Page 51 - Kutnar, Andreja, et al., eds., 2015. Proceedings of the 1st COST Action FP1307 International Conference - Life Cycle Assessment, EPDs, and modified wood. University of Primorska Press, Koper.
P. 51
erimental
characterization
of
the
mechanical
performance
of
wood
in
a
controlled
environment:
use
of
acoustic
emission
to
monitor
crack
tip
propagation
Malick
Diakhaté1,
Seif
Eddine
Hamdi2,3,
Emilio
Bastidas-‐Arteaga4,
Rostand
Moutou-‐Pitti2,3
1
Université
de
Bretagne
Occidentale,
LBMS,
43
Quai
de
Léon,
29600
Morlaix,
France,
malick.diakhate@univ-‐brest.fr
2
Clermont
Université,
Université
Blaise
Pascal,
Institut
Pascal,
BP
10448
Clermont-‐
Ferrand,
France,
seif_eddine.hamdi@univ-‐bpclermont.fr
3
CNRS,
UMR
6602,
Institut
Pascal,
63171,
Aubière,
France,
rostand.moutou_pitti@univ-‐
bpclermont.fr
4
LUNAM
Université,
GeM,
CNRS
UMR
6183/FR
3473,
44322
Nantes,
France,
emilio.bastidas@univ-‐nantes.fr
Keywords:
Acoustic
emission,
clustering,
data
mining,
probability
of
detection,
wood
material
Monitoring
material
damage
by
means
of
acoustic
emission
lies
in
the
ability
to
identify
the
most
relevant
descriptors
of
cracking
mechanisms.
The
sudden
release
of
stored
energy
during
the
damage
process,
known
as
the
acoustic
emission
(AE),
is
a
very
suitable
technique
for
in
situ
health
monitoring
applications.
In
the
present
study,
a
tensile
test
is
performed
on
a
DCB
(Double
cantilever
Beam)
wood
specimen
(Figure
1)
to
generate
mode
I
cracking.
The
AE
events
recorded
during
the
tensile
test
are
correlated
to
actual
damage
in
terms
of
cracking
length.
Various
signal
processing
and
pattern
recognition
techniques
have
been
performed
for
damage
feature
extraction
from
AE
signals.
In
this
study,
the
Hilbert–Huang
transform
(HHT,
Huang
2005)
is
used,
for
relating
a
specific
damage
mechanism
to
its
acoustic
signature.
The
applicability
of
the
HHT
based
on
damage
descriptors
of
AE
signals
is
discussed
(Hamdi,
2013).
First,
the
HHT
is
used
to
identify
the
damage
signature
by
correlating
the
measured
AE
signals
with
known
acoustic
sources.
Then,
the
performance
of
the
HHT
for
damage
propagation
monitoring
is
evaluated.
39
characterization
of
the
mechanical
performance
of
wood
in
a
controlled
environment:
use
of
acoustic
emission
to
monitor
crack
tip
propagation
Malick
Diakhaté1,
Seif
Eddine
Hamdi2,3,
Emilio
Bastidas-‐Arteaga4,
Rostand
Moutou-‐Pitti2,3
1
Université
de
Bretagne
Occidentale,
LBMS,
43
Quai
de
Léon,
29600
Morlaix,
France,
malick.diakhate@univ-‐brest.fr
2
Clermont
Université,
Université
Blaise
Pascal,
Institut
Pascal,
BP
10448
Clermont-‐
Ferrand,
France,
seif_eddine.hamdi@univ-‐bpclermont.fr
3
CNRS,
UMR
6602,
Institut
Pascal,
63171,
Aubière,
France,
rostand.moutou_pitti@univ-‐
bpclermont.fr
4
LUNAM
Université,
GeM,
CNRS
UMR
6183/FR
3473,
44322
Nantes,
France,
emilio.bastidas@univ-‐nantes.fr
Keywords:
Acoustic
emission,
clustering,
data
mining,
probability
of
detection,
wood
material
Monitoring
material
damage
by
means
of
acoustic
emission
lies
in
the
ability
to
identify
the
most
relevant
descriptors
of
cracking
mechanisms.
The
sudden
release
of
stored
energy
during
the
damage
process,
known
as
the
acoustic
emission
(AE),
is
a
very
suitable
technique
for
in
situ
health
monitoring
applications.
In
the
present
study,
a
tensile
test
is
performed
on
a
DCB
(Double
cantilever
Beam)
wood
specimen
(Figure
1)
to
generate
mode
I
cracking.
The
AE
events
recorded
during
the
tensile
test
are
correlated
to
actual
damage
in
terms
of
cracking
length.
Various
signal
processing
and
pattern
recognition
techniques
have
been
performed
for
damage
feature
extraction
from
AE
signals.
In
this
study,
the
Hilbert–Huang
transform
(HHT,
Huang
2005)
is
used,
for
relating
a
specific
damage
mechanism
to
its
acoustic
signature.
The
applicability
of
the
HHT
based
on
damage
descriptors
of
AE
signals
is
discussed
(Hamdi,
2013).
First,
the
HHT
is
used
to
identify
the
damage
signature
by
correlating
the
measured
AE
signals
with
known
acoustic
sources.
Then,
the
performance
of
the
HHT
for
damage
propagation
monitoring
is
evaluated.
39
