Structural Timber: Characteristics and Testing

ri() Ceccotti
Prof.. Univ. of Florence. Florence. Italy

Surnmar minish the effectiveness of adhesives

used for glued connections.
A basic introduction to timber as a structural material will be given, examining Environmental thermo-hygrometric
its physical parameters for both clear wood and structural timber and the influ- conditions are responsible for varia-
ence of these parameters on its structural applications. Current testing methods tions of moisture content in wood,
and criteria will also be presented. Finally, connections for timber structures will even if the wood has been 'seasoned".
he considered. The time necessary to reach the hygro-
metric equilibrium depends on the
permeability of the timber species and
on the environmental temperature, as
Characteristics places of origin, among different
well as the section dimensions. The
plants from the same place of origin
larger the section dimensions. the
Origin and even within the same plant.
smaller are the moisture content varia-
Therefore. when testing a specimen, it tions within a timber element. Con-
Timber is a natural material of biologi- trolling environmental conditions dur-
is crucial to have at least a preliminary
cal origin obtained from tree trunks.
statistical description of such scatter. ing testing is essential, especially when
Timber used for building purposes is attempting to analyse and interpret re-
This information can be found in the
obtained mainly from hardwoods
literature and gives average values. sults.
(oak. beech, chestnut. poplar, etc.) and
variation coefficients and similar data.
softwoods (spruce, pine, larch. douglas
In any case, there are considerable
fir. etc.) and it is identified according Strength Quality Grading
problems in choosing elements to he
to botanical species. The correct iden-
tested. Sampling criteria, numbers of
tification of timber species is critical to Timber in structural sizes can exhibit
samples. generalisation of the results.
the testing process and in most cases numerous anomalies, biological in ori-
limits of validity, all must be estab-
requires examination with a micro- gin and often not pathological, which
lished as part of the testing procedure.
scope. are considered "defects". For structur-
al purposes. knots (deriving from
branch outgrowths). deformations.
Anisotropv Visco-elasticitv
cracks, slope of grain, fungi and insect
The micro-structure of timber is Since timber is a visco-elastic material. attacks. etc. may all be considered im-
formed by cellular elements, arranged its characteristics of deformation and portant. Clear wood, without defects
differently according to species. with a strength strongly depend on the veloc- and with a straight grain, is normally
general orientation parallel to the ity and duration of an applied load, as found only in sizes too small for struc-
stem axis. This parallel orientation well as the obvious consideration of tural purposes.
largely determines timber's mechani- the direction of force, which is referred
The mechanical behaviour of structur-
cal properties. Attempts to assign val- to in terms of the anatomic directions
al timber is strongly influenced by pos-
ues for the elastic modulus of timber mentioned above.
sible defects, which can result in rup-
should consider its three anatomical
When testing the elastic modulus in ture. as defined by certain parameters.
directions: longitudinal (parallel to the Structural timber may he prone to brit-
bending of a particular piece of timber,
grain), tangential (tangent to the annu-
it is not possible to attain a useful re- tle failure in bending (according to
al growth rings) and radial (towards
sult. even for short-term purposes, if species and quality), while clear timber
the pith). It should he noted that the rigorous testing arrangements and a has a more ductile behaviour. Further-
ratio between the first two differs for strictly observed loading velocity are more. moisture-related resistance is
softwoods and hardwoods. Shear mod-
not defined. less accentuated in structural timber
ulus. which is much lower than longitu-
than in clear wood.
dinal elastic modulus (about 1/15). also
has an influence on test results that Structural timber may be classified
Hygroscopic Behaviour
should not he ignored. into various quality classes (grades)
Timber is a hygroscopic material and according to its strength. Such classifi-
its deformation and strength charac- cation is based on a variety of factors.
Variability teristics are therefore strongly influ- such as frequency and diameter of
The biological origin of timber results enced by its moisture content. Stiffness knots, slope of the grain. etc. This kind
in significant variability in all charac- and resistance particularly decrease as of visual analysis requires an accurate
teristics. Such variability can be seen the humidity of wood increases. Mois- survey of the specimen. Alternatively.
among different species (especially be- ture heavily influences the possible a grading machine provides several
tween softwoods and hardwoods). presence of fungi which are the cause nondestructive testing methods, for ex-
within the same species from different of decay. In addition. moisture can di- ample. the elastic modulus in bending.

Structural Engineering
International 2/93 Reports 95
 tensile and compressive stiffnesscs
perpendicular to the grain.
— prEN 384 Structural Timber: Deter-
mination of values of mechanical
properties and densities.
— CEN TC 124-206 Wood-based Mate-
rials: Determination of characteris-
tic values.
CEN TC 124-207 Glued-laminated
Timber: Strength classes and deter-
mination of characteristic proper-
ties.
1RII

— prEN 518 Structural Timber — Grad-

ing: Requirements for visual
strength grading standards.
8 0 prEN 519 Structural Timber — Grad-
ing: Requirements for machine
strength graded timber and grading
lIE I' machines.

Connections
Density
Connections between timber elements
can he glued or made with mechanical
fasteners. Mechanical joints are always
semi-rigid and their deformation and
AF strength are strongly dependent on the
- density of the wood. Embedding
I—. strength. timber thickness. fastener di-
ameter. and end- and edge-distance
are the most important factors for the
strength of the mechanical connec-
tions, apart from time and moisture.
Embedding strength is strongly influ-
enced by the density of the wood. Stiff-
Fig. 1: Simplified test fir a semi—rigid joint in a timber structure under cyclical loading ness, too, is influenced by density,
which reduces the deformation of the
wooden bed where the fastener acts.
For this reason, it is very important to
Testing Aims and Codes European Community and the Euro-
have an accurate idea of the
pean Free Trade Association) also ad-
dress the design of timber structures specimen's density if a correct inter-
Testing timber structures is undertak- pretation of test results is to he ob-
under Eurocode 5.
en not only for reasons common to the tained.
testing of structures in other materials. The following testing prestandards
i.e.. mechanical identification, model (prEN) and proposals of the European
validation, monitoring. etc. In addi- Committee for Standardisation (CEN) Slip
tion, timber structures are tested to de- will soon be available as supporting Slip occurs in all mechanical joints in
termine product homogeneity and per- standards for Eurocode 5: timber construction, both initial slip
formance. and as part of the design — prEN 789 Timber Structures (free) and elastic slip, according to the
—
process as well ('design by testing"). type of connection. Initial slip is due to
Wood-based panels: Test methods
for the determination of mechanical the closing of gaps between connec-
Testing Standards tion elements: elastic slip derives from
properties for structural purposes.
ISO standards for tests on clear and the deformation of fasteners and their
— prEN 408 Timber Structures — solid further embedding into the wood.
structural timber for structural purpos-
es, aimed at establishing common in- and glued-laminated timber: Deter-
mination of some physical and me- The slip modulus (i.e.. the slope of the
ternational practices for reporting test load-slip curve) of a joint strongly af-
chanical properties for structural
results, have existed for some time fects the deformation of the joint. A
(e.g.. ISO 8375). The Structural Euro- purposes (derived from ISO 8375).
glued connection is much stiffer than a
codes (comprehensive design stand- — ('EN TC 124-105 Solid Timber in mechanical one, Elastic deformation
ards for the 18 member nations of the Structural Sizes: Determination of of the wood in a joint is usually less im-

96 Reports Structural Engineering International 2/93

portant than the mechanical slip of the
whole joint.
Initial slip is widely present in older
structures due to the inaccuracy of the
tools and of the types of connections
used (i.e. non-drilled nails, stirrups.
and bolts, as well as wooden connec-
tors). Modern structures benefit from
more advanced connection techniques
and materials (pre-drilled and im-
proved nails, dowels. and toothed con-
ilectors, etc.).
In order to properly manage and inter-
pret testing. knowledge of both the cx-
pected initial slip and the values of the
slip modulus are necessary. Initial free
slip will not recover during an unload-
ing phase. For repeated loads, the hvs-
teresis cycle will increase as the load
increases (under the primary creep-
limit). The slip modulus will remain
unchanged or will increase under these
circumstances.
For cyclical tests with inversion of the
load sign, asymmetric pinched hystere-
sis loops occur, increasing with the cy-
cles. It is strongly advised to use stand-
ard cyclical testing methods in order
to obtain results comparable to those
of other researchers (Fig. 1).

Rupture
It must be emphasised that rupture be-
haviour for glued connections is very
different than for mechanical ones.
Glued connections are generally brit-
tle; mechanical ones max' he brittle or
ductile. depending on type. arrange-
ment of connection elements and rela-
tive dimensions.

Connection Standards Ii. 2: .Siuüc 1e,1 0/ Iiiie /)(II1 (() Ill) ,'III('(i—1(l11li!I(l1C(1 iintbt'r 10(1/

The following relevant CEN rules are

in preparation:
nation of strength and deformation reliability-based design codes. Tests on
— prEN 383 Timber Structures — Test characteristics. prototypes are considered by Euro-
methods: Determination of embed- code 5 for wall diaphragms in order to
ding strength and foundation values — EN 28970 Timber structures — Joints
determine load carrying capacity [8].
for dowel type fasteners. made with mechanical fasteners:
Requirements for wood density.
— prEN 409 Timber Structures — Test
methods: Determination of yield In addition. for cyclic loading, refer to Actual Structures
moment of dowel-type fasteners. RILEM TC 109 TSA. "Test Methods Load testing. static or dynamic. of
for Cyclic Loading of Timber Joints'. structures in situ usually is performed
— CEN TC 124-1 12 Timber Structures:
Testing of load-bearing nailed joints. in order to assess the mechanical state
of the structure (Fig. 2), as is the case
— CEN TC 124-I 16 Timber Structures: Structures with testing structures made of other
Testing of punched metal plate fas- materials. Old timber structures sub-
teners. Prototypes stantially retain their structural resis-
— EN 2ô$ I Timber Structures — Joints Tests on prototypes can be used for de- tance with the passage of time if stress-
made with mechanical fasteners: termining the characteristic values of es have remained below the creep-
General principles for the determi- stiffness and strength. especially for limit.

Siructural Engineering International 2/93 Reports 97

 10 prEN 594 Timber Structures — Tim-
her frame walls: Test for racking
I strength arid stiffness of structural
1 a
am
L

—
wall panels.
prEN 595 Timber Structures —
VP Trusses: Tests for determination of
strength and stiffness.
0 I
— CEN TC 124-109 Timber Structures
U
x — Test methods: The performance of
w
I!!' structural floor decking.

0
Also important is Appendix Al of Eu-
rocode 5: Determination of 5-
percentile characteristic values from
test results.

References
0
Time [1] ARRIAGA. F. et al. Grading and Load
Carrying capacity Determination of Old
Timber Beams. American Society br Agri-
cultural Engineers, paper n. 92-4(165. 1992.
Fig. 3: Loading procedure Jor ti,nber structures according to RILEM Recommendations L2] BODIG. J. et al. Load and Resistance
Factor Design Guidelines for Developing
Reference Resistance. Draft. E 1)M. Fort
Collins. Colorado. 1991.
It is essential to control old timber data on the conditions that preceded [3] CECCOTTI. A. and GIORDANO, S. .4
structures for the presence of fungi or the tests. Ongoing observation of the Device for Cyclic Load Testing of Glue-la-
insects, which can be especially dan- specimen's humidity is also important minated Ti,nher joints: Experimental Eva!-
gerous if they occur unseen inside the to a well managed testing procedure. nation of Hysteretic Ductility of Doweled
walls. It is also important with old tim- Connections. Proceedings of the Workshop
ber structures to make an accurate sur- The influence of slip in the connec- on Structural Behaviour of Timber Con-
vey of the structural elements, their di- tions must be considered, as it may be structions in Seismic Zones". Dept. of Civ-
responsible the unloading of some il Eng.. Univ. of Florence. 1989.
mensions and the connections, as well
as identifying the timber species and connections in a structure and increas- [4] GORLACHER. R. and KRO\ILR. \1.
strength quality class. Often, it will be es in the loading of others. In some Tragfahigkeit von Versatzansc/zlussen in
observed, earlier builders put the best cases. elements not normally consid- Izistorisehen !lol:konstrukiionen. l-Iolzhau -
ered load-bearing may share the loads. Statik-Aktuell Ausgabe, Düsseldorf.
pieces in the most stressed areas! 2/1992.
producing considerable alterations to
Numerous nondestructive testing the expected static scheme. [5] KUIPERS. J. Effect ofAge and/or Load
methods for evaluating in situ timber on Timber Strength. Paper no. 19-6-I, CIB
elements are available, including ultra- W18 Meeting. Florence. 1986.
Structural Testing Standards
sonic. stress wave velocity, vibration, 16] RUG. \V. and SEE\IAN. A. Strength of
drilling energy, and penetration test- It is advisable to follow established Old Timber. Building Research and Infor-
ing, etc. It should be noted that appli- and relatively simple structural testing mation. vol. 19 n.l.. 1991.
cation of a particular method requires procedures. adjusted to specific condi- I
17] SANDOZ. L. Grading of Construe-
that the operator has considerable tions and aims. A useful guideline is lion Timber by Ultrasound. Wood Science
knowledge of the limits of validity of RILEM 57 TSB-1: Recommendations and Technology 23. 95—lOS. 1989.
that method. for testing of timber structures (1985). [8] FN\' 1995-I-I. Eurocode 5— Design of
slightly revised in prEN 380: Timber Timber Structures. CEN. Brussels. 1992.
It is necessary to verify the environ- Structures — Test Methods: General
[91 RILEM Recommendation TSB-1. Test-
mental thermo-hygrometric conditions principles for static loading (Fig. 3). ing Methods or Timber Structures. Maté-
during testing. This information is Other relevant testing guidelines are riaux et Constructions vol. 23. no. 137.
even more useful if it is supported by found in: 1990.

98 Reports Structural Engineering International 2/93