Analysis of Fully Anchored Light-Frame Timber Shear Walls - Plastic Model

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 12

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/250058405

Analysis of fully anchored light-frame timber shear walls—elastic model

Article in Materials and Structures · September 2009


DOI: 10.1617/s11527-008-9463-x

CITATIONS READS

69 2,946

2 authors:

Bo Källsner Ulf Arne Girhammar


Linnaeus University Luleå University of Technology
46 PUBLICATIONS 476 CITATIONS 83 PUBLICATIONS 1,469 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Structural Robustness in Timber Buildings View project

All content following this page was uploaded by Ulf Arne Girhammar on 27 October 2014.

The user has requested enhancement of the downloaded file.


Engineering Structures 31 (2009) 2171–2181

Contents lists available at ScienceDirect

Engineering Structures
journal homepage: www.elsevier.com/locate/engstruct

Plastic models for analysis of fully anchored light-frame timber shear walls
Bo Källsner a,b , Ulf Arne Girhammar c,∗
a
School of Technology and Design, Växjö University, Växjö, Sweden
b
SP Wood Technology – Technical Research Institute of Sweden, Box 5609, SE-114 86 Stockholm, Sweden
c
Department of TFE – Civil Engineering, Faculty of Science and Technology, Umeå University, SE-901 87 Umeå, Sweden

article info abstract


Article history: In order to stabilise timber-framed buildings against lateral loads, the diaphragm action of roofs, floors
Available online 13 May 2009 and walls is often used. This paper deals with plastic analysis models for fully anchored sheathed shear
walls. The models are based on the assumption of plastic load–slip relations for the sheathing-to-framing
Keywords:
joints. Only static loads are considered. The basic structural behaviour and assumptions for the plastic
Shear walls
models are elucidated. Both upper and lower bound methods are applied. The load-bearing capacity and
Wall diaphragms
Plastic model the deformation of the shear walls in the ultimate and serviceability limit states, respectively, are derived.
Full anchorage Both a discrete point description and a continuous flow per unit length modelling of the fasteners are
Racking load capacity discussed. Also, the forces and displacements of the fasteners and sheathing are derived. The influence of
Shear wall displacement flexible framing members and shear deformations in the sheets, and also the effect of vertical loads on the
shear wall, both with respect to tilting and second order effects, on the horizontal load-bearing capacity
and displacement are evaluated. The stress distribution and the reaction forces at the ends of the different
framing members are derived. The elastic model is experimentally verified and an illustrative example is
given. The main objective of this work is to contribute to a better understanding of the structural behaviour
of these fully anchored walls and form the basis for establishing a new plastic design method for partially
anchored shear walls, i.e. a design method capable of analysing the more practical conditions of no or
partial anchorage of the studs and/or bottom rail in real structures.
© 2009 Elsevier Ltd. All rights reserved.

1. Introduction anchorage of the end studs of the shear wall. On other occasions
only the bottom rail is anchored to the floor or foundation. Due
In order to stabilise timber-framed buildings against lateral to economic reasons the building industry, at least in the Nordic
loads, the diaphragm action of roofs, floors and walls is often countries, desires to reduce the number of tie-downs, nails and
used. This paper is mainly focused on the structural behaviour screws to a minimum.
of shear walls. The purpose of it is to present the basic plastic This paper on fully anchored shear walls also serves as a
theories for fully anchored shear walls in order to elucidate the background for developing a new plastic design model for partially
fundamental behaviour of and assumptions for shear walls without anchored shear walls. By a fully anchored shear wall is meant a
imperfections. In a companion paper [1], these plastic models will wall where the vertical stud on the tension side is fully anchored
be used to evaluate the effect of imperfections such as gaps and to the floor or foundation. In a partially anchored shear wall, the
uplift. vertical stud is only partially anchored to the floor or foundation.
In the latter case, the bottom rail is usually assumed to be anchored
Shear walls are constructed in different ways depending on
to the floor or foundation.
local tradition. They can be prefabricated in a factory or built
For an extensive review of previous research on sheathed
on site. Prefabricated shear walls can consist of one or several
wood-framed shear walls and a more general description of the
wall assemblies of different sizes. The structural behaviour of
background, the reader is referred to Källsner & Girhammar [2].
shear walls is to a large extent determined by the sheathing-to-
The present and the companion [1] studies, where fully plastic
timber joints and how the walls are connected to the surrounding
properties of the mechanical sheathing-to-timber connections of
structure. Of particular importance is the anchoring of the shear the shear wall are assumed, are parallel with those studies where
wall to the floor or foundation. Sometimes tie-downs are used for elastic conditions are assumed for perfect [2] and for imperfect [3]
shear walls, respectively.
Expressions for both the horizontal load-bearing capacity and
∗ Corresponding author. Tel.: +46 90 786 74 95; fax: +46 90 786 64 69. the horizontal displacement of sheathed wood-framed shear walls
E-mail address: ulfarne.girhammar@tfe.umu.se (U.A. Girhammar). are derived.

0141-0296/$ – see front matter © 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.engstruct.2009.03.023
2172 B. Källsner, U.A. Girhammar / Engineering Structures 31 (2009) 2171–2181

a b (6) Edge distances of sheathing-to-framing joints are small


compared to the width and height of the sheets, i.e. the
fasteners are approximately located along the edges of the
sheets.
These assumptions will later be commented on.

3. Upper bound method—Kinematic theorem

The kinematic theorem for determining an upper bound of the


plastic load-carrying capacity is based on choosing a geometrically
possible pattern of deformations. By means of the principle of
virtual work, the internal work of all the fasteners is made equal
to the work of the external forces. In the formulation of the
internal work, each timber member can be regarded as a rigid body
rotating around its own centre of rotation relative to the sheet. The
true plastic load-carrying capacity of the shear wall is obtained
when the internal work of all the fasteners reaches its minimum
Fig. 1. (a) A shear wall unit built up of a timber frame and a sheet. The sheet value. The effect of assuming non-rigid conditions for the framing
is connected to the timber frame by means of mechanical fasteners. The centre
distances of the fasteners can vary in the general case. Here they are assumed to
members and sheets will also be evaluated.
be sr along the bottom and top rail, sps along the perimeter studs, and sis along the
intermediate stud, respectively; (b) Static model of a fully anchored shear wall unit 3.1. Horizontal load-bearing capacity—Discretely located fasteners
in a deformed state. The global coordinate system (x, y) has its origin in the lower
left corner, while the local coordinate system (x̂, ŷ) in the centre of gravity of the
fasteners (CG), usually in the middle of the sheet. For a wall unit according to Fig. 1(a) and the same wall
segment in a deformed state according to Fig. 1(b), the principle is
2. Plastic models—Upper and lower bound methods demonstrated in Fig. 2, where the assumed positions of the centre
of rotations (CR) and the matching fastener forces are shown. All
In the elastic analysis [2], a linear elastic load–displacement joints with a fastener shear capacity, Fv , have reached the plastic
relationship until failure for the sheathing-to-framing joints was capacity, Fv = Fp . Since the timber members are assumed to be
assumed. From tests of nailed joints it is well known that the connected to each other by hinges at the ends, each fastener in
load–slip curves are often characterised by plastic deformations. this position can be assumed to rotate either around the centre
Methods for determining the upper and lower bounds of the plastic of rotation of the adjacent horizontal timber member or around
load-carrying capacity of a shear wall have earlier been presented the centre of rotation of the adjacent vertical timber member. This
by Källsner & Lam [4] and Källsner et al. [5]. Here these plastic means that these two centres of rotation must fall on a straight line
models are generalised and extended. through the hinge under consideration. As in the elastic solution,
The upper bound method is based on the kinematic theorem the vertical frame members are assumed to rotate through the
and gives a load-bearing capacity that is higher than or equal to the angle γ and the sheet through the angle ϕ in relation to their initial
exact plastic load-bearing capacity and the lower bound method is positions. This means that the horizontal timber members rotate
based on the static theorem and gives a capacity that is lower than through the angle ϕ and the vertical members rotate through the
or equal to the exact value, Neal [6]. angle γ − ϕ in relation to the sheet.
Both these models are based on the assumption that the If the internal work is divided into contributions originating
load–displacement relationships of the sheathing-to-framing joints from the fasteners in the horizontal top and bottom rails (r) and in
are completely plastic. Except for this plastic assumption, all other the vertical perimeter (ps) and intermediate (is) studs, the relation
assumptions given for the elastic model are also applicable to the between the internal and the external work gives
plastic methods. X X X
A typical part of a fully anchored shear wall is shown in Fig. 1(a). 2 Fp rr,i ϕ + 2 Fp rps,i (γ − ϕ) + Fp ris,i (γ − ϕ) = Hhγ
This fundamental unit, consisting of a sheet fastened to a timber r ps is

frame, will be called a wall unit or wall segment. The width of the (1a)
segment is, b, and the height is, h. The spacing of the fasteners (si )
along the top and bottom rail is denoted sr , along the perimeter or
studs sps , and along the intermediate stud sis . The model is thus " X #
γ γ
X  X
based on the following assumptions: 2Fp rr,i + Fp −1 2 rps,i + ris,i = Hh (1b)
r
ϕ ps is
ϕ
(1) Framing members and sheets are rigid. However, the influence
of assuming flexible framing members and shear deformations where ri is the distance from each fastener along the different
in the sheets is also evaluated. framing members to its centre of rotation, Σr is the summation
(2) No contact between adjacent sheets or between sheets and over the fasteners on the horizontal top and bottom rails (r), Σps is
surrounding structure (sheets are free to rotate). the summation over the fasteners on the vertical perimeter studs
(3) Framing joints act as hinges. (ps) and Σis the summation over the fasteners on the vertical
(4) Sheathing-to-framing joints are completely plastic. The prop- intermediate stud (is).
erties are the same for all joints; they are independent of the The different distances to centres of rotation are given by
force direction and of the mutual orientation of the sheets and
framing members. + n2r q
X X
(5) Displacements of the wall are small compared to the width and rr,i = rr2,0 + (i · sr )2 (2a)
height of the sheets. r n
i=− 2r
B. Källsner, U.A. Girhammar / Engineering Structures 31 (2009) 2171–2181 2173

Fig. 2. Force distribution on the sheet according to the plastic upper bound method. Each of the five timber members rotates around its own centre of rotation (CR).

n 
+ 2
ps
−1 Insertion of Eq. (4) into Eq. (1) gives the racking load-carrying
q
capacity, H, as
X X
rps,i = 2
rps ,0 + (i · sps )
2 (2b)
ps
n  " #
ps
i=− −1
2 2r
Fp hr,0
P P P
2Fp rr,i + 2 rps,i + ris,i
r ps is

n

+ 2is −1
H =   . (5)
2rr,0
X X
ris,i = |i| sis (2c) h 1+ h
is

nis

i=− −1
2 The unknown quantity rr,0 /h in Eq. (5) must be estimated before
the capacity H can be calculated. A reasonable estimate of rr,0 /h
where rr,0 and rps,0 is the vertical distance between the top or can be made by determining the centres of rotation valid for elastic
bottom rail and the centre of rotation for the rail, and the horizontal conditions, i.e. using the γ /ϕ -value obtained by the elastic model
distance between the perimeter stud and the centre of rotation according to Källsner & Girhammar [2] and then calculating rr,0 /h
for the perimeter stud, respectively, according to Fig. 2, sr , sps , and from Eq. (4). The positions of the other centres of rotation can be
sis is the constant spacing between the fasteners along the top determined graphically as shown in Fig. 2. The two distances to the
and bottom rails, along the vertical perimeter studs, and along centres of rotation under elastic conditions are given by [2]
the intermediate stud, respectively, and nr = b/sr , nps = h/sps , n
and nis = h/sis is the number of fastener spacings along the top x̂2i
P
h i=1
and bottom rails, along the vertical perimeter studs, and along the r r ,0 = (6a)
n
intermediate stud, respectively. If sr = sps , then nps = nr · (h/b). 2P
ŷ2i
For reasons of symmetry, the vertical timber members must i =1

have their centres of rotation along a horizontal line through the n


ŷ2i
P
centre of the sheet, while the horizontal members must have
b i=1
their centres of rotation along a vertical line through the centre rps,0 = n
. (6b)
2P
of the sheet. From the condition that the hinge between the upper x̂2i
horizontal timber member and the central vertical member moves i=1

the same distance in relation to the two centres of rotation, a When all the centres of rotation are given, the ri -values can easily
relation between the angles ϕ and γ can be formulated as be determined and H can be calculated.
The ‘‘true’’ plastic load-bearing capacity of the wall is obtained
h when the internal work of all the fasteners reach its minimum
rr,0 ϕ = (γ − ϕ). (3) value, i.e. when the left side of Eq. (1) has its minimum. The worse
2
the locations for the different centres of rotation are chosen, the
Rearranging Eq. (3) gives greater the internal work done and, hence, also the greater the
load-carrying capacity.
γ 2rr,0 Now, consider a wall unit with three vertical frame members
=1+ . (4) with fasteners uniformly distributed along the frame members
ϕ h
2174 B. Källsner, U.A. Girhammar / Engineering Structures 31 (2009) 2171–2181

according to Fig. 1(a). The spacing between the fasteners along the sis = 200 mm. If the location of the centres of rotation according
perimeter of the wall unit is constant (sps = sr ). Also, the spacing to the elastic theory [2] is used, the following results are obtained
of fasteners along the intermediate stud is constant, but twice as γ
much as that along the other frame members (sis = 2sps ). For rr,0 = 0.391 m; rps,0 = 1.844 m; = 1.325 (10)
ϕ
such a wall with a height-to-width ratio of h/b = 2, we arrive
approximately at γ /ϕ = 1+rr,0 /b = 1.32, rr,0 = 7b/22 = 0.318b, H = 1.165fp b. (11)
and rps,0 = 11b/7 = 1.57b according to Källsner & Girhammar [2]. For the ‘‘true’’ or exact location of the centres of rotation under
plastic conditions, these parameters are given by
γ
3.2. Horizontal load-bearing capacity — Uniformly distributed rr,0 = 0.286 m; rps,0 = 2.517 m; = 1.238 (12)
fasteners
ϕ
H = 1.159fp b. (13)
Instead of using the fasteners as discretely located points as The load-carrying capacity according to Eq. (11) or (13) should
described above, it is often convenient for analysis purposes to be compared to the elastic solution according to Källsner &
smear them continuously along the framing members, i.e. to model Girhammar [2], i.e. H = 0.984fp b. As expected, the plastic solution
the shear forces of the fasteners as a shear force per unit length, gives a higher racking capacity, about 18% in this example.
a shear flow, f [N/m]. Then, the plastic capacity per unit length The upper bound method is not suitable as a hand calculation
method for practical design work, primarily for two reasons.
is given by fp = Fp /sr [N/m], i.e. the plastic capacity of the
One reason is that the load-bearing capacity is overestimated
fastener divided by the spacing sr between the fasteners along
and another is that the method is cumbersome. The reason for
the top or bottom rail (the reference spacing). For uniformly
discussing the method in this paper is to elucidate the behaviour
distributed plastic shear capacity of the fasteners, the summations of the composite wall structure in an illustrative way where each
are transferred into integrals, i.e. timber member has its own CR relative the sheet and to estimate
h i the displacements in a shear wall under ideally plastic conditions.
2rr,0
2 ssr sr
R R R
2fp r dx̂
r r ,l
+ fp h ps
r dŷ
ps ps,l
+ sis
r dŷ
is is,l
H = 2rr,0
(7) 3.4. Effect of flexibility of framing members on horizontal load-
h(1 + h
) bearing capacity
where
The analysis above was based on the assumption of rigid
Z Z + 2b q framing members. If full flexibility of a framing member is
rr,l dx̂ = rr2,0 + x̂2 dx̂ (8a) assumed, no forces perpendicular to that member develop. The
r − 2b Eqs. (2), (6), (8) and (9) for the shear wall change in such a way
that the terms associated with forces perpendicular to the framing
Z Z + 2h
members are omitted. The difference between the discrete and the
q
2 2
rps,l dŷ = rps ,0 + ŷ dŷ (8b) uniform fastener descriptions is how the fasteners in the corners
ps − 2h
are included in the different models. Eqs. (2) and (8) are now given
Z Z + 2h by
ris,l dŷ = ŷdŷ (8c) X b
is − 2h rr,i = (nr + 1)rr,0 ≈ nr rr,0 = r r ,0 (14a)
r
sr
where (according to the elastic approximation)
X h
2
R + 2b b2
R + 2h rps,i = (nps − 1)rps,0 ≈ nps rps,0 = rps,0
x̂2 2 sps
x̂ dx̂ + dx̂
R
h r+ps+is sl
dx̂ h sr − 2b 2sps − 2h ps
r r ,0 = = h nr h 2
ŷ2
R +h R + 2b
2 2
R 
dŷ 2 1 2 h2 = nr rps,0 = (14b)
r+ps+is sl + ŷ2 dŷ + dŷ b 4rr,0
sps sis − 2h 2sr − 2b
(9a)
X
ris,i = 0 (14c)
 R +h R + 2b
ŷ2 2 1 2 2 h2 is
+ ŷ dŷ + dŷ
R
b dŷ b sps sis − 2h 2sr − 2b
r+ps+is sl for discretely located fasteners, and
rps,0 = =
2 x̂2 2 R + 2b R + 2h
R
dx̂ b2
2
Z
r+ps+is sl x̂2 dx̂ + dx̂
sr − 2b 2sps − 2h rr,l dx̂ = brr,0 (15a)
r
(9b)
h2 b
Z
where sl is the spacing along the different frame members (l = r, rps,l dŷ = hrps,0 = (15b)
ps 4rr,0
ps, and is, respectively). Z
ris,l dŷ = 0 (15c)
3.3. Illustrative example with respect to horizontal load-bearing is

capacity for uniformly distributed fasteners. The influence of the corner


fasteners is approximated as is evident from Eqs. (14a) and (14b).
In Eqs. (14b) and (15b), the relations sr = sps and rr,0 · rps,0 =
No closed solutions can be given for the horizontal load-
hb/4 has been assumed. The horizontal load-carrying capacity then
carrying capacity according to Eqs. (5) and (7). The true solution becomes
can only be obtained by trial and error. To explicitly evaluate the
Fp
horizontal load-bearing capacity for a wall segment according to H = nr Fp = b = fp b. (16)
Fig. 1(a), it is necessary to know the actual dimensions of the wall sr
and the spacings of the fasteners. Therefore, consider a wall unit This capacity is the same as the one obtained by using the lower
with h = 2400 mm, b = 1200 mm, sr = sps = 100 mm and bound method, see Eq. (23).
B. Källsner, U.A. Girhammar / Engineering Structures 31 (2009) 2171–2181 2175

3.5. Effect of shear deformation of the sheet on horizontal load- 3.8. Effect of flexible framing members on horizontal displacement
bearing capacity
The analysis above was based on the assumption of rigid
The analyses in Sections 3.1 and 3.2 were based on the framing members. In the case of fully flexible framing members
assumption of rigid sheets. In determining the plastic collapse load and for sr = sps and h/b = 2, the horizontal wall displacement and
in the ultimate limit state, elastic deformations occurring in the fastener displacement are given by [2]
members of the shear wall up to that state have no influence
on the load-carrying capacity. Elastic properties of the sheet only sr H
uframe ≈ 6 = 4.24δcorner (20)
influence the horizontal displacement of the shear wall in the b k
serviceability limit state. For plastic solutions, only the extreme √ sr H sr H
cases of completely flexible or fully rigid sheets are of interest. In δcorner ≈ 2 = 1.41 . (21)
b k b k
case of completely flexible sheets, no plastic solution exists. Hence,
the horizontal load-carrying capacity remains the same according
to Eq. (5) or (7) even if shear deformations in the sheets take place. 3.9. Effect of shear deformation of the sheet on horizontal displace-
ment
3.6. Effect of vertical loads on horizontal load-bearing capacity
The analysis above was also based on the assumption of rigid
The influence of vertical loads on the horizontal load-bearing sheets. The influence of shear deformations in the sheathing
capacity is not treated in this section. In principle, it is the same as material can be estimated by introducing the shear angle γs =
is shown in Section 4 dealing with the lower bound method. H /Gbt [2]. The shear angle is assumed to be constant over
the entire sheet and the shear modulus the same in both the
3.7. Horizontal displacements of the shear wall and fastener displace- serviceability and ultimate limit states for the shear wall as a
ment whole. The displacement of the top rail of the frame according to
Fig. 1(b) is then given by [2]
As mentioned above, a reasonable estimate of rr,0 /h can be sr H

sr k h

H
obtained by using the value given by the linear elastic model [2]. uframe = (γ + γs )h = 4.52 + γs h = 4.52 + .
Thus, the displacement pattern (and force distribution), due to a b k b Gt b k
rotation of the frame through the angle γ and a rotation of the (22)
sheet through the angle ϕ (Fig. 1(b)) is approximately the same
For normally used sheathing materials and fastener distances, the
as for the elastic model (γ /ϕ = 1.32), i.e. the relation between
influence of the shear deformation on the displacements is often
the largest displacements in the fasteners in the corners and the
neglected.
horizontal displacement of the frame of the shear wall is the
same as given for the elastic model by Källsner & Girhammar [2].
This means that displacements calculated in this way are valid in 3.10. Effect of vertical loads on horizontal displacement
the elastic stage up to the plastic level is reached, usually within
the serviceability limit state. With an ideally plastic model, the The influence of vertical loads on the horizontal displacement
displacements in the plastic stage cannot be evaluated. However, of the shear wall is not treated in this section. In principle, it is
by using the secant values of the slip modulus (k) for the most the same as is shown in Section 4 dealing with the lower bound
strained fasteners (in the corners), using gradual values at chosen method.
progressive levels of the fastener force–displacement relationship,
the displacement of the shear wall in the ultimate limit state can 4. Lower bound method—Static theorem
be calculated. The horizontal displacement of the top rail of the
shear wall (uframe ) can be obtained by the geometry of the wall unit A lower bound of the plastic load-carrying capacity of a shear
according to Fig. 1(b) as wall can be obtained by assuming a force distribution that fulfils
the conditions of force and moment equilibrium and where the
h force on each fastener is at most equal to the plastic capacity of
uframe = γ h = 2ucorner + 2 vcorner (17)
b the fastener. At the same time, the constituent materials in the
where ucorner , vcorner are the displacements of the fasteners in the structure (timber members, sheets, and joints) must be able to
corners in each direction, respectively. For shear walls of common transfer the internal and external forces without exceeding the
design (h = 2b; sr = sps = sis /2), the horizontal displacement is strength of the materials. Also, the deformation capacity of the
given by [2] constituent materials and components needs to be sufficient.
In the present lower bound model, the framing members are
sr H
uframe ≈ 4.52 = 4.45δcorner (18) assumed to be completely flexible. This implies that the force
b k distribution from the fasteners will become parallel with the
where k is the slip modulus of the most strained fastener, and the framing members. The forces of the fasteners in the corners are
maximum displacement of the fasteners in the corners in the same here chosen as 0.5 F in each direction.
direction as that of the corresponding maximum fastener force
by [2] 4.1. Horizontal load-bearing capacity
Fcorner
q
δcorner = = u2corner + vcorner
2
Again, consider a shear wall with a fully anchored leading
k
stud, i.e. a wall where the end stud at the load-application-point
p sr H sr H
≈ 0.5452 + 0.8572 = 1.02 is fully anchored to the floor or foundation. For a wall unit as
b k b k in Fig. 1(a), where the distance of the fasteners along the edges
is constant, i.e. sr = sps , it is possible to find a simple force
p
= 0.121 + 0.190 uframe = 0.225uframe .
2 2
(19)
distribution according to Fig. 3. Each edge fastener, except those
Eq. (18) overestimates the displacements of the shear wall, since all in the corners, may be assumed to carry the same plastic load Fp
fasteners except those in the corners have a higher slip modulus. parallel with the edge. Each corner fastener is assumed to carry
2176 B. Källsner, U.A. Girhammar / Engineering Structures 31 (2009) 2171–2181

Fig. 3. Force distribution on the sheet according to the plastic lower bound method
assuming a pure shear flow.

one load component Fp /2 parallel with each of the associated sides Fig. 4. Stress distribution on the sheet and the framing members in a fully anchored
shear wall according to the plastic lower bound theory assuming a pure shear flow.
of the sheet. The fasteners in the centre stud are assumed not to
carry any load. For the chosen force distribution, there are no force
components perpendicular to the length or grain direction of the 4.3. Effect of shear deformation of the sheet on horizontal load-
timber members. As a consequence, there are no demands on any bearing capacity
capability of the framing joints between the timber members to
transfer shearing forces. As discussed in Section 3.5, the horizontal load-carrying
With the chosen force distribution, which corresponds to a pure capacity is not changed due to any shear deformations in the
shear flow along the edges of the sheet, fp = Fp /sr , the horizontal sheets.
load-carrying capacity, H, is given by
4.4. Effect of vertical loads on horizontal load-bearing capacity
b
H = nr F p = Fp = fp b for sps = sr (23)
sr The influence of unintentionally tilted vertical studs on the
horizontal load-carrying capacity can be accounted for by adjusting
where nr = b/sr is the number of fastener spacings along the top
the capacity with respect to a corresponding fictitious horizontal
rail and Fp the plastic load-carrying capacity of fasteners loaded in load given by [2]
shear. According to Källsner & Girhammar [2], the corresponding
elastic capacity is given by H = 0.984(b/sr )Fv = 0.984fp b. It is α1 m
 X
evident that the elastic capacity is almost the same as the plastic Htilt = α0 + √ Vi (24)
m i =1
capacity according to Eq. (23). The lower bound plastic capacity is
only about 2% higher than the elastic solution and about 16% lower where Vi (i = 1 to m) is the vertical load on the ith stud, m the
than the true plastic solution for the example according to Eq. (13). number of studs, α0 (=0.003) the systematic and α1 (=0.012) the
It is also clear from Eq. (23) that for a fully anchored shear wall, random angle of tilting of studs, respectively. Second order effects
the entire length of the wall (l = b) is effective in resisting the can in a corresponding way be written as
horizontal racking load, H, along the bottom rail.
m
Later a new plastic method of analysis will be developed uframe X
H2nd = Vi (25)
applicable to partially anchored shear walls, which is based on this h i =1
lower bound method. As a consequence only a part of the length
of the wall will be effective in resisting the horizontal racking load, where uframe is the horizontal displacement of the top rail of the
H, along the bottom rail, i.e. Hpartial = fp leff < Hfull = fp l. shear wall. The effective horizontal load-bearing capacity then
becomes

4.2. Effect of flexible framing members on horizontal load-bearing Heff = H − Htilt − H2nd
capacity
" √ m #
m
uframe X α0 + α1 / m X
= 1− Vi − Vi H (26)
Hh i=1
H i=1
The analysis above was based on the assumption of completely
flexible framing members, i.e. no forces develop perpendicular where uframe is given by Eq. (37) and H by Eq. (23). Usually, the
to that member. When using the lower bound method, it is not effect of tilted studs and, especially, the second order effects can
expedient to assume any rigidity of the framing members. be neglected.
B. Källsner, U.A. Girhammar / Engineering Structures 31 (2009) 2171–2181 2177

4.5. Vertical reactions at ends of wall due to overturning moment the studs. In the case of out of plane buckling of the trailing stud,
the external vertical load acting on the top of the stud together with
The reaction forces related to uplift of the leading stud the shear forces from the sheathing are taken into account in the
and compression of the trailing stud, respectively, due to the analysis. No composite action between the stud and the sheet can
overturning moment is according to Fig. 1(b) given by be utilised. Concerning buckling with respect to the weak axis, it
is reasonable to assume that the sheathing-to-framing joints will
h
Rt = Rc = H . (27) counteract this failure mode.
b
For a wall unit according to Fig. 1(a), the force distribution can be 4.6. Buckling of sheathing
chosen as shown in Fig. 4. All timber members and the sheet are in
force and moment equilibrium. The force distribution corresponds If the sheets are very thin, there is a risk of local buckling of the
to a pure shear flow (fp ). The spacing between the fasteners along sheets. If we consider the force distribution according to the plastic
the perimeter of the wall unit is constant (sps = sr ). Also, the lower bound method, see Fig. 3, where all fasteners along the edges
spacing of fasteners along the intermediate stud is constant, but except those in the corners carry the same load, it is reasonable
twice as much as that along the other frame members (sis = 2sps ). to determine the critical stress as if the sheet were subjected to a
However, the influence of the intermediate stud is not considered constant shear stress. The details of the calculations are given in
in this choice of plastic equilibrium. Due to the fact that the plastic Appendix A.
stress is distributed only parallel with the framing members and
no stress is acting perpendicular to them, no horizontal reaction 4.7. Shear walls comprising of more than one unit and/or of sheets on
forces develop in the studs and no vertical reaction forces on top of both sides
the studs.
The vertical force along the leading and trailing studs in Fig. 4 is The load-bearing capacity of a shear wall comprising several
given by wall segments is given as the sum of the load-bearing capacity of
the individual parts. For a wall panel consisting of only one sheet
Rstud = RV2 = fp h. (28)
fixed to one side of the timber frame the racking load-carrying
This is the highest value of the force that any stud as part of the capacity is given by Eq. (23). In Appendix B, the principles for
shear wall can transfer from one end to the other under pure plastic calculating the horizontal load-carrying capacity for shear walls
conditions. For example, that part of a vertical downward normal built up of more than one wall unit are given.
load acting on the top of the a stud that exceeds this ‘‘full shear For shear walls with sheathing (of the same design) on both
capacity’’ of the stud, will only pass through the stud down to the sides, the load-bearing capacity is twice the value given by the
floor or foundation without affecting the force distribution in the formulas above in a corresponding way as discussed in Källsner &
wall. Girhammar [2].
The condition of a fully anchored shear wall gives the highest
capacity that the shear wall can carry. 4.8. Horizontal displacement of the shear wall and fastener displace-
Moment equilibrium of the sheet requires that the shear ment
flow along the vertical (fV ) and horizontal (fH ) sides of the wall
(framing members and sheet) are equal to fp . The horizontal force Assuming elastic conditions up to the state of pure plastic shear
equilibrium of the top or bottom rail gives flow around the sheet (fH = fV = fp as shown in Fig. 4), the
H displacement of the fasteners in the x̂ and ŷ directions (ui , vi ) can
fp = . (29) be estimated to (cf. also Källsner & Girhammar [2])
b
The maximum forces on the fasteners along the perimeter are then FH fH sr f p sr sr H
ui = = = = (32a)
given by k k k b k
sr FV fV sps fp sr sr H
Fmax = fp sr = H. (30) vi = = = = (32b)
b k k k b k
The vertical force component between the stud and the bottom rail where k is the elastic slip modulus of the fasteners in the service-
is given by ability limit state. Thus, the absolute value of the displacement of
the fasteners can be written as
h
RV2 = fp h = H = Rt = Rc . (31) δ ≡ ui = v i . (33)
b
Comparing this result according to Eq. (31) with the previous result The displacement δ is parallel with the framing members around
according Eq. (27) shows that 100% of the overturning moment is the sheet. This holds true for all fasteners, except those in the
taken as a force couple in the perimeter vertical framing members, corners. In the present plastic model with a constant shear flow
i.e. the vertical force is transmitted fully to the floor or foundation along the edges of the shear wall, the force components of the
via the bottom rail and the tie-down, respectively. For comparison fasteners in the corners are equal to 0.5Fp and, thus, the total

with the elastic solution, it is shown by Källsner & Girhammar [2] force of the fasteners in the corners becomes only equal to Fp / 2
that 93% of the overturning moment is taken as a force couple in and they do not reach their plastic capacity. Their √ pertaining
the vertical frame members (the corresponding reaction forces), displacements become δcorner = (sr /b)(H /k)/ 2, now in a
while the remaining 7% of the moment is taken as fastener forces direction of 45◦ .
between the sheet itself and the bottom rails. A solution where the forces in the fasteners in the corners reach
Concerning the contact pressure on the bottom rail and anchorage their plastic capacity, Fp , is given in Appendix C. It is shown that
of the tensile stud, the same considerations as was discussed for the the components of the fastener forces in the corner (c) increase
elastic model still apply, see Källsner & Girhammar [2]. Buckling of to Fc,x̂ = 0.636Fp and Fc,ŷ = 0.772Fp , and that the horizontal
the trailing stud should be checked. Shear stresses acting along the load-carrying capacity is increased by an amount of 2(0.636 −
edges of the sheets introduce compression and bending stresses in 0.5)Fp = 0.272Fp . With this force distribution, the displacements
2178 B. Källsner, U.A. Girhammar / Engineering Structures 31 (2009) 2171–2181

of the fasteners in the corners become δcorner = (sr /b)(H /k) as 4.11. Effect of vertical loads on horizontal displacement
given in Eq. (22), but now in a direction of 39.5◦ to the horizontal
plane. The influence of unintentionally tilted vertical studs and second
order effects on the horizontal displacement of the shear wall can
It should be noted, however, that a force distribution where the
be accounted for by adding to the horizontal load on the shear wall
force components in the corners equal 0.5Fp is to be preferred from
the corresponding fictitious horizontal loads as given by Eqs. (24)
a strength point of view. The capacity of the fasteners in the corners and (25). Thus, the horizontal load (H) in Eq. (23) is replaced by the
can be expected to be less than the capacity of the other fasteners total load,
due the special edge distances occurring in the corners. From the
Htot = H + Htilt + H2nd (38)
elastic model [2] it is concluded that the fasteners in the corners
are those that are most strained. and the effective horizontal displacement can be written as (see
Källsner and Girhammar [2])
Assuming that ui = vi = δ , the total horizontal displacement of
the top rail of a shear wall with one segment is given by (cf. Fig. 1(b) √ m
1 + [(α0 + α1 / m)/H ]
P
Vi
and cf. also Källsner & Girhammar [2]) i=1
uframe,eff = m
uframe (39)
1 − (uframe /Hh)
    P
h h h sr H Vi
uframe = 2ui + 2 vi = 2 1 + δ =2 1+ . (34) i=1
b b b b k
where H is given by Eq. (23) and uframe by Eq. (35). Usually, the
For ordinary shear wall designs, h/b = 2, the relation between the effect of tilted studs and, especially, the second order effects can
displacements of the fasteners and the displacement of the wall as be neglected.
a whole is then given by
4.12. Illustrative example
sr H
uframe = 6δ = 6 (35) Consider a typical shear wall as shown in Fig. 1 and composed
b k
of a 2400 × 1200 × 8 mm Masonite fibre hardboard (C40) (mean
or shear modulus, G = 2100 MPa) nailed to the frame by using
annular ringed shank nails 50 × 2.1 (mean ultimate limit load,
δ = ui = vi = 0.167uframe . (36) Fv = 1 kN, according to Girhammar et al. [11]). In the serviceability
limit state (considered as a serviceability limit load corresponding
According to Källsner & Girhammar [2], the elastic solution gives, to 40% of the ultimate limit load), the mean secant slip modulus of
uframe = 4.52(sr /b)(H /k) which also corresponds to Eq. (18) the sheathing-to-framing joints is k = 800 kN/m (cf. Girhammar
for the approximate plastic upper bound solution assuming rigid et al. [11]). The aspect ratio is h/b = 2 and the spacing sr = sps =
sis /2 = 100 mm. Thus, the shear flow is fp = 10 kN/m. The wall is
framing members (and sheets). Thus, according to plastic lower
subjected to a vertical load of 6 kN on each stud (3 studs).
bound conditions, the displacement of the shear wall is 6 times that
According to the lower bound method, Eq. (23), the horizontal
of the displacement of the joints. load-carrying capacity of the shear wall is given by H = fp b =
As was pointed out in Section 3.7, the horizontal displacement 12.0 kN. The corresponding elastic value is H = 0.984(b/sr )Fv =
of the shear wall in the ultimate limit state can be obtained by 11.8 kN. The difference between plastic and elastic values is 1.6%.
using gradual values of the secant slip modulus k of the most The small difference between the elastic and plastic load-carrying
strained fastener up to the level where the displacement wants to capacity is due to assumption of rigid framing members in the
be calculated. elastic solution and fully flexible framing members in the plastic
solution. According to Eq. (35), the horizontal displacement of the
shear wall is given by uframe = 6(sr /b)(0.40H /k) = 3.00 mm. The
4.9. Effect of flexible framing members on horizontal displacement elastic value is uframe = 4.52(sr /b)(0.40H /k) = 2.22 mm, i.e. a
difference of 35% in the serviceability limit state. This big difference
with respect to the displacements is due to the assumption of fully
As mentioned in Section 4.2, the analysis above was based on flexible members in the plastic solution.
the assumption of completely flexible framing members, and when The influence of the shear modulus is evaluated by using
using the lower bound method, it is not expedient to assume any Eq. (37), i.e. uframe = [6 + (h/b)(k/Gt )(b/sr )](sr /b)(0.40H /k) =
rigidity of the framing members. (6 + 1.143)(sr /b)(0.40H /k) = 3.57 mm. Thus, the influence is
about 19%.
The effect of vertical loads on the horizontal capacity is given
by Eq. (26) and on the horizontal displacement by Eq. (39). To
4.10. Effect of shear deformation of the sheet on horizontal displace-
illustrate, assume a secant slip modulus in the ultimate limit state
ment
that is half the value valid for the serviceability limit state, i.e. k =
400 kN/m (in a real situation gradual values should be used for
Corresponding to the discussion in Section 3.9, the displace- the progressive displacement process). The effective horizontal
ment of the top rail of the frame is given by capacity √becomes, Heff = [1−(2×3.57/2400)(18/12.0)−(0.003+
0.012/ 3)(18/12.0)]H = (1 − 0.004465 − 0.014892)H =

sr k h

H 0.981H = 11.8 kN. Then, the effective capacity is 2.0% lower
uframe = 6 + . (37) than the capacity neglecting the influence of vertical loads, where
b Gt b k 0.5% refers to second order effects and 1.5% to effects of tilted
studs. Correspondingly, the effective horizontal displacement
For normally used sheathing materials and fastener distances, the
becomes, uframe,eff = [(1 + 0.014892)/(1 − 0.004465)]uframe =
influence of the shear deformation on the displacements is often 1.019uframe = 7.28 mm, i.e. 1.9% higher than the displacement
neglected. However, as is shown in the illustrative example in neglecting the influence of vertical loads, where 0.4% refers to
Section 4.12, the effect can be significant even for ordinary cases. second order effects and 1.5% to effects of tilted studs.
B. Källsner, U.A. Girhammar / Engineering Structures 31 (2009) 2171–2181 2179

4.13. Experimental verification Appendix A. Checking of buckling of wall panels

The critical stress for the sheets can be evaluated as follows.


To validate the proposed theory, shear walls are considered
Consider the force distribution according to the plastic lower
here according to Fig. 1 tested by Källsner [12]. Six full scale wall
bound method (Fig. 3), where all fasteners along the edges except
tests were conducted, five with different sheathing materials, of
those in the corners carry the same load. It is reasonable to
which one with two types of fasteners. Also, the sheathing-to-
determine the critical stress as if the sheet were subjected to a
framing joints for each wall design were tested. constant shear stress.
In Table 1, the experimental and theoretical results are For an isotropic, homogeneous, and elastic sheet material
compared with respect to the horizontal load-carrying capacity. and with dimensions according to Fig. 1(a), the critical stress is
It is evident that the applicability of the proposed analytical obtained as
procedure is very good. In addition, almost all values are on the
π 2E
 2
safe side. t
τcr = kτ (40)
Experimental and theoretical results with respect to the 12(1 − ν)2 b/2
horizontal displacement of shear walls are compared in a where t is the thickness of the sheet and ν is Poisson’s ratio. Half of
companion paper [1]. the width of the sheet (b/2) is used since the sheet is also supported
along the centre stud. The value of the coefficient kτ depends on the
length-to-width ratio of the sheet and on the boundary conditions.
5. Comments on the assumptions of the plastic models For shear walls in practice, the boundary conditions for the sheet
are somewhere between simply supported and clamped along all
All assumptions for the elastic and plastic models are the same, four edges. Expressions for the coefficient kτ are readily available
except for the properties of the sheathing-to-framing joints. For the in ordinary textbooks or e.g. in Eurocode No. 5, Appendix 7 [7].
plastic models, the characteristics of the joints are assumed ideally For an orthotropic, homogeneous, and elastic sheet material like
plastic. As for the elastic model, it is assumed that all joints have the plywood, the critical shear stress for a sheet with simply supported
same stiffness properties and that this stiffness is independent of boundaries along all edges is given by
the load direction and the mutual orientation between the sheets
π2 q
 2
t
and timber members. τcr = kτ ,ort 4
Ex Ey3 (41)
This means that the comments given for the elastic model; 3 b/2
see Källsner & Girhammar [2], essentially are also applicable to where the buckling coefficient, kτ ,ort , is given in, for example,
the plastic models, except for those referring to the sheathing-to- Halasz & Cziesielski [8], Larsson & Wästlund [9], Dekker et al. [10],
framing joints. However, when calculating the displacements of or Eurocode No. 5, Appendix 7 [7].
the frame and of the fasteners a secant modulus needs to be used
in the plastic models as well as in the elastic model. Also, for the Appendix B. Plastic lower bound racking capacity for shear
chosen force distribution in the plastic lower bound model, there is walls comprising of more than one unit
no force component perpendicular to the length or grain direction
of the timber members and, therefore, no shearing forces need to Here, the principles for calculating the horizontal load-carrying
be transferred by the framing joints between the timber members. capacity for shear walls built up of more than one wall unit are
given. An example of such a shear wall is shown in Fig. 5. In order
to illustrate the structural behaviour of the wall, the structure is
6. Summary divided into two parts, the timber frame and the sheets. If we
accept the model based on pure shear flow in the sheets, we obtain
the force distribution shown in the lower part of the figure, where
The basic principles for analysing fully anchored wood-
all timber members and sheets are in equilibrium. The racking
framed shear walls assuming plastic behaviour of the mechanical
capacity for this wall can be calculated as the sum of the capacities
sheathing-to-framing joints have been presented. The models can
of the individual wall units,
be applied to statically loaded shear walls in both the ultimate and
serviceability limit states to evaluate the horizontal load-bearing X bi
Htot = Fp . (42)
capacity and the horizontal displacements of the wall as well as slip sr
in the sheathing-to-timber joints. The plastic lower bound method
If a shear wall is built up of sheets with different widths, the
is suitable for hand calculation. capacity of the shear wall can be calculated as
X bi bi
Acknowledgements Htot = Fp (43)
sr bmax
where bmax is the width of the widest sheet. One reason for the
The authors express sincere appreciation for the financial
reduction factor bi /bmax is that a wall unit with a high height-
support from The Development Fund of the Swedish Construction
to-width ratio will reach its maximum load at a rather large
Industry (SBUF), The Swedish Research Council for Environment,
displacement and that the full plastic capacity is never reached.
Agricultural Sciences and Spatial Planning (FORMAS), The County
The proposed reduction is supported by test results and is intended
Administrative Board of Västerbotten, The European Union’s for the situation shown in Fig. 6(a) when bi < h/2. However, the
Structural Funds – The Regional Fund, SP Wood Technology – The reduction factor in Eq. (43) can lead to too conservative a result if
Technical Research Institute of Sweden, Umeå University, and Carl a sheet with a low height-to-width ratio is combined with a more
Wikström Foundation, Nordmaling, Västerbotten, together with normal sheet dimension (bi = h/2) as in Fig. 6(b). In this latter case
the timber and building industry. there is no need for a reduction. Consequently it is recommended
We also would like to thank the reviewers for their constructive that bmax in Eq. (43) be replaced by the fixed value h/2 and that the
comments and suggestions. reduction factor bi /bmax only should be used when bi < h/2.
2180 B. Källsner, U.A. Girhammar / Engineering Structures 31 (2009) 2171–2181

Table 1
Comparisons between experimental and analytical results for a number of shear wall designs, for details see Källsner [12].
Hexp
Sheathing Full scale wall test Hexp Fastener test Fv Analytical load Fastener spacing parameters Hanal Hanal
b h h
(N) (N) Hanal sr sps sis
(N)

Gypsum plasterboard (screws) 4555 642 Eq. (16) 6 12 8 3852 1.18


Gypsum plasterboard (nails) 4182 461 Eq. (16) 8 16 16 3688 1.13
Bitumen-impregnated fibreboard (nails) 5571 353 Eq. (16) 16 32 16 5648 0.99
Particleboard (nails) 7390 582 Eq. (16) 12 24 10 6984 1.06
Medium density fibreboard (nails) 7120 439 Eq. (16) 16 32 12 7024 1.01
Plywood (nails) 7304 887 Eq. (16) 8 16 16 7096 1.03

Fig. 5. Example of a wall panel consisting of more than one wall unit. Force
distribution on the sheets and the timber frame according to the plastic lower bound
method. Fig. 6. Examples of wall panels built up of sheets with different widths.

Fig. 7. Assumed force distribution on (a) the timber frame and (b) the sheet in a fully anchored shear wall according to the plastic lower bound method. The effect of the
fasteners along the intermediate stud is neglected.
B. Källsner, U.A. Girhammar / Engineering Structures 31 (2009) 2171–2181 2181

Appendix C. An alternative plastic lower bound racking capac- the fasteners in the corners equal Fc,x̂ = Fc,ŷ = 0.5Fp , is given by
ity
H = (nr − 1)Fp + 2 · 0.5Fp = nr Fp . (49)
In Fig. 7, a shear wall with the dimensions, h/b = 2, and the Thus, the difference between the two plastic capacities is 0.272Fp .
fastener spacing, sr , around the edge of the wall is shown. The
number of spacings between the fasteners along the horizontal References
and vertical framing members are then given by nr = b/sr and
nps = h/sr = 2nr , respectively. The plastic force distribution is [1] Girhammar UA, Källsner B. Elasto-plastic model for analysis of influence of
imperfections on stiffness of fully anchored light-frame timber shear walls.
shown where the forces in the fasteners in the corners are assumed Eng Struct 2009;31(9):2182–93.
to reach their ultimate plastic capacity, i.e. [2] Källsner B, Girhammar UA. Analysis of fully anchored light-frame
timber shear walls—Elastic model. Mater Struct. 2009;42:301–20.
Fc2,x̂ + Fc2,ŷ = Fp2 . (44) doi:10.1617/s11527-008-9463-x.
[3] Girhammar UA, Källsner B. Analysis of influence of imperfections on stiffness
The moment equilibrium for the sheet in Fig. 7(b) gives of fully anchored light-frame timber shear walls—Elastic model. Materials and
Structures 2009;42:321–37. doi:10.1617/s11527-008-9458-7.
[(nr − 1)Fp + 2Fc,x̂ ]h − [(2nr − 1)Fp + 2Fc,ŷ ]b = 0. (45) [4] Källsner B, Lam F. Diaphragms and shear walls. In: Holzbauwerke nach
Eurocode 5 – Step 3. Informationsdienst Holz. Germany: Arbeitsgemeinschaft
Eq. (31) gives Fc,ŷ = 2Fc,x̂ − Fp /2. Together with Eq. (30), the forces Holz e.V., Düsseldorf; 1995. [chapter 15].
[5] Källsner B, Girhammar UA, Wu L. A simplified plastic model for design of
in the fasteners in the corners can then be written as
partially anchored wood-framed shear walls. In: Proc CIB-W18 meeting. 2001.
√ Paper 34-15-1.
2+ 19 [6] Neal BG. Plastic methods of structural analysis. 2nd ed. London: Chapman &
Fc,x̂ = Fp = 0.636Fp (46) Hall and Science Paperbacks; 1978.
10
√ [7] Eurocode No 5. Common unified rules for timber structures. EUR 9887.
2 19 − 1 Commission of the European Communities, Brussel, Belgium.
Fc,ŷ = Fp = 0.772Fp . (47) [8] Halasz R, Cziesielski E. Berechnung und Konstruktion geleimter Träger mit
10 Stegen aus Furnierplatten. Berichte aus der Bauforschung, Heft 47. Berlin
(Germany): Wilhelm Ernst & Sohn; 1966.
The horizontal load-carrying capacity is then according to Fig. 7(a) [9] Larsson G, Wästlund G. Plywood as a structural material. Bulletin 1953:21,
given by Swedish National Committee for Building Research, Stockholm, Sweden. 1953
√ [in Swedish].
2+ 19 [10] Dekker J, Kuipers J, van Amstel HP. Buckling strength of plywood—Results of
H = (nr − 1)Fp + 2Fc,x̂ = (nr − 1)Fp + Fp tests and design recommendations. Heron 1978;23(4):1–59.
5 [11] Girhammar UA, Bovim NI, Källsner B. Characteristics of sheathing-to-timber
= (nr − 1)Fp + 1.272Fp joints in wood shear walls. 8th World conference on timber engineering, Lahti,
Finland, 2004.
= nr Fp + 0.272Fp . (48) [12] Källsner B. Panels as wind-bracing elements in timber-framed walls.
Wood Technology Report 56, The Institute for Wood Technology Research,
The plastic capacity for the force distribution where the forces in Stockholm, Sweden, 1984 (in Swedish).

You might also like