Wood Science and Technol ogy 30 (z996) 253-263 9 Springer-Verlag 1996

Determination of Young's modulus for spruce, fir and

isotropic m a te ria ls b y t h e resonance f l e x u r e m e t h o d w i t h
comparisons t o s ta tic f l e x u r e an d o t h e r d y n a m i c m e th o d s
D. W . Haines, J. M . L e b a n , C. H e r b e
Summa r y Dynami c met hods provi de r api d and accurat e means t o det ermi ne Young' s
modul us, i.e. t he modul us of elasticity, of wood. For dry, clear speci mens of ~picba
c ommun (Norway spruce, picea excelsa) and sapin pictin~ (silver fir, abies amabilis) we
present a compar i son of results f r om tests by a r esonance flexure met hod with results
obt ai ned f r om f our - poi nt static flexure tests. For a wide range of speci men size the
resonance flexure met hod provi des a simpler, mor e rapi dl y per f or med alternative to t he
classical static flexure met hod, giving Young' s modul us values whi ch are for t he spruce
and fir speci mens of this st udy, nearl y identical t o t hose calculated f r om t he static
flexure tests. Results are also pr esent ed whi ch show t hat a r esonance l ongi t udi nal
met hod yields hi gher values of Young' s modul us and an ul t rasoni c met hod yields still
hi gher values. We provi de also a compar i son of t he f our test met hods applied t o
i sot ropi c materials.
I n t r o d u c t i o n
Dynami c test met hods have devel oped i n parallel wi t h static test met hods for ma ny
types of solid materials i ncl udi ng wood. Static met hods are generally mor e familiar and
mor e widely used. However, dynami c tests can oft en be per f or med mor e rapidly,
part i cul arl y with t he availability of moder n electronic test equi pment . The availability of
digital st orage oscilloscopes with fast Fouri er t r ansf or m capability great l y simplifies
dynami c test analysis, part i cul arl y for resonance met hods.
Haines (1979, 1980) used resonance flexure in st eady state excitation for the
measur ement of Young' s modul us for t he charact eri zat i on of wood for musi cal
i nst rument s. Ot her appl i cat i ons of resonance flexure were r epor t ed for t he st udy of
253
Received 28 ]uly 1995
Daniel W. Haines
Mechanical Engineering Department
Manhattan College, Riverdale, NY 10471, USA
Jean-Michel Leban
Equipe de Recherches sur la Qualit~ des Bois
Institut National de la Recherche Agronomique
54280 Champenoux, France
Christian Herb~
Equipe de Recherches sur la Qualit~ des Bois
Institut National de la Recherche Agronomique
54280 Champenoux, France
The authors wish to thank Pierre Michel and Andrb Perrin for preparing
the test specimens and components of the test apparatus for this study
254
plywood and laminated lumber (Sobue 1983; Sobue and Iwasaki 1981a, b) and for the
study of the mechanical properties of epoxy-poplar composite materials (Moore et. al.
1983).
In this paper we present results obtained from vibrational resonance created by
impact of the wood specimen in states of free support conditions. The free support
conditions provide the most accurate support conditions achievable. Modern instru-
mentation enables this method to be applied with ease. Impact induced free-free
resonance flexure vibrations were employed for the determination of Young's modulus
of small clear specimens by Sobue (1986b) and Bordonne (1989), on samples of
commercial size logs (Arima et al. 1991), structural lumber (Sobue 1988) and clear and
knotty specimens by Chui (1991).
Several authors have reported on comparisons of dynamic and static tests of wood.
Comparison of cantilever resonance, static three-point bending and ultrasonic tests for
the determination of Young's modulus for Douglas-fir was reported by Sinclair and
Farshad (1987). Comparison of Young's modulus calculated from free-free resonance
flexure tests and static three and four-point bending tests were performed for 16 types
of wood by Curie (1989).
Hearmon (1948) reported on the determination of Young's modulus of beech and
pine by means of flexural and longitudinal tests. He found values calculated from
longitudinal tests to be 5% to 8% greater than those calculated from flexural tests.
Results of Brenndorfer (1972) and Radu and Brenndorfer (1976) also noted differences
between the Young's modulus deduced from longitudinal and flexural modes of
vibration.
Longitudinal vibrations in wood, produced by induced stress waves on which the
velocity was the measured parameter have been reported by Dunlop (1978 and 1980)
and Gerhards (1982). By tapping the specimen with a hammer and by receiving the
signal with a wireless microphone, Sobue (1986a) determined the Young's modulus of
various sizes of wood specimens. The resonances were determined from a frequency
analysis of the signal, and a microcomputer permitted treatment of the data.
Ultrasonic techniques have been applied to wood by Hearmon (1965) and Bucur
(1985) to determine the elastic constants of wood.
In this paper we present a discussion of the three dynamic tests and a comparison of
these results with tests of the same specimens by the four-point static flexure method.
The specimens tested are nominally 360 mm long with square cross section measuring
20 by 20 mm. This size is the French standard size for the four-point static flexure test; it
is also taken as the standard size for the dynamic tests. Throughout this report we note
nominal dimensions, however, all measurements were taken with accuracy of at least
three significant figures. All references to Young's modulus pertain to the measure of
stiffness in the fiber direction. The methods discuss in this paper may be applied to
determine the stiffness in directions other than the fiber direction, however this report
is restricted to discussion of the Young's modulus in the fiber direction.
It is shown that for the dry, standard specimens of both +pic~a commun (Norway
spruce, picea excelsa) and sapin pictin~ (silver fir, abies arnabilis) the mean value of
Young's modulus as calculated from resonance flexure tests results nearly matches that
calculated from the static flexure results. The mean value from the resonance
longitudinal results is above the resonance flexure and static results, and the mean value
of the ultrasonic longitudinal results is above those of all other results. All tests of the
same set of specimens were performed within a seven hour period in order to minimize
ambient effects. The moisture content at the time of the tests was approximately 12%.
In addition, full size specimens of construction lumber were tested by resonance
flexure and static flexure. The two sets of tests were performed several years apart.
These speci mens were of rect angul ar cross section and measur ed nomi nal l y three
met ers long. The results for Young' s modul us for these large speci mens show the
resonance flexure values t o be ver y close t o t hose of the static flexure met hod. The small
difference bet ween the results of the two types of tests is believed t o be due to di fferent
envi r onment al condi t i ons in existence dur i ng t he two tests.
Finally, we present a di scussi on and present at i on of t he results of resonance flexure
tests of small speci mens less t han 1 mm thick. These speci mens represent ed earl y and
late wood of spruce and had been tested in static t ensi on several years pr i or t o the
resonance flexure tests. Very close correl at i on is f ound bet ween t he results of t he two
types of tests.
Dynamic methods
The two resonance met hods empl oy excitation of t he speci men by a sharp bl ow with
a ha mme r at a known ant i node of vibration. The weight of the ha mme r must be great
enough t o deliver enough ener gy to excite t he vi brat i ons and light enough not t o deliver
so much mome nt um t o car r y the speci men with it as a rigid body. For ma ny situations
these requi rement s lead t o a hammer with a weight i n the same range as the specimen.
A mi cr ophone is pl aced in close pr oxi mi t y t o a second known ant i node t o receive the
radi at ed sound f r om the speci men due to t he bl ow of t he hammer . The electrical signal
generat ed in the mi cr ophone is sent to a digital storage oscilloscope with fast Fouri er
t r ansf or m processi ng capability. For our pur poses the LeCroy 9314M oscilloscope
served well. I n a pr oper l y posed experiment, the frequency of the st rongest r esonance
di spl ayed on t he screen of the oscilloscope is the resonance f r equency f r om whi ch
Young' s modul us may be calculated.
The t hr ough- t r ansmi ssi on ul t rasoni c test requires a pai r of piezoelectric t ransducers,
a signal gener at or to pr oduce t he ul t rasoni c pulse signal and a device to measur e t he
t i me of travel of t he ul t rasoni c pulse t hr ough the specimen. Often t he signal generat or
and t he t i me measur ement i nst r ument are cont ai ned in one unit. For t he tests r epor t ed
herei n the i nst r ument was the Sattec 1Mhz unit.
The f our - poi nt static flexure tests were per f or med in accordance with the French
st andar d NF B 51-016 "Det er mi nat i on of modul us of elasticity in static bendi ng usi ng
small clear speci mens. "
Resonance flexure vibration mode
The resonance flexure vi brat i on test utilizes t he f undament al mode of vi brat i on of
a free-free beam, i.e. a beam wi t h no const rai nt s. Figure 1 shows the test set-up. Light
suppor t may be pr ovi ded at t wo poi nt s l ocat ed 22% of t he length f r om each end
(Ti moshenko, 1955). This di st ance is not critical, pr ovi ded t he suppor t does not
unnecessari l y const rai n t he speci men. The poi nt s of suppor t are t he nodes of t he
f undament al mode of vibration. Wi t h t he st andar d 20 20 x 360 mm speci mens this
di mensi on is 79 ram. Support by light t hreads hel d in t ensi on by elastic bands serves
well for this purpose. The mi cr ophone is pl aced above and dos e to t he speci men near
one end. Wi t h appr opr i at e settings of the oscilloscope it is easy t o capt ure t he f r equency
of the f undament al mode of flexural vi brat i on when t he cent er of the speci men is st ruck
f r om above and perpendi cul ar to its length. The entire process requi res no mor e t han
a few seconds. Wi t h this f r equency (700 Hz- 900 Hz for the st andar d speci mens of
spruce and fir) Young' s modul us is calculated f r om t he equat i on
E = 0.946 p f 2L 4/h2 (1)
where E is Young' s modul us, p is t he densi t y of the wood speci men (mass per uni t
vol ume), f is t he measur ed frequency, L is the l engt h of the speci men (nomi nal l y
360 ram) and h is t he height, i.e. vertical thickness of t he speci men (nomi nal l y 20 mm) .
255
256
H a m m e > . ,,
I
Digital oscilloscope
with FFT
(Le Croy 9314M)
Microphone
(B+K4134)
I
022 1
Amplifier if
needed
(B+K 2610)
Fig. 1. Resonance flexure test
For all wood speci mens in this study, the density, p, represents the mean densi t y of the
entire speci men. The derivations of Eq. 1 and Eqs. 2 and 3 of following sections may be
f ound in Ti moshenko (1955) and Morse (1948).
Resonance l ongi t udi nal vi br at i on mode
The l ongi t udi nal mode of vi brat i on is pr oduced by striking one end of the speci men
di rect l y and parallel to the l ong di mensi on of the speci men while the speci men is hel d in
the hand close t o its center. Figure 2 shows t he test set-up. When pr oper l y executed this
act i on establishes the f undament al l ongi t udi nal vi brat i on of the speci men acting as
a bar. The t echni que t o excite this resonance requires practice, for example, the
speci men must be st ruck centrally at the end and directly in line with the speci men
length. Flexural vi brat i ons are also easily excited by this met hod, therefore, it is
i mpor t ant t o i dent i fy t he correct resonance. For t he st andar d speci mens of spruce and
fir, t he f undament al resonance l ongi t udi nal occurs between 5000 Hz and 9000 Hz. The
ant i nodes, i.e. the poi nt s of maxi mum mot i on are the two ends of the bar. The end
opposi t e t o t he st ruck end is available for pl acement of the mi cr ophone near by to
receive t he s ound radi at i ng f r om the free end. The signal is passed to the oscilloscope as
with t he flexure test and the resonance frequency is recorded. As for the resonance
flexure tests t he requi red t i me to per f or m the test is onl y a few seconds. Young' s
modul us is calculated f r om the equat i on
E = 4 p f 2 L 2, (2)
where f is the frequency of l ongi t udi nal vi brat i on and again E is t he Young' s modul us,
p is t he densi t y and L is the total length of the bar (nomi nal l y 360 mm).
Ul t r asoni c l ongi t udi nal t hr ough- t r ansmi ssi on mode
The ul t rasoni c met hod requires the pl acement of two piezoelectric t ransducers in
cont act wi t h t he ends, one serves t o emi t a l ongi t udi nal stress pulse of very short
dur at i on whi ch travels with the speed of sound in the wood to the t r ansducer at the
ot her end. Figure 3 shows t he set-up of this experiment. The time of pulse travel for t he
st andar d speci men of spruce, and fir falls in the range between 60 and 85 mi croseconds.
Wi t h this t i me of travel and the known distance separating the transducers, i.e. t he
l engt h of t he speci men (nomi nal l y 360 mm) , the vel oci t y of the stress pulse is calculated.
Young' s modul us is t hen calculated f r om the equat i on
E = pv ~, (3)
where v is t he calculated velocity and again 19 is t he densi t y of the wood specimen.
St at i c f l e x u r e t e s t
The static tests were per f or med in accordance with the French st andar d NF B 51-061
whi ch is the classical f our - poi nt bendi ng test. Deflection is measur ed in the central
Y
/ , ~ - 1 A
E )igital o s c i l l o s c - ~ e ]
w i t h F F T /
( L e C r o y 9 3 1 4 M ) J
A m p l i f i e r i f
n e e d e d
( B + K 2 6 1 0 )
Fig. 2. Resonance longitudinal test
E m itto r R e c e iv e r
IL,
U ltra s o n ic ~ - ~
GeneratodAnalyser
(Sattec/1 MHz)
Fig. 3. Ultrasonic longitudinal test
257
region, a zone of pure bending in the absence of shear deformation. Figure 4a shows the
support points separated by 320 ram, the load application points separated by 160 mm
and the contact points of the deflection measurement gauge separated by 120 mm.
Figure 4b shows the specimen deformed due to the application of load. Young' s
modulus is given by the equation
E = 3P (Ls-- a) m2/(8bh3u),
( 4 )
where P is the total force applied to the specimen, Ls is the support ed length of the
specimen (320 mm), a is the distance separating the points of load application
(160 mm), m is the distance separating the deflection measurement gauge contact
points (120 mm), b is the width of the specimen (nominally 20 mm), h is the height of
the specimen (nominally 20 mm) and u is the deflection measured at the mi dpoi nt
between the gauge contact points.
Test results of t he standard wood specimens ( 2 0 x 2 0 x 360 mm)
Figures 5 and 6 show the results of the static flexure tests performed on the 25
specimens of spruce and 80 specimens of fir, respectively. The results show the
anticipated rise of Young' s modulus with increase in density.
Figure 7 shows the plot of the data points for the spruce specimens, indicating the
Young' s modulus as found by static flexure and the percentage differences given by the
three dynamic tests.
Figure 8 is a similar plot for the data for the fir specimens. It is evident in Fig. 8 that
the spread widens considerably between the results of the four types of tests for Young' s
modulus below 9000 MPa. The cause of this interesting observation is under study.
Table 1 shows the compari son of the mean values of Young' s modul us and density for
spruce and fir as obtained by each of the four methods. In this study the results for the
resonance flexure met hod nearly mat ch the results for the static flexure method; the
mean Young' s modulus by resonance flexure for spruce and fir fall above those for static
flexure by 0.4% and 2.3%, respectively. The mean value of Young' s modul us from the
longitudinal resonance met hod is greater than the mean value obtained from the flexure
methods by 6% and 11% for spruce and fir, respectively. These results are similar to
those obtained by Hear mon (1948). The mean value of Young' s modulus from the
longitudinal ultrasound met hod exceed the mean value obtained from the flexure
methods by 17% and 22% for spruce and fir, respectively. The results report ed herein
are also consistent with the findings of Sinclair and Farshad (1987).
z58
a :,6 0 .I
m=120
L~=320
L=360
O
b=20~ ~ R=5 =;
Fig. 4. a Static flexure test, b Static flexure deflection measurement
20,000
15,000-
~IO,O00-
E
5,000
O
O
O
O
O
O
O
O
0
300 350 400 450 500 550
Density (kg/cu m)
Fig. 5. Young's modulus v. density for 25 specimens of spruce
600 650 700 750
20,000
15,000
v
~10, 000
E
5,ooo-
9 9 9 0 o e
o ~ 1 7 6 9 9 9 9
e ~ 9 9 e e e
~ ? ~
~ e ~ oo
o oS
0
3 0 0 3 ; 0 5 0 0
Density (kg/cu m)
Fig. 6. Young's modulus v. density for 80 specimens of fir
I
550 600 650
One ma y know t hat for hi gh r at i os of hei ght t o l engt h, shear def or mat i on is
si gni f i cant and r educes t he r esonance f r equency of beams vi br at i ng i n flexure. However,
t he di mens i ons of t he s t a nda r d speci men are such t hat t hi s effect is negl i gi bl e for t he
f undament al mode of vi br at i on, and Eq. 1 is val i d. Thi s fact was veri fi ed by exper i ment
in t hi s st udy.
5O
o 40
~ 30
~ 20
~ 1 0
~ 0
-10
-20
5000
A
A
9 A
0 [ ' I I Ol n l
o
70~00 9600 11000
[ ] Resonance f l exur e
9 Resonance l ongi t udi nal
z~ Ul t r asoni c l ongi t udi nal , ,
t ,
t .
t ,
A A 9
| . = - . ~ o o []
o o , ~
~ O 0 1 o I I I D
O
0
I I I
13000 15000 17000 19000
Young' s modul us by st at i c f l exur e (MPa)
Fig. 7. Comparison of Young's modulus found from dynamic methods with those obtained from
the static flexure method for the 25 specimens of spruce
259
100
o~ 8O
r
o~
~ 60
- O
g 40
c~
e-
9 20
a. 0
t ,
, , " [ ] Resonance f l exur e
, 9 Resonance l ongi t udi nal
! ,x Ul t r asoni c l ongi t udi nal
-20
6000 18000
9 o AAA a r e a l ~ A ~ - -
D 9 ~ "m 9149
80no0 i I ~ i i
10000 12000 14000 16000
Young' s modul us by st at i c f l exur e (MPa)
Fig. 8. Comparison ofYoung' s modulus found from dynamic methods with those obtained from
the static flexure method for the 80 specimens of fir
T e s t s o f 2 0 x 2 0 x 3 6 0 m m h o m o g e n e o u s , i s o t r o p i c s p e c i m e n s
Equat i ons 1, 2, 3 and 4 were der i ved for i deal el ast i c mat er i al s whi ch ar e al so
homogeneous , i.e., t hei r mat er i al pr oper t i es do not ar e cons t ant f r om poi nt t o poi nt ,
and i sot r opi c, i.e., t hei r pr oper t i es var y wi t h di r ect i on a r ound a gi ven poi nt . No mat er i al
ful l y achi eves t hi s i deal ; c ompa r e d t o mos t met al s and pl ast i cs, wood is hi ghl y non-
homogeneous a nd ani s ot r opi c because of i t s gr ai n st r uct ur e.
Whenever t est r esul t s are c ompa r e d of mat er i al as compl ex as wood it is i mper at i ve
t o ver i f y t he accur acy of t he t est met hods . One way to achi eve t hi s ver i f i cat i on is t o use
t he s ame met hod t o t est mor e i deal mat er i al s and compar e t he resul t s. A cl ose
c ompa r i s on i ndi cat es t hat t he t est met hods ar e accur at e. To so ver i f y t he t est met hods ,
each of t he f our met hods were appl i ed t o t wo speci mens each of dur al umi n, a mat er i al
of gr eat er st i ffness t han wood and two t ypes of nyl on, a mat er i al of l ess st i ffness t han
wood. Al l speci mens were nomi na l l y 20 20 360 mm, as for t he pr i nci pal seri es of
wood in t hi s st udy. These mat er i al s ar e near l y homogeneous and i sot r opi c. The r esul t s
gi ven i n Tabl e 2 show t hat wi t h t he except i on of t he hi gh ul t r asoni c l ongi t udi nal r esul t s
for nyl on, t he met hods gi ve r esul t s whi ch agree qui t e closely. These r esul t s pr ovi de
f ur t her conf i dence i n t he t est met hods , par t i cul ar l y t he r esonance fl exure met hod.
Table 1. Mean values of Young's modulus and density for wood
Young's modulus
Density Static Flexural Longitudinal
kg/cu m flexure resonance resonance
MPa MPa MPa
Longitudinal
ultrasound
MPa
Spruce 25 specimens 416 10600 10600 11300 12400
Fir. 80 specimens 465 12500 12700 14000 15400
260
Table 2. Young's modulus and density for isotropic materials. Means of tests on two
specimens each
Young's modulus
Density Static Flexural Longitudinal Longitudinal
kg/cu m flexure resonance resonance ultrasound
MPa MPa MPa MPa
Duralumin 2800 73600 71600 73400 73300
Nylon A 1410 3570 3600 3700 4450
Nylon B 1140 3300 3400 3500 4280
Tests of wood specimens of other sizes
The ease of application and accuracy of the resonance flexure method led to
experimentation with specimens of larger and smaller size than the standard size.
Large specimens of construction lumber
A total of 20 specimens of construction l umber spruce were available for dynamic tests
for which Young' s modul us had been determined previously by static flexure. The
weights of each specimen were nearly identical to those determined when the static tests
were conducted; the mean densities of the two sample sets were nearly identical
(412 kg/cu m for the static flexure tests and 411 kg/cu m for the resonance flexure tests).
Experience has shown that Young' s modulus remains constant if the mass (closely
related to moisture content) is constant. Therefore, tests by the resonance flexure
method could be readily compared with the earlier static flexure tests. The l umber
di mensi ons were a standard European construction size, nominally three meters long
with 50 by 150 mm rectangular cross section. Tested in flexure about the strong axis the
resonance flexure frequencies fell between 65 and 80 Hz.
Figure 9 shows the results of these tests. The Young' s modulus values for the
resonance flexure tests fall slightly below the values calculated from the static tests; the
mean values are 10300 MPa by the resonance flexure method and 10900 MPa by the
earlier static flexure method, a difference of less than 6%.
Small specimens
Following the application of the resonance flexure method to a number of specimens of
varying size, we pondered the question, "How small may the sample be to apply this
method. ~'' The smallest specimens available were those of a study by Chantre (1989) of
early and late wood within the tree ring of spruce. Chantre prepared specimens for
0
0
o -5
0
~: -10
~- 15
~- 20
~
-25
~. -30
-35
-40
4000
I 9 E 9 I 9 ?
Q
I I I
60LO0 80100 10000 12000 14000 16000
Young's modulus by static flexure (MPa)
Fig. 9. Comparison of Young's modulus found from the resonance flexure method with those
obtained from the static flexure method for 20 specimens of spruce construction lumber
261
40-
35-
30-
t - -
25-
20-
10
g
.o 5
d ~
a. 0
-5
-100
I
I
5000
I
I e I 9 [
10000 15000 20000 25000
Young's modulus by static tension (MPa)
Fig. 10. Comparison of Young's modulus found from the resonance flexure method with those
obtained from the static tension method for 8 spruce specimens less than 1 mm in thickness
tensile tests which were less t han 1 mm thick and nomi nal l y 5 mm wide and 30 mm
long. To test these specimens in flexural vibration it was necessary to hold the specimen
between a two pairs of tightly stretched threads spaced such that each was positioned
at or near a node (22% of the length from each end). We found it impossible to
strike a specimen of this size effectively, however it was found that the desired
fundament al mode of vibration could be excited by deflecting the center of the
specimen downward with an i nst rument having a pointed tip such as a pencil until
resistance is felt, then releasing the specimen as one would pluck a string. The frequency
range of the resonance flexure of these small specimens fell between 2500 Hz and
3500 Hz.
Figure 10 shows the results of these tests in comparison with the earlier static tensile
tests. Because of the time span of several years between the two sets of tests, the ambi ent
conditions were different, and therefore the density of each specimen had changed.
Therefore these results only yield a qualitative result. Nonetheless, the mean values are
13300 MPa for resonance flexure and 13600 MPa for static tension, a difference of less
t han 3%.
262
Conclusions
The resonance flexure met hod provi des a useful t ool to det ermi ne Young' s modul us of
wood. Wi t h the availability of moder n i nst rument at i on, this dynami c test can be
per f or med rapi dl y and Young' s modul us can be calculated from a simple formul a. The
st andar d speci men size adopt ed for static flexure tests is ideal also for the resonance
flexure test. For clear coni ferous wood specimens, resonance flexure test results provi de
a close mat ch t o the Young' s modul us expected f r om the static flexure results, and it is
mor e easily per f or med t han the static flexure met hod.
Compar i son of resonance flexure results with those f r om static tests for sizes ot her
t han the st andar d size show less t han a 6% difference for the large const ruct i on l umber
and less t han a 3 % difference for the small speci mens of less t han 1 mm thickness.
The differences bet ween the Young' s modul i obt ai ned f r om the flexure met hods and
t hose f r om t he l ongi t udi nal dynami c met hods require discussion. First, not e t hat the
l ongi t udi nal met hods cont ai n hi gh frequencies; the frequencies of t he ultrasonic
met hod are t he higher. Both l ongi t udi nal dynami c met hods yield high values for
Young' s modul us; t he ultrasonic values are the higher. Equat i ons 1, 2 and 3 were
deri ved on t he basis of the material being ideal on the macroscopi c level, i.e. isotropic,
homogeneous and elastic. The definition of Young' s modul us likewise assumes these
ideal propert i es. No material fully meets the ideal, but many are nearl y so, e.g., most
metals. Wood in this sense is less ideal t han metals. On the macroscopi c level, wood is
clearly nonhomogeneous and anisotropic, due pri mari l y to t he pat t ern of early and late
wood and the difference in their physical properties. The known capability for the
dampi ng of vi brat i ons (Haines 1979) establishes wood as mor e viscoelastic t han metals,
i.e. less elastic.
Viscoelasticity is the most likely source of the differences noted. Both l ongi t udi nal
met hods empl oy a pulse; t hat of the ul t rasoni c met hod is short er in time. Shorter pulses
are related t o waves of hi gher frequency content. Ferry (1961) report ed t hat bul k
l ongi t udi nal waves travel with greater velocity at higher frequencies in rubber and
relates this phenomenon t o the at t enuat i on (viscoelastic propert y) of rubber. Kolsky
(1963) pr ovi ded analysis of a compl ex viscoelastic model which predicts hi gher
velocities of l ongi t udi nal waves at hi gher frequencies. Since Young' s modul us in the
l ongi t udi nal modes is pr opor t i onal to the square of the longitudinal velocity, these
results of Ferry and Kolsky predi ct the hi gher value of calculated Young' s modul us for
wood as we obt ai ned f r om l ongi t udi nal resonance and the even hi gher values as we
obt ai ned f r om l ongi t udi nal ul t rasound.
Nonhomogenei t y may pl ay a role also. The hi gher densi t y and stiffer late wood passes
l ongi t udi nal stress waves with greater velocity. The longitudinal wave may follow the
pat h t hr ough t he late wood. Furt her st udy is requi red t o quant i fy the differences noted.
The mai n poi nt of this paper is t hat the flexural resonance met hod t o det ermi ne
Young' s modul us for wood is accurat e and easily performed.
References
Arima, T.; Nakamura, N.; Maruyawa, N.; Hayamura, S. 1991: Classification of logs based on sound
analysis and its application in product processing. Proc. Int. Timber Eng. Conf., London 2:
266-2,2.273
Bordonne, P. A. 1989: Module dynamique et frottement int~rieur dans le bois mesures sur poutres
flottantes en vibrations naturelles. Doctoral thesis, l'Institut National Polytechnique de Lorraine,
Nancy
Brenndorfer, D. 1972: Nondestructive testing of wood with acoustic vibrations. Industria Lemn. 6:
234-240
Bucur, V. 1985: Ultrasonic velocity, stiffness matrix and elastic constants of wood. CAS Journal 44:
23 28
Chantre, G. 1989: Liaison entre rigidit6 et densitb du bois a l' intbrieur du cerne. DEA Etude
(Masters Thesis) INRA, Nancy
Chui, Y. H. 1991: Simultaneous evaluation of bending and shear modul i of wood and the influence
of knots on these parameters. Wood Sci. Technol 25: 125-134
Curie, P. 1989: Essais compar at i f pour la mesure du modul e d'blasticitb. Provisional report of the
Centre Technique Forestier Tropical, Nogent-sur Marne
Dunlop, J. I. 1980: Testing of particleboard by acoustic techniques. Wood Sci. Technol 14: 69-78
Dunlop, ]. I. 1978: Dampi ng loss in wood at midkilohertz frequencies. Wood Sci. Technol. 12:
49-62
Ferry, ]. D., Viscoelastic Properties of Polymers, John Wiley & Sons, New York, 1961, p. 420
Gerhards, C. 1982: Longitudinal stress waves for l umber stress grading: factors affecting
applications. States of art. Forest Prod. J. 32(2): 20-35
Haines, D. W. 1979: On musical i nst rument wood, CAS Newsletter 31: 23- 32
Haines, D. W. 1980: On musical i nst rument wood - Part II, surface finishes, plywood, light and
water exposure. CAS Newsletter 3: 19 23
Hearmon, R. F. S. 1948: Elasticity of wood and plywood Forest Prod. Res. Spec. Rep. No. 7. His
Majesty' s Stationery Office, London, 87 pp
Hearmon, R. F. S. 1965: The assessment of wood properties by vibrations and high frequency
acoustic waves. Proc. 2nd Symp. on the NDT of wood, Washington State University, 49-65
Kolsky, H; Stress Waves in Solids, Dover, New York, 1963, p. 121
Moore, G. R.; Kline, D. E.; Blankerhorn, P. R. 1983: Dynamic mechanical properties of epoxy
poplar composite materials. Wood Fiber Sci. 15(4): 358-375
Morse, P.; Vibration and Sound, 2nd edition, McGraw-Hill, New York, 1948
Radu, A.; Brenndorfer, D. 1976: Zur zerstorungsfreien prufung des solzed dut ch schwingungsver-
suche. Holz als Rohund-Werkst off 34(6): 219-222
Sinclair, A. N.; Farshad, M. 1987: A compari son of three met hods for det ermi ni ng elastic constants
of wood. Journal of Testing and Evaluation JTEVA 15(2): 77 86
Sobue, N.; Iwasaki, Y. 1981a: Effect of veneer const ruct i on on anisotropy of dynamic visco-
elasticity of plywood. Mokuzai Gakkaishi 27(6): 457-462
Sobue, N.; Iwasaki, Y. 1981b: Effect of adhesive on dynamic viscoelasticity of l ami nat ed l umber of
two-ply sawn laminae. Mokuzai Gakkaishi 27(7): 597 601
Sobue, N. 1983: Prediction of dynamic Young' s modul us and loss tangent of plywood in bending.
Mokuzai Gakkaishi 29(1): 14 19
Sobue, N. 1986a: Measurement of Young' s modul us by the transient longitudinal vi brat i on of
wooden beams using a FFT spectrum analyser. Mokuzai Gakkaishi 32(9): 744 747
Sobue, N. 1986b: Instantaneous measurement of elastic constants by analysis of the tap tone of
wood. Application to flexurat vibration of beams. Mokuzai Gakkaishi 32(4): 274 279
Sobue, N. 1988: Simultaneous det ermi nat i on of Young' s modul us and shear modul us of structural
l umber by compl ex vibrations of bending and twisting. Mokuzai Gakkaishi 34(8): 652-657
Ti moshenko, S., Vibration Problems in Engineering, 3rd edition, D. Van Nostrand, New York,
1955
263