TIMBER DESIGN
Introduction
Wood is a very versatile raw material and is still widely used in construction. Timber can be used in
a range of structural applications including marine works: construction of wharves, piers,
cofferdams; heavy civil works: bridges, piles, shoring, pylons; domestic housing: roofs, floors,
partitions; shuttering for precast and in situ concrete; falsework for brick or stone construction.
Timber is naturally occurring which makes it a very difficult material to characterise and partly
accounts for the wide variation in the strength of timber, not only between different species but also
between timber of the same species and even from the same log. Quite naturally, this led to
uneconomical use of timber which was costly. However, this problem has now been largely
overcome by specifying stress graded timber.
The suitability of a particular timber type for any given purpose will depend upon various factors
such as performance, cost, appearance and availability. This makes specification very difficult. The
task of the structural engineer has been simplified, however, by grouping timber species into sixteen
strength classes for which typical design parameters, e.g. grade stresses and moduli of elasticity,
have been produced.
Design of timber elements is normally carried out in accordance with BS 5268: Structural Use of
Timber. This is divided into the following parts:
Part 2: Code of Practice for Permissible Stress Design, Materials and Workmanship.
Part 3: Code of Practice for Trussed Rafter Roofs.
Part 4: Fire Resistance of Timber Structures.
Part 5: Code of Practice for the Preservative Treatment of Structural Timber.
Part 6: Code of Practice for Timber Frame Walls.
Part 7: Recommendations for the Calculation Basis for Span Tables.
The design principles which will be discussed are based on the contents of Part 2 of the code. It
should therefore be assumed that all references to BS 5268 refer exclusively to Part 2. BS 5268 is
based on permissible stress design rather than limit state design. This means in practice that a partial
safety factor is applied only to material properties, i.e. the permissible stresses and not the loading.
The following general aspects are important to timber design
a) Stress grading
b) Grade stress and strength class
c) Permissible stress.
1
a) Stress grading
The strength of timber is a function of several parameters including the moisture content, density,
size of specimen and the presence of various strength-reducing characteristics such as knots, slope of
grain, fissures and wane.
The introduction of BS 5268 brought about a new way of assessing the strength of timber. The first
step involve grading structural size timber. Grading was carried out visually, although it is now
common practice to do this mechanically. Mechanical grading has the advantage of greater economy
in the use of timber since it took into account the density of timber which significantly influences its
strength.
When BS 5268 was published in 1984 the numbered grades (i.e. 75, 65, 50 and 40) were withdrawn
and replaced by two visual grades: General Structural (GS) and Special Structural (SS) and four
machine grades: MGS, MSS, M75 and M50.
b) Grade stress and strength class
Table 1 shows typical timber species/grade combinations and associated grade stresses and moduli of
elasticity. This information would enable the designer to determine the size of a timber member
given the intensity and distribution of the loads to be carried. However, it would mean that the
contractor’s choice of material would be limited to one particular species/grade combination, which
could be difficult to obtain. It would obviously be better if a range of species/grade combinations
could be specified and the contractor could then select the most economical one. Such an approach
forms the basis of grouping timber species/grade combinations with similar strength characteristics
into strength classes (Table 2).
In all there are sixteen strength classes, C14, C16, C18, C22, C24, TR26, C27, C30, C35, C40, D30,
D35, D40, D50, D60 and D70, with C14 having the lowest strength characteristics. The strength
class designations indicate the bending strength of the timber. Strength classes C14 to C40 and TR26
are for softwoods and D30 to D70 are for hardwoods. Strength class TR26 is intended for use in the
design of trussed rafters. The grade stresses and moduli of elasticity associated with each strength
class are shown in Table 3. In the UK structural timber design is normally based on strength classes
C16 to C27. These classes cover a wide range of softwoods which display good structural properties
and are both plentiful and cheap.
2
Table 1 Grade stresses for softwoods graded in accordance with BS 4978: for service classes 1 and
2 (Table 10, BS 5268)
Table 2: Softwood combinations of species and visual grades which satisfy the requirements for
various strength classes
3
Table 3 Grade stresses and moduli of elasticity for various strength classes: for service classes 1
and 2 (based on Tables 8 and 9, BS 5268)
c) Permissible stresses
The grade stresses given in Tables 1 and 3 were derived assuming particular conditions of service
and loading. In order to take account of the actual conditions that individual members will be subject
to during their design life, the grade stresses are multiplied by modification factors known as K-
factors. The modified stresses are termed permissible stresses.
4
BS 5268 lists over 80 K-factors. However, we shall consider only those modification factors relevant
to the design of simple flexural and compression members, namely:
K2: Moisture content factor
K3: Duration of loading factor
K5: Notched ends factor
K7: Depth factor
K8: Load-sharing systems factor
K12: Compression member stress factor.
a) MOISTURE CONTENT, K2
The strength and stiffness of timber decreases with increasing moisture content. This effect is taken
into account by assigning timber used for structural work to a service class. BS 5628 recognises three
service classes as follows:
Service class 1 is characterised 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% for a few weeks per
year.
Timbers used internally in a continuously heated building normally experience this environment. In
such environments most timbers will attain an average moisture content not exceeding 12%.
Service class 2 is characterised by a moisture content in the material corresponding to a temperature
of 20°C and the relative humidity of the surrounding air only exceeding 85% for a few weeks per
year. Timbers used in covered buildings will normally experience this environment. In such
environments most timbers will attain an average moisture content not exceeding 20%.
Service class 3, due to climatic conditions, is characterised by higher moisture contents than service
class 2 and is applicable to timbers used externally and fully exposed.
The grade stresses and moduli of elasticity shown in Tables 1 and 3. This apply to timber exposed to
service classes 1 and 2. According to clause 2.6.2 of BS 5268 where service class 3 exists, the values
in Tables 1 and 3 should be multiplied by a modification factor K2 given in Table 16 of BS 5268,
reproduced here as Table.4.
Table 4 Modification factor K2 by which stresses and moduli for service classes 1 and 2 should be
multiplied to obtain stresses and moduli applicable to service class 3 (Table 16, BS 5268)
5
ii) DURATION OF LOADING, K3
The stresses given in Tables 1 and 3 apply to long-term loading. Where the applied loads will act for
shorter durations e.g. snow and wind, the grade stresses can be increased. Table 17 of BS 5268,
reproduced as Table 5, gives the modification factor K3 by which these values should be multiplied
for various load combinations.
Table 5 Modification factor K3 for duration of loading (Table 17, BS 5268)
iii) NOTCHED ENDS, K5
Notches at the ends of flexural members will result in high shear concentrations which may cause
structural failure and must, therefore, be taken into account during design figure 1.
6
Figure 1 Notched beams: (a) beam with notch on top edge; ( b) beam with notch on underside ( Fig.
2, BS 5268).
In notched members the grade shear stresses parallel to the grain (Tables 1 and 3) are multiplied by a
modification factor K5 calculated as follows
Clause 2.10.4 of BS 5268 also notes that the effective depth, he, should not be less than 0.5 h,i.e. K5
≥ 0.5
iv) DEPTH FACTOR, K7
The grade bending stresses given in Table 3 only apply to timber sections having a depth h of 300
mm. For other depths of beams, the grade bending stresses are multiplied by the depth factor K7,
defined in clause 2.10.6 of BS 5268 as follows:
7
v) LOAD-SHARING SYSTEMS, K8
The grade stresses given in Tables 1 and 3 apply to individual members, e.g. isolated beams and
columns, rather than assemblies. When four or more members such as rafters, joists or wall studs,
spaced a maximum of 610 mm centre to centre act together to resist a common load, the grade stress
should be multiplied by a load-sharing factor K8 which has a value of 1.1 (clause 2.9, BS 5268).
vi) COMPRESSION MEMBERS, K12
The grade compression stresses parallel to the grain given in Tables 1 and 3 are used to design struts
and columns. These values apply to compression members with slenderness ratios less than 5 which
would fail by crushing. Where the slenderness ratio of the member is equal to or greater than 5 the
grade stresses should be multiplied by the modification factor K12 given in Table 22 of BS 5268,
reproduced here as Table 6. Alternatively Appendix B of BS 5268 gives a formula for K12 which
could be used. This is based on the Perry-Robertson equation which is also used to model the
behaviour of steel compression members previous topics. The factor K12 takes into account the
tendency of the member to fail by buckling and allows for imperfections such as out of straightness
and accidental load eccentricities.
The factor K12 is based on the minimum modulus of elasticity, Emin, irrespective of whether the
compression member acts alone or forms part of a load-sharing system and the compression stress,
σc,||, is given by:
σc,|| = σc,g,||K3
8
Table.6 Modification factor K12 for compression members (Table 22, BS 5268)
DESIGN OF TIMBER MEMBERS
FLEXURAL MEMBERS
Beams, rafters and joists are examples of flexural members. All calculations relating to their design
are based on the effective span and principally involves consideration of the following aspects
i) Bending
ii) Deflection
iii) Lateral buckling
iv) Shear
v) Bearing.
Generally, for medium-span beams the design process follows the sequence indicated above.
However, deflection is usually critical for long-span beams and shear for heavily loaded short-span
beams.
i) EFFECTIVE SPAN
According to clause 2.10.3 of BS 5268, for simply supported beams, the effective span is normally
taken as the distance between the centres of bearings (Figure 2).
9
Figure 2: Effective span of simply supported beams.
6.7.2 BENDING
If flexural members are not to fail in bending, the design moment, M, must not exceed the moment of
resistance, MR
M ≤ MR (6)
The design moment is a function of the applied loads. The moment of resistance for a beam can be
derived from the theory of bending and is given by
where
b breadth of section
d depth of section
Figure 4: Section modulus.
The permissible bending stress is calculated by multiplying the grade bending stress, σ m,g,||, by any
relevant K-factors
σm,adm,|| = σm,g,||K2K3K7K8 (as appropriate)…………………….eqn 8
For a given design moment the minimum required section modulus, Zxx req, can be calculated using
equation 6.9, obtained by combining equations 6 and 7:
10
A suitable timber section can then be selected from Tables NA.2, NA.3 and NA.4 of BS EN 336:
Structural timber. Sizes permitted deviations.
These tables give the commonly available sizes of, respectively, sawn timber, timber machined on
the width and timber machined on all four sides. Table NA.2 is reproduced here as Table 7. Table 8
is an expanded version which includes a number of useful section properties to aid design. Finally,
the chosen section should be checked for deflection, lateral buckling, shear and bearing to assess its
suitability as discussed below
Table 7: Commonly available target sizes of sawn softwood structural timber (based on Table
NA.2, BS EN 336)
Table 8 Geometrical properties of sawn softwoods
11
ii) DEFLECTION
Excessive deflection of flexural members may result in damage to surfacing materials, ceilings,
partitions and finishes, and to the functional needs as well as aesthetic requirements.
12
Clause 2.10.7 of BS 5268 recommends that generally such damage can be avoided if the total
deflection, δt, of the member when fully loaded does not exceed the permissible deflection, δp:
δt ≤ δp (10)
The permissible deflection is generally given by
δp = 0.003 × span (11)
but for longer-span domestic floor joists, i.e. spans over 4.67 m, should not exceed 14 mm:
δp ≤ 14 mm (12)
The total deflection, δt, is the summation of the bending deflection, δm, plus the shear deflection,
δv:
δ t = δm + δv (13)
Table 9 gives the bending and shear deflection formulae for some common loading cases for beams
of rectangular cross-section. The formulae have been derived by assuming that the shear modulus is
equal to one-sixteenth of the permissible modulus of elasticity in accordance with clause 2.7 of BS
5268.
For solid timber members acting alone the deflections should be calculated using the minimum
modulus of elasticity, but for load-sharing systems the deflections should be based on the mean
modulus of elasticity
13
Table 9: Bending and shear deflections assuming G = E/16
iii) LATERAL BUCKLING
If flexural members are not effectively laterally restrained, it is possible for the member to twist
sideways before developing its full flexural strength, thereby causing it to fail in bending, shear or
deflection. This phenomenon is called lateral buckling and can be avoided by ensuring that the depth
to breadth ratios given in Table 10 are complied with.
Table 10 Maximum depth to breadth ratios (Table 19, BS 5268)
14
iv) SHEAR
If flexural members are not to fail in shear, the applied shear stress parallel to the grain, τ a, should not
exceed the permissible shear stress, τadm:
τa ≤ τadm (6.14)
For a beam with a rectangular cross-section, the maximum applied shear stress occurs at the neutral
axis and is given by:
where τg is the grade shear stress parallel to the grain (Tables 1 and 3).
v) BEARING PERPENDICULAR TO GRAIN
Bearing failure may arise in flexural members which are supported at their ends on narrow beams or
wall plates. Such failures can be avoided by ensuring that the applied bearing stress, σ c,a,⊥, never
exceeds the permissible compression stress perpendicular to the grain, σc,adm,⊥:
15
The permissible compression stress is obtained by multiplying the grade compression stress
perpendicular to the grain, σc,g,⊥, by the K-factors for moisture content (K2), load duration (K3) and
load sharing (K8) as appropriate.
It should be noted that the grade compression stresses perpendicular to the grain given in Tables 1
and 3 apply to (i) bearings of any length at the ends of members and (ii) bearings 150 mm or more in
length at any position. Moreover, two values for the grade compression stress perpendicular to the
grain are given for each strength class (Table 3). The lower value takes into account the amount of
wane which is permitted within each stress grade (Fig. 6). If, however, the specification prohibits
wane from occurring at bearing areas the higher value may be used.
Example 1
A timber beam with a clear span of 2.85 m supports a uniformly distributed load of 10 kN
including self-weight of beam. Determine a suitable section for the beam using timber of
strength class C16 under service class 1. Assume that the bearing length is 150 mm and that
the ends of the beam are held in position and compression edge held in line.
Solution
EFFECTIVE SPAN
Distance between centres of bearing (l ) = 3000 mm
16
17
Example 3
A timber beam with a clear span of 3 m supports a uniformly distributed load of 8 kN
including self-weight of beam. Determine a suitable section for the beam using timber of
strength class C24 under service class 1. Assume that the bearing length is 175 mm and that
the ends of the beam are held in position and compression edge held in line.
18
DESIGN OF COMPRESSION MEMBERS
Struts and columns are examples of compression members. For design purposes BS 5268 divides
compression members into two categories
i) Members subject to axial compression only and
ii) Members subject to combined bending and axial compression.
The principal considerations in the design of compression members are:
1. slenderness ratio
2. axial compressive stress
3. permissible compressive stress.
The following subsections consider these more general aspects before describing in detail the design
of the above two categories of compression members
SLENDERNESS RATIO
The load-carrying capacity of compression members is a function of the slenderness ratio, λ, which
is given by
Le
λ= … … … … … … … .. (i )
i
Where
Le effective length
i radius of gyration
According to clause 2.11.4 of BS 5268, the slenderness ratio should not exceed 180 for compression
members carrying dead and imposed loads other than loads resulting from wind in which case a
slenderness ratio of 250 may be acceptable. The radius of gyration, i, is given by
i=
√ I
A
… … … … … … … … …(ii)
Where
I moment of inertia
A cross-section area.
For rectangular sections
b
i= … … … … … … .. ( iii )
√12
where b is the least lateral dimension.
19
The effective length, Le, of a column is obtained by multiplying the actual length, L, by a coefficient
taken from Table 11 which is a function of the fixity at the column ends.
Le =L× coefficient ( iv )
In Table 11 end condition (a) models the case of a column with both ends fully fixed and no relative
horizontal motion possible between the column ends. End condition (c) models the case of a pin
ended column with no relative horizontal motion possible between column ends. End condition (e)
models the case of a column with one end fully fixed and the other free. Figure 5 illustrates all five
combinations of end fixities.
Figure 5: End conditions
Table 11 Effective length of compression members (Table 21, BS 5268)
AXIAL COMPRESSIVE STRESS
The axial compressive stress is given by
20
…………(v)
Where
F axial load
A cross-sectional area.
PERMISSIBLE COMPRESSIVE STRESS
According to clause 2.11.5 of BS 5268, for compression members with slenderness ratios of less than
5, the permissible compressive stress should be taken as the grade compression stress parallel to the
grain, σc,g,||, modified as appropriate for moisture content, duration of loading and load sharing
……………………….(vi)
For compression members with slenderness ratios equal to or greater than 5, the permissible
compressive stress is obtained in the same way but should additionally be modified by the factor K 12.
……………..(vii)
MEMBER DESIGN
Having discussed these common aspects it is now possible to describe in detail the design of
compression members. As pointed out earlier, BS 5268 distinguishes between two categories of
members, that is, those subject to (a) axial compression only and (b) axial compression and bending.
Members subject to axial compression only
This category of compression member is designed so that the applied compressive stress, σc,a,||, does
not exceed the permissible compressive stress parallel to the grain, σc,adm,||.
σc,a,|| ≤ σc,adm,|| (viii)
The applied compressive stress is calculated using equation (v) and the permissible compressive
stress is given by equations (vi) or (vii) depending upon the slenderness ratio
Members subject to axial compression and bending
This category includes compression members subject to eccentric loading which can be equated to an
axial compression force and bending moment. According to clause 2.11.6 of BS 5268, members
which are restrained at both ends in position but not direction, which covers most real situations,
should be so proportioned that
21
……………(ix)
Equation (ix) is the normal interaction formula used to ensure that lateral instability does not arise in
compression members subject to axial force and bending. Thus if the column was subject to
compressive loading only, i.e. M = 0 and σm,a,|| = 0, the designer would simply have to ensure that
σc,a,||/σ c,adm,|| ≤ 1. Alternatively, if the column was subject to bending only, i.e. F = σc,a,|| = 0, the
designer should ensure that σ σ
m,a,||/ m,adm,|| ≤ 1. However, if the column was subject to combined
bending and axial compression, then the deflection as a result of the moment M would lead to
additional bending due to the eccentricity of the force F as illustrated in Fig. 6.8. This is allowed for
by the factor in the above expression
Figure 6: Bending in timber columns
Example 4
22
A timber column of redwood GS grade consists of a 100 mm square section which is restrained
at both ends in position but not in direction. Assuming that the actual height of the column is
3.75 m, calculate the maximum axial long-term load that the column can support
Example 5
23
Timber column resisting an axial load and moment (BS 5268) Check the adequacy of the
column in Example 4 to resist a long-term axial load of 10 kN and a bending moment of 350 kN
mm.
24
