Materials and Structures (2007) 40:151–161

DOI 10.1617/s11527-006-9158-0

O R I G I N A L A RT I C L E

Service life modelling of timber structures

J.-W. G. van de Kuilen

Received: 16 June 2005 / Accepted: 15 November 2005 / Published online: 18 October 2006

RILEM 2006

Abstract The service life of structures depends 1 Introduction

on many factors including the current status of
the structures and the age, as well as the expected Timber as a structural material may be found in
future use. In order to be able to make decisions many buildings and structures around Europe,
on possible interventions, an overview is given of with a current age of sometimes more than
a number of methods available to determine the 500 years. Already in the Roman era the foun-
current state of a structure. Data taken from dations were made with timber both in form of
existing structures (foundations) with and without horizontal beams under brick walls and as vertical
biological decay as well as literature have been piles for deeper foundations. But the increase in
analysed. It has been used to extent an existing building activities from the Renaissance period
damage model for timber structures that takes the until about 50 years ago has seen a great use of
load history into account into a model that allows timber. The use of timber is widely spread: timber
for service life calculations on partly deteriorated can be found in foundations of famous heritage
structures. This comprises structures or structural buildings such as the Royal Palace in Amsterdam,
elements that suffer for instance from (biological) the Netherlands and bridges in Venice (Siviero
decay or from drying cracks as is often found in and Murat 1993). Timber foundations are in these
foundations and historic timber structures. cases true foundations of cultural heritage and
therefore their economic and technical impor-
Keywords Timber piles Æ Durability Æ tance is large. A complication in the assessment is
Service life modelling Æ Damage that pile foundations are difficult to inspect and
maintain, and yet many thousands of buildings
and structures depend on them. The historic
centres of many Dutch cities were all built during
the Middle Ages and the golden age from about
1400 until recently. Besides the Netherlands and
J.-W. G. van de Kuilen
Faculty of Civil Engineering and Geosciences, Italy, piles have been used in countries like
Delft University of Technology, P.O. Box 5048, France, Germany and also in North America.
2600 GA, Delft, The Netherlands Failure of the foundation, either because of loss
of strength or because of excessive deformations,
J.-W. G. van de Kuilen (&)
CNR-Ivalsa, Florence, Italy may have severe consequences for the structure it
e-mail: j.w.g.vandekuilen@tudelft.nl is supporting. Timber with age of up to 500 years
152 Materials and Structures (2007) 40:151–161

can also be found in many floor and roof struc- could be re-used for new planned structures, and
tures. The Palazzo Vecchio in Florence, Italy is a consequently the structural performance has to be
fine example. Irrespectively of the applications, assessed.
the assessment of these structures for their safety Change of use in the upper structure: Residen-
and serviceability requires models that take into tial houses from centuries ago are often changed
account the history of the structure. Not only the into office spaces. The (design) loads on the
timber species, also the loads on the structure structural elements as well as on the foundation
over history and expected future loads are will increase accordingly. A typical ratio between
important in the assessment procedure. Although residential floor loads and office floor loads is
most of the timber foundations known worldwide 1.8–2.5 [Eurocode 1].
are in very good condition, assessments of the Alterations and changes in the upper structure:
quality and the load carrying capacity often have Floors may be added and attics are being put into
to be made. The objective of the research is to use, timber floors are reinforced with a concrete
determine the residual load carrying capacity of layer, walls are removed, replaced or added to
old foundation elements and to develop a model name just a few of possible alterations that
allowing predictions for the remaining lifetime. It may require an assessment of the structure and
will be shown that damage accumulation models, foundation.
which take into account age, load history and Underground building activities: The call for
current state of the material properties, can be underground building increases especially in
used in such an assessment procedure of timber densely populated areas. Often these densely
structures. Focus is given to the assessment of pile populated areas are near coastal regions where
foundations, since a reasonable amount of test soft soil conditions are present in many cases and
data is present. However, the method of service where structures are founded on timber piles.
life calculations can easily be extended to other Underground building activities generally cause
structures as will be shown. An example with some subsidence on ground level, which in turn
regard to the future lifetime of a timber pile may lead to structural damages (cracks). As a
foundation is given. result, disputes may arise between the building
owners and the contractors of the underground
building activities. An assessment of the founda-
2 Assessment of timber structures tion may be necessary to settle the disagreement.
Soil cleaning: Contaminated soil may be pres-
2.1 Reasons for assessment ent under houses in the vicinity of old factories
and on brownfield locations. When the contami-
There are several reasons why an assessment of nated soil has to be removed, the timber foun-
timber structures may be necessary or can be dation needs to be assessed for strength and the
justified at a certain moment in time. In the fol- possibility to withstand a temporary open air cli-
lowing a list is given which covers most of the mate. Removal of soil has a second negative
calls for assessment in the Netherlands, but the consequence. Due to the fact that vertical soil
list is probably valid for most areas in the world stresses reduce when soil is removed, the load
where timber is used in structural applications, carrying capacity of the foundation will conse-
from foundations to upper structures. quently be reduced.
Damage of the (upper) structure: When struc- Strategic maintenance planning: Maintenance
tures show signs of deterioration by means of schemes for other types of works of infrastructure
cracks, this might be due to diminishing perfor- may benefit from a combination of activities.
mance of the load bearing timber foundation Planned maintenance or renewal of cable
structure. Causes of performance loss can be long networks, sewage systems, road and pavement
term loads or excessive creep deformations. surfaces may benefit from combined works.
Re-use of timber: Structures are sometimes Redirection of traffic, losses in turnover of
demolished after their service life. The timber companies near the works and social problems
Materials and Structures (2007) 40:151–161 153

associated with the works are important param- not always optimised for the wood species under
eters in maintenance planning which could in- investigation.
clude the (regular) assessment of structures. Full scale pile testing: This is not always fea-
sible and can be costly, but when large renovation
2.2 Assessment techniques works are planned, the effort of testing a number
of piles is small compared to the potential bene-
Several possibilities exist for the assessment of fits. In addition, testing of the piles increases the
timber structures [2, 3]. It is outside the scope of knowledge on strength reduction as a function of
this paper to deal with all of the available tech- age and has therefore also benefits for other
niques, but they can be divided into the following projects. The last two are of particular interest
main areas. since they are the only measurement techniques,
Traditional inspection methods combining which give a direct indication of the residual
visual assessment and (sharp) tools. strength of the full-size timber. The determina-
NDT such as ultrasound and stress waves, but tion of the residual strength is of key importance
also Pilodyn hammer type equipment: The use of in the assessment procedure, when estimations
the Pilodyn hammer to determine the density is have to be made about the future use of the
well documented for measurements on standing foundation.
trees, sawn timber and timber suffering from Flow charts have been developed by a number
decay, [4]. Ronca and Gubana [5] developed of researchers and institutes in order to be able to
another ‘‘Pilodyn type’’, a so-called blow count make a reliable assessment [2, 3]. Generally, these
type apparatus. This apparatus counts the average flow charts require analysis of the original struc-
number of blows for penetrating a steel needle a ture, its uses, localization of defects/degradation
certain distance in timber. A severe change in the and classification. On the basis of an analysis of
number of blows indicates decay. The influence of the (historical) situation, the results of the
moisture content was also given, so this equip- inspection and the planned future use of the
ment might be used also for wet foundation tim- structure, an estimate is made on the residual
ber. The Pilodyn does not show an effect of lifetime and the necessity of repair or replace-
penetration angle on the density determined, ment. The age of a structure can often be esti-
while the blow count apparatus shows higher mated from historic files, but the load history is
resistance when used in tangential direction than sometimes more complex, although for founda-
in radial direction. The predicted strength tions the governing load can often easily be
increases accordingly, while this does not happen determined. Since timber has a load and time
with the Pilodyn hammer. dependent strength, the full short term resistance
Drilling and strength testing of bore cores: This based on modern design principles may no longer
technique allows estimating the residual strength be fully available, especially in the case of foun-
of piles or gluelines of glulam beams. The corre- dations, where the ratio between dead and live
lation between the strength of bore cores and the load is quite different from floors and roofs. In
compression strength is high and allows for the this respect, timber from floor and roof structures
determination of the residual strength of piles can be expected to have 100% strength capacity
without damaging the pile to an extent that they [8, 9]. The current state of the structure has to be
have to be replaced. Schwab et al. [6] and Rug determined. This will be discussed in the follow-
and Seemann [7] have extensively described and ing paragraphs. The expected future loads can be
tested the method. The results of Schwab et al. determined from the new building plans or from
will be used in the analysis of Dutch test results current design codes. Depending on the situation
on old piles. and legal requirements, deviations from loads
Drilling methods based on torque resistance specified in design codes could be accepted in
may also be used but are favoured for dry con- certain cases. The information obtained from the
ditions in buildings. In addition, these types of expected loads and the state of the structure
drills must be operated with care, since they are allows for calculations of the residual strength of
154 Materials and Structures (2007) 40:151–161

the structure and in conjunction with that the maximum load) was reached within 300 ± 120 s
expected ‘‘remaining’’ service life. with a constant loading head movement. The
values have been plotted as a function of the
distance to the pith. This is shown in Fig. 1. The
3 Strength research on old timber from pattern of lower strength around the centre of the
foundations piece and higher strength towards the outside is
similar to that found in new wood (Deresse, [11]).
3.1 Concert hall Amsterdam
3.2 Sluice at Schijndel
3.1.1 Pile tests and specimen preparation
3.2.1 Pile specimen
A number of spruce piles of the Concert hall in
Amsterdam were made available for research A 167 years old pile had been taken from a sluice
purposes at Delft University. The age of the piles in Schijndel, in the South-east of the Netherlands
was slightly over 100 years [10]. They were ram- during renovation works [12]. The sluice had been
med between 1880 and 1983 and were removed built in 1821 and during renovation works the
from the foundation in 1986, kept in a saturated upper part of a pile was sawn off and delivered to
environment and delivered to the laboratory. The the laboratory. The length of the piece was about
goals of the test programme were twofold. First, 750 mm. Visual inspection of the pile showed that
the strength of the piles was to be determined. In the uppermost layer of the pile and the pin for the
total seven full cross sections obtained from three connection with the head-beam suffered from
different piles were tested. The second goal was decay. The upper part of the pile was above
to determine a positive or negative effect from ground level, but below water level, for a number
temporary drying of the timber, for instance of years before the renovation works started and
during a period of restoration works. The ques- the oxygen content of the water was raised con-
tion was whether piles could be left to dry for a siderably due to the effect of ship propellers.
period of several months or whether protective
measures need to be taken. From the piles, discs 3.2.2 Material properties of cross sections
were sawn which in turn were divided into square
compression test specimens, for the determina- The pile was treated as follows: two discs were
tion of the cross sectional strength and the influ- sawn from the top and the bottom, each disc
ence of moisture content on the strength after a
period of drying.
The strength of the piles with full cross section
and saturated condition was determined in a
standard compression test set-up. The amount of
water pressed out during the tests, was almost
completely sucked into specimen again during
and after the unloading. The average, minimum,
and maximum values are presented in figure 5.

3.1.2 Material properties

The cross sectional strength of the spruce pile was

determined by sawing a disc into a squared pat-
tern of 35 · 35 · 210 mm specimens. These were
tested in compression in fully saturated condition.
The test procedure was according to ISO 3787 for Fig. 1 Compression strength of a cross section of a
compression parallel to the grain. Failure (or 100 year old pile without decay
Materials and Structures (2007) 40:151–161 155

having a thickness of about 100 mm. From the

two outer discs compression specimens were sawn
with the following dimensions: 30 · 30 · 90 mm,
from the two inner discs compression specimens
were sawn having dimensions of 20 · 20 ·
60 mm. From the squared pattern, the specimens
were either tested in the fully saturated condition,
or dried to approximately 13% moisture content.
The compression strength of the pieces was
determined in a standard testing machine. The
remaining middle piece of the pile was tested as a
full cross section. The compression strength was
10.5 N/mm2. Location of the pith was determined
and also the thickness of the sapwood. The latter
can be difficult, since sapwood is not always rec-
ognizable with the naked eye. Fig. 3 Compression strength of the same pile at 1 m depth
In Fig. 2 the upper part is shown, while Fig. 3 without decay
shows the same pile at one meter depth. Clearly,
on the upper part it can be seen that the outside is whether precautions need to be taken to prevent
deteriorated to a very low strength level, but the piles from drying, as there were indications
internal heartwood still has a resistance compa- that timber stored under water for very long
rable to that of new spruce piles [13]. At 1 m periods may sometimes disintegrate when dried.
depth, the strength is still equally distributed over Therefore, additional discs were sawn from the
the cross section. piles from the Amsterdam concert hall to study
the strength of dried specimens relative to wet
3.3 Drying of piles in open air specimens.
The discs were sawn again into square com-
As indicated earlier, piles sometimes have to be pression specimens with sizes 35 · 35 · 210 mm.
left in open air for a certain period of time, for In addition to the series that were tested directly
instance when contaminated soil has to be after removal of the piles, the following series
removed or cleaned. The question was raised were tested:
3 series in 20C and 80–90% RH, to be tested
after 3, 6 and 9 months;
3 series in a covered outdoor climate to be tested
after 3, 6 and 9 months,
3 series of new spruce, dry and kept under water
for 3 and 6 months.
In Fig. 4 the compression strength results of all
the tests are shown as a function of the moisture
content. Density of the spruce was about 420 kg/
m3, related to a moisture content of 12%. A
failure in the measurement of the specimens that
were kept in the 80–90% relative humidity
resulted in a moisture content of around 12%,
where 18–20% was expected. Despite this mois-
ture content error, the trend is clear: above 30%
Fig. 2 Compression strength of a cross section of a moisture content there are no hard conclusions to
167 year old pile with decay be drawn that timber from old piles is weaker
156 Materials and Structures (2007) 40:151–161

with age, but the cause of the strength decrease is

not known with certainty, but many factors play a
role [15]. Both load and decay will cause a
reduction in strength and since the load level
during the service life was not known, mechanical
damage will have accumulated to a certain
extend. The straight lines have been determined
on the average values of the test data of the
heartwood and sapwood data respectively. The
Fig. 4 Compression strength of temporary dried timber
test results on full piles from Delft University
from foundation piles have been added. These piles had an age of
100 years (six specimens) and one of 167 years.
than new timber, and drying of timber from old The strength ratio of these data has been calcu-
piles leads to a strength increase. Consequently, lated on the basis of the average strength value in
from these results it can be concluded that timber wet condition of new spruce piles which has been
free from decay can be left in a (covered) open air determined at 20.0 Mpa [13]. The results fit in
climate for at least a period of 9 months, without quite well and indicate that the residual strength
additional measures to be taken. It is assumed of piles is mainly governed by the amount of
that the possible occurrence of longitudinal heartwood in the cross section. It must be kept in
cracks as a result of drying is not influencing the mind that is not known to which load level the
compression strength parallel to the grain in a piles have been loaded. The construction of
negative way. foundations was based more on experience than
on engineering practise and therefore stress levels
3.4 Comparison with strength data from are difficult to be determined when structures of
literature one hundred years of age or more are studied.

In Fig. 5 the results of parallel to the grain tests

are shown, taken from small round specimen 4 Influence of decay on timber strength
drilled from piles [4]. The age of the piles varied
between 70 years and 640 years. The scatter is Timber in ground may suffer from decay by fungi
very large (Schwab et al. [6]), but still some con- or bacteria. Fungal decay may occur when the
clusions can be drawn. The strength decreases ground water level decreases below the upper

Fig. 5 Resistance of
timber from foundations
after service life
Materials and Structures (2007) 40:151–161 157

part of the pile. Bacterial decay is much slower might still have sufficient load carrying capacity.
but can also take place below the ground water Consequently, the ratio between decayed area
level. Clearly, in Fig. 2, the upper part of the pile and non-decayed area is an important parameter
shown shows severe decay on the outside. This is that can be used when the remaining lifetime of
generally the sapwood area, while the inner the structure has to be determined. In addition,
heartwood is much more durable. Liese and Sta- depending on the type of decay, the strength of
mer [16] studied the influence of brown rot on the decayed timber is not necessarily zero. The
compression strength. The fungi studied are the strength of decayed timber may still serve in
very common and comprised Merulius lacrimans, safety calculations.
Coniophora cerebella, Serpula himantioide and
Poria vaporaria. The relative strength of the
timber decreased by 30–100% within a time span 5 Service life and damage modelling
of just 300 h.
The size of the specimens used was not given, After an assessment of the foundation, the final
so the penetration rate of the decay as a function step is to predict the load carrying capacity and
of time could not be derived. This, from an the residual lifetime. The residual lifetime should
engineering point of view, is equally important as be estimated in order to be able to determine the
the rate of decay, since this influences the necessity of repair or replacement. A new ap-
remaining service life to a great extent. Wilcox proach for this is the assessment of the strength of
[17] as well as Kothe [18] have shown the influ- the pile by means of damage accumulation mod-
ence of certain types of decay on the residual els. Damage accumulated models have been used
strength by means of mass loss-strength relation- in timber research describing the strength devel-
ships and by reduced cross sections, see Fig. 6. opment of timber under long term loads. Several
This allows for assumption on the residual models have been developed for instance by
strength of structural elements when they have Gerhards [20] that will be used here. Other
been attacked by fungi. damage models are available, for instance those
More recently, Thornton et al. studied the rate of Foschi and Yao [21] and Nielsen [22]. Its
of decay in timber poles at ground level [19]. The parameters are relatively difficult to determine
rate of decay was found to be about 1.3 mm/year and show a large scatter between different timber
for Douglas fir in moderate decay hazard cir- species (Foschi et al. [21]) and joints [23]. The
cumstances. Foschi and Yao model was shown to be a special
Pile foundations where decay is present do not case of the damage model of Van der Put [24]
necessarily have lost their load carrying capacity which is based on true physical parameters
to such extend that replacement becomes (Krausz and Eyring 1975). Also the model of
unavoidable. If the sapwood of timber piles is Gerhards can be related to ‘‘true’’ material
deteriorated, the remaining heartwood section properties, since it leads to the same type of
equations as given by the damage model of both
Caulfield [25] and Van der Put. Since these
models only describe the strength development in
time they have to be modified to include the
influence of deterioration on the material prop-
erties. For instance biological degradation decay
of the cross section that may have considerable
effect on the strength capacity of members.
By combining durability models with strength
models, it becomes possible to model the
‘‘residual lifetime’’ of structures. Different app-
roaches are possible using probabilistic calcula-
Fig. 6 Relative compression strength [18] tions [26], of kinetic based models [27], but here
158 Materials and Structures (2007) 40:151–161

the exponential damage accumulation model of timber. Thus, a time dependent resistance fs(t) is
Gerhards is used with modified strength values. introduced in the model:
The limit state function on which most design
 
codes are currently based is generally written as: da rðsÞ
¼ exp
C1 þ C2 ð4Þ
dt fs ðtÞ
Z ¼R
S ð1Þ

with: Z = the limit state; R = the resistance; The damage a takes a value: 0  a  1. By
S = the solicitation or load. definition, failure occurs when a=1. This means
In reality, both the resistance and the solicita- that while 1– a > 0 the structure is still able to
tion are varying in time and the limit state func- carry the load at that point in time and Z(t) = 1–a.
tion can be written as: Consequently, the damage function can be used
to estimate residual lifetimes of structures when
ZðtÞ ¼ RðtÞ
SðtÞ ð2Þ the stress function r(s) representing the load path
from the time of erection of the structure until the
The solicitation S(t) concerns all load types end of the time span under consideration is
possible, such as mechanical or physical (tem- known as well the strength fs(t) representing the
perature, moisture). A structure is assumed not to time dependent load carrying capacity of the
have failed while Z(t) > 0. In Fig. 7 this is shown material.
schematically. With regard to the stress function r(s) it may
The probability of failure increases, since the be necessary to include other loads than
resistance of members generally decreases with mechanical. A ‘‘moisture load’’ due to humidity
age. In case of timber, the resistance R(t) depends variations may be included in cases where tension
on the load and the load history. Mositure content stresses perpendicular to the grain are expected.
and temperature also have an influence, but may For instance this is the case for curved or pitched
be assumed constant for piles in the ground. cambered glued laminated members, where
Consequently, Eq. 2 can be written as: cracks have been discovered in the area with high
stresses perpendicular to the grain, relatively
ZðtÞ ¼ RðsðsÞ; tÞ
SðtÞ ð3Þ early after erection of the structure [28]. The
parameters C1 and C2 are determined from stan-
For the resistance function R(s(s),t), the Ger- dard time to failure tests.
hards damage function is used here, which is easy For pile foundations the normal force has to be
and straightforward to use [20], but now the checked against the load carrying capacity. When
material strength is modified depending on the the cross section is denoted Atot, the value of the
amount of decay and residual strength of decayed pile resistance Fu is determined Fu = fc,0Atot.

Fig. 7 Schematic
representation of the
distribution of lifetimes of
structures (Siemes,
personal communication)
Materials and Structures (2007) 40:151–161 159

When it is assumed that the pile may suffer from

decay, after a number of years the non-decayed
cross section will be reduced to Arem (remaining
cross section). Denoting fc,0,dec the strength of
the decayed timber and Adec the cross section of
the decayed area, the pile resistance can be
written as:

Fu ¼ fc;0 Arem þ fc;0;dec Adec ð5Þ

Equation 5 can be further modified into a more

general form. When the ratio between the
remaining non-decayed cross section and the total
cross section is defined as:

Arem
d¼ ð6Þ
Atot
and when the strength ratio between the decayed
timber and the non-decayed timber is defined as:

fc;0;dec
b¼ ð7Þ
fc;0
Fig. 8 (a) Structure, (b) concrete extension pile, (c)
Eq. 5 becomes: timber pile (tapered), (d) ground water level, (e) negative
skin friction, (f) weak clay, (g) positive skin friction, (h)
Fu ¼ fc;0 Atot ðdð1
bÞ þ bÞ ð8Þ load bearing soil layer, (i) pile toe resistance

The value of fs(t) in Eq 4 can be replaced by Fu in a few years. In addition it is assumed that the
when the stress function r(s) is replaced by the ground water level records show that the level
load. Both b and d can be made time dependent, came below the top level of the piles 100 years
for instance when the decay rate is known. The after the foundation was built. Consequently,
parameters C1 and C2 in the damage model are the cross section is reduced to Arem after
determined on the basis of time to failure tests on 100 years in a way comparable to the reduction of
timber [20, 21] and joints [23]. the diameter as shown in Fig. 2, due to degrada-
An example will be shown on the basis of a tion of the outer layer.
problem regularly occurring with pile founda- In Figs. 9 and 10 the remaining service life
tions. A typical pile foundation is shown in Fig. 8. predictions are calculated shown for three values
The assumed pile diameter where the load of d(d = 0.8, 0.7 and 0.6) and two value of b
carrying capacity has to be verified is 160 mm, (b = 0.5 and 0.0). b = 0.0 gives always a safe
with a characteristic load of 100 kN. As a result of approximation since it neglects any contribution
unexpected decrease in ground water level the to the load carrying capacity of the decayed parts.
upper side of the pile was exposed to soil condi- In this example the pile cross section is checked
tions leading to severe decay of the sapwood of for safety against compression stresses parallel
the pile and after a number of years the cross to the grain. Also, in this example decay was
section was decreased to Arem. The lag time assumed to be initiated after 50 years of service
before decay is initiated in ground conditions life. In actual situations an estimate will have to
above the ground water level is taken to be zero be made about the load history of the structure
and it is assumed that the sapwood is deteriorated and the decay, to be fed into the model.
160 Materials and Structures (2007) 40:151–161

supports. The mathematical form of Eqs. 5–8

remains the same. For shear the cross section of a
beam b · h is needed in the calculations. When
the cross section is decreased by drying cracks,
width b decreases from the original state to the
cracked state bcrack increasing the relative load
level in Eq. 4. In most monuments, timber sizes
are large and the drying cracks are located near
the neutral axis, i.e. the location with the highest
shear stresses. Consequently, d as derived in Eq. 6
does not change and b is always zero in case of a
Fig. 9 Service life prediction for three levels of d and
b = 0.4 crack. The decrease in load carrying capacity as a
result of drying cracks is generally small, since the
actual shear resistance is often much higher than
code values [29].

7 Conclusions

A damage model from literature with parameters

derived from time to failure analysis of timber
and timber joints have been combined with
strength reduction models as a result of decay. An
example in which a strength reduction factor was
given for compression parallel to the grain was
Fig. 10 Service life prediction for three levels of d and b = 0
given and the influence on the residual lifetime
could be shown. However, the same principle can
Obviously, b values close to zero results in a be used to check bending stresses and support
rapid drop in remaining service life. Besides the conditions. The combination of a damage model
analysis of the influence of decay, the model can with deterioration influence on the strength
also be used to study the influence of safety fac- properties has resulted in a model which can
tors. Both load and material factors can be anal- easily be used to study the sensitivity of structures
ysed. Especially for existing structures that have for different load sets and material parameters.
to be assessed it may be necessary to have a closer When structures are found with decay, either
look on those values. active or inactive, the model can also be used to
analyse the effectiveness of different measures
that can be taken and the influence of the future
6 Other applications use of the structure. In this example the calcula-
tions have been made using fixed values, but
The approach given above may also be used for probabilistic calculations could be made when
the assessment of other type structures than both load and strength distributions are entered
foundations. In monuments, very often large in the model.
drying cracks have developed over time, espe- Regarding the influence of decay on the
cially when large beams have been used where strength it seems necessary that more knowledge
the pith is located in the centre of the beam. is developed on the residual strength of decayed
Depending on the sizes of the cracks, repair wood. May be even more important is the rate at
measures may be needed in areas where shear which the decay reduces the cross section of
stresses determine the design, i.e. near the timber elements.
Materials and Structures (2007) 40:151–161 161

References 16. Liese J, Stamer J, Angewandte Botanik Bd. 16, 1934

Page 363. [Referenced from Kollman]
1. Eurocode 1 (1994) Basis of design and actions on 17. Wilcox WW (1978) Review of literature on the effects
structures. Part 1: Basis of design, ENV 1991-1, sept. of early stages of decay on wood strength. Wood Fibre
2. Bonamini G (1995) Restoring timber structures – 9(4):252–257
Inspection and evaluation, STEP Lecture D3, Centrum 18. Kothe E (1988) Auswirking van Holzschäden durch
Hout, The Netherlands, ISBN 90-5645-002-6 Pilze und Insekten auf die Standsicherheit van Holz-
3. Görlacher R (1999) Historische Holztragwerke. bauwerken – eine Bestandsaufnahme. Bautechniek 75,
Untersuchen, Berechnen und Instandsetzen. Univer- Heft 8 pp 552–558
sität Karlsruhe (TH), Karlsruhe 19. Thornton J, Johnson G, Nguyen J-K (1994) Specimen
4. Görlacher R (1987) Zerstörungsfreie Prüfung van life statistics and mean rates of biodeterioration for
Holz: Ein in ‘‘situ’’-Verfahren zur Bestimmung untreated heartwood timbers exposed in-ground for 23
der Rohdichte. Holz als Roh- und Werkstoff 45, pp years. PTEC proc., pp 449–453
273–278 20. Gerhards CC, Link CL (1987) A cumulative damage
5. Ronca P, Gubana A (1996) Evaluation of actual ulti- model to predict load duration characteristics of lum-
mate strength in wooden structures by means of in situ ber. Wood Fibre Sci 19(2):147–164
penetration test. International Wood Engineering 21. Foschi RO, Yao RC (1986) Another look at three
Conference, Vol III, pp 226–232, New Orleans, 1996 duration of load models. CIB/W18 Timber Structures,
6. Schwab E, Waschau A, Willeitner H (1982) Bohrkerne Paper 19-9-1, Meeting 19, September 1986, Florence,
zur Beurteilung der Festigkeit hölzerner Rammpfähle, Italy
Bauen mit Holz 9, 84, 1982 22. Nielsen LF (1992) DVM Analysis of wood, lifetime,
7. Rug W, Seemann A (1988) Festigkeit von Altholz. residual strength and quality. CIB/W18 Timber Struc-
Holztechnologie 29(4):186–190 tures, paper 25-9-2, Ahus, Sweden
8. Kuipers J (1986a) Effect of age and/or load on timber 23. Van de Kuilen JWG (1999) The residual load carrying
strength. CIB-W18 Timber Structures, paper 19-6-1, capacity of timber joints, HERON Vol 44-3, pp 187–
Florence, Italy 214, ISSN 0046-7316, 1999
9. Kuipers J (1986b) Effect of age and/or load on timber 24. Van der Put (1986) TACM, A model of deformation
strength. CIB/W18 Timber Structures, Paper 19-6-1, and damage processes based on the reaction kinetics
Meeting 19, September 1986, Florence, Italy of bond exchange. IUFRO S5.02/CIB-W18/19-9-3,
10. Van Bueren R (1987) Druksterkte van 100 jaar oude Meeting 19 Florence, Italy, 1986
houten funderingspalen van het Amsterdamse con- 25. Caulfied DF (1985) A chemical kinetics approach to
certgebouw. TU Delft report 25-87-41/16-HE-13, the duration-of-load problem in wood. Wood and
October 1987 Fiber Sci 17(4):504–521
11. Deresse T (1988) The influence of age and growth rate 26. Leicester RH, Foliente GC, Cole IS, Wang C-H,
on selected properties of Main-grown red pine. Uni- Mackenzie C (2002) Prediction models for engineered
versity of Maine, Orono, ME. Ph.D. Thesis durability of timber in Australia, 9th Int. Conf. on
12. Kurstjens PBJ (1988) De druksterkte van een 167 jaar Durability of Building Materials and Components,
oude heipaal uit de sluis nr. 3 te Schijndel TU Delft March 2002, Brisbane, Australia
report 25-88-36/05-HE-18, June 1988 27. Winandy JF, Lebow PK, Murphy JF (2002) Predicting
13. Van de Kuilen JWG (1994) Bepaling van de kara- current serviceability and residual service life of ply-
kteristieke druksterkte van houten heipalen. TNO-re- wood roof sheathing using kinetics-based models. 9th
port 94-CON-R0271, 1994 (in Dutch) Int. Conf. on Durability of Building Materials and
14. Peek R-D, Willeitner H (1981) Behaviour of wooden Components, March 2002, Brisbane, Australia
pilings in long time service, Proc. Tenth Confer- 28. Morlier P, Ranta-Maunus A (1998) DOL effects of
ence International Society of Soil Mechanics and different sized timber beams. Holz als Roh- und
Foundations, Vol III, 9/23, pp 147–152, Stockholm, Werkstoff 56
Sweden 29. Frech P (1987) ‘Beurteilungskriterien fur Rissbildun-
15. Varossieau WW (1949) Opgegraven en aangetast hout gen bei Bauholz im konstruktiven Holzbau’. Bauen mit
uit biologisch oogpunt bezien. CIMO Delft, 1949 Holz, 9