Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

On The Plastic Design of Timber Beams With A Complex Cross-Section

Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

On The Plastic Design Of Timber Beams With A Complex Cross-Section

Maurice Brunner1

1. INTRODUCTION

For the design of steel and concrete structures, most modern international design codes allow the use of plastic stress-
strain-diagrams for the analysis of the bending resistance of the cross-section in the ultimate limit state. Curiously, in
Europe and most other parts of the world, timber structures are still designed for the ultimate limit state with the old,
conservative concept of full elastic behavior till failure. The situation is rather funny for two main reasons. First, it is well
known that under compressive loading, timber exhibits a much better ductile behavior than concrete, for example.
Secondly, between the two World Wars, timber scientists developed refined calculation methods to explain test results
which showed that the bending strength of small, clear timber specimens lies well above the compressive strength. It is
curious that these findings have never been developed to the stage where they could be integrated into modern timber
design codes.

There are of course reasons why the Plastic Design Approach is not yet used for timber structures. The main problem is
the unreliable tensile strength. Even small, clear timber specimens exhibit brittle behavior under tensile loading. In the
case of large specimens of the size used in structures, the tensile strength can be considerably reduced by defects and by
the so-called “volume effect”. A good description of the volume effect – or “weakest link concept”, as it is also known –
is given in [1]. One of the most important factors involved is the type of stress distribution: the failure stress in the case of
a bending moment is much higher for timber of the same graded quality and dimensions than the failure stress under
tensile loading (Fig. 1). In the latter case a greater area is under high tension, hence the probability of early failure due to
the so-called “weakest link”.

Fig. 1: Influence of the stress distribution on the magnitude of the failure stresses

In the European Design Code EC5 for timber structures, the question of plastic design has been indirectly taken into
account by the introduction of an apparent bending strength fM, which has been determined directly in loading tests. The
calculation of the bending resistance MR is based on the full elastic approach. The concept of the fictitious bending
strength is equivalent to full plastic design for solid, rectangular timber cross-sections and also for the glulam types listed
in the Code, for which extensive studies have also been carried out. Since most structural timber used in the construction
industry has a solid, rectangular cross-section, the concept of the fictitious bending strength will in effect lead to full
plastic design for most practical purposes.

There is a small but quietly growing market for highly stressed, high-tech timber cross-sections ([2]). For such
applications, the design method based on the fictitious bending strength is inadequate, because the true failure mode of
timber beams as depicted in fig. 1 is not evident. Indeed, the theory of the fictitious bending strength may in fact be
misleading sometimes, since the impression is created that there is no difference between the compression and the tension

1
Professor, Swiss School of Engineering for the Wood Industry, Biel, Switzerland
zones of a beam. It is common knowledge among engineers that the weakness of a timber beam lies in the brittle tension
zone, yet when engineered, complex cross-sections are necessary, the solution is typically a symmetrical strengthening of
both tension and compression zones – glulam is a typical example. In comparison, the concrete engineer, aware that the
brittle tensile strength of concrete is too risky, typically prescribes reinforcement for the tension zone only. In a similar
way, it might make engineering and economic sense to strengthen future engineered timber beams much more in the
critical tension zone than in the compression zone. Fig. 2 shows some simple examples:

• Grading techniques should be used to strengthen in particular the tension side of glulam beams.
• In the case of flanged timber beams, the boards on the tension flange should exhibit the best material properties;
alternatively, the tension flange could be made thicker than the compression flange.
• Existing beams could be strengthened in-situ by gluing a flat steel bar on the tension side only.

Fig. 2: Vision for high-tech timber cross-sections: unsymmetrical strengthening of the tension side

The present widely used design concept of the fictitious bending strength does not permit the engineer to take advantage
of the good plastic properties of timber under compression in order to design highly stressed, high-tech cross-sections of
the future. The purpose of this paper is to present simple models to illustrate how the existing Eurocodes for timber could
be supplemented – not changed! - in order to let this happen.

2. OVERVIEW PLASTICITY OF MATERIALS

Fig. 3: Stress-strain-diagrams for common building materials

Like most metals, a steel bar under tensile loading initially exhibits a linear relationship between the stress σ and the
strain ε (Fig. 3). After the yield stress fy has been attained, the stress remains constant even when the strain increases
considerably. In the case of European Standard Steel Fe235, failure occurs only after an enormous strain of about 250%o
has been attained, compared to the yield strain of about 1.1%o. Materials which exhibit an extensive yield plateau are
called “ductile”; their structural behavior in the ultimate limit state is characterized by the following advantages:

• Failure occurs only after large deflections have been attained: there is visible warning of danger
• The bending resistance is increased by the so-called “plastification” of the cross-section.
• In the case of statically indeterminate structures, the failure loads are higher than in the case of purely elastic
behavior because the internal forces can be redistributed by plastic joints.

The Theory of Plasticity is ideal for the design of steel structures. It is however also widely used today for the design of
reinforced concrete structures. In this composite building material, steel bars are placed to take up the internal tensile
forces, whilst the concrete itself deals with the compressive forces. The concrete engineer uses the so-called “condition
of ductility” to make sure that the steel is the weakest link in reinforced concrete: in the ultimate limit state the steel yields
in tension before the concrete compression zone fails, hence the ductile behavior of reinforced concrete structures. In
comparison to steel, concrete cubes under compression exhibit fairly brittle behavior. Failure occurs at a relatively low
strain of about 3.5%o. However, in the ultimate limit state, most design codes permit the plastification of the concrete
compression side, thus helping to ensure economic design.

Clear timber specimens exhibit very different behavior under tensile and compressive loading. The behavior depicted in
fig. 3 has been known since the 1920’s ([3]). Under tensile loading, the stress-strain-diagram runs a straight line till brittle
failure. Under compressive loading, the stress-strain-diagram exhibits an initial linear stage. At a strain level of about
3%o, the buckling of the timber fibers leads to the attainment of a yield plateau. The stress remains fairly constant whilst
the strain increases to about 12% at failure.

Structural timber is characterized by defects such as knots and resin pockets. The tensile strength in particular is very
sensitive to such defects and may be sharply reduced. The compressive strength on the other hand is usually much less
affected. This phenomenon is evident in Eurocode EN338: between the timber grades C22 and C40 the compressive
strength rises from 20 to 26 N/mm2, whilst the tensile strength for an axial load changes from 13 to 24 N/mm2 (Fig. 4).

3. SIMPLE STRESS-STRAIN-DIAGRAMS FOR THE PLASTIC DESIGN OF TIMBER BEAMS WITH


COMPLEX CROSS-SECTIONS

Fig. 4: Tensile (ft), compressive (fc) and bending(fM) strengths according to Eurocode EN 338

The table of fig. 4 shows in simplified form the system of Timber Strength Classification according to the European
Standard EN 338. All values refer to stresses acting in the direction of the fibers. The fictitious bending stress fM plays a
central role in the system and in fact the timber grade is directly named after the value of the bending strength. As shown
in fig. 1, however, the bending stress fM is not identical with the real stress conditions in the ultimate limit state. It is a
fictitious value derived mathematically from tests by assuming a linear stress distribution across the cross-section:

fM = [ Mf / W ] / f with Mf = Bending moment at failure


W = Moment of resistance (Rectangle: W = b.h2/6)
f = Safety factor

As mentioned above, the concept of the fictitious bending strength fM is very practical and it is equivalent to the plastic
design of rectangular beams, which form the vast majority of structural timber members in use. For many complex, high-
tech cross-sections which may be needed for special applications, however, a more refined approach is called for.
There are many calculation models for the non-elastic distribution of stresses in a timber beam. One of the simplest and
best-known, developed in 1939 by Thunell ([3]), is shown in fig. 5. The tension zone is characterized by a linear, brittle
model, whereas the compression zone exhibits an initial linear stress rise followed by a clear yield plateau. The E-moduli
for both tension and compression are the same.

Fig. 5: Trapezoid-like stress distribution over the cross-section according to Thunell ([3]).

The standard stress values listed in the Eurocodes can be used to obtain working values for the Thunell model if the
following simple assumptions are taken:

• The so-called “volume effect” applies to the tension zone: under bending conditions, the true failure stress ft,M is
much higher than the value ft under pure tension.

• Research work indicates that the compression zone is much less influenced by the volume effect ([1]). For
simplicity, the failure stress fc under normal compressive force is assumed to remain unchanged in the
compression zone of the beam under bending. Plastification of the cross-section is possible.

The value ft,M, the “true” failure stress on the tension side of the beam, is not listed in the Eurocodes. It can however be
determined by comparing the equations for the bending resistance MR of beam with a rectangular cross-section, calculated
according to Thunell ([3]), and then according to the Eurocode.

• Thunell: MR = fc . (b.h2/6) . c with c = [3 + 8.m + 6.m2 – m4] / (1 + m)4


and m = fc/ft,M

• Eurocode: MR = fM . (b.h2/6)

The following equation is obtained which can be solved for m, and then for ft,M (Fig. 6):

• (3 + 8.m + 6.m2 – m4) = fM


(1 + m)4 fc
Fig. 6:Characteristic values (fc and ft,M) for Thunell’s stress diagram, derived from the characteristic stress values of the
Eurocode

4. CALCULATION EXAMPLE

A simple example will now be presented to illustrate how a Plastic Design Approach could change the outlook for high-
tech, timber cross-sections of the future.

A timber beam, Grade C27, has a width of 100mm and a height of 300mm. Its bending resistance is:

W = 100 x 3002 / 6 = 1.5 x 106 mm3 → MR = fM x W = 27 x 1.5 x 106 / 106 = 40.5 kNm

Since this resistance is inadequate, it is to be strengthened by gluing a steel plate (Fe335) on the tension side. The material
properties involved can be listed as follows:

Timber: C27 Steel: Fe335


ET = 12 kN/mm2 ES = 210 kN/mm2
fM = 27 N/mm2 fy = 335 N/mm2
fc = 22 N/mm2 εy = 335 / 210’000 = 1.60%o
εc = 22 / 12’000 = 1.83%o

The calculation according to Eurocode 5 is illustrated in fig. 7. Due to the high E-modulus of the steel plate, the neutral
axis is shifted below the geometrical center of the 305mm high beam. The relevant characteristic values of the composite
cross-section are:

Relative stiffness of steel: n = ES / ET = 210 / 12 = 17.5


Position of neutral axis: zU = 121mm (zO = 184mm)
Moments of resistance: Timber compression zone: WO = 2.07 x 106 mm3
Timber tension zone: WU = 3.31 x 106 mm3
Steel tension zone: WS = 3.16 x 106 mm3

The calculation suggests that the timber compression zone would be responsible for failure because the fictitious bending
strength fM = 27 N/mm2 would be attained there first before the tension zone. The steel yield stress of 335 N/mm2 would
not have been reached when the timber compression zone fails, hence the excellent plastic qualities of steel would be
useless according to this calculation approach. The bending resistance would be:

MR = WO x fM = 2.07x 106 x 27 / 106 = 56 kNm


Fig. 7: Timber beam with an attached steel plate: bending resistance MR calculated in accordance with the fictitious
bending strength approach

The calculation of the bending resistance of the same beam according to the Plastic Design Approach described under
chapter 3 is illustrated in fig. 8. The neutral axis is shifted even further down to zU = 115mm, and the critical strains are
about 4.0%o at the top compression zone of the timber, 2.3%o at the bottom tension zone of the timber, and 2.4% at the
very bottom of the steel. The compressive and tensile forces corresponding to these strains are in balance. The steel plate
and the timber compression zone have both reached the plastic state, but failure is actually triggered when the tension
zone of the timber attains the “true” brittle tensile strength ft,M. The bending resistance can be calculated to be:

MR = 228 x 0.081 + 154 x 0.131 + 168 x 0.170 = 67 kNm

Fig. 8: Timber beam strengthened with an attached steel plate: calculation with a Thunell-like stress distribution

The bending moment thus derived with the plastic design method is 20% higher that the result obtained using the
fictitious bending strength fM according to the Eurocodes. The Plastic Design Method also points out that the brittle
tension zone of the timber, and not the ductile compression zone, will remain the weakest link in the beam.

5. FURTHER READING

[1] Steiger Rene “Mech. Eigenschaften von Schweiz. Fichten-Bauholz bei Biege-, Zug-, Druck- and
komplizierter M/N-Beanspruchung“, IBK-Bericht Nr. 221, June 1996, Birkhäuser Verlag Basel

[2] Steurer Anton „Holzkonstruktionen mit Stahl- und Kunststoffverstärkung”, 31. SAH-Tagung, 1999,
Lignum, Zürich

[3] Kollmann/Côté „Principles of Wood Science and Technology, Volume 1“, 1984, Springer Verlag,
Berlin, Chapter 7.4
[4] Eurocodes EC5 Design and Construction of Timber Structures
EN338 Strength Grades of Structural Timber

You might also like