Journal of the Korean Physical Society, Vol. 49, No. 5, November 2006, pp.

1961∼1966

Measurement of the Resonance Frequency, the Loss Factor, and

the Dynamic Young’s Modulus in Structural Steel and Polycarbonate
by Using an Acoustic Velocity Sensor

Sung Soo Jung,∗ Yong Tae Kim and Yong Bong Lee
Mechanical Metrology Group, Korea Research Institute of Standards and Science, Daejeon 305-340

Su Hyun Shin and Duckwhan Kim

Department of Mechanical Engineering, Pusan National University, Busan 609-735

Ho Chul Kim
Department of Physics, Korea Advanced Institute of Science and Technology, Daejeon 305-701

(Received 1 August 2006)

An acoustic velocity sensor, micro-flown, was used to monitor the vibration signal in measure-
ments of the resonance frequency, the loss factor, and the dynamic Young’s modulus of structural
steel and polycarbonate by using a resonance method. The results showed that the measured loss
factors of the structural steel and polycarbonate used in the present specimen were 0.0023 and
0.056, respectively, and the Young’s moduli were 198.4 GPa and 2.4 GPa. The values measured
with micro-flown agreed with those obtained with a conventional displacement sensor and with the
theoretically estimated and reference values. The agreements ensure the reliability of the use of a
micro-flown acoustic velocity sensor in measuring the resonance frequency, the loss factor, and the
dynamic Young’s modulus.

PACS numbers: 43.40.Cw, 43.40.Yq, 43.58.Fm

Keywords: Loss factor, Young’s modulus, Micro-flown, Structural steel, Polycarbonate

I. INTRODUCTION instantaneous time-independent elastic strain, as well as

a time-dependent strain of relaxation or anelasticity that
lags behind the applied load, exhibiting viscoelastic or
The elastic modulus of materials is the measure of the relaxation. To account for the non-linearity encountered
resistance to elastic deformation and one of the funda- in a real material, one introduces a complex modulus,
mental mechanical properties in engineering design for comprised of an elastic (storage) modulus and an imagi-
prediction of deflection under a static load and of natu- nary (loss) modulus. The ratio of the loss to the storage
ral frequencies of vibration under dynamic load. In elas- modulus is defined as the loss factor.
tic materials, the force and displacement or the stress ASTM E 756 [1] recommends non-contact excitation
and strain are linearly related by a proportionality con- and detection of the vibration signal, and various meth-
stants, the elastic modulus, and obey Hooke’s law. In ods have been proposed, such as audio frequency excita-
the process, the applied stress and strain are assumed to tion [2] and laser displacement sensing [3]. The conven-
be linear and perfectly in phase. The condition is valid in tional method for obtaining the loss factor and the dy-
the low-strain range for low loading rates, and the static namic Young’s modulus is to use an electromagnetic in-
deformation process is considered to be instantaneous. ductive excitation with a capacitive displacement sensor
However, the mechanical response in real materials, for electromagnetic conductive specimen to comply with
such as structural materials, polymers and compos- the recommendation in view of non-invasiveness. How-
ites, varies markedly with temperature and loading rate ever, the method cannot be adopted to non-conductive
(strain rate or frequency), and Hooke’s law does not ac- polymer and polymer composites. In that case, a small
count for the nonlinear behavior. The loading and un- light-weight metal disk, length and mass not to exceed
loading are not identical, with an accompanying hystere- 2 % and 1 % of the testing specimen, is attached to the
sis loss. Thus, the response of a real material involves an bottom part of the specimen to excite the vibration sig-
nal.
∗ E-mail: jss@kriss.re.kr We report the first use of a micro-flown non-contact
-1961-
-1962- Journal of the Korean Physical Society, Vol. 49, No. 5, November 2006

acoustic velocity sensor [4] to detect the vibration signal

in the measurement of the loss factor and the dynamic
Young’s modulus of conductive and non-conductive spec-
imens. The micro-flown was originally devised to mea-
sure the acoustic flow based on the temperature differ-
ence caused by the acoustically induced air movement
between two closely spaced heated resistive sensors by
which the particle velocity was measured. The results
were compared with those theoretically calculated and
with those obtained conventionally by using an electro-
magnetic excitation with a capacitive displacement sen-
sor for validation.

II. BASIC THEORY TO DETERMINE THE

LOSS FACTOR AND THE DYNAMIC
Fig. 1. Frequency response function curve.
YOUNG’S MODULUS

A one-dimensional simple harmonic system can be rep- where 4ω = |ω1 − ω2 |. If the damping is very small,
resented by a mass m and a spring constant k. For a η << 1, Eq. (5) can be approximated as
real system with hesteretic damping, the stiffness k is
replaced by the complex stiffness k ∗ (= k(1 + jη)), where 4f 4ω η η
= ' (1 + ) − (1 − ) = η. (6)
η is the loss factor. When a steady-state harmonic force fr ωr 2 2
F ejωt with angular frequency ω is applied to the hes- For a cantilever-beam type of specimen for which one
teretic damping ststem, Newton’s second law yields a end is clamped and the other end is free to vibrate, Hmax
differential equation for the displacement x [5]: between the applied force and the detected vibration sig-
d2 x nal of the specimen can be represented as in Fig. 1. Also,
m + k(1 + jη)x = F ejωt . (1) the loss factor is determined by using the resonance fre-
dt2
quency, and the frequency difference between the –3 dB
The particular solution of the Eq. (1) can be expressed reduction points, so-called, 3 dB method. [5]
as For a cantilever beam with a clamped-free boundary
x = A cos(ωt − ), (2) condition, the Young’s modulus (E) is related to the res-
onance frequencies (frn ) as [1]
√ F
where A = and  is the phase difference
(k−mω 2 )2 +k2 η 2
12ρL4 frn2
between the force and the displacement and is given by E= 2 2
, (7)
 = tan−1 [ (k−mω
k an t
2 ) ]. When ω is low, the response of the

mass is in phase with the excitation,  ' 0, and Eq. (2) where L is the beam length, ρ is the mass density, an
is reduced to is the characteristic number of the nth mode, and t is
x 1 the beam’s thickness. The values of an depend on the
H= =p , (3) boundary conditions, and a1 = 0.5596, a2 = 3.5070, a3
F (k − mω 2 )2 + k 2 η 2
= 9.8194,· · ·, etc.
where H is the frequency response function and is called
the compliance (x/F ) or the impedance (v/F : v is the
velocity). H has a maximum
p value, Hmax , at the reso- III. MEASUREMENT OF THE RESONANCE
nance frequency (ωr = k/m). The frequencies at the FREQUENCY, THE LOSS FACTOR, AND
points ω1 and ω2 at which the H is reduced to 1/N times THE DYNAMIC YOUNG’S MODULUS
Hmax can be expressed as
r
k p The resonance frequency method was applied for mea-
ω1,2 = ( )[1 ± η N 2 − 1]. (4) suring the loss factor and the dynamic Young’s modulus
m
as a function of the frequency at room temperature. The
Consider the points A and B located at the
√ frequencies resonance measurements were made on the first three res-
where the amplitude of the response is 1/ 2 times that onances for the rectangular cross section of a cantilever-
of Hmax as shown in Fig. 1. Then, Eq. (4) can be type specimen by using the flexural vibration of the free
simplified to bottom end, while the top end was clamped, as in the
4ω p p arrangement in Fig. 2. The most significant frequency
= 1 + η − 1 − η, (5) ranges were less than 1000 Hz in the case of vibration
ωr
Measurement of the Resonance Frequency, the Loss Factor,· · · – Sung Soo Jung et al. -1963-

Table 1. Dimensions of the structural steel and the poly-

carbonate specimens.

Free length Width Thickness Density

(mm) (mm) (mm) (kg/m3 )
Structural Steel 144 10 0.97 7846
Polycarbonate 180 8.8 2.85 1195

Fig. 3. A metal disk is attached to the polycarbonate to

excite the specimen with an electromagnetic exciter.

Fig. 4. A micro-flown was used tor detect vibrations of the

specimen.

Fig. 2. Measurement system and block diagram. the vibration-induced displacement in the present work.
The sound velocities in the structural steel and the poly-
carbonate were determined as 5000 m/s and 1800 m/s,
damping, so the vibration signals used in the present respectively, by measuring the pulse delay time between
work were generated with frequencies up to 800 Hz by two accelerometers placed on the specimen.
using a frequency analyzer (B&K 3550) and were ampli-
fied with a power amplifier (B&K 2706). The dimensions
and the density of the structural steel and the polycar-
bonate specimens tested are listed in Table 1. IV. RESULTS AND DISCUSSION
For non-metallic polycarbonate (Sehwa Polytec Co.,
Ltd.), a circular metal disk (8.8-mm diameter, 0.4-mm 1. Theoretically Estimated and Experimentally
thickness, and 0.07-gram mass) was attached at the free Determined Resonance Frequency
end of the specimen to excite a vibration electromagnet-
ically and measure the capacitance change due to dis-
placement during flexural vibration, as shown in Fig. 3. The theoretical fundamental resonance frequency of
A capacitive displacement transducer (B&K MM0004) a bar-type specimen with clamped-free boundary condi-
was used to measure the vibration motion of the spec- tion is given as [6]
imen. Alternatively, as shown in Fig. 4, a non-contact √
micro-flown acoustic velocity sensor was used to monitor fr1 = 1.1942 πtc/(8 12L2 ), (8)
-1964- Journal of the Korean Physical Society, Vol. 49, No. 5, November 2006

Fig. 5. Frequency response function of a structural steel Fig. 6. Frequency response function of polycarbonate for
specimen for a micro-flown and a capacitive transducer. a micro-flown and a capacitive transducer.

Table 2. Theoretically calculated and experimentally de-

termined (the numbers in parenthesis) resonance frequencies.

fr1 fr2 fr3

(Hz) (Hz) (Hz)
Structural Steel 38 (38) 239 (238) 663 (668)
Polycarbonate 26 (26) 160 (166) 449 (478)

where c is the sound velocity in the sample. The higher-

order resonance frequencies are related to the funda-
mental frequency as fr2 = 6.267fr1 , fr3 = 17.55fr1 ,
fr4 = 34.39fr1 ,· · ·, etc. Substituting the respective
sound velocity and the specimen length, the resonance
frequencies of the structural steel and the polycarbonate
were calculated and are listed in Table 2. Fig. 7. Magnified view of the second resonance frequency
The frequency response function of the structural steel region of the frequency response function for the structural
measured with a micro-flown sensor in terms of v/F steel specimen with a micro-flown.
and with a capacitive transducer in terms of x/F are
shown in Fig. 5. The variations of the frequency re-
sponse functions with frequency as obtained by using 449 Hz, are also in agreement with the experimental val-
the micro-flown and the capacitive sensor are virtually ues within maximum relative error of 7 %. The fact that
identical. Resonance frequencies higher than the fourth the measured resonance frequencies are higher than the
are not present because they are out of our experimental theoretical values is attributed to the mass-adding effects
frequency range; i.e., the fourth resonance frequency is caused by shifting the resonance toward lower frequen-
around 1300 Hz. The resonance frequencies measured cies due to the presence of the small metal disk attached
using both sensors, 38, 238, and 668 Hz, agree exactly. to the bottom of the specimen for electromagnetic exci-
The first resonance frequency of the structural steel is tation and displacement monitoring.
estimated as 38 Hz with Eq. (8), and the higher-order
frequencies, fr2 = 239 Hz and fr3 = 663 Hz, listed in Ta-
ble 2 are in good agreement with the experimental values 2. Comparison of the Loss Factor and Dynamic
within a maximum relative error of 1 %. Young’s Modulus Determined by Using the Micro-
flown and Capacitive Sensors
The measured resonance frequencies for the polycar-
bonate by using both sensors, 26, 166, and 478 Hz, are
again in good agreement, as shown in Fig. 6. The the- The loss factor was determined by using the ‘3-dB
oretical resonance frequencies determined by using the method’. For the determination of 4f , the frequency re-
values in Table 2, fr1 = 26 Hz, fr2 = 160 Hz, and fr3 = sponse function curve was magnified and averaged 2000
Measurement of the Resonance Frequency, the Loss Factor,· · · – Sung Soo Jung et al. -1965-

Fig. 8. Loss factor of structural steel vs frequency with a Fig. 10. Dynamic Young’s modulus of structural steel vs
micro-flown and a capacitive transducer. frequency with a micro-flown and a capacitive transducer.

Fig. 9. Loss factor of polycarbonate vs frequency with a Fig. 11. Dynamic Young’s modulus of polycarbonate vs
micro-flown and a capacitive transducer. frequency with a micro-flown and a capacitive transducer.

times to improve the signal-to-noise ratio. The magni- difference may arise from microstructure variation asso-
fied frequency response function curve around the second ciated with molecular weight and residual stress varying
resonance frequency region is shown in Fig. 7 as an ex- with the supply source. The present results assure that
ample, and the frequency resolution was 0.025 Hz. The a micro-flown, acoustic velocity sensor can be used as an
loss factors measured by using the micro-flown and the alternative vibration detection sensor instead of a con-
capacitive sensor agrees well, 0.023 up to 478 Hz for the ventional sensor, such as a capacitive transducer, in view
structural steel and 0.056 up to 668 Hz for the poly- of non-invasiveness.
carbonate, nearly independent of frequency, as shown in
Figs. 8 and 9, respectively.
The measured dynamic Young’s modulus for the struc-
tural steel by using the micro-flown and the capacitive V. CONCLUSION
sensor was 198.4 GPa up to 668 Hz in Fig. 10, agrees
with the reference value of 200 GPa [7] within a relative The resonance frequency, the loss factor, and the dy-
error of 1 %. The dynamic Young’s modulus for poly- namic Young’s modulus of structural steel and polycar-
carbonate determined by using the micro-flown and the bonate were measured by using a resonance method with
capacitive sensor was 4 GPa, however, there is a marked a micro-flown acoustic velocity sensor to detect the vibra-
deviation from the reference value of 2.4 GPa [8]. The tion signal. The values measured with the micro-flown
-1966- Journal of the Korean Physical Society, Vol. 49, No. 5, November 2006

sensor agree well with those obtained with the conven- bration Damping Properties of Materials (1998).
tional capacitive sensor. The measured resonance fre- [2] S. H. Shin, Y. B. Lee, S. S. Jung and D. H. Lee, Key
quencies of the structural steel and the polycarbonate Engineering Materials 297, 1958 (2005).
specimen agree well with the theoretically calculated val- [3] M. D. Rao, J. Sound Vib. 262, 457 (2003).
ues within relative errors of 1 % and 7 %, respectively. [4] H. E. de Bree, P. Leussink, T. Korthorst, H. Janson, T.
Lammerink and M. Elwenspoek, Sensors and Actuators:
The present work suggests that the loss factor and the
Physical 54, 552 (1996).
dynamic Young’s modulus are nearly frequency indepen- [5] A. D. Nashif, D. I. G. Jones and J. P. Henderson, Vibra-
dent in the frequency range up to 800 Hz at room tem- tion Damping (Wiley, New York, 1985), Chap. 4.
perature. The agreements ensure the reliability of using [6] L. E. Kinsler, A. R. Frey, A. B. Coppens and J. V.
the micro-flown sensor for vibration measurements. Sanders, Fundamentals of Acoustics, 4th ed. (Wiley, New
York, 2000), Chap. 3.
[7] D. Halliday, R. Resnik and J. Walker, Fundamental of
Physics, 6th ed. (Wiley, New York, 2000).
REFERENCES [8] F. Cardarelli, Material Handbook (Springer Verlag, Lon-
don, 2000), Chap. 10.
[1] ASTM E 756, Standard Test Method for Measuring Vi-