-
NDTCE’09, Non-Destructive Testing in Civil Engineering Nantes,
France, June 30th – July 3rd, 2009
Ultrasonic indirect method for evaluating clear wood strength
and stiffness
José MACHADO1, Pedro PALMA1, Sofia SIMÕES1
1 Laboratório Nacional de Engenharia Civil, Lisboa, PORTUGAL,
e-mail [email protected]
Abstract Strength and stiffness assessment of timber members
in-service is a crucial task in the
rehabilitation of a timber structure. Unsatisfactory results
given by traditional procedures applied for appraisal of in-service
structural timber members are a strong motivation for research on
non-destructive techniques (NDT). For assessing strength and
stiffness of clear wood zones of structural members an indirect
ultrasonic method is proposed, given the advantage of requiring the
access to only one surface of the element.
This paper presents results of a study comprising three
different wood species (oak, maritime pine and european spruce)
with two different cross-sections and 150kHz transducers. Different
arrangements of the transducers on the same surface were tested and
the results compared with the ones obtained with a direct method
(transducers placed at opposite ends of the specimens). The results
obtained are discussed having in mind the ones already obtained in
almost clear structural timber elements as well as in small size
clear wood specimens.
Résumé L'évaluation de la résistance et de la rigidité des
éléments en bois en service est une tâche
cruciale dans le processus de réhabilitation d’une structure en
bois. Les résultats insuffisants donnés par des procédures
traditionnelles pour l'évaluation des membres en bois en service
sont une motivation forte pour la recherche sur les techniques non
destructives (TND). Pour évaluer la résistance et la rigidité des
zones en bois sans défaut des éléments structurels, une méthode
ultrasonore indirecte est proposée ; elle présente l'avantage
d'exiger l'accès à seulement une surface de l'élément.
Ce document présente des résultats d'une étude comportant trois
espèces en bois différentes (chêne, pin maritime et sapin européen)
avec deux sections transversales différentes et utilisant des
sondes de 150kHz. Différentes implantations des sondes sur la même
surface ont été examinées et les résultats ont été comparés à ceux
obtenus avec une méthode directe (sondes placées aux extrémités des
corps d’épreuve). Les résultats obtenus sont discutés en ayant en
tête les résultats déjà obtenus en éléments de bois de construction
presque sans défaut aussi bien que dans les petits échantillons en
bois sans défaut.
Keywords non-destructive testing; wave propagation; pine ;
spruce ; oak
1 Introduction Timber inspection is conducted with the aim of
providing valuable information for safety
analysis of timber structures. Reliable tools for assessing
mechanical behaviour of timber members are specially needed in the
case of old timber structures where no information is available on
wood species and characteristic strength and stiffness values.
Long-established procedure for assignment of strength and stiffness
values to structural timber members is
-
NDTCE’09, Non-Destructive Testing in Civil Engineering Nantes,
France, June 30th – July 3rd, 2009
based on application of visual strength grading standards. This
approach underestimates the actual strength capacity of timber
members and gives unsatisfactory answers.
Assessment of strength and stiffness of timber members is
therefore a challenge. On this topic a national research project
began in 2008, having in mind the development of a probabilistic
safety analysis model for old timber structures. The project is
composed of several tasks, being one related with the development
of tools for predicting the strength and stiffness capacity of
timber elements, as a joint distribution of weak and clear wood
zones.
Evaluation of strength and stiffness of clear wood zones will be
studied by using non-destructive (velocity of ultrasonic wave
propagation), and semi-destructive testing (extraction of wood
material to perform tension testing, identification of wood species
and determination of density). Ultrasounds application to
in-service timber members should have in mind that in most
situations only one face of the element will be available and
therefore standard direct method (longitudinal wave propagation
between ends of the element) used for sawn timber grading is not
applicable and an alternative method should be envisaged, Fig.
1.
Figure 1. Indirect (left) and direct (right) methods applied to
timber elements
Indirect methods are often used with concrete. PUNDIT manual
mentions that for concrete the pulse velocity determined by the
indirect method of testing will be lower than that using the direct
method, being possible however to point out a conversion factor of
1.05 (Vdirect ∼ 1.05 Vindirect).
Indirect method was already used in some studies carried out by
the research team. Application of ultrasounds on the assessment of
clear wood zones combined with information from knots
characteristics show to improve the capacity of predicting strength
and stiffness of structural dimension pine elements [1]. Ultrasonic
testing of small clear pine specimens also show a good capability
of predicting the compression strength parallel to the grain [2]
and an acceptable capacity for predicting stiffness in tension and
compression parallel to grain of clear wood chestnut specimens
[3].
Although the positive results obtained so far with the
application of indirect method for assessing the stiffness of clear
wood zones, questions as the effect of the distance between
transducers, effect of specimen volume and dimensions and wave
propagation mode, among others, have to be studied in respect with
the equipment available (PUNDITplus) and test-setup proposed.
The objective of this paper is to analyze the characteristics of
the signal (velocity and frequency) that is captured by the
receiver transducer, evaluate the effect of distance between
transducers (in order to maintain the actual procedure or to
modified it) and to evaluate the influenced that wood surface and
deeper layers have in respect to propagation mode.
-
NDTCE’09, Non-Destructive Testing in Civil Engineering Nantes,
France, June 30th – July 3rd, 2009
2 Experimental work The test pieces, Table 1, were conditioned
in a 20ºC±2ºC temperature and 65%±5%
relative humidity environment until constant mass was achieved
(less than 1% variation in 2 hours), prior to began ultrasonic
testing.
At one edge of each test piece eight marks were drawn (four
marks placed symmetrically to the other four in relation to the
middle of the test piece, Fig. 2), corresponding to a distance
between transducers of 10cm, 20cm, 30cm and 40cm.
Table 1. Characteristics and number of test pieces used Number
of test pieces (N)
dimensions (mm) Wood species 50 x 150 x 500 25 x 150 x 500
Maritime pine (Pinus pinaster Ait.) N = 3 N = 3
European spruce (Picea Abies L.) N = 3 N = 3
White oak (Quercus alba L.) N = 3 N = 3
Figure 2. Setup used for indirect testing. Detail of compression
spring placed on top of the
transducer
Ultrasound waves were generated using a PUNDITplus equipment, on
one-shot basis as pulse mode, with two flat contact 150kHz
transducers (transmitter and receiver) under a constant pressure of
60N assured by a controlled deformation of a compression spring
placed on top of the each transducer, Fig. 2. The equipment was
combined with a digital oscilloscope and a laptop allowing the
signal acquisition at the receiver transducer. Time-of-flight was
taken directly from the display of PUNDITplus equipment. Ultrasound
velocity (V) was determined according with Eq. 1, taking the
time-of-flight (t) and distance (d) between transducers. In each of
the four positions five readings were determined.
tdV = (1)
After performing this first stage, in each test piece at the
middle of its length a 10mm
depth cut was performed, Fig 3, and all measurements were
repeated.
-
NDTCE’09, Non-Destructive Testing in Civil Engineering Nantes,
France, June 30th – July 3rd, 2009
Figure 3. 10mm cut made at the middle-length of the edge of each
test piece
3 Results and discussion Fig 4 shows typical frequency spectrums
obtained for a distance between transducers of
10cm and 40cm. In the first case it is recognized from the
figure a strong peak around 25kHz and small peak around 130kHz
whereas for a distance of 40cm only a peak at 25kHz is visible.
0 0.5 1 1.5 2 2.5x 105
0
0.01
0.02
0.03
0.04
0.05
Frequency (Hz)
FFT Amplitude
Teste piece S1d = 10cm
0 0.5 1 1.5 2 2.5x 105
0
0.02
0.04
0.06
Frequency (Hz)
FFT Amplitude
Test piece S1d = 40cm
Figure 4. Characteristic frequency spectrums obtained for spruce
test piece S1
These results are inline with previous results [4] using a
standard PUNDIT equipment and 37kHz, 54kHz and 150kHz transducers
that have shown that actual frequency of transducers could be quite
different from those expected in advance (from information on
resonant frequency of piezoelectric crystals), Fig. 5. The
significant loss of high frequency content even for a distance
between transducers of only 10cm is related with the dispersive
nature of wood material and the fact that most of the signal energy
is transmitted to the back-wall (inferior edge) of the test piece
being only a small part of the energy of the signal transmitted
along the edge surface of the test piece. For a distance of 40cm
all the high frequency of the signal is lost (for all the wood
species tested).
0.5 1 1.5 2 2.5x 10
5
0
2
4
x 1012
Frequency (Hz)
FFT Amplitude
Figure 5. Frequency spectrum of a 150kHz transducer(right) as
result of a direct contact
(gel couplant) between transmitter-receiver transducers [5]
-
NDTCE’09, Non-Destructive Testing in Civil Engineering Nantes,
France, June 30th – July 3rd, 2009 Fig. 6 shows the relative
differences obtained between the velocity using indirect method
(for all the distances between transducers) and longitudinal
velocity, given by the direct method.
Test pieces 50mmx150mmx500mm
δ (P1,50) δ (P2,50) δ (P3,50) δ (S1,50) δ (S2,50) δ (S3,50) δ
(O1,50) δ (O2,50) δ (O3,50)
10 20 30 40
Distance betweentransducers (cm)
-12
-8
-4
0
4
8
12
Rel
ativ
e di
ffere
nce
(%)
Test pieces 25mmx150mmx500mm
δ (P1,25) δ (P2,25) δ (P3,25) δ (S1,25) δ (S2,25) δ (S3,25) δ
(O1,25) δ (O2,25) δ (O3,25)
10 20 30 40
Distance betweentransducers (cm)
-8
-6
-4
-2
0
2
4
6
8
Rel
ativ
e di
ffere
nce
(%)
Test pieces 50x150x500mm
δ (P1,50) δ (P2,50) δ (P3,50) δ (S1,50) δ (S2,50) δ (S3,50) δ
(O1,50) δ (O2,50) δ (O3,50)
10 20 30 40
Distance betweentransducers (cm)
-35
-30
-25
-20
-15
-10
-5
0
5
10
Rel
ativ
e di
ffere
nce
(%)
Test piece 25mmx150mmx500mm
δ (P1,25) δ (P2,25) δ (P3,25) δ (S1,25) δ (S2,25) δ (S3,25) δ
(O1,25) δ (O2,25) δ (O3,25)
10 20 30 40
Distance betweentransducers (cm)
-35
-30
-25
-20
-15
-10
-5
0
5
Rel
ativ
e di
ffere
nce
(%)
Figure 6. Relative differences (mean and 95% confidence limits)
observed between velocity
values obtained by the indirect and the direct method (above -
before making the cut; below – after making the cut; P – pine; S –
spruce; O - oak)
For the first part of the study (test pieces without cut) the
relative difference between velocities (indirect versus direct
method) can be considered for all test pieces and distances between
transducers as constant in a range of ± 10%. The results obtained
after making the 10mm cut on the edge (eliminating surface
transmission) show that as the distance between transducers
increases the effect of the cut decreases and for a distance of
40cm the results with or without cut could be considered the same.
This results seems to indicate that as the distance becomes larger
the influence of deeper wood layers in velocity of wave propagation
velocity increases. These findings support the use of indirect
method for predicting the longitudinal velocity moreover the use of
40cm distance between transducers, distance already used in a
previous study [1], as it provides a good correlation, Fig. 7,
between the indirect and direct velocity readings.
-
NDTCE’09, Non-Destructive Testing in Civil Engineering Nantes,
France, June 30th – July 3rd, 2009
VL = 1351.5 + 0.75 Vind,40
3000 4000 5000 6000
Indirect velocity d=40cm - Vind,40 - (m/s)
3000
4000
5000
6000
Long
itudi
nal v
eloc
ity -
VL -
(m/s
)
r2 = 0.90
Figure 7. Correlation between indirect velocity and longitudinal
velocity adopting a distance between transducers of 40cm (pieces
without cut) (dash lines – 95% confidence
interval limits)
4 Conclusions The results obtained show the feasibility of using
a indirect transmission method with
transmitter and receiver transducers 40cm apart for predicting
the longitudinal velocity in cases where the ends of the timber
elements are not available. The study also shows that indirect
testing modifies significantly the characteristics of the frequency
spectrum of the wave received. Although more studies are needed for
a full comprehension of the propagation mode in the indirect
method, the results now obtained give a strong support to the use
of the method within the project for assessment of strength and
stiffness of clear wood zones of structural dimension timber
elements.
Acknowledgements The authors gratefully acknowledged the support
given by Fundação para a Ciência e a
Tecnologia (FCT) to the project PTDC/ECM/66527/2006 entitled
“Safety evaluation of timber structures through non-destructive
methods and stochastic analysis”.
References 1. Machado, J. S. (2007) “Strength appraisal of
wooden members in service by combining
new and old technology”, Proc. of Structural Studies, repairs
and maintenance of heritage architecture X, Ed by C. A. Brebbia,
4-6 July 2007, Prague, Czech Republic, pp. 279-287.
2. Machado, J. S., Costa, D., Cruz, H. “Evaluation of pine
timber strength by drilling and ultrasonic testing”, Proc. of the
International Symposium Non-destructive Testing in Civil
Engineering, Ed by German Society for Non-Destructive Testing and
Federal Institute for Materials Research and Testing, 16-19
September 2003, Berlin, Germany, CD-ROM.
3. Artur, F., Lourenço, P., Machado, J. S. (2007)
“Non-destructive evaluation of the mechanical beavior of chestnut
wood in tension and compression parallel to grain”, International
Journal of Architectural Heritage, Vol. 1, Issue 3, pp.
272-292.
4. Machado, J. S. (2001) “ Avaliação da variação das
propriedades mecânicas de pinho bravo (Pinus pinaster Ait.) por
meio de ultra-sons” (in Portuguese), PhD Thesis, Universidade
Técnica de Lisboa, Lisboa, Portugal, 265p.
IntroductionIndirect (left) and direct (right) methods applied
to timber
Experimental workCharacteristics and number of test pieces
usedSetup used for indirect testing. Detail of compression
sprin10mm cut made at the middle-length of the edge of each
test
Results and discussionCharacteristic frequency spectrums
obtained for spruce test Frequency spectrum of a 150kHz
transducer(right) as result oRelative differences (mean and 95%
confidence limits) observCorrelation between indirect velocity and
longitudinal veloc
Conclusions
/ColorImageDict > /JPEG2000ColorACSImageDict >
/JPEG2000ColorImageDict > /AntiAliasGrayImages false
/DownsampleGrayImages true /GrayImageDownsampleType /Bicubic
/GrayImageResolution 300 /GrayImageDepth -1
/GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true
/GrayImageFilter /DCTEncode /AutoFilterGrayImages true
/GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict >
/GrayImageDict > /JPEG2000GrayACSImageDict >
/JPEG2000GrayImageDict > /AntiAliasMonoImages false
/DownsampleMonoImages true /MonoImageDownsampleType /Bicubic
/MonoImageResolution 1200 /MonoImageDepth -1
/MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true
/MonoImageFilter /CCITTFaxEncode /MonoImageDict >
/AllowPSXObjects false /PDFX1aCheck false /PDFX3Check false
/PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true
/PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ]
/PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [
0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile ()
/PDFXOutputCondition () /PDFXRegistryName (http://www.color.org)
/PDFXTrapped /Unknown
/Description >>> setdistillerparams>
setpagedevice