Ni and Mr. Sylvain Gagnon of FPInnovations, Canada.
71
Chapter
6
Steel or Reinforced Concrete
Frameworks with Timber Diaphragms:
Special Considerations
Summary: This chapter discusses the structural performance issues associated with employ-
ing relatively lightweight timber floors and roof slabs in high-rise buildings, as substitutes for
reinforced concrete (RC) slabs. The focus is specifically on hybrid construction systems where
structural steel or RC framework substructures are the primary means of resisting effects of
gravity forces associated with the occupied building space. RC building core substructures
are the primary means of resisting effects of lateral forces associated with wind or seismic
loading. Use of massive engineered timber composites known generically as cross laminated
timber (CLT) is discussed as a substitution for RC slabs. CLT is approximately mechanically
equivalent to RC slabs of equal thickness, but with the advantage of being only around one
third of the weight. The lightness of CLT reduces demands on the structural framework and
foundation substructures with respect to both gravitational and lateral loads. These advan-
tages are the greatest when CLT is used in combination with structural steel frameworks,
because the proportional reduction in total mass is much greater than with RC frameworks.
In practical terms, the advantage of using CLT instead of RC slabs is that either the structural
framework and foundation materials can be significantly reduced or the structural perform-
ance of the entire superstructure system can be improved without extra cost. The design per-
formance of a 24-storey building with alternative steel and RC framework substructures is
used to illustrate the typical magnitude of structural gains associated with substituting CLT
for RC slabs.
6.1 Introduction
The discussion below addresses the design of superstructure systems having massive
timber floor and roof slabs, supported by steel or RC frameworks that act compositely with
RC building cores. Specifically, such systems employ massive timber slabs as horizontal
diaphragms at all floor levels and at the roof level or combine them with some RC fire
floors. By employing a 24-storey (96 m) building, a comparison is made between using
massive timber floor slabs and using normal weight RC slabs, so that readers can obtain
a sense of the relative structural performance of alternative systems under various design
load scenarios.
72 CHAPTER 6. STEEL OR REINFORCED CONCRETE FRAMEWORKS WITH TIMBER DIAPHRAGMS
(a) (b)
Fig. 6.1: Massive timber plate elements: (a) PLT; (b) CLT
Shear walls
8 × 3.2 m
X 6 × 6.4 m
(b)
Framework sections
Steel Columns 1st-8th: W14 × 500
framework* 9th-16th: W14 × 342
17th-24th: W14 × 159
Main beams W24 × 94
Secondary W16 × 49
beams
24 × 4 m = 96 m
RC Columns 1st-12th:
framework** 800 mm × 1200 mm
13th-24th:
600 mm × 800 mm
Main beams 350 mm × 650 mm
Secondary 200 mm × 400 mm
Z
beams
X * AISC sections [95]
38 ** Dimensioned based on ACI [96]
.4
m m
.6
25 Y
Fig. 6.2: Twenty-four storey building: (a) typical floor layout; (b) three-dimensional system
model
optimized for the application. Here the notations CLTf and RCf are used to distinguish sys-
tems embodying CLT and RC floor/roof slabs, respectively. Firef is used to denote mixed
CLT–RC slab situations where most floors have CLT slabs but every fourth floor has a RC
slab. This arrangement achieves a total floor area between adjacent fire floors consistent
with accepted sizes of fire compartments within low-rise timber buildings (Chapter 3). As
discussed by Smith and Frangi [20], this approach applies when regulatory authorities will
not permit fully CLTf systems.
74 CHAPTER 6. STEEL OR REINFORCED CONCRETE FRAMEWORKS WITH TIMBER DIAPHRAGMS
(a) (b)
Cast in-situ concrete t =150 mm (floor), CLT plate t =190 mm (floor),
110 mm (roof) 112 mm (roof)
Shear studs
Screw fasteners
Corrugated steel
permanent formwork
Fig. 6.3: Floor systems used with steel frameworks: (a) RCf (RC floor) system; (b) CLTf (CLT
floor) system
(a) (b)
Steel column
Steel girders
Fig. 6.4: Typical CLT slab arrangement and connection methods: (a) CLT panel layout over
supporting steel floor framing; (b) schematic CLT connections
The load combinations considered included the effects of gravity (self-weight and imposed floor
and roof) loads, seismic excitation (peak ground acceleration 0.5 g), and wind forces (basic
wind speed 240 km/hour). Details of the load data and factor combination are drawn from
contemporary US Load and Resistance Factor Design (LRFD) practices, with details given in
Ref. [89]. The load magnitudes and combinations used are not for a particular site. Instead, they
approximate a worst-case wind and seismic design scenario for continental North America. This
approach gives an indication of how buildings of the type studied herein are expected to perform
in general, rather than at a specific construction site, for the alternative CLTf, RCf, and Firef
systems. Members in steel and RC system frameworks were the same irrespective of floor slab
6.3 TWENTY-FOUR-STOREY CASE STUDIES 75
type and corresponded to the necessary minimum dimensions for systems with only RC slabs
(RCf systems, Fig. 6.2). Thus, the results show only the effects of altering the slab type, and in
the case of CLTf systems introducing flexibility to the slab connections.
Fully three-dimensional finite element (FE) models using SAP2000 [97] predicted structural
response. Modelling approaches were directly equivalent, except with respect to the physical
and mechanical properties for CLTf and RCf systems. Floor and roof slab components were
modelled using four-node shell elements. Framework members were modelled with two-
node frame elements having six degrees of freedom per node (i.e. three translations and three
rotations). Table 6.1 summarizes the material properties used in the analyses. Link elements
represented flexibilities of inter-element connections between consecutive CLT slab elements
and between them and other parts. A linear dynamic analysis based on the response spectrum
method was used to apply seismic displacements at the base of the buildings (Recommendations
2–4 and 6 from Chapter 2).
RC building cores were modelled as four-node shell elements having thickness from 250 to
300 mm. In the analysis, it was assumed that CLT slabs were connected to frames with trans-
lational link elements spaced at 800 mm and having stiffness of 8 kN/mm for forces parallel
to the slab planes and 10 kN/mm for forces normal to the slab planes. The stiffness values
were estimated based on connection tests done in the work of Ref. [98]. Abutting edges of the
CLT panels interlock via lap joints fastened using long wood screws to ensure slab continuity
as shown in Fig. 6.4b. Both for the RCf and CLTf systems, the connections between the steel
skeleton and shear wall cores were modelled as rigid [95]. Pinned joints were used to simulate
the connections between the RC or CLT slabs and the RC cores. In the case of the CLT slabs,
this corresponded to fastening them to the RC using light steel angles, whereas for RC slabs it
corresponded to continuous steel bar links that run from the slabs to the cores (no heavy reinfor-
cing anchorages). The column frames were modelled as rigidly fixed at the foundation level.
The foundations were excluded from the analyses. All parts of superstructures were assumed to
respond as crack-free bodies.
76 CHAPTER 6. STEEL OR REINFORCED CONCRETE FRAMEWORKS WITH TIMBER DIAPHRAGMS
Fig. 6.5: The most deformed frame in steel or RC systems during seismic event
With the steel frameworks, the total weight of the RCf system was 115.9 MN, and the total
weight of the all CLTf system was about 50% of that (57.2 MN). The maximum gravity load
floor diaphragm deflections (i.e. effects of dead, superimposed plus live loads) were 23.1 mm
and 19.0 mm, respectively for the RCf and CLTf systems (i.e. less than span/300). Peak stress
levels in bending were far below tolerable levels, thereby indicating acceptability of floor dia-
phragms as distribution elements for floor loads. The superior performance of the CLTf system
was a direct consequence of the comparative lightweight of the CLT slab relative to that of the
RC slab (Table 6.1). Significant benefits also accrued with respect to reduced impacts of the
gravity force demands on frameworks, cores, and foundations for the CLTf systems. However,
the lightness of the CLT slabs yielded the largest benefits when the effects of the lateral loads
were analysed.
From that analysis, the critical lateral load for the RCf and CLTf systems was as a result of the
combination of gravity and earthquake loads. Figure 6.5 shows the typical deformed shape of
the frame under the critical earthquake load (30%-X + 100%-Y), and Fig. 6.6 shows the pre-
dicted peak lateral displacements (drifts) at each storey, along with the associated inter-storey
drifts [30%-X indicates 30% of the peak ground acceleration applied at the building base in
the X-direction (parallel to the long plan axis), and 100%-Y indicates 100% of peak ground
acceleration applied in the Y-direction (normal to the long plan axis)]. At the roof level, the
predicted drift of the RCf system was 240 mm, and 149 mm (about 40% less) for the CLTf
6.3 TWENTY-FOUR-STOREY CASE STUDIES 77
(a) (b)
24 24
18 18
Storey number
Storey number
RCf
12 12 CLTf
RCf Firef
CLTf
6 Firef 6
0 0
0 50 100 150 200 250 300 0 3 6 9 12 15
Drift (mm) Interstorey drift (mm)
Fig. 6.6: Peak drift and inter-storey drift in steel frameworks during seismic event: (a) lateral
displacement; (b) inter-storey drift
system. The maximum inter-storey drift was 12.0 mm for the RCf system, and 7.9 mm for the
CLTf system, both occurring in the 17th storey. In both cases the inter-storey drift was less than
0.5% of the storey height, thereby achieving a performance level typically required of a tall
building intended to be fully functional after a major earthquake [17,99]. Using CLT as floor
slabs components in the tall building in lieu of RC slabs was advantageous in terms of total
system response and mitigation of the potential for damage. As would be expected, the peak lat-
eral and inter-storey drift of the Firef system was intermediate to the matching responses of the
CLTf and RCf systems (Fig. 6.6). However, it should be noted that the vertical mass irregularity
between floors in the Firef system causes irregularities in mode shapes. Given that the analysed
case represents a building with simple overall geometry and construction, the results shown here
exemplify why simplified assumptions about design loadings and structural responses (e.g. as
embedded in equivalent static load analyses of superstructures) are not accurate or reliable in
general, and can be the source of unexpected damage or other manifestations of poor building
performance.
Peak horizontal shearing forces carried by fasteners around the perimeters of CLT slab panels
were about 15 kN/m. This equates to using screws of around 10 to 12 mm diameter at a spacing
of 200 mm. The largest forces on the fasteners occurred under load combinations involving seis-
mic excitation of the fifth floor slab, which is where the predicted torsional accelerations were
the greatest. Analyses compared the response of a fully CLTf building having realistic, semi-
flexible connections with the response of an otherwise similar building having slabs rigidly
connected together and rigidly connected to the framework. The maximum difference in shear
flows in the slab to the framework connections was less than 4%, indicating that neglecting con-
nection flexibilities does not necessarily result in serious errors. Reductions in computing effort
can be significant by assuming rigid connections (as is often done when designing steel framed
buildings), particularly when the buildings are large and have complex shapes and framing lay-
outs. Notably, the slab connection force demands need to be calculated by accounting for forces
developed at nodal points linking specific CLT shell and steel frame elements together.
78 CHAPTER 6. STEEL OR REINFORCED CONCRETE FRAMEWORKS WITH TIMBER DIAPHRAGMS
Fig. 6.7: Mode shapes and periods for steel frameworks (illustrated mode shapes are for CLTf
case)
In the dynamic analysis underpinning this discussion of buildings with steel frameworks, 24
mode shapes were extracted to ensure that modal participation factors reached 90% (i.e. to
ensure they were valid for modal combination analysis) [20]. Figure 6.7 illustrates the first
three mode shapes and associated structural periods for RCf, CLTf, and Firef systems. As the
modal frequencies in the table embedded in the figure show, the lower order frequencies were
not widely separated, thereby implying the strong possibility of motion amplification during
normal service events such as low-intensity winds and abnormal service events such as hurri-
canes. No significant differences in mode shapes were predicted for the three systems. However,
this should not be construed as implying that the time history responses will match in shape or
that the peak drift profiles will be similar in “roughness”. As expected, given that modal stiff-
ness values were derived essentially from the RC building core and steel framework, the lateral
predicted structural periods were distinctly higher for the heavier systems, (t1 = 1.96, 2.29, and
2.90 seconds for CLTf, Firef, and RCf systems, respectively). The first mode shape was similar
to a simple cantilever deflection, whereas the second and third modes had strong torsion com-
ponents. The mode shapes of the quite simple case study buildings and the possibility of modal
clustering are further indications that even relatively simple, tall, hybrid structures can exhibit
complexities in responses and need to be analysed with care.
Further analysis not reported in this chapter shows that structural demands on the steel skeleton are
always much less in CLTf systems than in comparable RCf systems, irrespective of the design load
combinations. On the basis of simple maximum sums of stress ratio criteria (i.e. Σ design stress
to design strength ratios), “sum values” are 20% to 30% higher for framework members in RCf
systems than in CLTf systems [89]. Although also not presented in this chapter, steel reinforcement
demands in the RC building cores and foundations will tend to be minimized by CLTf systems.
In a broad sense, much of the discussion of RC framework systems parallels the discussion
related to steel framework systems. The discussion in this chapter is, therefore, focussed on the
consequences of different construction details and the heavier weight of RC frameworks.
6.3 TWENTY-FOUR-STOREY CASE STUDIES 79
Steel angle
and fasteners RC beam
RC beam
Fig. 6.8: Floor system used with RC framework: (a) RCf; (b) CLTf
(a) (b)
24 24
20 20
16 16
Storey number
Storey number
RCf
12 12
CLTf
RCf Firef
8 8
CLTf
Firef
4 4
0 0
0 50 100 150 200 250 300 0 3 6 9 12 15
Drift (mm) Interstorey drift (mm)
Fig. 6.9: Peak drift and inter-storey drift in RC frameworks during seismic event: (a) lateral
displacement; (b) inter-storey drift
Figure 6.8 gives detailed schematics of the construction for RCf and CLTf systems involv-
ing RC frameworks and RC cores. Dimensions of framework members are given within the
information box in Fig. 6.2. RCf floors slabs are cast monolithically with horizontal frame-
work members (Fig. 6.8a), whereas CLT slab panels are mechanically attached using flexible
fasteners in ways closely paralleling practices for steel frameworks (Fig. 6.8b). In practice, it
is impossible to make completely rigid connections/to rigidly bond timber members to other
materials using mechanical devices, even with adhesives because of the shear lag through the
member depths. Also, long-standing experience teaches that structural timber substructures
do not function well if the local ability of parts to “adjust themselves” during service is con-
strained excessively. The FE modelling details for the CLT slabs and associated connections
were the same as in the steel framework system analyses.
The total weight of the RCf system was 166.1 MN, whereas the total weight of the CLTf
system was about 35% less (108.4 MN). Maximum gravity load floor diaphragm deflections
80 CHAPTER 6. STEEL OR REINFORCED CONCRETE FRAMEWORKS WITH TIMBER DIAPHRAGMS
System First mode (seconds) Second mode (seconds) Third mode (seconds)
CLTf 2.67 2.47 1.74
Firef 2.87 2.59 1.86
RCf 3.25 2.87 2.10
Table 6.2: First three periods for RC framed systems
(i.e. effects of dead, superimposed plus live loads) were 20.1 mm and 18.0 mm, respectively
for RCf and CLTf systems. Notably, those values are slightly lower than for the steel framed
systems reflecting mostly the greater torsional rigidities of the RC framework members. Con-
clusions match those from prior discussion regarding acceptability of CLT slabs for tall hybrid
construction. Figure 6.9 shows predicted peak lateral displacements and inter-storey drifts due
to the most critical load combination involving seismic excitation. Critical locations are once
again along the outer plan edges (as shown in Fig. 6.5). Predicted drift at the roof level for
the RCf system is 274 mm, and 228 mm (i.e. 17% less) for the CLTf system. The maximum
inter-storey drift was 13.5 mm for the RCf system occurring in the 13th storey, and 11.5 mm
occurring in the 16th storey for the CLTf system. In both instances, the levels of inter-storey
drift were less than that typically permitted (around 0.5% of the height). As in the case of steel
framework systems, using CLT slabs as floor components in tall building in lieu of RC slabs
was advantageous in terms of the total system response, as a direct reflection of lower modal
masses. Magnitudes of predicted movements were not problematic in this particular instance,
irrespective of slab type. However, that the storey level that will govern any aspect of a sys-
tem’s design cannot be known in advance is another indicator of the pitfalls that could beset
those who attempt to oversimplify calculations that support design. As with steel framework
systems, demands on the RC frameworks are always lower under other loading scenarios too,
for the same reason. The difference between the situations is, however, the gains are propor-
tionally smaller when frameworks themselves are relatively heavy. Arguably, design choices
of framing and other construction materials should be market driven. However, the greater the
amount of vibrating mass involved, the harder it is to achieve economy without risking the
possibility of damage during an extreme event. Approximately 10% more reinforcing steel
is estimated to be required in the superstructure employing the only RCf solution. The Firef
system, as expected, had a seismic response lying between the other systems with matching
practical design implications.
Delving more deeply into the dynamic responses of RC framed systems’ is informative.
The first three mode shapes are similar to those illustrated in Fig. 6.7, with the associated
natural periods given in Table 6.2. Cantilever action dominates the fundamental mode shape
for all of the RCf, CLTf, and Firef systems, but their responses can be classified differently.
According to practices in the Canadian National Building Code [17] only the RCf system
can be considered to be responding like a very slender structure. The predicted fundamental
period of 3.25 seconds nearly matches the upper limit of 3.45 seconds in that code, for cases
that can be considered to be RC moment frames in simplified analysis. This indicates that
it is highly feasible that hybrid design situations (involving the use of timber as a primary
structural material) will lie on the margins of where simplified design practices are allowed
by codes.
6.4 GENERAL IMPLICATIONS OF USING CLT SLABS 81
Quite high stresses in CLT components are predicted at the 23rd floor level for the deadweight,
live load plus earthquake load combination. Even then, the stresses generated were much less
than typical CLT strength properties (Table 6.1). Peak forces in the CLT slab to framework
connections occurred in the shear around the slab perimeter of the fifth floor. However, simple
construction details like the metal angle connection in Fig. 6.8b were found adequate based on
data in the literature [100,101]. As when modelling steel framework systems, modelling CLTf
system connections as rigid did not significantly alter the total structural response (i.e. drift,
mode shapes, and modal periods). However, care should once again be taken regarding extrapo-
lation of this observation.
d1 d2
dmid
100% EQ-x
x
30% EQ-y
Fig. 6.10: Deformed shape of a horizontal diaphragm: Note: davg is the average storey drift or
(d1 + d2)/2, and dm is the mid-span deflection
82 CHAPTER 6. STEEL OR REINFORCED CONCRETE FRAMEWORKS WITH TIMBER DIAPHRAGMS
of the general level of flexibility of diaphragms of the types discussed is handled typically as
part of the overall structural concept. For example, flexibility can be controlled through provi-
sion of adequate frames/chords at the building perimeter to resist the flexural action and provi-
sion of collector frames capable of transmitting lateral forces to vertical lateral force resisting
systems. Mention of diaphragm flexibility here is simply to indicate that it need not be radically
different from familiar RC slab approaches. In the “short span” direction, diaphragm flexibilities
were in the order of 1.1 for both CLTf and RCf systems (i.e. quite rigid).
Analyses of case study systems were performed to determine the impact of simply assuming
diaphragms are rigid in-plane, instead of accounting for their actual flexibility. In such sim-
plified analysis, there was no need to account for details of either the diaphragms or framing
components to which they are directly attached. All that was required was to consider two trans-
lational and one rotational degrees of freedom of a master node at each storey. Weight (mass) of
a particular storey was lumped at its master node. Other elements at each level were modelled
to deform horizontally with the same magnitude and direction as the master degrees of freedom.
When this approach was employed, results indicated significant differences in predictions of
drift and periods (some larger than 100%) between flexible (actual properties) and rigid models.
This was especially pronounced for RC framework systems. The CLTf system with a RC frame-
work exhibited largest differences, relative to the corresponding RCf system. Notably, this was
in part because of the inability of the simplified analysis to incorporate the flexibility of connec-
tions between CLT plates and the RC building core. As a broad finding of the particular com-
parative analyses, rigid diaphragm response assumptions yield results similar to more detailed
modelling in terms of the overall structural response when the structural frameworks act alone
to provide the rigidity of a relatively tall building via moment frame action. However, when such
frameworks act compositely with RC tower-like building cores and shear walls, the simplified
approach can be unreliable. Although far from being specific to the instances discussed here
[103], it is prudent to reemphasize that making simplifying assumptions about behaviours of
timber diaphragms may or may not yield accurate results. Thus, extrapolation should always be
undertaken with all possible care. The only robustly reliable approach is detailed analysis of a
system, combined when appropriate with full dynamic analysis.
This Structural Engineering Document (SED) intentionally omits any discussion of Load-Delta
(i.e. P-D) or other secondary effects, because these are applicable broadly to any structural
analysis. The premise herein is that engineers will be cognizant of the generic issues that domin-
ate the practices of structural engineering.
the construction of tall hybrid buildings. This is the reason why it is repeated in this chapter and
emphasized elsewhere in this SED that analysis, design, and construction practices that elimin-
ate the possibility of doubts about the reliability of proposed designs must be adopted. This is
essential as engineers shun the adoption of structural analysis and situation-specific design code
rules of unproven robustness. Obviously, they should apply only fundamental structural engin-
eering concepts whenever in doubt. Here dynamic response analysis of the case study systems
is a metaphor illustrating the need for this prudent approach. Modern engineers have readily
available analysis tools and know-how gained over more than a century of practical learning
with other materials and computational developments of the last about half century to support
them in this. Or more precisely, what they need is to apply these skills and tools to explore the
use of timber.
The case study analyses presented demonstrate very clearly that timber, and potentially other
lightweight materials, can occupy niche applications that exploit their mechanical characteris-
tics and ancillary features with high efficiency. From a construction viewpoint, the exploited
attributes of CLT are essentially the same as those widely taken advantage of to construct large
buildings from ancient to Edwardian times (Chapters 1 and 5). CLT panels are lightweight,
mechanically efficient, “hold” fasteners well, and at the time of writing are becoming widely
available in world markets as commodity or specialty products. Hence, their discussion is an
example, as opposed to being prescriptive or exclusionary.
Acknowledgements
The analysis on which this chapter is based was done by Professor Andi Asiz, Prince Mohammad
Bin Fahd University, Kingdom of Saudi Arabia.
85
Chapter
7
Platform Construction Using Timber
Plates: Special Considerations
Summary: Since around the beginning of the 21st century there has been strong and growing
interest in using timber plates as substitutes for reinforced concrete (RC) and masonry elements
in medium-rise building superstructures in European countries like Austria, Germany, Italy,
Sweden, and the UK. The concept has also begun to spread to regions like North America and
The Antipodes. The generally preferred types of timber plates are known collectively as cross
laminated timber (CLT or XLAM). They consist of layers of relatively small rectangular lumber
arranged such that the axes of these pieces “cross-reinforce” pieces in other layers. There are
many proprietary CLT products available in thicknesses that range from around 50 to 500 mm.
Their layers are bonded together using rigid adhesives or aluminium nails. Some products have
stiffness and strength properties comparable with those of normal-weight RC slabs of equal
thickness but at only one quarter of the RC mass. Apart from low mass to mechanical property
ratios, the growing popularity of CLT products reflects the ease in cutting and shaping to form
complex geometries and that prefabricated elements can be rapidly and easily joined together
using simple methods. When encapsulated in non-combustible materials, CLT elements perform
excellently and predictably during building fires. This chapter emphasizes the use of CLT in
residential construction based on the so-called platform construction method. In such build-
ings, CLT panels are interconnected using simple mechanical fasteners like long slender screws
and simple anchor brackets and ties. This chapter also discusses the key aspects of structural
analysis and design of 40 to 50 m high buildings.
7.1 Introduction
Since the millennium, there has been strong and growing European interest in CLT or XLAM as
superstructure elements in low-rise and medium-rise buildings. More recently interest has spread
to places as diverse as North America and The Antipodes. Such products are manufactured as
large plates/panels having three or more layers of finger-jointed structural grade softwood lum-
ber, with layer thicknesses ranging from 17 to 38 mm (Fig. 7.1). Lumber pieces in some layers
are arranged orthogonally to pieces in other layers, such that when layers are bonded together
using rigid adhesives or aluminium nails, the composite arrangement is cross-reinforced in
all directions in the same manner as plywood. Thus, when loaded in-plane, CLT panels are
toughened against splitting in a manner that eliminates structural design problems associated
with more traditional timber products. Depending on the layering employed, CLT plates can be
86 CHAPTER 7. PLATFORM CONSTRUCTION USING TIMBER PLATES: SPECIAL CONSIDERATIONS
Also, gaps initially existent between pieces of lumber in layers, or those that occur because
of shrinkage, result in some degree of permeability. This permeability has important implica-
tions for fire performance with respect to the integrity requirement that hot gases could not
escape from the non-exposed side of an assembly during a fire resistance rating period (Sec-
tion 3.5). Consequently, encapsulation of CLT plates that form fire separation elements using
non-combustible material is necessary. Similarly, layers of other materials are often added
for thermal and sound insulation, such that CLT is the structural layer of composite wall and
floor plates.
Adhesively laminated CLT products are not intended for use in wet service conditions and
should only be used in service environments such as Service Classes 1 or 2 as defined by Euro-
code 5 [4]:
Service Class 1 corresponds to average moisture contents in most softwood not exceeding 12%,
and Service Class 2 corresponds to average moisture contents in most softwood not exceeding
20%. For most multi-storey buildings Service Class 1 is applicable.
To date, there is no generic production standard for CLT manufactured in Europe and all prod-
ucts manufactured there are proprietary. However, a production standard has been developed in
North America [104], which acknowledges that products can be generic or proprietary. To date,
the only design properties that exist are ones recommended by manufacturers or third-party
technical organizations (e.g. Canadian industry led R&D organization FPInnovations, USA-
based APA – The Engineered Wood Association). In European Union countries, design proper-
ties are published by third-party notified European authorities, in accordance with procedures
that apply to all member states.
From a structural engineering perspective, the crucial characteristic of CLT is that the cross-
lamination of the lumber in layers reinforces it against splitting. This means that at any point of
time, material in some of the layers is resisting the stiff and strong in-plane force flows in the
parallel to grain orientation. Therefore, unlike most other structural products made from timber,
it is possible to make reliable tension and shear connections using mechanical fasteners installed
through the thickness, irrespective of the in-plane loading direction. This does not negate the
need to select connection methods and fasteners carefully, but it does translate into possibilities
of using timber in novel ways (Section 7.5).
vary depending on the orientations that plates will have when installed in buildings (Fig.
7.2). Consistent with materials like plywood and oriented strand board (OSB), properties are
specified according to the planes in which stresses are expected to be applied. Table 7.1 shows
5th-percentile characteristic strength and stiffness properties for representative European CLT
products. Strength classes C24 and C16 denoted in that table correspond to grades of lumber
from which European CLT is often manufactured [105]. Properties in Table 7.1 are intended
to be used in conjunction with the Eurocode 5 timber design code [4]. This means that char-
acteristic strength properties are adjusted during design for combined effects of load duration
and service class according to Eurocode 5. The importance of this is that the values in the
table are not directly equivalent to CLT design properties adopted elsewhere. For example,
it can be assumed roughly that reference strength properties applicable in North America are
80% of those in Table 7.1. Irrespective of that, it is however valid to assume that similar thick-
nesses of CLT are required in similar situations in various countries when sizing is controlled
by similar design event levels (e.g. similar peak seismic ground acceleration and similar peak
wind speed).
Strength class of
Property timber in laminates
(MPa) C16 C24
Strength properties (5-percentile characteristic values based on short-term load)
Bending fm,k 16 24
Tension parallel to grain in face layers ft,0,k 10 14
Tension perpendicular to grain in face layers ft,90,k 0.4 0.4
Compression parallel to grain in face layers fc,0,k 17 21
Compression perpendicular to grain in face layers fc,90,k 2.2 2.5
Shear in the plane of the plate fv,k 1.8 2.5
Shear through thickness of the plate fr,k 0.7 0.7
Stiffness properties (mean values)
Modulus of elasticity in bending parallel to grain in face layers E0,mean 8000 11,000
Modulus of elasticity in bending perpendicular to grain in E90,mean 270 370
face layers
Shear in the plane of the plate Gv,mean 500 690
Shear through thickness of the plate Gr,mean 50 50
Table 7.1 Example of strength and stiffness properties for proprietary European CLT products
directly on a RC foundation or on top of one or more RC plinth storeys. At the time of writing,
the tallest building was reported in Victoria Harbour, Melbourne, Australia: the 10-storey
Forte Apartments building, of which nine storeys are made of CLT. The approach is generally
economically and technically suitable in situations where room sizes are quite small, which
means that walls are numerous and floor spans are limited. In such cases, plate thicknesses
need not be large. Most of these buildings are for residential or mixed commercial and resi-
dential occupancies. Wall and floor elements arrive on-site, accurately precut to their final
dimensions, with the cutting normally done by computer numerical control (CNC) machines.
Sometimes windows, doors, and other non-structural items are preinstalled. Total construc-
tion periods are often very short once the foundation is ready, and site equipment required is
not very specialized.
In cases where rooms are relatively large, panels sometimes have preinstalled timber stiffening
ribs and double-skins, to economically increase feasible spans/lengths (Fig. 7.5). When double-
skin wall or floor panels are used, it is necessary to adopt construction details that prevent unde-
tected fires to spread through the void areas (Section 10.4).
Screws
at joints
between
walls
1. 2.
3. 4.
5. 6.
...the construction is
then completed very
quickly
7.
Fig. 7.3: Assembly of a typical timber plate superstructure employing the platform construction
method (LVL = laminated veener lumber)
1. High rigidity of the panels prevents development of significant in-plane warping at abutting
panel edges, which means that most panels move relative to the joint interfaces through
rigid body translation and rotation.
2. Line contacts that exist where plates abut are long, making it possible to create sufficiently
rigid and strong connections without concentrating the internal force flows that might frac-
ture CLT or foundation elements.
7.4 CONNECTION METHODS 91
3. The general rigidities of each storey are sufficient such that even when hold-down connec-
tions between walls in adjacent storeys cannot be directly aligned or when wall openings are
not the same at each storey, the superstructures respond as stacked rigid layers. This means
that such systems are normally very robust and not typically prone to so-called soft-storey
(a) problems.
Experience to date is that slender screws (usually proprietary self-tapping types) having diam-
eters 8 to 12 mm are suitable for line connections between CLT panels, with such screws being
often available in lengths up to at least 600 mm. Screw installation can be achieved without pre-
drilling in most softwood CLT products. Such screws efficiently transfer forces either laterally
or parallel to their axes. Sometimes self-tapping screws are used to locally strengthen panels in
compression, by inserting them as reinforcement akin to steel reinforcement in concrete. Such
92 CHAPTER 7. PLATFORM CONSTRUCTION USING TIMBER PLATES: SPECIAL CONSIDERATIONS
A C
Fig. 7.6: Examples of metal fasteners, ties, and connectors (A = shear connector; B = anchor
tie, anchor bolt not shown; C = fasteners for attaching shear connectors/ties to CLT; D = self-
tapping metal screws)
reinforcement can be used to locally strengthen floor plates against crushing under pressures
beneath supported walls, as an economical alternative to increasing wall thicknesses.
The assumption here is that such buildings are designed according to contemporary Partial
Coefficient Limiting States Design practices, such as those specified in the model codes of the
European Committee for Standardization [4,18,106–108]. Consistency with such codes requires
that linear elastic response analysis be used to estimate internal forces in components that are
the resulting effects of individual or combined design load cases; and that characteristic strength
and stiffness properties of the CLT are apparent values used in conjunction with geometric prop-
erties determined from gross-dimensions of elements (Section 2.1). For example, the second
moment of area to be used in combination with E0,mean or E90,mean from Table 7.1 is the element
width multiplied by h3/12, and the section modulus used in combination with fm,k from the same
table is the element width multiplied by h2/6 (where h is the panel thickness).
Load paths in timber plate superstructures are quite transparent. Vertical force transfers occur
directly from floors and roof plates to plates that form walls, with force in walls accumulating
7.5 STRUCTURAL ANALYSIS AND DESIGN 93
through the heights of superstructures. According to the platform construction method, horizon-
tal force transfers also occur cumulatively from level to level through the heights of superstruc-
tures. Two important aspects to consider with respect to the load flows between storeys are the
rigidity of those storeys in plan and in elevation. Once assembled, unless plan shapes are irregular
or very elongated, storeys within superstructures like those in Figs. 7.3 and 7.4 are quite rigid
both in plan and in elevation. This means that floor platforms tend to behave as rigid, rather than
as flexible diaphragms. The simplified approaches outlined in Subsections 7.5.1.3 and 7.5.1.4
for designing floors and walls are based on not making specific presumptions about the rigidity
of horizontal diaphragms and adopts the worst interpretation of how forces may flow from floor
platforms to walls. The question of how to consider the behaviour of walls and the effect of their
behaviour on rigidities of complete storeys is more complicated. In-plane rigidities of individual
wall segments, and therefore complete walls and assembles of walls, depend on the following:
Often, it is conservatively assumed that walls or segments within them are not stiffened or
strengthened by transverse walls. Two extreme cases are considered. In the first, walls are cut
from single pieces of CLT (Fig. 7.7). In the second, they are assembled from CLT plate segments
(Fig. 7.3). These two scenarios can result in distinctly different behaviour mechanisms in terms
of distortion, and therefore flexibility and internal force flows within walls subjected to rack-
ing forces that result from lateral design loads on superstructures. Figure 7.8 illustrates the two
extreme cases. Intermediate cases will also exist depending on factors such as whether or not
edges of abutting CLT panels in segmented walls are mechanically interconnected. Therefore,
when performing wall design (as discussed in Subsection 7.5.1.4) the influences of construction
details on the racking deformation mechanism and strength of CLT panels and their perimeter
connections must be considered.
Top CLT
CLT wall panel connection
Base CLT
connection
Floor platform
Top CLT
CLT wall panel CLT wall panel connection
Base CLT
connection
Floor platform
Fig. 7.8: Racking behaviour of CLT walls: (a) one-piece wall; (b) segmented wall
formed by floor and wall plates promotes even redistribution of forces flowing from one sto-
rey to another. Therefore most, if not all, walls in the various storeys participate in resisting
effects of other than locally applied design loads. Superstructures like those in Figs. 7.3 and
7.4 contain much structural redundancy and forces would be redistributed relatively benignly
were particular structural elements to fail during any type of loading event. Consequently, the
general likelihood of progressive collapse is very limited, and timber plate systems such as the
ones discussed in this chapter are highly robust, if properly designed and constructed. The most
likely system-level vulnerabilities are connections that link storeys together, such as the roof
to the walls or the superstructure to the foundation. Therefore, close attention should be given
to selection and design of shear and hold-down connections resisting sliding and uplift due to
wind and seismic loads. Connections used at level-to-level interfaces should contain redundancy
and ductility. In the case of hold-down anchoring between elevated storeys, the flow of forces
should be whenever possible directly from wall to wall, rather than from walls to floor platforms
sandwiched between them.
Discussion here applies directly to situations where elevator and staircase shafts are constructed
from CLT or comparable timber panels. In such instances, no specific distinctions need to be
made for the purposes of structural analysis between wall elements forming those shafts and
other wall elements.
Floor plates are designed considering them to act as one-way spanning elements (i.e. ignoring
that panels are interconnected at abutting side edges, as seen in Fig. 7.9). This consideration
reflects that the edge-to-edge joints are made using simple carpentry lap joints fastened by
screws or just screws. Such joints are intended to transfer shear forces but not moment forces
that result from vertical loads on floors. This approach tends to be conservative. Most often, ser-
viceability performance-related bending deflection criteria control the required plate thickness.
The deflection criteria take the form of the maximum acceptable ratio of deflection to span and
are selected to avoid damage to non-structural elements of buildings (e.g. 1/500). In some cases
7.5 STRUCTURAL ANALYSIS AND DESIGN 95
Rolling-shear
M M
Rolling-shear
V V Traction
Fig. 7.10: Analogies used to represent internal stress distributions due to moment and shear
forces
such criteria are intended as indirect control of floor motions resulting from building use, like
effects of footfall impacts. However, neither static deflection nor simplified dynamic analyses
are robustly reliable ways of ensuring satisfactory vibration serviceability of CLT floors [109].
For the types of buildings discussed, installation of floating floors over CLT slabs is common
and is an effective solution to vibration serviceability problems.
When designing floor plates for strength it is necessary to consider both bending and shear
strength. Design strengths such as those in Table 7.1 take into account the layered nature of
plates (Fig. 7.10). Bending strengths are based on the assumption that only CLT laminations
with lumber oriented parallel to the plane in which bending moments occur resist forces. How-
ever, such complexity is ignored during normal design, because it is integrated into the apparent
design properties (Table 7.1).
The explicit design consideration required with respect to ensuring floors behave adequately as
horizontal diaphragms depends on factors like floor layouts and plate element connections as
discussed in Subsection 7.5.1.1.
Using CLT panels as walls provides interesting architectural options, because their in-plane
stiffness and strength allow them to resist gravity, uplift, and racking design forces; to span gaps
in facades; and to create overhanging storeys (Fig. 7.7). These applications have often been
beyond the capabilities of more traditional timber construction methods (e.g. light-frame, post
and beam) because of the requirement of large dimensions of elements, difficulties connecting
elements, and affiliated high costs. CLT walls can consist of panels with or without stiffening
ribs or as double-skin box elements with multiple glulam webs (Fig. 7.5). The use of stiffened
96 CHAPTER 7. PLATFORM CONSTRUCTION USING TIMBER PLATES: SPECIAL CONSIDERATIONS
plates usually permits economical construction of walls that can resist high vertical compression
forces at the lower storeys of tall and slender superstructures.
When superstructures are not structurally slender, it is often possible to use a number of quite
simple approaches to estimate the flow of the horizontal forces (i.e. wall racking forces) that are
caused by lateral design loads (i.e. equivalent static wind or seismic forces) to the walls. In cases
where floor plans and wall openings are replicated between storeys, and the extent of openings
is limited, buildings are essentially symmetrical about a vertical plane passing through the cen-
troid of the building’s footprint. In such cases, it is often adequate to:
• ignore the resistance of walls that are not parallel to the vertical plane in which the super-
structure’s response is being evaluated and
• assume that the floor platforms and roof behave as completely flexible diaphragms (pro-
jected load area method) or as perfectly rigid diaphragms (relative wall stiffness method),
and to take the worst outcome from those extreme case assumptions as the design force.
In other instances, like when systems are not essentially symmetric on plan but storeys are
replicated, the same approach can be taken to estimate components of horizontal force flows
to the walls. However account has to be taken of interactions between those forces and ones
that are associated with bending and shear distortion in planes coincident with a vertical plane
passing through the centroid of the building’s footprint. In such instances it is also necessary to
add components of horizontal force flows that occur because of torsional distortion about the
superstructure’s vertical axis (i.e. effects of torsion force due to plan eccentricities).
The approaches outlined above, or similar ones, will usually provide a sufficient basis for decid-
ing whether a building design concept is feasible, or if a more refined follow-up structural
analysis is necessary. When further investigation is necessary, a finite element analysis may
be required to represent CLT panels in walls and floors, the roof substructure and connections
between elements, and any other substructures in the superstructure and foundation. A refined
analysis will, for example, account for interactions between walls that do not lie in the same
plane (Fig. 7.11). In many instances a refined follow-up analysis will not be required, but it is
unwise to assume that is the case.
The combined effects of vertical and horizontal force flows for various design load combina-
tions (e.g. effects of factored dead and seismic loads) will determine the minimum required
dimensions of wall panels and connections. Unless a specific rule is mandated by locally appli-
cable design codes, use of the summation of the ratios of the factored force effects to the fac-
tored resistances approach is suggested:
Tf Rf
Tension + Racking: ___ + ___ ≤ 1.0 (7.1)
Tr Rr
Cf
( ) Rf
2
Compression + Racking: ___ + ___ ≤ 1.0 (7.2)
Cr Rr
where Tf is the factored tension force, Tr is the factored tension resistance, Cf is the factored
compressive force, Cr is the factored compressive resistance (accounting for either crushing or
buckling), Rf is the factored racking force, and Rr is the factored racking resistance.
7.5 STRUCTURAL ANALYSIS AND DESIGN 97
Seismic actions:
Realistic analogue: Effective tube
Foundation
Fig. 7.11: Interaction of wall panels directly in load paths with other walls to resist effects of
lateral loads
Screw connections:
1 wall panel-to-panel
2 floor panel-to-panel
4 9 1
3 3 wall corner
4 floor-to-wall junction
6 9 2 7 Hold-down connections:
4
4 5 wall-to-foundation
6 wall-to-floor-to-wall (using anchors)
3 7 wall-to-wall (using tie strap)
5 8 Shear connections:
8 wall-to-foundation
9 wall-to-floor
Fig. 7.12: Example locations and types of connections requiring structural design
The foregoing does not apply without modification to hybrid superstructures in which multi-
storey assemblies of CLT plates work in combination with substructures of other types. This
exclusion applies, for example, when a building has primary lateral load-resisting systems con-
structed from RC or reinforced masonry.
Figure 7.12 shows an illustrative example of connection types and locations in a timber plate
superstructure that require structural design.
Because suitable metal fasteners, connectors, and anchors are mostly proprietary products,
their design properties must usually be acquired from manufacturers or third-party technical
organizations that conducted tests on behalf of manufacturers. Consequently, engineers must
98 CHAPTER 7. PLATFORM CONSTRUCTION USING TIMBER PLATES: SPECIAL CONSIDERATIONS
(a)
A B C D E
(b)
F G H I J
Fig. 7.13: Selection of connections joining CLT plates: (a) inappropriate and appropriate wall
panel to wall panel connections (methods A and B are unsuitable, method C is viable but not
recommended, methods D and E are recommended); (b) inappropriate and appropriate floor
panel to wall panel connections (methods F and G are unsuitable, method H is viable but not
recommended, methods I and J are recommended)
assure themselves that the available information is consistent with the needs of specific design
projects and applicable design codes. This necessity will, for example, ensure that available
design information is based on the appropriate definition of capacity (in terms of the associ-
ated failure mechanism), deformation at failure, ductility ratio, and indexing effects of factors
like duration and loading-type. Data in the public domain mostly indicate that connections
in CLT made using slender self-tapping screws have high ductility when used to resist shear
flows between panels [110], but it is prudent to check that such behaviour apply to particular
product brands.
When installed in systems of the type discussed in this chapter, screws and other dowel-type
fasteners usually have design capacities that are higher than for products like glulam that is
manufactured from the same grade and species of lumber, because of the already mentioned
toughening that CLT possesses from cross-laminating of lumber.
Figure 7.13a shows the suitability of sample connectors for joining wall panels at building cor-
ners, and Fig. 7.13b shows the same for attaching floor or roof plates to tops of walls. In each
application, desirable connections minimize the likelihood of causing delamination type split-
ting of CLT (i.e. avoid wedging the layers apart).
where simply satisfying the requirement of preventing building collapse is insufficient. The fol-
lowing paragraphs discuss how well-designed and properly constructed CLT plate superstruc-
tures can be expected to perform during seismic events.
Because of the way in which CLT plate superstructures are assembled, there are many frictional
interactions between wall plates, and between plates in walls and plates in floor platforms.
Those interactions dissipate kinetic energy flowing through such systems. The frictional damp-
ing potential of the wall to floor platform contacts increases towards the ground, because of the
increasing overburden mass of the supported floors. Mechanical connections are also primary
energy sink sites, but their potential cannot be mobilized without damage (e.g. plastic yielding
of fasteners and/or crushing of CLT beneath them). Consequently the amount of damping super-
structure mobilization would be considerably greater than that affiliated with material damping
(which is widely reported to be in the region of 1% of viscous damping [15]). That CLT platform
construction has a high inherent damping was confirmed by full-scale shake-table testing of a
100 CHAPTER 7. PLATFORM CONSTRUCTION USING TIMBER PLATES: SPECIAL CONSIDERATIONS
Consequently, seismic design strategies and methods for timber plate superstructures must rec-
ognize the potential influences that construction site characteristics, structural form, construc-
tion material, construction details, and building occupancies have on the overall responses of
particular foundation and superstructure systems. An example of such a suitable design strategy
and method is presented in Section 7.6.
In general, superstructure responses can be determined through linear dynamic analysis by apply-
ing seismic displacements at the base of a building according to the response spectrum method.
In Italy, as elsewhere, provided some of the restrictions are met, design codes also allow the use
of simpler equivalent static force methods to estimate factored force flows in the elements of a
Seismic Force Resisting System (SFRS). In essence, the restrictions applicable to the simpler
method are that buildings must not be structurally slender and must have regular geometry plans
and elevations. To date, relatively tall multi-storey CLT plate superstructures constructed in Italy
and other seismically active locations have conformed to these restrictions, thus permitting the
use of equivalent static force methods of seismic design. This approach is discussed below.
7.6 EXAMPLE OF SEISMIC DESIGN PRACTICES 101
According to Eurocode 8 [18], the Factored Based Shear Force for a SFRS of a superstructure
having total weight W is:
where Sb(T1) is the ordinate of the design spectrum at the building period T1 (i.e. at the funda-
mental period of the SFRS) and taken to be:
2.5
Sb(T1) = agS ___ (7.4)
q
In these equations gI is the importance factor for the building (1.0 for residential buildings and
1.5 for strategic buildings), ag is the design ground acceleration assuming T1 equals 0.2 seconds,
S is the soil factor, and q is the behaviour factor. The assumption that T1 is 0.2 seconds corre-
sponds to the plateau of the European design spectrum [18]. Factor q is also known variously as
the q-Factor, action factor, or force modification factor. It has the function of reducing design
forces obtained from linear structural analyses to ones that account for capabilities of the SFRS
to dissipate energy (Section 2.6).
Apart from q, the variables in Eq. (7.3) are specified independently of the materials and type of
SFRS. Therefore, q is the quantity that requires to be quantified for CLT plate superstructures.
This was done based on the calibration of q such that the results of simplified equivalent static
force analyses match the results of detailed non-linear analysis studies and shake-table tests on
isolated wall assemblies and complete building superstructures. That calibration exercise indi-
cated that taking q as equal to 3.0 achieves acceptable design solutions, with values estimated
during detailed background studies falling in the range of 2.5 to 4.6 [15,113].
Level
South view 23.20 7
Storey 7
18.55 6
Storey 6
15.46 5
Storey 5
12.37 4
23.5 m
Storey 4
9.28 3
Storey 3
6.19 2
Storey 2
3.09 1
13.5 m Storey 1
7.5 m 0.00 0
OST
EG 1. OG - 6. OG
x z
13,44
NORD
13,44
SÜD
1,60
7,68 7,68
1,60
y WEST
records of relatively low intensity to detect whether the system’s response changed during “real”
earthquakes. The artificial tests indicated limited changes in the lowest order, natural frequen-
cies of the system, throughout the complete series of earthquakes. This finding was taken as
prima facie evidence of robustness and that any structural damage was localized. For the X-z
and Y-z planes (as defined in Fig. 7.15) the lowest natural frequencies were 2.34 Hz and 3.32 Hz
prior to application of earthquake records. After application of 10 earthquake records the fun-
damental natural frequencies had changed to 1.95 Hz and 2.93 Hz respectively. The system also
remained self-centring throughout (i.e. free from residual distortions). Application of seismic
events much stronger than those believed to be credible for Bolzano or other parts of Italy dem-
onstrates the practicality of designing and constructing systems capable of sustaining no more
than superficial damage during design level earthquakes (Recommendation 3 from Chapter 2).
S = 1.25, which corresponds to type B soil (very dense sand or very stiff clay),
q = 3.0.
The values of ag used here correspond to peak values used during shake-table tests (i.e. not those
applicable to design of the building in Bolzano).
The mode shapes for lowest order beam modes are assumed to result in linear increases in
horizontal displacement amplitude from the bottom to the top of the SFRS, thus resulting in the
horizontal equivalent lumped-static-forces values as follows:
zi mi
Fi = Fb 6
∑z m
j =1
j j
(7.6)
where i signifies the level of a lumped mass, and zi is the height of a lumped mass of magnitude
mi above the base of the SFRS. Figure 7.16 shows the lumped mass values, associated zi values,
and resulting individual and accumulated Fi values at various storeys. The accumulated Fi values
(∑Fi values) at any level are the shear flows to be resisted by the SFRS below each lumped mass
position.
As can be deduced from horizontal shear flows (∑Fi values) in Fig. 7.16, the largest structural
demand in terms of horizontal shear to be resisted occurs at the base of the SFRS, in the nar-
rower plane direction (X-z plane). This is a consequence of both the relative slenderness of the
building in that plane, which leads to the longest actual beam mode period and the relatively
low lengths of CLT walls that are available to act as shear walls in the X direction. Practices for
determining force flows to individual wall panels and connections in the lower and other storeys
follow the principle outlined in Subsection 7.5.1.4. Because floor plans were quite simple and
reasonably close to symmetric (Fig. 7.15), the design results were not very sensitive to the type
of force flow analysis used to convert Fi values into force flows in wall elements and their con-
nections.
Figure 7.17 shows the types of shear connectors and anchors used to resist sliding and over-
turning at elevated floor levels and at the base of the test system. The hardware used was a
mixture of commercially available products and specialty anchors. The building in Bolzano
has less heavy-duty connections, because the design peak ground acceleration is less. Detailed
analysis of test information suggests that connections in the tested SFRS were overdesigned
relative to actual demands on their capacities by around 30%. As is normal for timber build-
ing superstructures, the connections were primary in defining the behaviour of the SFRS,
and their selection and detailed design was crucial to obtaining and exceeding the desired
structural performance.
The maximum lateral drift observed during tests was in the order of H/80 for the X-z plane
and H/130 for the Y-z plane, where H was the total height of 23.5 m. However, as noted previ-
ously, the earthquake records used in these tests included 100% of the ground accelerations
observed during the actual earthquakes with one being the highly destructive Kobe earth-
quake on January 17, 1995. The maximum peak ground acceleration for Italy is only 0.35 g
104 CHAPTER 7. PLATFORM CONSTRUCTION USING TIMBER PLATES: SPECIAL CONSIDERATIONS
Fi Fi
Fi
Fi
Fi
Fi
Fi
Fi
Fi
Fi
Fi
Fi
(a) (b)
Level 7 Level 7
Fi Fi
Level 6 Level 6
Level 5 Fi Level 5 Fi
Level 4 Fi Level 4 Fi
Level 3 Fi Level 3 Fi
Level 2 Fi Level 2 Fi
Level 1 Fi Level 1 Fi
Level 0 Level 0
Level zi (m) mi (tonne) Fi (kN) Fi (kN) Level zi (m) mi (tonne) Fi (kN) Fi (kN)
6 18.55 20.1 488 488 6 18.55 20.1 357 357
5 15.46 48.1 974 1462 5 15.46 48.1 713 1070
4 12.37 51.3 831 2293 4 12.37 51.3 608 1678
3 9.28 53.5 650 2943 3 9.28 53.5 476 2154
2 6.19 55.4 449 3392 2 6.19 55.4 328 2482
1 3.01 61.2 247 3639 1 3.01 61.2 181 2663
Fig. 7.16: Lumped mi and associated zi values, and calculate Fi and ∑ Fi values: (a) X-z plane;
(b) Y-z plane
[113], and drift levels that might occur in a building like the one in Bolzano are in the order
of H/200 or less. Therefore, lateral drift levels would actually be within the range of target
limits typically specified by design codes. Drift predictions for the tested system obtained via
7.6 EXAMPLE OF SEISMIC DESIGN PRACTICES 105
IVALSA
hold-down
142 (d)
IVALSA
hold-down
Fig. 7.17: Connection hardware in SFRS test: (a) IVALSA hold-down anchor used at bottom
storey; (b) Simpson HTT22-hold-down used at elevated storeys; (c) shear connector used at
upper storeys; (d) shear connector used at bottom storey
relatively complex dynamic analysis were accurate, but a pushover analysis would also result
in a satisfactory SFRS. Pushover analyses should account for flexural and shear deformation
in CLT wall elements and any flexibility of connections that might lead to horizontal rela-
tive sliding or vertical opening (relative rotations of storeys because of uplift). Although not
elucidated in detail here, the design of this seven-storey SFRS’ elements was otherwise quite
straightforward.
Finally, the admissibility of equivalent static force design depends upon the estimation of T1.
Therefore, engineers are required to estimate the fundamental natural periods associated with
each design orientation (e.g. planes X-z and Y-z in the discussed building). Usually this cannot
be done with high accuracy, and opinions differ on acceptable methods. Consequently, given
that experience with design and performance of CLT plate superstructure is so far quite limited,
the surest fall-back approach is to base estimates on dynamic analysis. However, even if that
is done, the details of the SFRS will not be known at first, and a seed estimate of T1 is required
as the basis to begin iterative component sizing. For this, various empirical formulas have been
suggested, with an example being:
For the tested seven-storey system, this yields the estimate of T1 = 0.05 × 23.50.75 = 0.53 seconds
(fundamental frequency f1 = 1.9Hz). This compares with the actual measured T1 values of the
system prior to the application of a series of earthquake records of 0.43 and 0.30 seconds for
X-z and Y-z planes, respectively. Approximate formulas like Eq. (7.7) have limitations, and
engineers need to satisfy themselves concerning what depth of analysis is appropriate for par-
ticular projects.
106 CHAPTER 7. PLATFORM CONSTRUCTION USING TIMBER PLATES: SPECIAL CONSIDERATIONS
The rough rule of thumb is that for residential and similar building occupancies the minimum
wall thickness for three-storey systems is about 90 mm. Each additional storey requires a 10 mm
increase in wall thickness per storey (except for the top three floors). Therefore, a 14-storey all-
timber plate building superstructure would require 200 mm thick walls in the lowest storeys.
50 m
27
m
25 m
Vista 3D Seizone
Struttura in legno X-LAM Struttura in acciaio
Struttura in cemento armato
Fig. 7.18: Fifteen-storey CLT building concept designed for Northern Italy (Dante O. Benini &
Partners, Milano)
7.7 ADDITIONAL COMMENTS 107
The practical limit for the number of storeys or height for timber plate buildings cannot be stated
exactly. To date, buildings up to nine CLT plate superstructure storeys have been successfully
constructed in diverse locations like Berlin (Germany), London (UK), Milano (Italy), and
Melbourne (Australia). A 15-storey CLT plate superstructure design concept has been created
for Northern Italy (Fig. 7.18). However, as the cutaway schematic for that building design shows,
the building’s superstructure consists of a 13 storey CLT plate assembly that wraps around
a structural framework at the building’s core and sits on top of two above ground RC plinth
storeys. This reflects that, as has already been conjectured, using only timber plates becomes
impractical at around that height. Economic and other non-technical considerations suggest that
the maximum number of storeys that will ever be constructed from timber plates alone lies in
the range 12 to 15 (40 m and 50 m).
Acknowledgements
This chapter incorporates ideas, information, diagrams, and photographs supplied by Professor
Dr. Ario Ceccotti, who at the time of writing was the Director of the Italian National Research
Council Trees and Timber Institute (CNR-IVALSA).
Thanks are also due to Dr. Andrea Polastri of CNR-IVALSA who assisted with finalization of
this chapter.
109
Chapter
8
Example Project 1: Six-Storey Hybrid
Building in Quebec City, Canada
8.1 Background
There are many fairly tall historical buildings in Canada that have primary timber-frame super-
structure systems, with the tallest being the McLennan and McFeely Building in Vancouver
(Fig. 5.5). However, during most of the last century, building regulatory provisions explicitly
prohibited or limited the use of timber in multi-storey buildings other than very low-rise ones,
with many jurisdictions applying a limitation equivalent to four storeys. Publication of the
2005 edition of the National Building Code (NBC) of Canada recognized that “Acceptable
Solutions” and “Alternative Solutions” are suitable ways of designing buildings to achieve spe-
cific performance objectives [17]. The latter approach was introduced to facilitate and encour-
age the use of more technological innovations based on R&D, advanced engineering design
tools, and other demonstrably reliable approaches. In the remainder of this chapter, mention
of “accepted design” and “alternative design” are synonymous with the design classifications
Acceptable and Alternative Solutions as defined by the NBC of Canada. Under the Canadian
building regulatory system, the provisions of the NBC only become legally binding after they
are adopted by provincial or municipal governments. In some instances there are deviations
between the NBC model language and the actual building code requirements in a particular
jurisdiction, but such deviations are relatively minor. Although the NBC was revised in 2010,
110 CHAPTER 8. EXAMPLE PROJECT 1: SIX-STOREY HYBRID BUILDING IN QUEBEC CITY
those revisions had not been adopted in Quebec, and the discussed building was designed
according to the 2005 edition.
When a timber building is not permitted under Acceptable Solutions provisions, most often
it is because combustible materials are used as the load-bearing elements/assemblies within
superstructure systems. Architects and engineers who propose Alternative Solutions must, as
a minimum, demonstrate to building regulatory officials that the proposed solution will have a
performance that is at least equivalent to prescriptively accepted solutions. Building regulatory
authorities have the legal responsibility to analyse the affiliated technical documentation that
underpins each proposed alternative design and decide whether it meets the NBC objectives.
As discussed in Chapter 1, this is fostering a revival of interest in using timber and other new
or modified products (e.g. structural glass and plastics) for diverse construction purposes from
which Acceptable Solutions requirements had barred them previously.
The remainder of this chapter discusses the design and construction of a six-storey building
superstructure in Quebec City, Canada that was justified to the relevant regulatory authorities
based on a special design concept and the use of timber construction products. As with many
timber buildings, the main challenge was to address concerns of regulatory officials related
to fire performance objectives. Consequently, advanced fire modelling and performance-based
techniques became the main tools for demonstrating equivalency to accepted solutions. The
structural system was designed via Acceptable Solutions methods.
The building was constructed in stages, wherein the RC wall elements were cast in situ,
prefabricated glulam framework and diaphragm elements were added, and floors were com-
pleted structurally. Although floors were added and completed approximately in harmony
with the addition of the framework for different storeys, the construction sequence was coor-
dinated such that two-storey framework segments could be preassembled and then lifted into
place (Fig. 8.2). These segments are located in vertical planes parallel to the minor building
plan axis. Therefore, within the superstructure there is continuity of column and girder ele-
ments across some connections. The framework functions such that main floor and roof gird-
ers are connected to prefabricated frame segments. Once the system was completed, floors
functioned as continuous diaphragms. All glulam elements were cut precisely to length and
shape, and holes for dowel fasteners were drilled using computer numerical controlled (CNC)
8.2 SUPERSTRUCTURE SYSTEM 111
(a)
(b)
2m
20.
22 m
52.6
m
Fig. 8.1: Six-storey office building in Quebec City, Canada: (a) completed building (courtesy of
FPInnovations); (b) framework elements (courtesy of Nordic Engineered Wood)
machines. This resulted in less workmanship skill sets and was essential for ease and speed
of construction. Precise cutting of glulam elements avoided the possibility of distorting or
damaging the superstructure by force-fitting parts. Columns have one piece, and girders are
made from either one piece or two pieces of glulam. In some cases, girders fit together as
composite elements that wrap around the column, thereby resulting in desirable connection
and framework actions. As seen in Fig. 8.2, secondary framework elements were connected
to primary framework elements by brackets; floor slab elements were connected to the frame-
work elements to achieve composite action; and in some locations special framework details
were adopted (Fig. 8.2f).
After the structural parts were added to the superstructure, the roof was made watertight, and
the exterior claddings were added. Finally the building was finished internally. The construction
process is illustrated schematically in Fig. 8.3. In the finished state, the glulam structural ele-
ments are visible inside the building.
112 CHAPTER 8. EXAMPLE PROJECT 1: SIX-STOREY HYBRID BUILDING IN QUEBEC CITY
(a) (b)
(c) (d)
(e) (f)
Fig. 8.2: Installation of glulam superstructure framework and slab elements: (a) two-storey
segment; (b) interconnection of segments; (c) primary glulam framework elements; (d) glulam
floor slab elements; (e) interior column supporting girders in and transverse to the planes of
framework segments; (f) special framework of sixth storey on the curved and sculpted facade
(courtesy of FPInnovations)
RC slab above below ground parking garage First storey RC walls cast
First level of glulam framework installed First elevated slab structurally complete
and second level RC walls cast
Second level glulam framework installed Second elevated slab structurally complete
and third level RC walls cast
Fig. 8.3: Selected stages of the supertstructure construction sequence (courtesy of Nordic
Engineered Wood)
low parapet walls and minor obstructions (Fig. 8.3, bottom-right diagram), leading to the pos-
sibility of snow accumulations, which is accounted for via standard provisions according to the
NBC. Structural steel and RC framework buildings of similar general size and shape in Quebec
114 CHAPTER 8. EXAMPLE PROJECT 1: SIX-STOREY HYBRID BUILDING IN QUEBEC CITY
City can have superstructure elements sized for strength based on dominant effects of gravity
or lateral loads within the design loading combinations. The overall superstructure shape, local
topographical features, and the non-symmetric arrangement of superstructure elements of the
discussed building indicate that it sways laterally and twists about its vertical axis under the
effects of wind and would do so during seismic events.
The framework of glulam columns and girders are the primary parts of the above-ground system for
resisting effects of gravity forces associated with the self-weight of building elements, roof snow
load, and occupancy floor loads (e.g. office and document storage). As already discussed in Sub-
section 8.2.1, the framework itself is structurally complex because of the in-plane continuity that
exists in the prefabricated framework segments, and articulations/rotations that can occur where
framework segments join together or are joined to RC substructures. Connections within and at the
boundaries of framework segments are also complex in their design, construction, and behaviour.
Elevated floors and the roof incorporate glulam elements that are arranged to make those substruc-
tures function as horizontal diaphragms capable of collecting the effects of external wind pressures
or seismic ground accelerations (i.e. lateral load effects) and transferring them to RC shear walls.
Figure 8.4 shows typical details of connections within the superstructure. The connections were
designed such that glulam framework elements that are continuous through connections lock
together (in the style of carpentry lap joints), and when glulam elements are not continuous
through connections, their ends join precisely with steel plate linking and seating elements.
Dowel fasteners keep elements in place, besides having force transfer functions. Where pos-
sible, transfer of forces is by direct bearing on glulam or steel plate elements rather than fasten-
ers, which minimizes the possibility of splitting members. In all cases, the intimate nature of the
connection between elements means that force transfers are achieved by combination of glulam-
to-glulam, dowel fastener-to-glulam, dowel-fastener-to-steel linking element, and glulam-to-
steel seating element bearing and friction. Close attention was paid to avoiding situations where
moisture content-related dilations of glulam members might strain connections. Specifically,
care was also taken to avoid placing fasteners that joined glulam to steel plates too far apart in
the transverse direction, in which glulam could shrink and lead to splitting (Subsection 8.5.1).
The Canadian timber design code [114] and similar international best practice specifications
emphasize the need for such considerations.
8.2 SUPERSTRUCTURE SYSTEM 115
(a)
Connection Trp. SOLD
Connection Trp. SOLD
(b) (c)
(d)
(e) (f)
Fig. 8.4: Typical primary connection between glulam girder and column elements: (a) scheme
for arrangement of interlocking glulam elements showing steel plate linking and girder seating
elements at a location where an edge column is continuous between storeys (courtesy of Nordic
Engineered Wood); (b) two-storey column element precut and drilled to receive dowel fasten-
ers; (c) multi-bay girder element precut and drilled to receive dowels (one half of the complete
built-up girder); (d) steel plate linking and seating elements ready to receive the preassembled
framework segment for the next two storeys; (e) shear keys inserted in slots to create compos-
ite diaphragm action between glulam slabs elements in a floor; (f) attachment of glulam slab
elements to RC shear wall using steel angle sections, self-tapping wood screws, and concrete
anchor bolts (photographs (b) to (f) courtesy of FPInnovations)
116 CHAPTER 8. EXAMPLE PROJECT 1: SIX-STOREY HYBRID BUILDING IN QUEBEC CITY
• Canadian NBC [17], that details Acceptable Solutions/Practices for engineering design of
building superstructures (Chapter 4, of NBC). This includes definition of the load combina-
tions to be considered according to a Load and Resistance Factor Design (LRFD) format,
and incorporating the use of the so-called “companion loads approach” to account for the
relative likelihood of simultaneous occurrence of different catastrophic events (Table 8.1).
The LRFD loading combinations presume linear elastic analysis of the internal force ef-
fects that various components of the total load have on individual superstructure elements.
The NBC specifies the minimum roof snow, floor occupancy loads, wind velocity, and
peak ground acceleration applicable to the design of buildings in specific locations. Static
analysis is applicable for determination of internal forces in superstructure elements that
result from gravity loads. Dynamic response analysis approaches are the default methods
for determining internal forces in superstructure elements that result from wind and seis-
mic events. However, for building superstructures that do not exceed 60 m in height and
conform to certain requirements (e.g. simple building shape and regularity of the structural
arrangement), equivalent static load analysis may be performed for effects of wind and
Table 8.1: LRFD load combinations specified in the Canadian NBC [17]
8.3 STRUCTURAL DESIGN 117
seismic loads. In the case of equivalent static seismic load effects, it is permitted to reduce
the magnitudes of calculated elastic internal element forces if elements in the Seismic Force
Resisting System (SFRS – elements in the designed load paths to resist effects of seismic
forces) belong to substructures that exhibit ductile behaviour and/or can redistribute forces
to parallel substructures. The NBC provisions listed in Chapter 4 of that document apply
to superstructure systems constructed from all types of materials for which the parallel Ca-
nadian material design code exists; the Code de construction du Québec applicable at the
local level of the city of Quebec is equivalent to the NBC.
• Canadian Timber Design Code [114] gives detailed provisions related to the design of ge-
neric structural timber elements. This includes definition of design properties of glulam ele-
ments and detailed guidance on design of connections that join structural timber elements to-
gether, or join them to structural elements made from other materials. In the present context,
provisions of the Canadian Timber Design Code were employed to size glulam elements and
for certain aspects of the connection design. This took into account all potentially applicable
ultimate and serviceability limiting states for the load combinations in Table 8.1.
• Canadian RC Design Code [115] gives detailed provisions related to the design of RC
elements using steel reinforcement. In the present context, the provisions of this standard
were used to design the foundation substructure and shear walls based on the need to ensure
sufficient strength and control deflections. This took into account all potentially applicable
ultimate and serviceability limiting states.
• Canadian Structural Steel Design Code [116] gives detailed provisions related to the design of
structural steel elements and connections. In the present context, the provisions of this standard
were used to design steel plate linking and girder seating elements, and other steel elements in
the superstructure; accounting for applicable ultimate and serviceability limiting states.
In the case of each of the timber, concrete, and steel design codes, it is required that elements
be manufactured according to traceable specifications and standards under recognized quality
assured schemes. When steel is employed, it is also necessary that any welding be done under a
specified quality control regime. Technical design provisions embedded in the Canadian mate-
rial design codes are similar to provisions in comparable international material-specific design
codes conforming to LRFD formats.
The building is classified as belonging to Group D major occupancy according to the NBC
(i.e. business and personal services occupancies). Design dead loads were 1.5 kPa for the roof
and 2.45 kPa for the floors, including 1.0 kPa for effects of partitions. The minimum design
roof snow load is 3.5 kPa, plus allowance for accumulation of snow behind the parapet wall and
other obstructions. The design live loads level for floors associated with the type of occupancy
is 3.6 kPa. Code-specified adjustments were made to loads that account for tributary areas sup-
ported by various columns and girders. The design wind pressure was calculated using a refer-
ence velocity pressure of 0.53 kPa (1 in 50-year return period), a composite external pressure
and gust factor (CpCg) of 1.3, and a topography-related exposure factor (Ce) of 0.84 for building
facades that face oncoming winds [17].
emphasis on the behaviour of the glulam framed substructure under wind load. The concern
was that as the total mass of the six-storey hybrid superstructure is about 1500 tonne, it
was possible that wind-induced accelerations associated with lateral sway of the building
could reach levels that disturb its human occupants. This is because the vibration period
was expected to be much lower than for similar RC buildings (i.e. total superstructure mass
would have been about 2800 tonne for an RC building). However, it is well known that tim-
ber structures typically embody relatively high modal damping because they often contain
significant sources of damping beyond material damping (e.g. frictional damping) [20].
Therefore, close attention was given to prediction of the likely vibration period, which was
also an important input for determining the admissibility of the equivalent static seismic
load design approach in the NBC. Relative to the seismic design, it was determined that the
building met requirements for application of the equivalent seismic load approach. There-
fore, in the seismic design it was assumed that the ductility-related force modification factor
(Rd) and the overstrength-related force modification factor (Ro) could be assigned values of
1.5 and 1.3 respectively, yielding an RdRo (= q Factor in Europe) factor of 1.95. This means
that the estimated internal forces in elements associated with an elastic structural response
of the system were divided by 1.95 to deter the earthquake load (E) components within
the load combinations listed in Table 8.1. Adopted values of Rd and Ro are those assigned
to RC shear walls by the NBC. As a comparison, RdRo of 2.25 is permitted for braced
timber frameworks with limited ductility, and RdRo of 3.0 is permitted for braced timber
frameworks with moderate ductility. The practice adopted for seismic design was probably
conservative.
Main floor and roof girders are made from a grade of glulam intended for continuous beam
applications (i.e. the bending moment capacity is the same for positive and negative bending).
Girders are coincident with the planes of prefabricated framework segments and are parallel to
8.4 FIRE DESIGN 119
the minor building plan axis that is continuous over three spans approximately 6.0 m long, and
have cross sections of 362 × 527 mm2. Girders spanning transversely to the main ones (parallel
to the major building plan axis) are spaced 2 m apart and were designed as simple spans of 9 m,
resulting in cross sections of 190 mm × 527 mm. For simplicity of fabrication, construction, and
aesthetics, cross sections of columns are constant from the ground to the roof. Interior columns
have cross-section dimensions of 362 mm × 480 mm, and perimeter columns have cross-section
dimensions of 260 mm × 362 mm. The nominal dimensions of the glulam slab members are
89 mm × 450 mm with a maximum bending span of 2 m. These members span parallel to the
major building plan axis and are the primary elements of horizontal diaphragms at each elevated
floor and the roof. As shown in sub-diagrams (e) and (f) of Fig. 8.4, those slab elements are
interconnected and attached to RC shear walls.
(a)
(b)
Fig. 8.5: Measurement of relative settlement of the glulam framework: (a) expected deformed
shape of eight storey superstructure due to shrinkage of glulam; (b) attachment of potentiometer
device at a column base
Logic of this approximate calculation is that RC walls would shrink by negligible amounts;
creep deformations of the superstructure would be negligible; all framework columns would
shrink by the same amount; shake-down of the structural system (i.e. seating of structural ele-
ments into their permanent positions) would occur during construction because of self-weight
of the building elements and floor and roof loading associated with construction processes.
Figure 8.5a illustrates the expected deformed shape of the six-storey hybrid building due to
timber shrinkage. Transverse shrinkage of glulam was not expected to influence the frame-
work settlement because of the manner in which the connections were made. The likelihood
of an about 7 mm relative settlement of the timber substructure at the roof was considered
negligible for practical purposes of design and construction of the superstructure. Such a value
would be far less than the typically assumed 25 mm relative settlement of column bases of
RC or structural steel frameworks because of post-construction foundation movements. Nev-
ertheless, the unusual nature of the hybrid construction method demanded that the estimate of
the roof settlement be validated by field observations. Two potentiometer-type displacement
measuring devices were installed by a leading organization (R&D institute FPInnovations) in
the building after the superstructure was structurally complete in order to continuously mea-
sure changes at the level of the roof relative to the foundation level (Fig. 8.5b). A vertical line
of columns located at the extremity of the building and close to the curtain wall was chosen
for making observations. Measurements of relative roof level settlement indicated that actual
movements during construction and after more than 1 year were about half the estimated
possible deformation due to glulam column shrinkage. Most of the recorded settlement was
thought to be associated with closure of gaps between column segments rather than shrinkage.
The conclusion reached, therefore, was that differential movements in similar buildings can
be expected to be negligible provided that careful attention is paid to the design and construc-
tion of connections and that glulam is installed dry and protected from getting wet during
construction.
In Canada and elsewhere, lower fire resistance rating requirements are applicable to office
buildings than to residential or other occupancies where building occupants sleep. Therefore,
meeting Alternative Solutions fire performance requirements for residential buildings may be
more challenging.
Acknowledgements
This chapter is based on information, diagrams, and photographs supplied by Mr. Sylvain
Gagnon and Dr. Mohammad A. H. Mohammad of FPInnovations, Canada, Mr. Stéphane Rivest
of Bureau d’études spécialisées Inc., Canada, and Nordic Engineered Wood, Canada.
125
Chapter
9
Example Project 2: Fire Design of
a Seven-Storey Hybrid Building in
Berlin, Germany
Summary: The seven-storey superstructure discussed herein is for residential occupancy and
utilizes massive timber and reinforced concrete (RC) wall segments, composite timber-and-
concrete floor slabs, and steel framing members. Structural and acoustical performance design
requirements were satisfied relatively easily, and the main technical challenges centred on the
definition of a fire design concept, and selection of construction details and technical measures
that result in an effective fire resistance rating of 90 minutes. The building demonstrates the fea-
sibility of using timber as a modern, high-performance construction material based on hybrid
construction methods, and employing a fire design that utilizes a combination of passive fire
resistance of building elements and automated fire detection and suppression systems.
9.1 Background
Generally, the taller the buildings, the more stringent are the fire protection requirements. Until
2004, the above-ground height of top floors of timber building superstructures in Germany was
prescriptively limited to a maximum of 7 m, which translated to a maximum of three storeys.
This traditional limitation was associated with achieving buildings with 60-minute passive fire
resistance and fires that could be fought externally with scaling ladders. Since 2004, the top
storey height limit for timber buildings relying on passive fire resistance to achieve satisfactory
fire performance has been increased to 13 m (i.e. maximum of five storeys; [120]). As illustrated
in Fig. 9.1, buildings with upper floors more than 13 m above ground level must be designed
based on the assumption that fires will be fought via internal access to upper floors. Also build-
ings with more than four storeys must have 90-minute fire resistance, which can be achieved
prescriptively with non-combustible (class A) construction materials. Buildings with upper floor
levels more than 13 m above ground must have 90-minute passive fire resistance and employ
technical measures for fire detection and suppression [121].
Building classes
1 2 3 4 5
Buildings
Middle height Other
Stand-alone Low height floor* ≤ 7 m
building building
13 m
7m
RC firewall REIM90-A
Auxiallary Main
building RC wall building
REI 90-A
Solid-wood wall
REI90 K60
13.90 m
Elevator RC stairwall
RC firewall REIM90-A
Fig. 9.2: Plan arrangement showing the vertical structural components and firewalls
access that are physically separated from the main floor area by a distance of 3.1 m, as part of
the architectural and fire design strategies. Although the auxiliary superstructure is physically
adjacent to the main superstructure, the two are separated by a wall that meets the 90-minute fire
resistance requirements of DIN EN 13501-2 [121].
9.3 FIRE COMPARTMENTALIZATION OF THE BUILDING 127
Massive timber
wall panel
RC foundation slab
RC shear wall and RC first storey
The first storey is constructed from RC for reasons that include building security against acci-
dental mechanical damage and vandalism, and separation of timber elements from the ground
for durability. The overall structural concept combines massive timber and RC wall segments,
composite timber-and-concrete floor slabs and steel framing members, with massive timber and
concrete elements that also serve passive fire resistance functions. As shown in Figs. 9.2 and
9.3, the vertical structural elements were simple, relatively few, and designed to resist effects
of gravitational forces associated with self-weight of the construction, occupancy floor loads,
and roof snow load, as the main controls on dimensions of the columns. Composite timber-
and-concrete elevated floor slabs were designed to satisfy both structural and acoustical mass
performance requirements, as well as achieving fire performance objectives. Those floor slabs
are relatively heavy, compared with many traditional timber floor construction methods, but pre-
sented no special difficulties from a structural design perspective. Because peak seismic design
ground accelerations and wind loads for Berlin are low, there were also no special problems
associated with achieving satisfactory in-plane design resistances of shear walls and floor slabs.
The most challenging design decisions were those associated with providing 90-minute fire
resistance, which is why the remainder of this chapter is focused on those aspects.
The footprint of elevated floors in the main area of the primary superstructure is 174 m² includ-
ing public spaces, with apartment footprints being 148 m². According to the German building
128 CHAPTER 9. FIRE DESIGN OF A SEVEN-STOREY HYBRID BUILDING IN BERLIN
1600 m2 148 m2
1600 m2 148 m2
REI 90 K60
REI 90A
Fig. 9.4: Unlikely case of all floors burning to limits for prescriptive building classes
code, building class 4 can have apartment areas up to 400 m2. This translates to a total allowable
fire load of:
where qf,k is the characteristic fire load density of residential buildings [123].
Employing the actual apartment dimensions, the design fire load is 161 × 103 MJ. However,
there is also the danger posed by the building construction itself fuelling a fire. Therefore, a
smoke detection system was installed to reduce the level of risk, by guaranteeing early alerting
of the fire brigade and mitigating the likelihood that the structural system of the building would
burn (Chapter 3).
The German code [120] permits a maximum fire compartment area of 1600 m2, with enclos-
ing fire walls required to prevent fire spread for at least 90 minutes. This is achievable by a
REIM90-A wall [121] with Euroclass A covering material. Fire walls must have the ability
to bear additional loads caused by falling debris during fires as protection for firefighters.
Figure 9.4 compares the potentially affected fire compartment areas for the actual seven-
storey building in the case of all apartments burning at once to compartment areas permit-
ted for building classes 4 and 5. As shown, the very unlikely scenario of fires on all floors
results in a total affected floor area of about 1000 m² (7 × 148 = 1036 m²). This contrasts
with and lies between the respective limits for class 4 and 5 buildings of 60-minute fire
containment in an area of 400 m² and 90-minute fire containment in an area of 1600 m².
This effectively means that the total area of the worst envisaged case for the actual building
is far less critical than the limit for any one compartment of a class 5 building. Taking into
account all implemented measures, the fire compartmentalization strategy for the building
is highly conservative.
9.4 DETAILED ASPECTS OF THE DESIGN 129
Local regulations (applicable in Berlin) require that the maximum distance between any loca-
tion in a fire compartment and the wall of the escape staircase be 35 m. The actual maximum
distance is about 20 m. Therefore, the time it should take for building occupants to reach a
protected escape route is roughly half of what is allowed. The external location of the escape
staircases further reduced the fire risk and minimizes the potential for hot flue gas contaminating
the escape route. Other special measures were taken to ensure adequate fire performance of the
building as summarized in Table 9.1. The table also compares those measures with requirements
for building classes 4 and 5.
(1)
(2)
(3)
(4)
(5)
Fig. 9.5: Composite timber-and-concrete floor slabs; layers: (1) flooring, (2) 80 mm particle-
board subflooring, (3) 70 mm insulation, (4) 100 mm concrete, (5) 160 mm laminated timber
completed slab construction that also includes non-structural layers of insulation, subflooring,
and flooring. The timber layer was installed as prefabricated panels that subsequently acted as
permanent formwork for the RC layer (Fig. 9.6).
Fire design calculations allowed for reduction of the depth of the timber layer, which would
be exposed during an apartment fire below. The RC layer’s full thickness was used in the fire
design calculations associated with fires above or below the slab, as is normal for fully RC
elements. Table 9.2 summarizes the design parameters used for structural assessment of the
complete floor assembly for normal (no fire) and after 90 minutes of fire exposure situations.
The partial coefficients employed to assess the residual design load and deflection resistance
capabilities of floor slabs after a fire are less stringent than the partial coefficients applicable
9.4 DETAILED ASPECTS OF THE DESIGN 131
Load effects
to normal situations. Nominally this implies allowance of larger deflections and stresses in
buildings that have experienced fires. However, it should be borne in mind that buildings that
have been damaged by fire are unlikely to be in normal use prior to their full repair. So in
practice, safety levels for fire damaged buildings are not compromised.
Undersides of floors are coated with a fire retardant and the building is fitted with fire detection
and sprinkler systems. The fire retardant coating is transparent with a thickness of 150 μm and
meets the requirements of the Euroclass B (hardly ignitable; [121]).
7,8
Girder 16/30
Reinforced concrete
10
16
Mineral wool 6
Mineral wool 4
Fig. 9.7: Inter-element junction of a floor slab and exterior massive timber wall (dimensions in cm)
fire resistance of their own that would be activated if the gypsum plasterboard layers are destroyed.
In the case of the building discussed, an important critical inter-element junction situation occurs at
locations where floor slabs met exterior massive timber wall panels (Fig. 9.7). In the design of those
junctions, it was assumed that undersides of floor slabs will passively resist a full 90-minute fire. How-
ever, the massive timber walls are protected against interior fires by two layers of plasterboard. This
means that they only require 30-minute passive resistance (i.e. total resistance = 60 + 30 = 90 min-
utes). Exterior surfaces of massive timber walls are protected by 12.5-mm gypsum plasterboard and
100 mm of mineral wool beneath an 8-mm surface layer of mineral stucco. Such external protection
combats the possibility of inter-storey spread of fire via combustible building facades (Section 3.8.3).
Electrical Mechanical
design expectation was that fire engines in Berlin will reach any building within 15 minutes
of being alerted. On that basis, selection of the 60-minute fire resistance is believed eminently
reasonable. Additionally it is not to be forgotten that a number of special fire measures were
employed (Table 9.1).
The possibility of between-storeys spread of fire in the complete building is guarded against by
ensuring that electrical, plumbing, and other services for apartments only connect to vertical
RC shafts that run the complete height of the building (Fig. 9.1). Those RC shafts are part of
the overall structural system of the building (for resisting gravity forces), as well as being self-
capable of attaining a 90-minute passive fire resistance.
Laminated timber slabs and panels are typically only partially effective as containment barriers
for hot gases produced during building fires. Therefore, the composite floor slabs rely on the
concrete layer rather than the timber, as a barrier to transmission of gases from an apartment(s)
on fire.
Acknowledgements
This chapter is based on input provided by Professor Dr. Dirk Kruse and Professor Dr. Ing.
Bohumil Kasal of the Fraunhofer Wilhelm Klauditz Institute, Braunschweig, Germany. Kaden,
Klingbeil Architects, Berlin provided the diagrams and the image in Fig. 9.6; and Dehne, Kruse
Brandschutzingenieure, Gifhorn gave permission to describe the fire safety concepts.
135
Chapter
10
Example Project 3: Limnologen—
Block of Four Eight-Storey Residential
Buildings in Växjö, Sweden
Summary: Since 1994 performance-based design codes have been used in Sweden that allow
use of timber to construct residential buildings having any number of storeys, provided general
safety and serviceability requirements are met. At first this resulted in five- and six-storey resi-
dential buildings being constructed using timber, but in 2007 four eight-storey buildings having
mostly cross laminated timber (CLT) load-bearing walls were constructed in Växjö. Floors of
those buildings are of CLT plates with glued laminated timber (glulam)-stiffening ribs having
spans up to 9.7 m. Apart from structural stability, the main performance concerns during design
and construction were ensuring satisfactory sound and vibration serviceability. The discussion
in this chapter addresses architectural considerations, wall and floor design, construction meth-
ods, and R&D studies based on the Limnologen buildings.
10.1 Background
Four eight-storey residential buildings in Limnologen are the tallest, modern, completely timber
building superstructures in Sweden. They are located on a narrow site on the western shore of
Lake Trummen, between the city centre of Växjö and the campus of Linnaeus University, on
a parcel of land known as the Wälle Broar. The site is a showcase for the use of timber as a
high-performance construction material, as an initiative of the county of Småland and the city
of Växjö. The local action is consistent with the national desire to increase timber usage in the
building sector that is embodied by the Mer trä i byggandet (greater use of wood in the build-
ing sector) programme launched by the Swedish parliament in 2004. The basic idea behind
that national strategy is that the forestry-based sector, as the most important industrial sector in
Sweden, should be given opportunities to create jobs, and increase net export income. Another
consideration was redressing the effects of a ban on that industry, which had prohibited the use
of timber to construct residential buildings with more than two storeys for 122 years. Technical
enablement of construction of taller timber buildings flows from changes to the Swedish build-
ing code in 1994 that made it possible to construct residential buildings having any number
of storeys and using combustible materials, provided safety and serviceability performance
requirements are met.
The remainder of this chapter presents details about design, construction, and performance
of the Limnologen project, that has the general characteristics summarized in Table 10.1.
136 CHAPTER 10. LIMNOLOGEN—BLOCK OF FOUR EIGHT-STOREY
The buildings are owned by the tenants, who form a tenant-ownership community responsible
for the management and maintenance of the buildings.
A typical floor plan consisting of five apartments of different sizes, two staircases, and elevator
shafts is shown in Fig. 10.1b. The three apartments closest to the lake have balconies on two
facades, whereas the two furthest from the lake have only one balcony each. The shape of the
building plan privacy for the building occupants. Another important aspect of the design is that
in making the balconies on the southern side of the buildings contiguous, they could function
10.3 STRUCTURAL DESIGN 137
(a) (b)
B305
B304
HISS
B306
B307
HISS
B308
Hus B
Plan 3
Fig. 10.1: Architectural layout: (a) site plan showing four residential and ancillary structures;
(b) typical floor layout with five apartments
as an emergency escape route in the event of a building fire. The balconies could function as
projections that prevent or mitigate vertical spread of fire on facades. Those measures in com-
bination with the installation of automated fire detection and internal sprinkler systems enabled
the use of timber boards as cladding on the facades. Portions of the facades that could not be
reached from balconies for maintenance and repair are rendered.
Fig. 10.2: Example of (a) exterior CLT wall; (b) light-frame separating wall; and (c) interior
CLT wall (covering layers are insulation and/or gypsum plasterboard)
Internal walls (Fig. 10.2b and c) also perform structural functions related to stabilization of the
superstructures. External walls mostly have exterior insulation and timber cladding or rendering
on the exterior (Section 10.2). Room linings are made of two layers of gypsum plasterboard in
order to address fire performance requirements (Section 3.6).
The largest floor elements are 11.7 m long and 3.7 m wide and continuous over three supports.
With minor enhancements of the basic floor construction some elements have spans of 7.9 m
and 9.7 m. Floor elements were joined together to create diaphragms that channel the force
flows caused by wind pressures on the facades to the shear walls.
Besides the passive fire protection in the form of gypsum, the primary buildings are equipped
with automated fire detection and sprinkler systems. According to Swedish building regulations,
these additional technical measures are not required. However, adoption of them facilitated
acceptability of the overall design concept to regulatory officials, building owners, and building
occupants. As discussed in Section 3.8, the use of sprinklers minimizes the chance that fire will
spread through windows and burn facia materials and was necessary because the south facade is
clad entirely with wood. Other measures taken include that any spread of fire on the underside
of balconies should be easily visible, and therefore is likely to be detected early. Each apartment
has two exits and two fire escape routes, with one being a fire protected route inside a building
and the other external to the building via the balconies.
140 CHAPTER 10. LIMNOLOGEN—BLOCK OF FOUR EIGHT-STOREY
Fig. 10.4: Protection of elements: (a) during delivery; (b) during early stage construction; (c)
late stage construction showing an elevator segment being lifted into position by an overhead
crane
installations running in across-the-ribs direction were installed on-site. Work sequences were
organized to minimize the amount of time spent on-site, which minimized the overall construc-
tion period, and facilitated improved quality control, which is maximized when manufacturing
and construction are done in factory settings. Installation of the under-floor heating system took
a considerable portion of the total construction time (Fig. 10.5). Grooves in floor elements into
which tubes were placed were only partially prepared in the factory. Sheet metal to distribute
heat laterally was installed on-site, as were the floor finishing layers. Each apartment has a cen-
tralized heating system control that enables occupants to regulate the temperature.
The overall project was undertaken in two stages. Stage one involved construction of the two
northern-most primary buildings, which were finished during spring and early summer 2008.
The second stage comprised the two southern primary and ancillary structures and was fin-
ished during late spring 2009. The southern facade of the southern-most building is shown in
Fig. 10.6a, whereas Fig. 10.6b shows the collection of buildings viewed from the north-west.
Heat and water consumption is measured continuously in each apartment and is viewable by
occupants/tenants through private web pages. It was expected that the energy consumption
would not exceed 90 kWh/m2 per year and that the individual monitoring would result in con-
sumption reductions by up to 30%. Measurements made during the first 2 to 3 years of occu-
pancy show that excluding staircases and other communal areas the average heating energy was
only 55 kWh/m2 per year. The combination of energy-efficient construction techniques and a
consumption feedback system proved highly effective to limit energy usage.
142 CHAPTER 10. LIMNOLOGEN—BLOCK OF FOUR EIGHT-STOREY
(a) (b)
Fig. 10.6: Completed buildings: (a) most southerly building; (b) collection of buildings viewed
from north-west
– Planning documentation associated with suggestions put forward and decisions taken
during the early stages of the project.
– Quality control documentation of errors made and problems encountered.
– Inventories of the technical and environmental performance of the chosen solutions,
including measurements taken by instruments installed in the buildings.
– Economic and marketing documentation, including customer surveys of perceptions of
timber construction.
– Building process documentation, such as data from time studies and logistics of site operations.
– General information documentation about the project.
The research projects are described in detail elsewhere [128–136]. Here, two examples are
briefly presented that relate to the physical performance of the building and the efficiency of the
construction practices employed.
10.7 RESEARCH STUDIES 143
The relative, within-storey displacements were measured using reference points located directly
on an external wall. Actual displacement measurements were taken using a displacement
CLT-panel ~ 100
of installation of equipment.
temp. and R.H., storey 7
Bar of invar with potentiometers
~ 60
installed at the end placed on top
of cantilevers and supported
laterally by concentric bearings.
Storey 6
1 2 3
Measuring length 2995 mm
Cantilever used
as support
Wood material
Measuring length 2970 mm
Displacement measurement 4 5 6
potentiometer
Reinforced concrete
slab and wall
Fig. 10.7: Displacement measuring devices on northern facade of the northern-most building
144 CHAPTER 10. LIMNOLOGEN—BLOCK OF FOUR EIGHT-STOREY
30
2nd
3rd
25 4th
5th
6th
Vertical deformation (mm)
7th
20 Total
15
10
0
24-Dec- 11-Jul- 27-Jan- 15-Aug- 03-Mar- 19-Sep- 07-Apr- 24-Oct-
2007 2008 2009 2009 2010 2010 2011 2011
Fig. 10.8: Vertical settlements per storey and total settlement of second to seventh storeys
(September 2007 to December 2011)
potentiometer (Regal – WPL 50EFZ) placed at one end of a 20-mm diameter rod made of invar,
a nickel – iron alloy notable for its unique low coefficient of thermal expansion, which mini-
mizes extraneous thermal influences on observed displacements. As indicated in Fig. 10.7 the
invar rods have rounded ends and connect to cantilever brackets via concentric bearings, which
further minimized the possibility of extraneous influences on observations. The reference and
measurement end brackets were attached to CLT panels using screws. Potentiometers were sep-
arately connected to a data acquisition system (Datataker DT85). Synchronized measurements
were taken of the external surface temperature and relative humidity at the second and seventh
storey ceiling levels using Vaisala model HMP50 devices.
Figure 10.8 shows the displacement recorded between September 2007 and December 2011.
The maximum observed cumulative settlement for second to seventh storeys was about 24 mm
over a distance of 17.95 m and was recorded during the summer of 2011. This corresponded to
0.13% of the total measuring length. The major part of this displacement occurred during the
first 9 months that data were collected. During the same period, the estimated average moisture
content decreased by roughly 10%, which is normally expected to result in shrinkage of wood
walls by about 0.04% (Section 8.5.1). There is no clear sequencing of displacement magnitudes
according to the storey level. Results suggest that the main causes of settlement are shrinkage
of floor layers and shake-down settlements, as the components bed themselves in and as build-
ings adjust to the surrounding climate and the influence of in-service building loads. The effect
of cyclic seasonal variation of moisture contents of timber components was clearly apparent,
and can be seen in Fig. 10.8. Extrapolating the observed trends, it is reasonable to conclude that
a maximum settlement of about 25 mm would be reached around 6 years after construction.
Although this value would not be replicated exactly in other circumstances, the results do show
that with proper attention to delivery, storage, and installation, settlements in buildings of the
type discussed herein can be expected to be of magnitudes that will not cause building perfor-
mance problems, even in high-rainfall climate zones such as southern Sweden.
10.7 RESEARCH STUDIES 145
Analysis of the data determined that the average installation time for wall elements was about
22 minutes and that the average installation time for floor elements was about 26 minutes. This
suggests that under normal circumstances the floor elements are slightly more time-consuming
to install than the wall elements. However, this reflected some difficulties in fitting the flooring
elements together that could be overcome by design and construction process modifications.
Figure 10.10 provides a more detailed analysis of the average installation time as a function of
which floor was involved. The highest and the lowest average duration per storey for installing
a wall element were approximately 29 minutes and 18 minutes, with both being for building
1 (called house 1 in the diagram). The corresponding durations for the flooring elements were
30 minutes and 17 minutes for buildings 1 and 3, respectively. Such differences suggest that the
installation process is very sensitive to disturbances and that installation times for both floor and
wall elements could be shortened appreciably.
Linear regression analysis showed that there was a tendency in both buildings studied for instal-
lation of wall elements to take a shorter time higher up in the building, despite the fact that the
(a) (b)
Average installation time [minutes]
35:00 35:00
25:00 25:00
20:00 20:00
Walls house 1 Flooring house 1
15:00 Walls house 3 15:00 Flooring house 3
Walls house 1 linear regression Flooring house 1 linear regression
Walls house 3 linear regression Flooring house 3 linear regression
10:00 10:00
1 2 3 4 5 6 7 8 2 3 4 5 6 7 8
Storey number Flooring number
Fig. 10.10: Average installation times in houses 1 and 3: (a) wall elements; (b) floor elements
time required for transportation of each element from the ground level to the storey on which
it was to be installed increased with height above the ground. Potential explanations are that
the time required was affected by: (i) adjustments made at the manufacturing plant or at the
construction site that improved the efficiency with which elements could be fitted together on-
site as the project progressed; and (ii) workers became increasingly familiar with the building
system and, thereby, became more efficient in execution of their tasks as the project progressed.
There was a likewise decrease in installation times for floor elements in building 3. However, for
building 1 the linear regression analysis indicated a marginal increase in the time required for
installation of floor elements as the working above-ground increased. The likely correct overall
interpretation of what was observed is that experiential learning associated with manufacture of
elements and site practices yields productivity gains, but when practices are stable, productivity
decreases slightly with each change in the lifting distance from the ground to the work level.
By implication, the project undoubtedly did not achieve optimal efficiency in terms of practice-
based learning or storage and handling of elements on-site, but given the uniqueness of the
project this is not surprising.
Acknowledgements
This chapter is based on information, diagrams, and photographs supplied by Professor Erik
Serrano and Dr. Johan Vessby of Linnaeus University in Växjö, Sweden.
147
Chapter
11
Example Project 4: Björkbacken,
a 10-storey hybrid building
in Stockholm, Sweden
Summary: The building discussed herein is for residential occupancy by the elderly and has
four reinforced concrete (RC) plinth superstructure storeys and additional six modular timber-
framed storeys above. Timber-framed modules are of a type that has been previously proven sat-
isfactory for construction of an “all” timber six-storey superstructure. The primary reason for
constructing lower storeys from RC was that such an approach was acceptable to building regu-
latory officials in Stockholm with respect to fire performance expectation of the complete build-
ing. Such a requirement reflected that no similar buildings had been previously constructed in
Sweden. The other primary technical issues associated with design of the building were related
to control of expected vibrations and static deflection responses of timber floors for serviceabil-
ity. Force flows resulting from lateral external wind loads on upper storeys are channelled from
the building facades to the timber floors (which act as diaphragms), then to timber shear walls,
and then into the RC storeys, which are monolithic with a RC foundation. Timber modules are
anchored to the storey below against uplift, overturning, and sliding due to wind by hold-down
and shear connectors at their bases.
11.1 Background
Until 1994, the maximum height of timber-framed buildings in Sweden was prescriptively lim-
ited to two superstructure storeys, because of concerns about the potential risk of fire spread-
ing between buildings in urban areas. This mirrored restrictions that existed in many countries
until quite recently (Chapter 3). Since 1994, the Swedish building code permits buildings of
any height to be constructed provided that it can be demonstrated that the design meets mini-
mum safety and serviceability performance requirements related to the intended type of building
occupancy. In early 2011, Sweden adopted the Eurocode building codes with specific national
requirements being contained in annexes to those documents. Design and construction of the
10-storey Björkbacken building discussed here was required to achieve a 90-minute fire resis-
tance rating. Other general technical requirements were that the structural system (superstruc-
ture and foundation) be able to resist potential ultimate limiting states effects of the self-weight
of the building and design loads associated with a residential building occupancy, wind and
snow, and perform satisfactorily with respect to serviceability limiting states associated with the
acoustical and vibration performance of the superstructure.
148 CHAPTER 11. BJÖRKBACKEN, A 10-STOREY HYBRID BUILDING IN STOCKHOLM, SWEDEN
The timber modules arrived at the construction site with windows, doors, services, and internal
linings already installed and their interiors painted. The structural concept is that the building
system minimizes concentrated force transfers as much as possible; meaning that relatively small
FA FA
Fa Fa
FA FA
Fa Fa
Pa
FA FA
Fa Fa
FA FA
Fa Fa
FA FA
Fa Fa
FA FA
Fa
FA Ground level
9026
9026
2000
2000
11559 11759
7026
7026
3060
3060
3060
3060
4904 5104
17927
17927
0 0
11559 11759
4901
4901
4766 3360
3360
4766 3360
6496
4000
4000
2625
8901
8901
8901
Fig. 11.3 Floor plan for upper storeys (black dots are locations of steel linking elements in the
lateral load-resisting system)
150 CHAPTER 11. BJÖRKBACKEN, A 10-STOREY HYBRID BUILDING IN STOCKHOLM, SWEDEN
Fig. 11.1, the site on which the building is located has a sloping terrain, such that one facade of
the building has only eight above-ground storeys.
Floor joists spans do not exceed 4.0 m because module widths are limited to 4.15 m for transpor-
tation reasons. The joists are made from glued laminated timber (glulam) and have cross-section
dimensions of 42 × 225 mm2 and are placed at a spacing of 600 mm, apart from bathroom floors
where the spacing is 300 mm in order to control deflection.
The most challenging aspect of designing the timber vertical load-resisting system was provid-
ing for design load effects to be spread evenly from supported elements to supporting elements,
11.6 CONSTRUCTION OF BUILDING 151
so that construction methods could be consistent throughout the superstructure. This was
an issue, for example, where large openings were required in facades for windows. In such
instances, the adopted solution was to frame the openings with glulam elements that were larger
than other framing elements. Glulam elements were also incorporated inside exterior walls to
support loads from the balconies. Roof snow load and self-weight is carried by timber trusses
supported by the outer walls of the modules. Close attention was paid to avoid problems in situ-
ations where wall studs were loaded with horizontal framing members in compression, perpen-
dicular to the grain. In some cases, use of double or triple wall studs was required to keep the
design stress transfers at acceptable levels.
The RC walls in the lowest four storeys act as shear walls that stabilize that portion of the
building against the effects of design wind loads. The light-frame perimeter walls of the timber
modules were designed to act as shear walls of the upper six storeys. Where possible, vertical
planes of RC walls are aligned with vertical planes of the timber shear walls, to avoid eccentrici-
ties in flows of wind-induced shear forces from one level to another, down through the lateral
load-resisting system elements and eventually to the foundation and the ground. The maximum
shear flow through the timber shear walls can be resisted by two layers of sheathing. Using a
combination of one cement-bonded layer and one gypsum plasterboard layer was found to be
sufficient for this, and also met the 90-minute fire resistance requirement for the building. Uplift,
overturning, and shear flow force transfers are handled by steel plate link elements, as illustrated
in Figs. 11.2 and 11.4, and placed at locations as shown in Fig. 11.3.
Floor assemblies act as rigid horizontal diaphragms for the purposes of transferring wind forces
from building facades to shear walls and maintain the building shape as it sways because of
wind pressures on the facades. The connections between horizontal timber diaphragms and
shear walls had to be designed to be sufficiently strong and stiff to transfer forces effectively but
sufficiently flexible to avoid creating monolithic joints that would transmit sound and vibration
directly or by flanking between storeys. The lowest timber modules are therefore connected to
the RC storeys by steel angles, as shown in Fig. 11.3.
2 100
280
A
120 A
Fig. 11.4 Example of hold-down devices (dimensions vary with load levels), units: mm
by tarpaulins. As there was no storage space at the site, the modules were lifted from the deliv-
ery trucks directly into their final positions by a tower crane. This required that the modules be
fabricated to correct design dimensions to avoid unnecessary site work. Dimensions were cor-
rect in all cases relative to the requirement that gaps left between adjacent modules measured
between 20 and 25 mm (i.e. 5 mm tolerance per outside module dimension). Contacts between
modules were initially through simple male–female joints, with steel link elements being added
after final placement of particular modules (Figs. 11.2 to 11.4). Gaps between modules were
made airtight and fireproof by filling them with insulation material. As timber storeys were
added, the storeys themselves became the working platform for the next storey, with each level
being made watertight as quickly as possible. This ensured that finishing work below the current
working platform was done in dry conditions. Activities were planned such that a full storey was
added in a single day. The first three of those storeys were assembled during Tuesday to Thurs-
day of one week and the remaining three during the following week. Vapour barriers in exterior
walls were made continuous on-site by securing overlapping projecting pieces of barrier mate-
rial that had been preinstalled in the modules at the manufacturing plant. Building services (e.g.
plumbing and electrical) were also preinstalled in the modules, and on-site internal work was
restricted to installing various systems in service shafts and coupling those to modules and to
external services. Fireproof sealants were used to block openings around pipework.
The building is shown in nearly completed form in Fig. 11.5. As seen in that picture, the facades
are clad in material that, except for exposed concrete in lowest storeys, gives no hint of the
nature of the structural system beneath.
11.7 ADDITIONAL COMMENTS 153
11.7 Additional
comments
The Björkbacken project demonstrates
that it is possible to construct quite tall
buildings in congested urban locations
rapidly and efficiently, based on a com-
bination of cast-in-place RC storeys
below and modularized prefabricated
timber storeys above. Although not dis-
cussed in detail, it should be mentioned
that construction of this particular build-
ing was cost-competitive and provided a
high quality living environment for the
elderly. The low mass of the light-frame
timber modules greatly reduced the
structural demands on the RC substruc-
ture compared with other options. This
approach can in general lead to substan-
tial cost savings, especially in the case
of foundations, and in some instances
the incorporation of light-frame or other
timber storeys will enable relatively
Fig. 11.5 Building close to completion tall buildings to be constructed on sites
where ground conditions would other-
wise preclude similar height buildings
from being constructed. This project is consistent with the philosophy embodied by this Struc-
tural Engineering Document (SED) that timber should be thought of by architects and engineers
as a material that is suitable for use on its own or in combination with other materials.
Acknowledgements
This chapter is based on information, diagrams, and photographs supplied by Professor Erik
Serrano and Dr. Johan Vessby of Linnaeus University in Växjö, Sweden.
155
Chapter
12
Looking to the Future
Summary: It is impossible to be exact about what types of buildings will be constructed in the
future, what types of superstructures they will have, or what construction materials will be
used. However, it is possible to say with a high degree of certainty what will be technically pos-
sible in those respects. This chapter discusses likely achievable building heights for multi-storey
superstructure systems of different types that employ primary elements made from timber and/
or timber-based composite materials. The LifeCycle Tower concept for constructing timber
framed buildings with up to 20 superstructure storeys is mentioned as a metaphor for the types
of systems that are already being envisioned in various parts of the world. Also discussed in
this chapter is the nature of changes to structural and fire design codes that would facilitate
engineering design of building superstructures containing timber elements.
For timber structural elements, as with concrete or masonry ones, the ultimate capacity attain-
able in compression depends on the degree of peripheral containment provided by other struc-
tural parts or internalized containment by embedded steel or other types of lateral reinforcing
materials. In the case of material like glued laminated timber (glulam), measures as simple as
inserting self-tapping reinforcing screws could increase axial compressive capacities of col-
umns. Therefore, once the shape of any superstructure has been selected, the primary operable
question in determining its maximum possible height is how porous it can be near the base
(what the minimum void proportion can be) without disrupting the functionality of the building
to which it belongs. This statement presupposes that designing portions of superstructures to
resist combined tension and shear force flows is simpler than stabilizing them to resist combined
compressive and shear force flows. Usually this is a quite reasonable presumption, and that is
assumed in comments below.
In lightweight plate timber construction, wall and floor plate substructures act together as assem-
blies of folded and interlocking elements, with the plate elements consisting of structural wood-
based composite sheathing panels that are stiffened by edge framing and intermediate framing
components that are attached by mechanical fasteners (e.g. nails and screws) or glue. From a
structural engineering perspective, the individual plate elements are rib-stiffened plates capable
of efficiently resisting in-plane forces and out-of-plane bending forces. Many variations in the
basic theme are possible with constructed systems mostly being compartmentalized assemblies
that are similar in form to internally divided cardboard boxes. Structurally, the systems can be
very strong and robust, but only as long as the substructures remain interconnected.
Smith [137] addressed this issue of where the limits on the heights of timber lightweight plate
superstructures lie from the structural engineering perspective. He examined two classes of
construction methods suitable for multiple residential occupancies or mixed office–residential
occupancies. The construction methods considered were the platform and balloon methods,
as illustrated in Fig. 12.1. Within each method, possibilities considered are use of “standard”
timber framing materials like sawn lumber and use of modern timber-based composite products
like laminated veneer lumber (LVL) that have enhanced engineering properties. Also considered
was how supplementing timber lightweight plate assemblies with “building cores” consisting
of reinforced concrete (RC) shear walls, or other relatively massive systems capable of resisting
lateral wind and seismic loadings on taller superstructures, would affect achievable superstruc-
ture heights. Table 12.1 summarizes those results, based on the relationship that each storey
corresponds to a superstructure height increment of 3 m.
The findings are that 10- to 20-storey structures are theoretically feasible, if best possible construc-
tion methods are employed, especially if there is a RC core(s) in the buildings to stiffen them later-
ally. In most countries, buildings with more than three above-ground storeys are required to have
12.1 LIKELY LIMITS ON HEIGHTS OF MULTI-STOREY SUPERSTRUCTURE SYSTEMS 157
(a) (b)
Discontinuous Continuous
wall plates wall plate
Fig. 12.1: Construction methods for lightweight plate assemblies: (a) platform construction;
(b) balloon construction method (inserted photograph courtesy of Canadian Wood Council)
elevators, and most buildings with more than four storeys must have fire escapes and firefighting
access staircases protected by non-combustible material. Therefore, all modern tall buildings must
already have RC or other structurally substantial walls. Therefore, these could satisfy structural
system requirements without the need for unusual measures. Platform construction has lesser
height capabilities compared with balloon construction, because platform construction involves
the transfer of compressive forces between storeys and from the lowest storey to the foundation
by compression perpendicular to the weak horizontally oriented layers of timber (Fig. 12.1a). The
balloon construction method eliminates the source of weakness inherent to platform construction
by aligning vertical wall framing members such that force transfers are only parallel to their axes
(i.e. parallel to grain). In principle, it is possible to reinforce weak zones in platform construction
systems, but normally it is simpler to not select an approach that creates weak zones.
158 CHAPTER 12. LOOKING TO THE FUTURE
Despite what Table 12.1 suggests, there is much practical experience that indicates that
the achievable number of superstructure storeys/heights using lightweight plate assemblies
is constrained mostly by the ability to economically counteract problems associated with
vibration and sound transmissions between building occupancy units (e.g. between residen-
tial apartments). Also, there are public and firefighter concerns about fire performance of
lightweight plate assemblies. Consequently it is speculated that building superstructures con-
structed from timber lightweight plate assemblies will never be more than around 10 storeys
(i.e. 30 m high).
(a) (b)
Fig. 12.2: Massive CLT plate assembles: (a) CLT plates being hoisted for installation; (b) hotel
extension under construction
12.1 LIKELY LIMITS ON HEIGHTS OF MULTI-STOREY SUPERSTRUCTURE SYSTEMS 159
being prepared for buildings with around 15 to 20 storeys built mainly from CLT, for diverse
geographic locations (e.g. Milan in Italy and Vancouver in Canada). The ultimate limit on struc-
tural height will depend on the slenderness and regularity of shape of the superstructures. Most
likely, the critical factors will be the ability to design connections capable of handling wind and
seismic shear and overturning force flows between storeys and between the superstructure and
foundation, when superstructures have regular geometries. In cases where the superstructure
geometry is irregular, achievable heights are likely to be considerably less. This reflects the
need to pay close attention to avoiding out-of-plane instability of perimeter walls and floor dia-
phragms, and controlling lateral sway and acceleration rates near the perimeter walls.
Within the limits of the above comments and the discussion in Chapter 7, it is speculated to be
unlikely that many massive timber-plate superstructures will ever exceed the originally envis-
aged height of 10 or perhaps 12 storeys (circa 35–40 m), but rare examples of buildings will
probably reach heights of around 40–50 m. The suggested range of maximum heights does not
include systems where massive timber plates are installed as floor slab and wall panels of build-
ings with frameworks of materials such as steel, RC, or glulam.
There exist many other examples, both historical and modern, of modern multi-storey super-
structures that employ moment resisting timber frameworks as their primary structural system
(Fig. 12.3). However, such buildings are not common because it is difficult and expensive to
make moment transferring connections between framework members. Difficulties are associ-
ated with the tendency of framework members to split, unless they have complex carpentry
junction arrangements, and/or members are reinforced at junctions, and/or use special con-
nection hardware. Also, because timber is quite compliant and shrinks perpendicular to grain,
even well-made carpentry joints do not tend to be reliable in creating rigid connections in
frameworks. Therefore, it is most normal in multi-storey heavy timber framed superstructures
to have separate gravity and lateral load-resisting systems in which framework connections
approximate hinges or are simple bearing joints. As discussed in Chapter 5, lateral rigidity in
complete systems is provided by cross-bracing, shear panels of various materials, or masonry
or RC building cores. Floors and roofs act as horizontal diaphragms that prevent in-plan distor-
tion of the frameworks.
160 CHAPTER 12. LOOKING TO THE FUTURE
(a) 11.456
10.747
7.820
8.635
8.855
14.608
(b) (c)
(d) (e)
(f)
Fig. 12.3: Timber heavy-framed assemblies: (a) Sakyamuni Pagoda of Fogong Temple in
Yingxian, China; (b) Venetian example of carpentry tying parts together under heavy grav-
ity load; (c) Bolognese knee braces in a medieval porticoed residence; (d) Claremont Hotel
under construction in 1906 (source: University of California Berkeley Library); (e) modern
multi-storey car-park in Mura, Austria, with separate timber gravity and lateral load-resisting
systems; (f) modern five-storey building in Baar, Switzerland, with separate gravity (timber
frame with timber–concrete composite slabs) and lateral load-resisting systems (RC core)
12.1 LIKELY LIMITS ON HEIGHTS OF MULTI-STOREY SUPERSTRUCTURE SYSTEMS 161
Late 19th- and early 20th-century hotels in North America are believed to be the largest resi-
dential timber-frame buildings ever constructed, with the now demolished Yellowstone Canyon
Hotel having been the largest [8]. The largest remaining structure is the Claremont Hotel in
Oakland, California, which has 279 rooms and conference facilities that together occupy 10
floors (Fig. 12.3d). The ridge of the upper roof is approximately 36 m above the foundation, and
a central tower rises about 49 m above the foundation. Tall timber-frame commercial buildings
were constructed in the USA in the 19th century, with many having six or seven timber storeys
and the tallest having nine storeys. Such buildings were filled with vibrating heavy machinery.
Reportedly, sprinklers for fire suppression were invented to protect them. Another interesting
aspect of the history of tall commercial buildings in the USA is that it was common until about
the mid-19th century for cast iron columns to be used. However, following some catastrophic
brittle failures of such elements they were widely replaced by timber columns, which a half-
century later were in turn widely supplanted by steel elements [138].
What can be said for certain is that heavy-frame, timber buildings perform well at heights of
at least 50–60 m (i.e. in the range of 55 m which is the height of the Sakyamuni Pagoda minus
its steeple). This is without requiring more than simple construction details. However, the theo-
retical maximum height using modern structural engineering approaches is much greater. For
squat shaped buildings, the structural limit on height is in the order of 160 m, with the ultimate
capacities of the columns in lowest storeys being the critical issue. This estimate is based on
1% of the lowest floor plan being occupied by columns. Slender shaped buildings of types more
likely to be built will have maximum possible heights nearer to 80 m. This relates to inability to
economically constrain magnitudes of translational and torsion movements caused by wind as
the limiting factor. As discussed in Section 12.2, the LifeCycle Tower concept is proposed as a
suitable way to economically construct superstructures to about 64 m (20 storeys).
The suggestion that achieving heights of 160 m is technically possible without consideration
of non-structural superstructure performance criteria may appear unrealistic. However, readers
should bear in mind that on February 21, 1872, William Ferguson measured the stem of a fallen
eucalyptus tree (Eucalyptus regnans) at Watts River, Victoria, Australia to be 133 m long. That
dimension did not include the tree top that had broken off. At the break, the diameter of the stem
was about 1 m and it was estimated that the tree had probably been at least 152 m tall. Structur-
ally speaking, tree stems are not very efficient compared to what engineers can achieve using
the same mass of material, because tree stems have architectures that support life functions of
the trees, apart from achieving structural efficiency at various stages of tree growth. The 118 m
Gliwice Radio Tower, shown in Fig. 1.2, demonstrates what humans have achieved even without
timber construction capabilities as sophisticated as those existing now. Future timber heavy-
frame building superstructures will never achieve heights that take them into the skyscraper
classification (> 100 m), because account must be taken of the relative proportion of the total
supported weight that is non-structural components, and because humans have quite a low toler-
ance for building movements during events like wind storms.
Roof
Secondary structural system:
• Preventing disproportionate collapse
• Fire protected
Superstructure to
foundation isolation layer
Composite isolating layers for:
• Fire separation
• Vibration & sound separation Foundation
• Structural damping
Important to note is that when multiple functional objectives are involved, achieving high design
efficiency does not necessarily mean that different materials will be combined to create struc-
turally efficient monolithic constructions, because maximizing structural efficiency can impair
attainment of other objectives. It should also be recognized that particular attributes and proper-
ties of combined materials have to be accommodated, or there will be problems of force-flow
concentrations associated to differences in compliance and different sensitivity to temperature
and moisture flows (e.g. associated with processes like curing, drying, and changes in outside-
and inside-building environments). The term composite system/assembly refers to situations
where two or more construction materials are designed to act as one unit. This can be synonym-
ous with hybrid systems.
From a purely structural perspective, it is known that using relatively low-mass timber com-
ponents instead of some traditional heavy components within tall multi-storey framed
12.2 EXAMPLE OF PROPOSED SYSTEMS: LIFECYCLE TOWER CONCEPT 163
18
Storey number
12
SRC_centre (C)
= 96 m
6 SRC_corner (B)
SXL_centre (C)
SXL_corner (B)
24 × 4 m Ground 0
0 50 100 150 200 250 300
Drift at top of storey (mm)
2.5 SRC
2 SXL
1.5
1
25 m
.4 0.5
.6 38
m 0
1 5 9 13 17 21
Mode number
Fig. 12.5: Steel frame with XLAM massive timber or RC slabs [89]: (a) structural represen-
tation of 24-storey building; (b) comparison of peak lateral seismic drifts; (c) comparison of
modal periods (SRC denotes steel framework with RC slabs, SXL denotes steel framework with
CLT, XLAM, slabs; locations centre and corner are alternative plan positions on floor and roof
slabs)
superstructures can yield significant gains. As discussed in Chapter 6 and illustrated in Fig. 12.5,
using massive timber floor slabs instead of RC floor slabs in a 24-storey building markedly alters
the dynamic response and reduces levels of lateral drift during strong earthquakes by about 40%
[89]. Potentially, there is reason to believe that if properly designed and constructed, future
building superstructures containing substantial amounts of timber could surpass the heights of
current super tall structures by a considerable margin (Fig. 12.6). This possibility reflects that
the self-weight of the system would greatly reduce effects of gravity and wind-induced stresses
in steel or possibly other materials in the primary framework, and help control lateral motions
by passive methods.
800 m
600 m
200 m
0m
The illustrated completed system has 20 above-ground storeys of a 24.8 by 38.3 m rectangular
footprint and a total height of 76 m, including two below-ground plinth storeys. Design of the
building core assumes it will be erected in two 32 m high above-ground construction stages,
which facilitates maximization of lateral building stiffness and provides localized ductility, and
12.2 EXAMPLE OF PROPOSED SYSTEMS: LIFECYCLE TOWER CONCEPT 165
(a) (b)
Fig. 12.8: Superstructure arrangement of LifeCycle Tower: (a) typical floor layout; (b) cross
section parallel to minor plan axis
economic competitiveness with alternative high-rise building systems. Having only two hori-
zontal interfaces (i.e. core to foundation and between construction stages) permits a high degree
of prefabrication and shortens construction time. Although employing central building cores
has well-defined structural benefits (Chapter 5), it can have disadvantages in terms of efficient
use of floor area.
A ribbed composite glulam rib and concrete slab diaphragm system was selected to transfer
the very large forces from the facade columns to the building core, and secondarily for fire
separation between floors. In addition to increased resistance to the spread of fire beneath it,
the diaphragm arrangement provided space between the ribs to accommodate technical build-
ing services within the depth of the deck. Facade columns were designed to satisfy structural
requirements associated with gravitational loads and to transfer facade forces directly to floor
diaphragms. Notably, avoidance of application of compressive forces to horizontal layers in
timber superstructures is critical, because materials like glulam are weak and compressible
when stressed in the perpendicular to the grain direction. Therefore, the LifeCycle Tower
facade column connections were designed to completely avoid such situations in the transfer
of gravitational and other vertical forces between storeys and across the depths of the floor
diaphragms. Attachment of the columns to the diaphragms was via single mortice carpentry
joints that were not dependent on metal components for force transfer. The adopted connec-
tions produced close to pin-end conditions for the columns, thereby rendering their effective
buckling lengths close to their physical lengths. During construction, several pairs of columns
could be installed as a prefabricated facade unit, with secondary facade elements preinstalled,
so that site work could progress rapidly compared with alternative high-rise construction
systems.
According to the fire protection approach the system had to be designed and constructed to meet
regulations applicable to various building occupation categories (e.g. private residential, hotel
166 CHAPTER 12. LOOKING TO THE FUTURE
Offices
Corridor
Stairwell
Primary escape
distance 22.5 m
Secondary escape
distance 34 m
Fig. 12.9: Fire design of LifeCycle Tower: (a) fire zoning strategy (schematic), blue zone = fully
fire protected building core, yellow zone = escape staircase; (b) escape route lengths within a
storey (schematic)
and office) in various jurisdictions. Consequently, the building fire performance concept was
designed to address the following considerations:
• Creating a simple and comprehensible layout of escape routes and provisions for firefight-
ing access routes.
• Using standardized prefabricated elements, so that design complexity will not lead to
construction errors.
• Provision of 90-minute fire resistance for exposed timber elements (e.g. glulam ribs in dia-
phragms and glulam facade columns).
• Use of only mineral-based building materials to create junction interfaces between fire
walls and ceilings (bottom surface of floor diaphragms) or tops of floors.
• Ensuring that horizontally orientated building service conduits connect only to non-
combustible building elements.
• Encapsulation of building services within non-combustible materials.
• Not creating enclosed spaces adjacent to glulam or other combustible materials that would
facilitate undetected fire spread.
• Not constructing facades in ways that facilitate fire growth between storeys (e.g. extension
of concrete slabs in floor diaphragms right up to the facade).
• Dividing floors into zones that permit escape of building occupants and safe access to each
floor level for firefighters (Fig. 12.9).
12.3 REFOCUSING DESIGN CODES 167
Notably, the choice of 90-minute fire resistance may not be consistent with requirement in
various regulatory jurisdictions. However, satisfying other fire resistance requirements is not
problematic. Taking contemporary Austrian building regulation requirements as an illustrative
example, the maximum permitted escape route length within a storey is 40 m measured from any
point in an occupied space to the stairwells of primary and second escape routes. Figure 12.9(b)
shows application of this requirement to the LifeCycle Tower. As is clear, from this example,
fire performance requirements may sometimes impose limitations on adoption of the discussed
approach, but not necessarily on using timber to construct tall buildings in other ways. In this
example, the building services include an integrated high-pressure water mist extinguishing
system that self-activates to extinguish and suppress fire based on three principles of physics: (1)
cooling, which is the most important property of water in firefighting; (2) inertisation (sudden
local evaporation of the water mist inside the flames), which results in the water mist expanding
to about 1700 times its original volume to displace oxygen inside the flames and suppression of
fires; and (3) blocking/shielding of radiated heat, which because of the high number and density
of extremely small mist droplets (micro-water droplets from 20 to 200 μm) absorbs heat very
effectively. In addition to extinguishing the fire, the discharged water mist binds the fumes from
fires, washes them out of the air, and carries them to the floor. Self-activating extinguishing sys-
tems are a primary tool for mitigating loss of life and property damage (Chapter 3).
The LifeCycle Tower concept clearly demonstrates the feasibility from structural and fire engi-
neering perspectives of constructing high-rise building superstructures that incorporate frame-
works made from glulam members or other large dimension structural timber elements. As with
any alternative system(s), the key to success is creation of a holistic overall concept and paying
close attention to technical details during implementation of such a concept.
The above-listed examples of reorientations of general aspects of structural design code require-
ments are likely to benefit engineers who design superstructures made of any material(s), and
owners and occupiers of buildings who experience the results of design decisions.
Additionally, it is highly desirable that design guidelines/manuals give explicit information rel-
evant to the calculation of deformations in structural systems and substructures, which is relevant
to the estimation of internal force distribution and deflections. Especially, there is a need for
guidance on the behaviour of connections that typically behave as semi-rigid spring elements. In
the absence of other guidance, designers would be wise to satisfy themselves that they adopt a
conservative interpretation of connection stiffness during estimation of internal force flows.
12.4 FINAL COMMENTS 169
Acknowledgements
The information and diagrams for the LifeCycle Tower concept was provided by Jakob Bonomo
and Harald Professner of Cree GmbH, Bregenz, Austria.
171
References
[1] Fux, X., Guo, D., Liu, X., Pan, G., Qian, Y., Sun, D. 2002. Chinese Architecture. Yale
University Press, New Haven, CT, USA.
[2] Smith, I., Snow, M. 2008. Timber: an ancient construction material with a bright future.
Forest Chron., 84(4): 504–510.
[3] Hu, S. 1991. The earthquake resistant-resistant properties of Chinese traditional construc-
tion. Earthquake Spect., 7(3): 355–389.
[4] European Committee for Standardisation (CEN). 2004. Eurocode 5—Design of Timber
Structures, Part 1-1: General—Common Rules and Rules for Buildings. EN 1995-1-1:2004,
CEN, Brussels, Belgium.
[5] Yenne, B. 2003. Seaplanes and Flying Boats: A Timeless Collection from Aviation’s
Golden Age. BCL Press, New York, NY, USA.
[6] Rihll, T.E. 2007. The Syrakosia or Alexandris, Department of Classics and Ancient
History. University of Swansea, UK, http://www.swan.ac.uk/grst/What’s%20what%20
Things/Syrakosia.htm.
[7] Lam, F., He, M. 2008. Example of traditional tall timber buildings in China – The Yingxian
Pagod. Struct. Eng. Int., 18(2): 126–129.
[8] Langenbach, R. 2008. Building tall with timber: a paean to wood construction. Struct.
Eng. Int., 18(2): 130–132.
[9] Foliente, G.C. 2000. History of timber construction. In: Wood Structures; An East-West
Forum on the Treatment, Conservation, and Repair of Cultural Heritage. Kelley, S.J.,
Loferski, J.R., Salenlikovich, A.J., Stern, E.G. (eds.), ASTM STP 1351, American Society
for Testing and Materials, West Conshohocken, PA, USA, 3–22.
[10] Canadian Standards Association (CSA). 2009. Engineering Design in Wood. CSA Stan-
dard 086-09, CSA, Toronto, Canada.
[11] Smith, I., Landis, E., Gong, M. 2003. Fracture and Fatigue in Wood. John Wiley and Sons,
Chichester, England.
[12] Smith, I. 1999. Joints in timber structures: state-of-the-art knowledge in North America.
COST C1 Int. Conf. Control Semi-rigid Behaviour Civil Eng. Struct. Connections. Liège,
Belgium, September 17–19, 1998, Office Publications European Communities, Luxem-
bourg, 255–264.
[13] Foliente, G.C. 1998. Design of timber structures subjected to extreme loads. Prog. Struct.
Eng. Mater., 1(3): 236–244.
172 REFERENCES
[14] Langenbach, R. 2008. Resisting earth’s forces: typologies of timber buildings in history.
Struct. Eng. Int., 18(2): 137–140.
[15] Ceccotti, A. 2008. New technologies for construction of medium-rise buildings in seismic
regions: the XLAM case. Struct. Eng. Int., 18(2): 156–165.
[16] Buchanan, A., Deam, B., Fragiacomo, M., Pampanin, S., Palermo, A. 2008. Multi-storey
prestressed timber buildings in New Zealand. Struct. Eng. Int., 18(2): 166–173.
[17] Institute for Research in Construction (IRC). 2005. National Building Code (updated
2010). IRC, National Research Council, Ottawa, Canada.
[18] European Committee for Standardization (CEN). 2003. Eurocode 8—Design of Structures
for Earthquake Resistance—Part 1: General Rules, Seismic Actions and Rules for
Buildings. EN 1998-1:2003, CEN, Brussels, Belgium.
[19] Smith, I. 2006. Reaching for the Limits with Timber Construction. Reports, Vol. 92.
IABSE Symposium on Responding to Tomorrow’s Challenges in Structural Engineering,
September 13–15, 2006, Budapest, International Association for Bridge and Structural
Engineering, Zurich, Switzerland.
[20] Smith, I., Frangi, A. 2008. Overview of design issues for tall timber buildings. Struct. Eng.
Int., 18(2): 141–147.
[21] Knoll, F., Vogel, T. 2009. Design for Robustness, Structural Engineering Document 11.
International Association for Bridge and Structural Engineering, Zurich, Switzerland.
[22] Smith, I., Snow, M., Asiz, A., Vasic, S. 2007. Failure mechanisms in wood-based materials:
discrete, continuum, and hybrid finite element representations. A review. Holzforschung,
61(4): 352–359.
[23] Smith, J.W. 2006. Structural robustness analysis and the fast fracture analogy. Struct. Eng.
Int., 16(2): 118–123.
[24] Östman, B., Rydholm, D. 2002. National fire regulations in relation to the use of wood
in European and some other countries, Swedish Institute for Wood Technology Research,
Trätek Rapport P0212044, Stockholm, Sweden.
[25] Mikkola, E. 1990. Charring of Wood, VTT Research Notes 689. Espoo, Finland.
[26] König, J., Walleij, L. 2000. Timber Frame Assemblies Exposed to Standard and
Parametric Fires, Part 2: A Design Model for Standard Fire Exposure. Report I 0001001,
Trätek – Swedish Institute for Wood Technology Research, Stockholm, Sweden.
[27] Young, S.A., Clancy, P. 2001. Structural modelling of light-timber framed walls in fire.
Fire Safety J., 36(3): 241–268.
[28] Frangi, A., Fontana, M. 2005. Fire performance based design of multi-storey timber build-
ings, Proceedings of the IABSE Symposium on Structures and Extreme Events, Lisbon,
Portugal, September 14–17.
[29] Fontana, M., Maag, T. 2004. Fire risk assessment based on Bayesian networks, Proceed-
ings of the 5th International Conference on Performance Based Codes and Fire Safety
Design Methods, Luxembourg, October 6–8, 2004, Society of Fire Protection Engineers,
Bethesda, MD, USA.
[30] Fontana, M., Favre, J.P., Fetz, C. 1999. A survey of 40,000 building fires in Switzerland.
Fire Safety J., 32(3): 137–158.
[31] Schweizerische Brandschutzvorschriften VKF, 2003. Vereinigung Kantonaler Feuerversi-
cherungen VKF. Bern, Switzerland.
[32] Buchanan, A.H. 2000. Fire performance of timber construction. Prog. Struct. Eng. Mater.,
2(3): 278–289.
173
[51] Schmid, J., König, J., Köhler, J. 2010. Fire-exposed cross-laminated timber—modelling
and tests, Proceedings von 11. World Conference on Timber Engineering, Riva del Garda,
Italy, June 20–24.
[52] British Standards Institution (BSI). 1990. Structural use of timber - Section 4.2, Recom-
mendations for calculating fire resistance of timber stud walls and joisted floor construc-
tions, Standard S 5268-4:1990, BSI, London, UK.
[53] Östman, B., König, J., Norén, J. 1994. Contribution to fire resistance of timber frame
assemblies by means of fire protective boards, Proceedings of the 3rd International Fire
and Materials Conference, Washington, DC, USA.
[54] König, J., Oksanen, T., Towler, K. 2000. A review of component additive methods used
for the determination of fire resistance of separating light timber frame construction,
Proceedings of CIB W18, Meeting thirty-three, Delft. Lehrstuhl für Ingenieurholzbau,
University of Karlsruhe, Karlsruhe, Germany.
[55] Schleifer, V. 2009. Zum Verhalten von raumabschliessenden mehrschichtigen Holzbau-
teilen im Brandfall, PhD Thesis - No. 18156, Eidgenössische Technische Hochschule,
Zürich, Switzerland.
[56] Frangi, A., Schleifer, V., Fontana, M. 2010, Design model for the verification of the
separating function of light timber frame assemblies. Eng. Struct., 32: 1184–1195.
[57] European Committee for Standardization (CEN). 2003. Fire Classification of Construc-
tion Products and Building Elements—Part 2: Classification using Data from Fire Resis-
tance Tests, Excluding Ventilation Services. EN 13501-2:2003, CEN, Brussels, Belgium.
[58] Dock Tower. 2002. Urbane Visionen an der Swissbau 02, Architektur & Technik, B+L
Verlags AG, Schlieren, Switzerland.
[59] Frangi, A., Fontana, M. 2005. Fire performance of timber structures under natural fire
conditions. Fire Safety Sci., 8: 279–290.
[60] Frangi, A., Bochicchio, G., Ceccotti, A., Lauriola, M.P. 2008. Natural full-scale fire test
on a 3 storey XLam timber building, 10th World Conference on Timber Engineering,
Miyazaki, Japan, June 2–5.
[61] Lennon, T., Bullock, M., Enjily, V. 2000. Medium rise timber frame 2000 stair fire test,
BRE Report No 200-711, Building Research Establishment, Cardington, UK.
[62] Hakkarainen, T. 2002. Post-flashover fires in light and heavy timber construction compart-
ments. J. Fire Sci., 20: 133–175.
[63] Standards Australia (SA). 2005. Specification for Preservative Treatment. Part 1: Sawn
and Round Timber. Australian Standard 1604.1-2005, SA, Sydney, Australia.
[64] International Organization for Standardization (ISO). 2008. General Principles on the Design
of Structures for Durability. ISO Standard 13823:2008(E), ISO, Geneva, Switzerland.
[65] Australian Building Codes Board (ABCB). 2002. Durability in Buildings. Guideline
Document, ABCB, Canberra, Australia.
[66] Bennett, A., Clark, S., Benge, C. 2001. Design for durability: A review of New Zealand
practice, Proceedings World Building Congress, Conseil International du Bâtiment,
Rotterdam, The Netherlands (CD-Rom).
[67] Foliente, G.C., Leicester, R.H., Wang, C-H., Mackenzie, C., Cole, I.S. 2002. Durability
design for wood construction. Forest Prod. J., 52(1): 10–19.
[68] Leicester, R.H., Wang, C-H., Nguyen, M.N., Foliente, G.C., Cole, I.S., MacKenzie, C. 2008.
Engineered durability: completion of a 10-year project, Proceedings World Conference
Timber Engineering, Paper No. 403, Wood Utilization Research Center, Miyazaki, Japan
(CD-Rom).
175
[69] Nguyen, M.N., Leicester, R.H., Wang, C-H. 2008. Embedded corrosion of fasteners
in timber structures. Manual No. 6, Commonwealth Scientific and Industrial Research
Organization - Sustainable Ecosystems, Highett, Australia.
[70] Leicester, R.H., Nguyen, M.N., Wang, C-H. 2008. Derivation of design equations for use
with AS1720.1. Manual No. 2, Commonwealth Scientific and Industrial Research Organi-
zation - Sustainable Ecosystems, Highett, Australia.
[71] MacKenzie, C.E., Wang, C-H., Leicester, R.H., Foliente, G.C., Nguyen, M.N. 2012.
Timber Service Life Design Guide. Forest and Wood Products Australia, Melbourne,
Australia (ISBN: 978-1-920883-16-4).
[72] Zabel, R.A., Morrell, J.J. 1992. Wood Microbiology; Decay and Its Prevention. Academic
Press, New York, NY, USA.
[73] MacKenzie, C.E., Wang, C-H., Leicester, R.H., Foliente, G.C., Nguyen, M.N. 2007. Tim-
ber Service Life Design Guide. Research Report, Forest and Wood Products Australia,
Melbourne, Australia.
[74] Leicester, R.H., Wang, C-H., Cookson, L.J. 2008. A reliability model for termite attack on
housing. Reliab. Eng. Syst. Safety, 93(3): 468–475.
[75] Grace, J.K., Abdallay, A., Farr, K.R. 1989. Eastern subterraneum termite
(isoptera:rhinotermitidae) foraging territories and population growth in Toronto. Canad.
Entomool., 121: 551–556.
[76] Su, N-Y., Scheffrahn, R.H. 2000. Termites as pest of buildings. In: Abe, T., Bignell, D.E.,
Higashi, M. (eds). Termites: Evolution, Sociality, Symbioses, Ecology. Kluwer Academic
Press, Dordrecht, The Netherlands, 437–453, Chapter 20.
[77] Nguyen, M.N., Leicester, R.H., Wang, C-H. 2008. Atmospheric corrosion of fasteners
in timber structures. Manual No. 5, Commonwealth Scientific and Industrial Research
Organization - Sustainable Ecosystems, Highett, Australia.
[78] Nguyen, M.N., Leicester, R.H., Wang, C-H., Foliente, G.C. 2008. A Draft Proposal of
AS1720.5 – Timber Service Life Design Code. Research Report, Commonwealth Scientific
and Industrial Research Organization: Sustainable Ecosystems, Highett, Australia.
[79] Kear, G., Wu, H-Z., Jones, M.S. 2007. Corrosion of ferrous- and zinc-based materials in CCA,
ACQ and CuAz timber preservative aqueous solutions. Mater. Struct., 41(8): 1405–1417.
[80] Leicester, R.H., Wang, C-H, Nguyen, M.N, MacKenzie, C.E. 2009. Design of exposed
timber structures. Aust. J. Struct. Eng., 9(3): 217–224.
[81] Canada Housing and Mortgage Corporation (CMHC). 1999. Wood Frame Envelopes in
the Coastal Climate of British Columbia. Best Practice Guide, CMHC, Ottawa, Canada
(ISBN 0-660-17765-X).
[82] Bassett, M.R., McNeil, S. 2005. The theory of ventilation drying applied to New Zealand
cavity walls, Proceedings Conference of Institution of Refrigeration, Heating and Air
Conditioning Engineers (IRHACE), IRHACE, Auckland, New Zealand (CD-Rom).
[83] Burnett, E., Straube, J. 1995. Tests, Ventilation Drying and Pressure Distribution.
Research Report, Building Engineering Group, Department Civil Engineering, University
of Waterloo, Waterloo, Canada.
[84] Cole, I.S., Chan, W.Y., Trinidad, G.S., Paterson, D.A. 2004. Holistic model for atmo-
spheric corrosion: 4 – Geographic information system for predicting airborne salinity.
Corros. Eng. Sci. Technol., 39(1): 89–96.
[85] Nguyen, M.N., Leicester, R.H., Wang, C-H., Foliente, G.C. 2013. Corrosion effects in
the structural design of metal fasteners for timber construction, Struct. Inf. Eng., 9(3):
275–284.
176 REFERENCES
[86] Weckendorf, J., Smith, I. 2012. Multi-functional interface concept for high-rise hybrid
building systems with structural timber, 12th World Conference on Timber Engineering.
July 16–19, Auckland, New Zealand (CD-Rom).
[87] Langenbach, R. 2010. Better than steel? The use of timber for large and tall buildings.
In: Cruz, P.J.S. (ed.) Structures and Architecture, Proceedings of the First International
Conference on Structures and Architecture, Guimaraes, Portugal, July 21-23, 2010).
CRC Press, Leiden, The Netherlands, 103–104 (extended version on CD-Rom).
[88] Macdonald, A.J. 2010. The changing relationship between architects and structural
engineers. In: Cruz, P.J.S. (ed.) Structures and Architecture, Proceedings of the First
International Conference on Structures and Architecture, Guimaraes, Portugal, July
21–23, 2010). CRC Press, Leiden, The Netherlands, 21 (extended version on CD-Rom).
[89] Asiz, A., Smith, I. 2009. Demands placed on steel frameworks of tall buildings having
reinforced concrete or massive wood horizontal slabs. Struct. Eng. Int., 19(4): 395–403.
[90] Madsen, B. 1998. Reliable timber connections. Prog. Struct. Eng. Mater., 1(3): 245–252.
[91] Madsen, B. 2000. Behaviour of Timber Connections. Timber Engineering Ltd., North
Vancouver, Canada.
[92] Canadian Wood Council (CWC). 2010. Wood Design Manual. CWC, Ottawa, Canada.
[93] Larsen, H.J., Enjily, V. 2009. Practical Design of Timber Structures to Eurocode 5.
Thomas Telford Ltd, London, UK.
[94] Pampanin, S. 2005. Emerging solutions for high seismic performance of precast-
prestressed concrete buildings. J. Adv. Concrete Technol., 3(2): 202–222.
[95] American Institute of Steel Construction (AISC). 2006. Steel Construction Manual, 13th
ed. AISC, Chicago, IL, USA.
[96] American Concrete Institute (ACI). 2008. Building Code Requirements for Structural
Concrete and Commentary, Standard 318–08, ACI, Farmington Hills, MI, USA.
[97] Wilson, E. 2004. Static and Dynamic Analysis of structures, 4th ed. Computer and
Structures Inc., Berkeley, CA, USA.
[98] Asiz, A., Smith, I. 2009. Structural connections for massive timber elements in hybrid
structures. Proceedings of 33rd IABSE Symposium, Bangkok, Thailand.
[99] International Code Council (ICC). 2012. International Building Code. ICC, Washington,
DC, USA.
[100] Lukaszewska, E., Johnsson, H., Fragiacomo, M. 2008. Performance of connections for
prefabricated timber-concrete composite floors. Mater. Struct., 41: 1533–1550.
[101] Newcombe, M.P., Carradine, D., Pampanin, S., Buchanan, A., van Beerschoten, W.A.,
Fragiacomo, A. 2009. In-plane experimental testing of timber-concrete composite floor
diaphragms, Proceedings of 2009 New Zealand Soc. Earth. Eng., Paper No 19, p8.
[102] American Society of Civil Engineers (ASCE). 2005. Minimum Design Loads for Build-
ings and Other Structures. Standard ASCE-7-05, ASCE, Reston, VA, USA.
[103] Ju, S.H., Lin, M.C. 1999. Comparison of building analysis assuming rigid or flexible
floors. J. Struct. Eng., 125 (1): 25–31.
[104] American National Standards Institute (ANSI). 2012. Standard for Performance-Rated
Cross-Laminated Timber, Standard ANSI/APA PRG 320-2012, ANSI, New York, NY, USA.
[105] European Committee for Standardization (CEN). 2012. Structural Timber—Strength
Classes—Assignment of Visual Grades and Species. EN 1912:2012, CEN, Brussels,
Belgium.
177
[123] Deutsches Institut für Normung e.V. (DIN). EN 1991-1-2/NA. 2010. Nationaler Anhang
- National festgelegte Parameter - Eurocode 1: Einwirkungen auf Tragwerke 1-2:
Allgemeine Einwirkungen – Brandeinwirkungen, German Edition, December 2010,
Beuth Verlag GmbH.
[124] Deutsches Institut für Normung e.V. (DIN). 1052. 2004. Entwurf, Berechnung und
Bemessung von Holzbauwerken – Allgemeine Bemessungsregeln und Bemessung-
sregeln für den Hochbau, August 2004, Beuth Verlag GmbH.
[125] Deutsches Institut für Normung e.V. (DIN). 4102-2. 1977. Brandverhalten von Baus-
toffen und Bauteilen, Bauteile: Begriffe, Anforderungen und Prüfungen. Beuth Verlag
GmbH.
[126] European Committee for Standardization (CEN). 2012. Fire resistance tests - Part 1:
General Requirements. EN 1363-1:2012, CEN, Brussels, Belgium.
[127] Gesetz über die Feuerwehren im Land Berlin (Feuerwehrgesetz – FwG), vom 23.
September 2003, Senatsverwaltung für Inneres, Berlin.
[128] Bard, D., Sjökvist, L-G. 2008. Sound transmission through a complete wood cross junc-
tion in a light-weight building. Proceedings of Internoise 2008, October 26–29, Shangaï,
China.
[129] Bolmsvik, Å. 2008. Evaluation of vibration distribution from a full scale measurement
in an eight storey wooden house. Proceedings of International Conference on Noise and
Vibration Engineering, September 15–17, Heverlee, Belgium, 2008, 593–607.
[130] Gustavsson, L., Joelsson, A., Sathre, R. 2009. Life cycle primary energy use and car-
bon emission of an eight-storey wood-framed apartment building. Energy Build., 42(2):
230–242.
[131] Klostermeier, C. 2008. Europameister - “Limnologen”: Achtgeschössige Holzhäuser in
Växjö/Schweden. Bauhandwerk, 7(8): 28–36.
[132] Petersson, Å., Holterman, L. 2008. Vibrations in a 7-Storey Wood Building, Report
TVSM-5157, Division of Structural Mechanics, Lund University, Lund, Sweden.
[133] Rosenkilde, A., Jarnerö, K., Axelson, M. 2008. Flervåningshus med trästomme –
Uppföljning av Kv Limnologen och Kv Rya, Rydebäck. Rapport 2008:18, SP Technical
Research Institute of Sweden, Borås, Sweden.
[134] Serrano, E. 2009. Documentation of the Limnologen project - Overview and summaries
of sub projects: Results. Report 56/2009, School of Technology and Design, Växjö Uni-
versity, Växjö, Sweden.
[135] Serrano E., Enquist B., Vessby J. 2010. Vertical relative displacements in a medium-rise
CLT-building, Proceedings of 1st International Conference on Structures and Architec-
ture. CRC Press, London, UK, 388–395.
[136] Schauerte, T. 2010. Consumer perceptions on wooden multistory houses: Segmenting
international markets. Proceedings of the International Convention of Society of Wood
Science and Technology and United Nations Economic Commission for Europe – Timber
Committee, October 11–14, Geneva, Switzerland, paper MA-1, 11.
[137] Smith, I. 2006. Reaching for the limits with timber construction, Proceedings of IABSE
Symposium Reaching to Tomorrow’s Challenges in Structural Engineering, Budapest,
Hungary, September 13–15, 384–385.
[138] Swailes, T., Marsh, J. 1998. Structural Appraisal of Iron-Framed Textile Mills. Thomas
Telford Ltd., London, UK.
Structural Engineering Documents
About the Authors: Objective:
To provide in-depth information to practic-
ing stuctural engineers in reports of high
Dr. Ian Smith is Lifetime Professor scientific and technical standards on a wide
Emeritus of Structural Engineering range of structural engineering topics.
at the University of New Brunswick in SED Editorial Board:
Canada, where he leads a research J. Sobrino, Spain (Chair); H. Subbarao,
India (Vice Chair); M. Bakhoum, Egypt; C.
group in hybrid construction. He holds Bob, Romania; M. Braestrup, Denmark;
Doctor of Philosophy and Doctor of M.G. Bruschi, USA; R. Geier, Austria; N.P.
Hoej, Switzerland; S. Kite, Hong Kong; D.
Science Degrees from London South Laefer, Ireland; R. Mor, Israel; H.H. (Bert)
Bank University in the United Kingdom. Snijder, The Netherlands; R. von Woelfel,
Germany.
Topics:
The International Association for Bridge
and Structural Engineering (IABSE) oper-
ates on a worldwide basis, with interests
of all type of structures, in all materials. Its
members represent structural engineers,
Dr. Andrea Frangi is Professor for employed in design, academe, construc-
Structural Engineering at the Depart- tion, regulation and renewal. IABSE or-
ganises conferences and publishes the
ment of Civil, Environmental and quarterly journal Structural Engineering
Geomatic Engineering at ETH Zurich, International (SEI), as well as reports and
monographs, including the SED series, and
where he leads the research group presents annual awards for achievements
of Timber Structures. He received his in structural engineering. With a member-
ship of some 4,000 individuals in more than
diploma in civil engineering and his 100 countries, IABSE is the international
doctoral degree from ETH Zurich. organisation for structural engineering.
Readership:
Practicing structural engineers, teachers,
researchers and students at a university
level, as well as representatives of owners,
operators and builders.
Publisher:
The International Association for Bridge
With Contributions From: and Structural Engineering (IABSE) was
founded as a non-profit scientific associa-
G.C. Foliente, R.H. Leicester, S. Gagnon, M.A.H. Mohammad, C. Ni, tion in 1929. Today it has more than 3900
M. Popovski, A. Asiz, A. Ceccotti, A. Polastri, S. Rivest, B. Kasal, members in over 90 countries. IABSE’s
D. Kruse, E. Serrano, J. Vessby, J. Bonomo, H. Professner. mission is to promote the exchange of
knowledge and to advance the practice of
structural engineering worldwide. IABSE
organizes conferences and publishes the
quarterly journal Structural Engineering In-
ternational, as well as conference reports
and other monographs, including the SED
series. IABSE also presents annual awards
for achievements in structural engineering.
For further Information:
IABSE
c/o ETH Zürich
CH-8093 Zürich, Switzerland
Phone: Int. + 41-44-633 2647
Fax: Int. + 41-44-633 1241
E-mail: [email protected]
Web: www.iabse.org
13
Structural Engineering Documents