You are on page 1of 23

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/356918313

Prediction of Elastic Constants of Timoshenko Rectangular Beams Using the


First Two Bending Modes

Article  in  Structural Engineering & Mechanics · December 2021


DOI: 10.12989/sem.2021.80.6.000

CITATIONS READS
0 35

2 authors:

Hung-Liang Chen Guadalupe Leon


West Virginia University Doane University
77 PUBLICATIONS   806 CITATIONS    8 PUBLICATIONS   36 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Mass Concrete View project

Mass Concrete View project

All content following this page was uploaded by Hung-Liang Chen on 02 April 2022.

The user has requested enhancement of the downloaded file.


To be published in Structural Engineering and Mechanics, Vol. 80, No. 6 (2021) 000-000
https://doi.org/10.12989/sem.2021.80.6.000

Prediction of Elastic Constants of Timoshenko Rectangular


Beams Using the First Two Bending Modes
Hung-Liang (Roger) Chen1a and Guadalupe Leon2b
1
Department of Civil and Environmental Engineering, West Virginia University, Morgantown, WV 26505,
USA

(Received keep as blank , Revised keep as blank , Accepted keep as blank )

Abstract. In this study, a relationship between the resonance frequency ratio and Poisson’s ratio was
proposed that can be used to directly determine the elastic constants. Using this relationship, the
frequency ratio between the 1st bending mode and 2nd bending mode for any rectangular Timoshenko
beam can be directly estimated and used to determine the elastic constants efficiently. The exact solution
of the Timoshenko beam vibration frequency equation under free-free boundary conditions was
determined with an accurate shear shape factor. The highest percent difference for the frequency ratio
between the theoretical values and the estimated values for all the beam dimensions studied was less
than 0.02%. The proposed equations were used to obtain the elastic constants of beams with different
material properties and dimensions using the first two measured transverse bending frequencies. Results
show that using the equations proposed in this study, the Young’s modulus and Poisson’s ratio of
rectangular Timoshenko beams can be determined more efficiently and accurately than those obtained
from industry standards such as ASTM E1876-15 without the need to test the torsional vibration.
Keywords: Timoshenko beam, Dynamic elastic modulus, Poisson’s ratio, bending mode, non-destructive
testing

1. Introduction

The elastic modulus and Poisson’s ratio are important material constants needed to predict the
behavior of a homogenous isotropic material. These elastic constants can be measured using non-
destructive testing (NDT) methods. Material properties measured with these methods are called dynamic
elastic constants. Non-destructive techniques are mostly based on the propagation of waves or vibration.
Most wave propagation techniques are based on the propagation of either longitudinal waves (P-wave),
Rayleigh waves, or transverse waves (S-wave). Lee et al. (Lee et al. 2017) found the dynamic elastic
modulus using S-waves. Based on their experimental measurements, they stated that the wave
propagation techniques produced slightly higher values for the dynamic modulus compared to the
vibration techniques (Lee et al. 2017). However, in their analysis a Poisson’s ratio of 0.2 was assumed
since it was not measured independently. Herein lies a disadvantage of using a single wave propagation
method because Poisson’s ratio must be known in advance. Researchers have used forced vibration to
measure the dynamic elastic constants of steel plates (Alfano and Pagnotta 2007) as well as impact-echo
resonance and Rayleigh wave velocity (Medina and Bayón 2010), and Wang et al. (Wang et al. 2012)
evaluated Poisson’s ratio of rods using longitudinal and cross-sectional resonant frequencies from
impact-echo. Quaglio et. al. (2020) used the longitudinal oscillations to determine the Young’s modulus
of different aggregate. Safari et al (2021) used the natural frequency to study the stiffness of Timoshenko

a
Corresponding author, Professor, E-mail: roger.chen@mail.wvu.edu
b
Ph.D. Student, E-mail: gleon@mix.wvu.edu

1
beams with porous cores.

Researchers have studied the accuracy of using the frequency ratio to predict the shear modulus
and elastic modulus of beams (Larsson 1991). Ip et al. used an iterative method to calculate the elastic
and shear modulus of thick composite beams under vibration. Their method usually converged after
seven iterations, and they chose the first two bending modes because it had the lowest measurement
error (Ip and Tse 2001). Kolluru et al. showed a method to evaluate the elastic constants of standard test
cylinders using the longitudinal resonance frequencies of concrete, steel, and aluminum (Kolluru et al.
2000). The ratio between the 2nd and 1st longitudinal resonance frequencies was used to determine
Poisson’s ratio for a variety of length-to-diameter ratios (L/D) and the elastic modulus was estimated
afterward. Brancheriau (2014) and Bosomworth (2010) developed methods to improve the calculation
of the elastic constants using the resonant frequencies of a free-free Timoshenko beam. Carneiro et al.
developed an indirect method using vibration tests to determine the modulus of cellular solids (Carneiro
et al. 2021). Recently, Chen and Leon (2019) showed a method to use the frequency ratio between the
1st torsional mode and 1st bending mode to determine the elastic constants for Timoshenko prisms and
rods (Chen and Leon 2019; Leon and Chen 2019). They also developed a method using the ratio between
the first two bending frequencies for a rod with 𝐿/𝐷 ≥ 2 (Leon and Chen 2019). Park et. al. (2020) used
the resonance frequency measured with ASTM C215 to predict the static modulus and compressive
strength of concrete.

ASTM E1876 (2015) and ASTM C215 (2019) are widely used to determine the dynamic elastic
modulus using the fundamental resonance frequencies with both the torsional mode and bending mode.
The equations provided were formulated by simplifying the free-free vibration solution of a Timoshenko
beam with the effects of shear and rotational inertia reduced to a simple correction constant. However,
the value of the elastic modulus cannot be directly calculated without knowing the material’s Poisson’s
ratio. Therefore, in order to calculate the dynamic modulus and Poisson’s ratio, a lengthy iteration
process is needed.
In this study, the exact solution of the Timoshenko beam vibration frequency equation under
free-free boundary conditions was determined. The fundamental transverse frequencies were used to
determine the dynamic elastic modulus and Poisson’s ratio of any rectangular beam with a length/height
ratio (L/t) > 2 and a width/height (b/t) < 1. A relationship between the transverse resonance frequencies
and Poisson’s ratio was proposed that can directly determine the elastic modulus and Poisson’s ratio
simultaneously, without the need for iteration, using the frequency ratio between the 2nd bending mode
and the 1st bending mode which can be quickly obtained from a single transverse vibration test.

2. Analysis
In the early 20th century, Timoshenko developed a fourth-order differential equation to consider
shear and rotary effects in the beam vibration problem. Using the Timoshenko beam vibration equation,
the vibration of non-slender beams could be predicted with more accuracy (Rossit et al. 2018). For a
homogenous and isotropic material, the Timoshenko beam vibration equation can be expressed as:
𝜕4 𝑦 𝜕2 𝑦 𝐸 𝜕4 𝑦 𝜌2 𝐼 𝜕 4 𝑦 (1)
𝐸𝐼 + 𝑚 − 𝜌𝐼 (1 + ) + =0
𝜕𝑥 4 𝜕𝑡 2 𝑘 ′ 𝐺 𝜕𝑥 2 𝜕𝑡 2 𝑘 ′ 𝐺 𝜕𝑡 4
The third and fourth terms were added by Timoshenko to consider the shear and rotary effects
in the beam (Timoshenko 1937). In the equation, 𝑦 is the transverse displacement; t is time; 𝑘 ′ is a shape
factor; G is the shear modulus; E is the Young’s modulus; I is the moment of inertia of the cross section;
ρ is the mass density and m is the mass per unit length. In Timoshenko’s publication (Timoshenko 1937),

2
2
a shape factor of 3 was assumed for a rectangular cross-section. In 1966, Cowper also derived shape
factors for different geometries (Cowper 1966). However, Hutchinson derived more accurate shape
factors for a variety of geometries which match more closely with experimental results (Hutchinson
2000). In this study, Hutchinson’s shape factor, Eq. (2), was used to calculate the shape factor. The shape
factor for a rectangular cross-section depends on the dimensions and Poisson’s ratio. In Eq. (2), t is the
height of the beam and b is the width as shown in Fig. 1.
2(1 + 𝑣) (2)
𝑘′ = −
144 𝑏2
[ 5 𝐶4 + 𝑣 (1 − 2 )]
𝑡 𝑏 𝑡
where:
∞ 𝑣 2 𝑏 5 (𝑛𝜋𝑡 − 𝑏 𝑡𝑎𝑛ℎ (𝑛𝜋𝑡 )) (3)
1 3 𝑏
𝐶4 = 𝑡 𝑏(−12𝑡 2 − 15𝑣𝑡 2 + 5𝑣𝑏2 ) + ∑ 5
720 4(𝑛𝜋) (1 + 𝑣)
𝑛=1

Fig. 1 Dimensions of rectangular Timoshenko beam

𝜕𝜓
For a Timoshenko beam with free-free boundary conditions, the bending moment, 𝑀 = 𝐸𝐼 𝜕𝑥 and the
𝜕𝑦
shear force, 𝑉 = 𝑘 ′ 𝐴𝐺(𝜕𝑥 − 𝜓) at the ends of the beam (𝑥 = 0, 𝑥 = 𝐿) are zero, where 𝜓 is the cross
sectional rotation due to bending.
Using these boundary conditions for a nontrivial solution, the frequency equation can be written
as (Chen and Kiriakidis 2005; Goens 1931):

𝛾𝑖2 (𝛾𝑖4 𝑟 2 (𝑟 2 − 𝑠 2 )2 + (3𝑟 2 − 𝑠 2 )) (4)


2 − 2 cos ℎ(𝛼𝑖 ) cos(𝛽𝑖 ) + sinh(𝛼𝑖 ) sin(𝛽𝑖 ) = 0
√1 − 𝛾𝑖4 𝑟 2 𝑠 2

𝐼 𝐸𝐼 𝜌𝐴𝐿4 2
where 𝑟 2 = 𝐴𝐿2 , 𝑠 2 = 𝑘 ′𝐺𝐴𝐿2 , 𝛾i4 = 𝜔𝑖 and A is the cross-sectional area; α and β represent the
𝐸𝐼
eigenvalues of the frequency equation. The eigenvalues α and β can be expressed as:

𝛾𝑖2 4
𝛼𝑖 = √−(𝑟 2 + 𝑠 2 ) + √(𝑟 2 − 𝑠 2 )2 + (5)
√2 𝛾𝑖4

3
and

𝛾𝑖2 4
𝛽𝑖 = √(𝑟 2 + 𝑠 2 ) + √(𝑟 2 − 𝑠 2 )2 +
√2 𝛾𝑖4

A MATLAB program was used to solve for the eigenvalues of the frequency equation. The
natural frequencies (Hz) can be expressed using the eigenvalues (Huang 1961):

𝛾𝑖2 𝐸𝐼
𝑛𝑖 = √
2𝜋 𝜌𝐴𝐿4 (6)

Let i = 1 to obtain the 1st bending mode and i = 2 for the 2nd bending mode.

In addition, using ABAQUS, a commercially available finite-element analysis (FEA) software,


Chen and Leon (2019) showed that the analytical Timoshenko beam solution and FEA for the 1st and
2nd bending modes had less than 0.1% difference for six different sets of elastic constants(Chen and Leon
2019; Leon and Chen 2019). The mode shapes obtained from the FEA are shown in Fig. 2. Therefore,
the assumed shear shape factor,𝑘′, and the roots of Eq. (4) are accurate.

(a) (b)
Fig. 2 Mode shapes from FEA: (a) 1st and (b) 2nd bending modes

3. Frequency Ratio Analysis

Since ASTM E1876-15 requires an iterative process to calculate the dynamic elastic modulus
and Poisson’s ratio using the 1st bending and torsional mode, it would be beneficial to develop a method
based on only the transverse bending modes, which can be measured simultaneously using just one test
setup. In this study, the ratio between the 2nd and 1st bending mode was used to directly calculate
Poisson’s ratio and the dynamic elastic modulus.
The ratio between the 2nd and 1st fundamental transverse frequency, 𝜒𝑏 can be expressed with
the following relationship:

𝑛2 2 𝛾2
1 (7)
𝜒𝑏 = ( ) =
𝑛1 𝛾1

4
where, 𝑛1 and 𝑛2 are the 1st and 2nd fundamental transverse frequency (Hz). 𝛾1 and 𝛾2 are the roots from
the frequency equation, shown in Eq. (4), for the 1st and 2nd bending mode. For a very slender beam, the
frequency ratio, 𝜒𝑏 will approach a constant, and the relationship shown in Eq. (8) can be derived to
express 𝜒𝑏 using the Euler-Bernoulli beam equation with free-free boundary conditions.
7.853 (8)
𝜒𝑏 = = 1.66
4.73
However, for the vibration of a non-slender beam, 𝜒𝑏 is no longer a constant and needs to be
expressed with the roots of the Timoshenko frequency equation. Hence, Eq. (7) was evaluated for a
variety of Poisson’s ratios and beam dimensions. Table 1 shows the calculated values of 𝛾1 and 𝛾2 due
to different Poisson’s ratio and length/height (L/t) for a width/height (b/t) ratio of 0.5. 𝜒𝑏 can be
calculated from the tabulated values. As expected, as the L/t ratio becomes very large, the values of 𝛾1
and 𝛾2 approach 4.73 and 7.853. They are the same values shown in Eq. (8) and correspond to a Euler-
Bernoulli beam. For a very slender beam (e.g. L/t = 100), 𝜒𝑏 is no longer dependent on Poisson’s ratio
and 𝛾𝑖 approaches a constant value as seen in Table 1. A relationship between 𝜒𝑏 and Poisson’s ratio
with different beam dimensions can be further observed in Fig. 3. From the figure, a slightly parabolic
relationship between Poisson’s ratio, v, and the ratio, 𝜒𝑏 can be observed. The y-intercept,𝐸1 ,
corresponds to the value of 𝜒𝑏 when v = 0. It was found that the frequency ratio, 𝜒𝑏 is independent of
the elastic modulus and the mass density, and 𝜒𝑏 only depends on the roots of Eq. (4). Hence, the ratio,
𝜒𝑏 can be expressed as Eq. (9).

𝑛2 2 𝛾2
1 (9)
𝜒𝑏 = ( ) = = 𝐶1 𝑣 2 + 𝐷1 𝜈 + 𝐸1
𝑛1 𝛾1
𝐶1 , 𝐷1 and 𝐸1 are the coefficients of the 2nd order polynomial describing the relationship
between 𝜒𝑏 and Poisson’s ratio.

(a) (b)
Fig. 3 Relationship between χb and Poisson’s ratio with different beam dimensions (width x height x length): (a)
50 mm x 100 mm x 800 mm and (b) 6 mm x 20 mm x 40 mm

Table 1 Example calculation of 𝛾1 and 𝛾2 for beams with different length/height (L/t) dimensions (for a beam with
width/height, b/t = 0.5)

L/t
Poisson's
2 4 8 10 20 50 100
ratio
γ1 γ2 γ1 γ2 γ1 γ2 γ1 γ2 γ1 γ2 γ1 γ2 γ1 γ2
0 3.804 5.069 4.361 6.571 4.618 7.392 4.656 7.539 4.711 7.767 4.727 7.839 4.729 7.850
0.1 3.791 5.006 4.353 6.537 4.615 7.376 4.654 7.527 4.710 7.763 4.727 7.838 4.729 7.849
0.2 3.778 4.945 4.344 6.504 4.612 7.360 4.652 7.516 4.710 7.760 4.727 7.838 4.729 7.849
0.3 3.765 4.887 4.336 6.471 4.608 7.344 4.650 7.504 4.709 7.756 4.727 7.837 4.729 7.849

5
0.4 3.752 4.832 4.328 6.439 4.605 7.328 4.648 7.493 4.709 7.753 4.727 7.837 4.729 7.849
0.5 3.739 4.778 4.320 6.407 4.602 7.313 4.646 7.482 4.708 7.749 4.726 7.836 4.729 7.849
Poisson’s ratio can then be calculated using the equation below:

1 (10)
𝑛 2
−𝐷1 − √𝐷12 − 4𝐶1 (𝐸1 − (𝑛2 ) )
1
𝑣=
2𝐶1
Therefore, Poisson’s ratio can be determined using the experimentally measured transverse
vibration frequencies of the first two bending modes, 𝑛2 and 𝑛1 . Furthermore, after knowing Poisson’s
ratio, the elastic modulus can be determined using the 1st bending mode as shown in Eq. (11). To find
𝛾1 , one can use the relationships for 𝐴1 and 𝐵1 developed (Chen and Leon 2019), where the frequency
ratio between the 1st torsional and the 1st bending was shown as 𝜒 = 𝜋 2 𝐴𝐿2 ⁄𝑅𝐼𝛾1 4 = 𝐴1 𝑣 + 𝐵1 .
Hence, using Eq. (6) and the calculated Poisson’s ratio from Eq. (10), the elastic modulus can
now be calculated directly as:
𝐸 = 4𝜌𝐿2 𝑛12 𝑅(𝐴1 𝑣 + 𝐵1 ) (11)
where
(1) 𝑅 = (𝑏⁄𝑡 + 𝑡⁄𝑏 )/[(4𝑏⁄𝑡) − 2.52(𝑏⁄𝑡)2 + 0.21(𝑏⁄𝑡)6 ] (12)
With the equations presented above, the dynamic elastic modulus and Poisson’s ratio can be
directly calculated using the experimentally measured resonance frequencies provided that the
coefficients 𝐴1 , 𝐵1 , 𝐶1 , 𝐷1 and 𝐸1 are known beforehand. The procedures to directly obtain these
coefficients are shown in the following section.

4. Empirical Equations for 𝑪𝟏 , 𝑫𝟏 𝐚𝐧𝐝 𝑬𝟏

To determine the elastic constants for a rectangular beam, the coefficients 𝐶1 , 𝐷1 and 𝐸1 are
needed for any combination of width/height (b/t) and the length/height (L/t) ratios. Although it is
possible to solve the roots of Eq. (4) to obtain the exact values of 𝐶1 , 𝐷1 and 𝐸1 for every beam cases, it
is very time consuming. Therefore, empirical equations were developed to make it simpler and more
efficient. The coefficients for different b/t and L/t ratios for a rectangular beam obtained using Eq. (4)
are shown in Table 2-
Table 4. The coefficients were calculated by evaluating the frequency ratio, Eq. (7), at different
Poisson’s ratio and fitting the best-fit second-order polynomial. It can be seen from the tables that as the
beam gets slenderer (e.g. L/t = 100), the value of C1 and D1 approach zero. Therefore, 𝜒𝑏 becomes a
constant equal to only E1. It should be noted that E1 is only dependent on the length/height ratio unlike
the other two coefficients.
Table 2 Values of C1 for different rectangular beams

Width/ Length/height
height

2 4 8 10 20 50 100
0.1 2.74E-02 4.36E-03 1.21E-03 6.62E-04 6.58E-05 1.97E-06 1.28E-07
0.2 2.74E-02 4.37E-03 1.21E-03 6.64E-04 6.67E-05 2.11E-06 1.64E-07
0.3 2.75E-02 4.39E-03 1.22E-03 6.75E-04 7.01E-05 2.71E-06 3.16E-07

6
0.4 2.76E-02 4.47E-03 1.26E-03 7.03E-04 7.90E-05 4.26E-06 7.09E-07
0.5 2.79E-02 4.61E-03 1.34E-03 7.59E-04 9.71E-05 7.39E-06 1.50E-06
0.6 2.83E-02 4.86E-03 1.47E-03 8.56E-04 1.28E-04 1.28E-05 2.88E-06
0.7 2.91E-02 5.25E-03 1.67E-03 1.01E-03 1.77E-04 2.13E-05 5.03E-06
0.75 2.96E-02 5.52E-03 1.81E-03 1.11E-03 2.10E-04 2.70E-05 6.47E-06
0.8 3.02E-02 5.83E-03 1.97E-03 1.23E-03 2.49E-04 3.37E-05 8.17E-06
0.9 3.17E-02 6.62E-03 2.38E-03 1.54E-03 3.47E-04 5.08E-05 1.25E-05
1 3.37E-02 7.67E-03 2.93E-03 1.94E-03 4.77E-04 7.34E-05 1.82E-05

Table 3 Values of D1 for different rectangular beams

Width/ Length/height
height

2 4 8 10 20 50 100
0.1 -0.123001 -0.049254 -0.023969 -0.017497 -0.005387 -0.000922 -0.000233
0.2 -0.122998 -0.049253 -0.023968 -0.017496 -0.005387 -0.000922 -0.000233
0.3 -0.122985 -0.049249 -0.023967 -0.017495 -0.005386 -0.000921 -0.000233
0.4 -0.122954 -0.049238 -0.023962 -0.017492 -0.005385 -0.000921 -0.000233
0.5 -0.122891 -0.049218 -0.023953 -0.017485 -0.005384 -0.000921 -0.000233
0.6 -0.122781 -0.049182 -0.023936 -0.017473 -0.005380 -0.000920 -0.000232
0.7 -0.122609 -0.049126 -0.023911 -0.017455 -0.005375 -0.000920 -0.000232
0.75 -0.122495 -0.049088 -0.023894 -0.017443 -0.005371 -0.000919 -0.000232
0.8 -0.122360 -0.049044 -0.023874 -0.017429 -0.005367 -0.000918 -0.000232
0.9 -0.122018 -0.048931 -0.023822 -0.017392 -0.005356 -0.000916 -0.000231
1 -0.121569 -0.048783 -0.023755 -0.017344 -0.005342 -0.000914 -0.000231

Table 4 Values of E1 for different rectangular prisms

Width/ Length/height
height
2 4 8 10 20 50 100
0.1 1.3324 1.5067 1.6007 1.6190 1.6487 1.6584 1.6598
1 1.3324 1.5067 1.6007 1.6190 1.6487 1.6584 1.6598

In order to develop empirical equations for 𝐶1 and 𝐷1 , the values shown in Table 2-
Table 3 were normalized and fitted with high-order polynomials. It was found that 𝐶1 and 𝐷1
can be calculated from the normalized values, ̅̅̅
𝐶1 and ̅̅̅
𝐷1 , using Eqs. (13) - (14). The coefficients were
normalized by subtracting the value of the coefficient at b/t =0.1 and dividing by that value at b/t = 1.0.
Note, since 𝐸1 only depends on the length/height ratio, it does not need to be normalized. As shown in
Fig. 4, the normalized coefficients are only a function of the width/height ratio. Therefore, if the
coefficients at b/t = 0.1 and b/t = 1.0 are known, 𝐶1 and 𝐷1 can be calculated for any set of beam

7
dimensions. Fourth-order polynomials can be used to express the normalized values, ̅̅̅
𝐶1 and ̅̅̅
𝐷1 , shown
in Eqs. (15) - (16).
(1) 𝑏 𝑏 𝑏
̅̅̅1 [𝐶1 ( = 1.0) − 𝐶1 ( = 0.1)] + 𝐶1 ( = 0.1) (13)
𝐶1 = 𝐶
𝑡 𝑡 𝑡

(1) 𝑏 𝑏 𝑏 (14)
𝐷1 = ̅̅̅
𝐷1 [𝐷1 ( = 1.0) − 𝐷1 ( = 0.1)] + 𝐷1 ( = 0.1)
𝑡 𝑡 𝑡

𝑏 4 𝑏 3 𝑏 2 𝑏 (15)
̅̅̅
𝐶1 = 0.25828 ( ) + 1.16632 ( ) − 0.51591 ( ) + 0.09761 ( ) − 0.00604
𝑡 𝑡 𝑡 𝑡
𝑏 4 𝑏 3 𝑏 2 𝑏 (16)
̅̅̅
𝐷1 = 0.25784 ( ) + 1.16708 ( ) − 0.51630 ( ) + 0.09769 ( ) − 0.00604
𝑡 𝑡 𝑡 𝑡

(a) (b)

(c) (d)
Fig. 4 Coefficients at different length/height ratios: (a) 𝐶1, (b) ̅̅̅
𝐶1 , (c) 𝐷1 and (d) ̅̅̅
𝐷1

8
𝑏 𝑏 𝑏 𝑏
As shown in Fig. 5-Fig. 7, 𝐶1 ( 𝑡 = 0.1), 𝐶1 ( 𝑡 = 1.0), 𝐷1 ( 𝑡 = 0.1), 𝐷1 ( 𝑡 = 1.0) and 𝐸1 can be
expressed as a function of t/L (height/length) using polynomial equations, and to increase the accuracy
of the proposed method, two L/t intervals are specified. A total of 97 data points were used to determine
the polynomials. The proposed equations can be used to find the coefficients for any rectangular beam
with 𝐿/𝑡 ≥ 2.

(a) (b)
𝑏 𝑏
Fig. 5 Plot of C1 versus the height/length (t/L) ratio: (a) 𝐶1 ( 𝑡 = 0.1) and (b) 𝐶1 ( 𝑡 = 1.0).

(a) (b)
𝑏 𝑏
Fig. 6 Plot of D1 versus the length/height (L/t) ratio: (a) 𝐷1 ( = 0.1) and (b) 𝐷1 ( = 1.0).
𝑡 𝑡

9
Fig. 7 Plot of E1 versus the height/length (t/L) ratio.

𝐿
For < 10
𝑡
𝑏 𝑡 6 𝑡 5 𝑡 4 𝑡 3 (17)
𝐶1 ( = 0.1) = −12.07366 ( ) + 21.89841 ( ) − 13.32730 ( ) + 3.72722 ( )
𝑡 𝐿 𝐿 𝐿 𝐿
𝑡 2 𝑡
− 0.50491 ( ) + 0.05495 ( ) − 0.00241
𝐿 𝐿
𝑏 𝑡 5 𝑡 4 𝑡 3 𝑡 2 (18)
𝐷1 ( = 0.1) = 22.23715 ( ) − 30.92178 ( ) + 14.27138 ( ) − 2.55722 ( )
𝑡 𝐿 𝐿 𝐿 𝐿
𝑡
− 0.06673 ( ) + 0.00334
𝐿
𝑏 𝑡 6 𝑡 5 𝑡 4 𝑡 3 (19)
𝐶1 ( = 1.0) = −12.13328 ( ) + 20.36119 ( ) − 11.37704 ( ) + 2.91476 ( )
𝑡 𝐿 𝐿 𝐿 𝐿
𝑡 2 𝑡
− 0.38773 ( ) + 0.06715( ) − 0.00289
𝐿 𝐿
𝑏 𝑡 5 𝑡 4 𝑡 3 𝑡 2 (20)
𝐷1 ( = 1.0) = 22.01168 ( ) − 30.57988 ( ) + 14.10124 ( ) − 2.52228 ( )
𝑡 𝐿 𝐿 𝐿 𝐿
𝑡
− 0.06748 ( ) + 0.00335
𝐿
𝑡 6 𝑡 5 𝑡 4 𝑡 3 (21)
𝐸1 = −19.155091 ( ) + 56.925868 ( ) − 62.997361 ( ) + 32.386821 ( )
𝐿 𝐿 𝐿 𝐿
𝑡 2 𝑡
− 7.931555 ( ) + 0.132169 ( ) + 1.658512
𝐿 𝐿

10
𝐿
For ≥ 10
𝑡
𝑏 𝑡 6 𝑡 5 𝑡 4 𝑡 3 (22)
𝐶1 ( = 0.1) = 456.499058 ( ) − 193.954796 ( ) + 24.953345 ( ) − 0.419613 ( )
𝑡 𝐿 𝐿 𝐿 𝐿
𝑡 2 𝑡
+ 0.007496 ( ) − 0.000062 ( )
𝐿 𝐿
𝑏 𝑡 5 𝑡 4 𝑡 3 𝑡 2 (23)
𝐷1 ( = 0.1) = −327.078249 ( ) + 94.390941 ( ) − 0.325165 ( ) − 2.334495 ( )
𝑡 𝐿 𝐿 𝐿 𝐿
𝑡
+ 0.000054 ( )
𝐿
𝑏 𝑡 6 𝑡 5 𝑡 4 𝑡 3 (24)
𝐶1 ( = 1.0) = 362.548041 ( ) − 133.715418 ( ) + 13.339241 ( ) − 0.275862 ( )
𝑡 𝐿 𝐿 𝐿 𝐿
𝑡 2 𝑡
+ 0.186427 ( ) − 0.000040 ( )
𝐿 𝐿
𝑏 𝑡 5 𝑡 4 𝑡 3 𝑡 2 (25)
𝐷1 ( = 1.0) = −325.280476 ( ) + 93.761143 ( ) − 0.315497 ( ) − 2.315721 ( )
𝑡 𝐿 𝐿 𝐿 𝐿
𝑡
+ 0.000056 ( )
𝐿
𝑡 5 𝑡 4 𝑡 3 𝑡 2 (26)
𝐸1 = −316.007488 ( ) + 111.483533 ( ) − 0.794082 ( ) − 4.843080 ( )
𝐿 𝐿 𝐿 𝐿
𝑡
− 0.000122 ( ) + 1.660283
𝐿
𝛾
To check the accuracy of the above coefficients, 𝛾2 was calculated using Eqs. (13) - (26) for the
1
beams with a constant width/height (b/t = 0.5) used in Table 1. The results are shown in Table 5. The
estimated values using the proposed equations are denoted as “-E”. The higher-order polynomials were
found to produce a maximum percent difference, Δ (%) equal to 0.017%, compared with the theoretical
𝛾 𝛾
values of 𝛾2 , denoted as “-T”, for all the b/t = 0.5 cases. Using these equations, the prediction of 𝛾2 for
1 1
all the other cases with different b/t were also carried out, and the results for b/t = 0.1 and b/t = 1.0 are
shown in Table 6-Table 7. The highest percent difference between the theoretical values and the
estimated values for all the beam dimensions studied (2 < L/t < 100 and 0.1 < b/t < 1) was less than
0.02%. This indicates that the proposed polynomial relationships can be used to adequately estimate the
coefficients in Eq. (9), and the proposed equations can accurately predict the frequency ratio between
the 2nd and 1st bending modes.

11
Table 5 Comparison of theoretical and estimated 𝛾2 /𝛾1 for a width/height (b/t = 0.5)

L/t
Poisson's 2 4 8 10 20 50 100
ratio
γ2/γ1-T γ2/γ1-E Δ (%) γ2/γ1-T γ2/γ1-E Δ (%) γ2/γ1-T γ2/γ1-E Δ (%) γ2/γ1-T γ2/γ1-E Δ (%) γ2/γ1-T γ2/γ1-E Δ (%) γ2/γ1-T γ2/γ1-E Δ (%) γ2/γ1-T γ2/γ1-E Δ (%)
0 1.3325 1.3324 0.014% 1.5068 1.5067 0.004% 1.6007 1.6006 0.003% 1.6192 1.6190 0.010% 1.6487 1.6487 0.002% 1.6583 1.6584 0.000% 1.6600 1.6598 0.010%
0.1 1.3205 1.3203 0.012% 1.5017 1.5018 -0.007% 1.5983 1.5983 0.000% 1.6173 1.6173 0.002% 1.6482 1.6481 0.004% 1.6581 1.6583 -0.008% 1.6598 1.6598 -0.001%
0.2 1.3089 1.3089 0.001% 1.4972 1.4971 0.013% 1.5958 1.5959 -0.004% 1.6156 1.6156 0.005% 1.6476 1.6476 -0.002% 1.6581 1.6582 -0.002% 1.6598 1.6598 0.000%
0.3 1.2980 1.2980 0.003% 1.4924 1.4924 0.002% 1.5938 1.5936 0.011% 1.6138 1.6139 -0.006% 1.6471 1.6471 0.000% 1.6579 1.6581 -0.009% 1.6598 1.6597 0.002%
0.4 1.2878 1.2876 0.017% 1.4878 1.4878 -0.001% 1.5913 1.5913 0.002% 1.6121 1.6122 -0.004% 1.6464 1.6465 -0.007% 1.6579 1.6580 -0.004% 1.6598 1.6597 0.003%
0.5 1.2779 1.2778 0.003% 1.4831 1.4833 -0.011% 1.5891 1.5890 0.005% 1.6104 1.6105 -0.004% 1.6459 1.6460 -0.005% 1.6581 1.6579 0.010% 1.6598 1.6597 0.005%

Table 6 Comparison of theoretical and estimated 𝛾2 /𝛾1 for a width/height (b/t = 0.1)

L/t
Poisson's 2 4 8 10 20 50 100
ratio
γ2/γ1-T γ2/γ1-E Δ (%) γ2/γ1-T γ2/γ1-E Δ (%) γ2/γ1-T γ2/γ1-E Δ (%) γ2/γ1-T γ2/γ1-E Δ (%) γ2/γ1-T γ2/γ1-E Δ (%) γ2/γ1-T γ2/γ1-E Δ (%) γ2/γ1-T γ2/γ1-E Δ (%)
0 1.3324 1.3324 0.0016% 1.5067 1.5067 0.0011% 1.6007 1.6006 0.0006% 1.6190 1.6190 0.0000% 1.6487 1.6487 0.0000% 1.6584 1.6584 0.0000% 1.6598 1.6598 0.0000%
0.1 1.3204 1.3203 0.0029% 1.5018 1.5018 0.0011% 1.5983 1.5983 0.0005% 1.6173 1.6173 0.0000% 1.6481 1.6481 0.0000% 1.6583 1.6583 0.0000% 1.6598 1.6598 0.0000%
0.2 1.3089 1.3088 0.0031% 1.4970 1.4970 0.0010% 1.5959 1.5959 0.0004% 1.6156 1.6156 0.0000% 1.6476 1.6476 0.0000% 1.6582 1.6582 0.0000% 1.6598 1.6598 0.0000%
0.3 1.2979 1.2979 0.0035% 1.4923 1.4923 0.0008% 1.5936 1.5936 0.0003% 1.6138 1.6138 0.0000% 1.6471 1.6471 0.0000% 1.6581 1.6581 0.0000% 1.6597 1.6597 0.0000%
0.4 1.2876 1.2875 0.0044% 1.4877 1.4877 0.0007% 1.5913 1.5913 0.0001% 1.6121 1.6121 0.0000% 1.6465 1.6465 0.0000% 1.6580 1.6580 0.0000% 1.6597 1.6597 0.0000%
0.5 1.2777 1.2777 0.0050% 1.4832 1.4832 0.0006% 1.5890 1.5890 0.0000% 1.6105 1.6105 0.0000% 1.6460 1.6460 0.0000% 1.6579 1.6579 0.0000% 1.6597 1.6597 0.0000%

Table 7 Comparison of theoretical and estimated 𝛾2 /𝛾1 for a width/height (b/t = 1.0)

L/t
Poisson's 2 4 8 10 20 50 100
ratio
γ2/γ1-T γ2/γ1-E Δ (%) γ2/γ1-T γ2/γ1-E Δ (%) γ2/γ1-T γ2/γ1-E Δ (%) γ2/γ1-T γ2/γ1-E Δ (%) γ2/γ1-T γ2/γ1-E Δ (%) γ2/γ1-T γ2/γ1-E Δ (%) γ2/γ1-T γ2/γ1-E Δ (%)
0 1.3324 1.3324 0.0016% 1.5067 1.5067 0.0011% 1.6007 1.6006 0.0006% 1.6190 1.6190 0.0000% 1.6487 1.6487 0.0000% 1.6584 1.6584 0.0000% 1.6598 1.6598 0.0000%
0.1 1.3205 1.3205 -0.0014% 1.5019 1.5019 -0.0001% 1.5983 1.5983 0.0000% 1.6173 1.6173 -0.0004% 1.6481 1.6481 -0.0001% 1.6583 1.6583 0.0000% 1.6598 1.6598 0.0000%
0.2 1.3094 1.3094 0.0019% 1.4973 1.4973 0.0007% 1.5960 1.5960 0.0003% 1.6156 1.6156 -0.0001% 1.6476 1.6476 0.0000% 1.6582 1.6582 0.0000% 1.6598 1.6598 0.0000%
0.3 1.2990 1.2989 0.0063% 1.4928 1.4928 0.0018% 1.5938 1.5938 0.0008% 1.6140 1.6140 0.0002% 1.6471 1.6471 0.0001% 1.6581 1.6581 0.0000% 1.6597 1.6597 0.0000%
0.4 1.2892 1.2891 0.0074% 1.4884 1.4884 0.0017% 1.5916 1.5916 0.0007% 1.6124 1.6124 0.0002% 1.6466 1.6466 0.0001% 1.6580 1.6580 0.0000% 1.6597 1.6597 0.0000%
0.5 1.2800 1.2800 0.0011% 1.4842 1.4842 -0.0003% 1.5895 1.5895 -0.0003% 1.6108 1.6108 -0.0004% 1.6461 1.6461 -0.0001% 1.6579 1.6579 0.0000% 1.6597 1.6597 0.0000%

12
5. Implementation of Method

To demonstrate the usage of the proposed method, the proposed equations were used to predict
the Poisson’s ratio and Young’s modulus for two different materials (steel and aluminum beams) using
their first two bending frequencies. Three different sets of dimensions were analyzed for each material.
The dimensions of each set can be found in Table 8. The 1st and 2nd bending modes were obtained with
ABAQUS finite-element analysis using the material properties shown in Table 9. All the sets utilized
8-node linear brick with reduced integration and hourglass control element (C3D8R) in ABAQUS using
at least 350,000 elements for each set. The exact frequencies obtained from the roots of Eq. (4) are
compared to the FEA results. The percent difference is small for all cases with the largest difference
being 0.055% for Set 6. With the first two bending frequencies given, the elastic constants can be
calculated using the proposed equations. As described in the flow chart in Fig. 8.
Table 8 Example set of beams

Width Height Length FEA Exact % Difference


Set Material b/t L/t
(mm) (mm) (mm) 𝑛1(𝐻𝑧) 𝑛2(𝐻𝑧) 𝑛1 (𝐻𝑧) 𝑛2(𝐻𝑧) 𝑛1 𝑛2
1 15.0 30.0 120 0.50 4.0 9129.2 20400 9129.7 20397 0.005% -0.015%
2 Steel 2.18 10.91 120 0.20 11.0 3833.5 10080 3834.1 10081.6 0.016% 0.016%
3 25.0 50.0 1000 0.50 20.0 257.87 699.77 257.99 700.09 0.047% 0.046%
4 11.99 27.0 100 0.44 3.7 11507 25080 11506.7 25072 -0.003% -0.032%
5 Aluminum 4.81 32.05 500 0.15 15.6 655.94 1762.0 656.05 1762.23 0.017% 0.013%
6 1.125 1.5 150 0.75 100.0 345.86 952.73 346.05 953.256 0.055% 0.055%

Table 9. Material properties used in FEA

Material Properties Steel Aluminum


3
Density (𝑘𝑔/𝑚 ) 7800 2700
Elastic modulus (GPa) 200 68.9
Poisson’s ratio 0.25 0.33

13
Fig. 8 Flowchart to calculate elastic modulus and Poisson's ratio using the 1 st and 2nd bending mode

14
First, based on the dimensions of each beam, the normalized values of 𝐶1 and 𝐷1 were calculated
𝐿 𝑏 𝑏 𝑏
using Eqs. (15)-(16). Then, depending on the length/height, 𝑡 ratio, 𝐶1 ( 𝑡 = 0.1) , 𝐶1 ( 𝑡 = 1.0) , 𝐷1 ( 𝑡 =
𝑏
0.1) and 𝐷1 ( 𝑡 = 1.0) were calculated using Eqs. (17) - (20) and Eqs. (22) - (25). Finally, using Eqs.
(13) - (14), the actual values of 𝐶1 and 𝐷1 were estimated. Likewise, 𝐸1 was obtained using Eq. (21) or
𝐿
Eq. (26). Derivation of the equations to estimate 𝐴1 and 𝐵1 for any rectangular beam with 𝑡 ≥ 2 can be
found in Chen and Leon (Chen and Leon 2019). In this paper, the equations have been modified to
𝐿
achieve better accuracy using higher order polynomials depending on two 𝑡 intervals as shown in Eqs.
(27) - (34).

𝑏 6 𝑏 5 𝑏 4 𝑏 3 𝑏 2 (27)
̅̅̅
𝐴1 = −1.5768 ( ) + 5.1004 ( ) − 3.1984 ( ) − 4.6752 ( ) + 5.414 ( )
𝑡 𝑡 𝑡 𝑡 𝑡
𝑏
− 0.0671 ( ) + 0.0032
𝑡

𝑏 6 𝑏 5 𝑏 4 𝑏 3 𝑏 2 (28)
̅̅̅
𝐵1 = −1.4307 ( ) + 4.569 ( ) − 2.5779 ( ) − 4.7736 ( ) + 5.2836 ( )
𝑡 𝑡 𝑡 𝑡 𝑡
𝑏
− 0.0739 ( ) + 0.0035
𝑡

𝐿
For 𝑡 < 10
𝑏 𝐿 4 𝐿 3 𝐿 2 𝐿 (29)
𝐵1 ( = 1.0) = −9.8078 × 10−5 ( ) + 3.0185 × 10−3 ( ) + 0.16427 ( ) + 0.19894 ( )
𝑡 𝑡 𝑡 𝑡 𝑡
+ 0.83547
𝑡 6 𝑡 5 𝑡 4 𝑡 3 𝑡 2 (30)
𝐴1 = −2.8127 ( ) + 10.039 ( ) − 13.487 ( ) + 8.9672 ( ) − 2.9095 ( )
𝐿 𝐿 𝐿 𝐿 𝐿
𝑡
+ 0.0982 ( ) + 0.3917
𝐿

𝐿
For 𝑡 ≥ 10
𝑏 𝐿 4 𝐿 3 𝐿 2 (31)
−10 −7
𝐵1 ( = 1.0) = −2.6824 × 10 ( ) + 1.2300 × 10 ( ) + 0.19991 ( )
𝑡 𝑡 𝑡 𝑡
𝐿
+ 1.1717 × 10−3 ( ) + 1.2927
𝑡
𝑏 𝑡 6 𝑡 5 𝑡 4 𝑡 3 𝑡 2 (32)
𝐴1 ( = 1.0) = −73.361 ( ) − 14.569 ( ) + 16.077 ( ) − 0.0248 ( ) − 1.5839 ( )
𝑡 𝐿 𝐿 𝐿 𝐿 𝐿
𝑡
− 8.886 × 10−7 ( ) + 0.3946
𝐿

𝑏
̅̅̅1 ∗ 𝐴1 ( = 1.0) (33)
𝐴1 = 𝐴
𝑡

15
𝑏
𝐵1 = ̅̅̅
𝐵1 ∗ 𝐵1 ( = 1.0)
𝑡 (34)
The calculated coefficients for the six sets of dimensions can be found in Table 10. Using Eqs.
(10) - (11), the elastic modulus and Poisson’s ratio were calculated for each set using the estimated
values for 𝐴1 , 𝐵1 , 𝐶1 , 𝐷1 and 𝐸1 . The percent difference between the calculated elastic constants and the
properties input in the finite-element analysis are shown in Table 11. The elastic modulus is predicted
with less than 0.2% error for all cases while Poisson’s ratio is predicted with less than 1.4% error for all
cases except Set 6 which has a 3.05% error. For Set 6, the length/height ratio is 100; it is a very slender
beam, so the bending frequency ratio is no longer sensitive to Poisson’s ratio. As shown in Table 11, the
𝐿 𝐿
proposed method works very well for short (𝑡 < 10) or intermediate length (𝑡 < 20) beams because 𝜒𝑏
is more sensitive to Poisson’s ratio. A similar conclusion was obtained by Leon and Chen regarding
Timoshenko rods (Leon and Chen 2019). For a very slender beam, 𝜒𝑏 is only slightly affected by
Poisson’s ratio, therefore, the bending frequency ratio must be measured with high accuracy otherwise
a large error could occur when it is used to estimate Poisson’s ratio.
Further verification of the method was made using the experimental measurements of Diaz de
Anda et al. (Díaz-de-Anda et al. 2012). In their paper, the bending frequencies of a 25.2-mm x 50.4-mm
x 500-mm aluminum beam with a mass density of 2699.04 (kg/m3) was tested using an electromagnetic
acoustic transducer. Poisson’s ratio was calculated to be 0.25223 using the ratio between the reported
elastic modulus (𝐸 = 67.42 𝐺𝑃𝑎) and shear modulus (𝐺 = 26.92 𝐺𝑃𝑎). The measured 1st and 2nd
bending modes (denoted as “-M”) are shown in Table 12. Following the method described in this paper,
the 1st and 2nd bending modes were estimated (denoted as “-E”) based on the reported elastic modulus
and Poisson’s ratio values using Eqs. (10) - (11). The difference between the estimated frequency and
the measured frequency is about 2.0% for both modes as shown in Table 12. One can use the measured
bending frequencies of the aluminum beam to estimate the Young’s modulus and Poisson’s ratio, and
the results are 70.2 GPa and 0.266, which shows a percent difference of 3.96% and 5.17%, respectively,
in comparison with their reported values as shown in Table 12.

6. Experiment
The 1st and 2nd bending modes of a 78.83-mm x 101.54-mm x 407.99-mm concrete prism were
measured. The mass density of the concrete was 2339.5 𝑘𝑔/𝑚3 . The concrete proportions for large
aggregate, fine aggregate, Portland cement, ground-granulated blast furnace slag and water were 1065,
809, 151, 151, and 128 𝑘𝑔/𝑚3 , respectively, with a water-cementitious material ratio of about 0.42. The
28-day compressive strength of the concrete mix was measured to be approximately 41.3-MPa. After
curing in the water tank for 28 days, the concrete specimen was stored for about six months at a
laboratory under air-dry condition at about 50% humidity. The free-free bending modes were measured
following ASTM E1876-15. A B&K Type 4374 accelerometer was glued at approximately 0.25L using
rubber cement. A sampling frequency of 50 kHz was used to measure the acceleration-time history of
the beam. The accelerometer was connected to a Tektronix TDS-2024B for data acquisition. The 1st and
2nd bending modes were excited using an instrumented impact hammer (PCB Model 086C03) at 0.75L.
Supports were placed to provide minimal restrain and allow free-free vibration (ASTM E1876-15).
Afterwards, fast Fourier transform (FFT) was performed on the acceleration-time histories to obtain the
1st and 2nd bending modes. A picture and schematic drawing of the experimental set-up showing the
instruments and the concrete specimen is displayed in Fig. 9.

16
Table 10 Estimated coefficients for six sets of dimensions
̅̅̅
𝐶1 ̅̅̅
𝐷1 𝑏 𝑏 𝑏 𝑏 𝐶1 𝐷1 𝐸1 𝐴1 𝐵1
Set 𝐶1 ( = 0.1) 𝐶1 ( = 1.0) 𝐷1 ( = 0.1) 𝐷1 ( = 1.0)
𝑡 𝑡 𝑡 𝑡
1 0.0757200 0.0757300 0.0043862 0.0076877 -0.0492507 -0.0487875 0.0046362 -0.0492156 1.5067078 0.2229294 2.8782129
2 0.0025822 0.0025880 0.0004989 0.0016156 -0.0151186 -0.0149878 0.0005018 -0.0151182 1.6252971 0.0631366 4.0495032
3 0.0757200 0.0757300 0.0000657 0.0004768 -0.0053865 -0.0053416 0.0000968 -0.0053831 1.6486680 0.2631970 52.837398
4 0.0477490 0.0477571 0.0048989 0.0083993 -0.0526180 -0.0521188 0.0050661 -0.0525942 1.4929246 0.1891180 2.2268758
5 0.0010626 0.0010681 0.0001594 0.0007964 -0.0084346 -0.0083635 0.0001600 -0.0084345 1.6417069 0.0380631 4.7118189
6 0.3507308 0.3507528 -0.0000001 0.0000181 -0.0002323 -0.0002304 0.0000063 -0.0002317 1.6597978 0.3722025 1840.9278

Table 11 Estimated elastic Young's modulus and Poisson's ratio

Exact FEA Estimated Error (%)


1 𝛾2 E 𝛾2 E 𝛾2 𝛾2
𝑛2 2 𝛾2
Set ( ) = 𝛾1 (GPa) v 𝛾1 (GPa) v 𝛾1 𝛾1 E v
𝑛1 𝛾1 −𝐹𝐸𝐴 −𝐸𝑠𝑡
1 1.494703 1.494854 200.0 0.25 1.494694 199.94 0.2466 0.0101% -0.0006% -0.03% -1.37%
2 1.621560 1.621558 200.0 0.25 1.621549 199.94 0.2494 -0.0001% -0.0007% -0.02% -0.25%
3 1.647311 1.647317 200.0 0.25 1.647328 199.81 0.2521 0.0004% 0.0011% -0.09% 0.82%
4 1.476111 1.476327 68.9 0.33 1.476120 68.877 0.3258 0.0146% 0.0006% -0.17% -1.27%
5 1.638939 1.638970 68.9 0.33 1.638941 68.792 0.3266 0.0019% 0.0001% -0.16% -1.04%
6 1.659722 1.659720 68.9 0.33 1.659722 68.826 0.3401 -0.0001% 0.0000% -0.11% 3.05%

Table 12 Comparison with Diaz de Anda et al. (2012)

Material b/t L/t 𝐶1 𝐷1 𝐸1 𝐴1 𝐵1 𝑛1 𝑛1 Error 𝑛2 𝑛2 Error E Error ν Error


(Hz)-M (Hz)-E (%) (Hz)-M (Hz)-E (%) (GPa) (%) (%)

Aluminum 0.50 9.92 0.000753 −0.0177 1.6185 0.2559 13.634 1,021.1 1000.8 1.99% 2,659.4 2607.3 1.96% 70.20 3.96% 0.266 5.17%

17
Fig. 9 Experimental set-up picture (a) concrete specimen, (b) accelerometer, (c) instrumented harmer,
and (d) digital oscilloscope, and schematic drawing.

An example of the acceleration-time histories and frequency domain is shown in Fig. 10.
The peaks in the frequency domain correspond to the bending modes. Signal processing was used
to obtain a 0.49 Hz resolution for the bending frequencies. Three separate tests were conducted
with identical results.

(a) (b)
Fig. 10 Examples: (a) Acceleration-time history and (b) Frequency domain

The Poisson’s ratio and elastic modulus of the concrete beam were found using the
proposed method. First, the five coefficients were calculated based on the measured width, height,
and length. Then, using the frequency ratio, Eqs. (10) - (11) were used to find the elastic constants.
A summary of the calculations can be found in Table 13.
Table 13 Summary of calculations for concrete prism

Material b/t L/t 𝐶1 𝐷1 𝐸1 𝐴1 𝐵1 𝑛1 (Hz) 𝑛2 (Hz) E (GPa) ν


Concrete 0.78 4.02 0.00567 -0.0489 1.5075 0.3174 4.1711 2254.16 5085.10 42.11 0.115

To demonstrate the sensitivity of the proposed method, a sensitivity analysis was


performed using the concrete beam. As an example, a resolution of ±0.1-mm was considered for
the width, height, and length. While a frequency resolution of ±0.5-Hz was investigated. The mass
of the beam was measured with a ±0.01-kg accuracy. As shown in Table 14, the proposed method
is more sensitive to the height and frequency variation. A ±0.1-mm change to the height can change
Poisson’s ratio by ±0.0037 and the elastic modulus by ±0.114-GPa. The resolution of the 1st

18
bending mode is shown to be more important than the resolution of the 2nd bending mode. A ±0.5-
Hz in the 1st bending mode can change Poisson’s ratio by ±0.0035 and the elastic modulus by ±0.03-
GPa.

Table 14 Sensitivity analysis of concrete prism

Measured E (GPa) v
Properties
Mass (±0.01-kg) 46.06±0.055 0.115±0.0000
Width (±0.1-mm) 46.06±0.053 0.115±0.000004
Height (±0.1-mm) 46.06±0.114 0.115±0.0037
Length (±0.1-mm) 46.06±0.028 0.115±0.00092
𝑛2 (±0.5-Hz) 46.06±0.005 0.115±0.0016
𝑛1 (±0.5-Hz) 46.06±0.030 0.115±0.0035

7. Conclusions

In this study, the solution of the frequency ratio, 𝜒𝑏 between the 2nd bending mode and the
st
1 bending mode of any rectangular Timoshenko beam under free-free boundary conditions was
determined. The following conclusions can be drawn from this study:
The relationship between 𝜒𝑏 and Poisson’s ratio was found to be slightly parabolic with
coefficients 𝐶1 , 𝐷1 , and 𝐸1. These coefficients depend only on the dimensions of the beam
and can be tabulated for different specimen sizes.
Empirical relationships for 𝐶1 , 𝐷1 , and 𝐸1 at different length/height and width/height ratios
were found, allowing the usage of any rectangular beam dimensions.
As the length/height increases, the values of 𝐶1 and 𝐷1 approach zero, and 𝜒𝑏 approaches
a constant value.
Equations for the elastic modulus, E, and Poisson’s ratio, v, were proposed in this paper,
allowing for a quick and direct calculation of both E and v, simultaneously, for any
rectangular beam with a width/height ratio (b/t) < 1 and a length/height ratio (L/t) > 2.
Using the proposed methods, the elastic constants can be quickly estimated from just a
single transverse vibration test, unlike the current ASTM E1876-15 method, and do not
require any iterations.
Results show that, using the equations proposed in this study, the Young’s modulus and
Poisson’s ratio can be determined efficiently and accurately

Acknowledgments

The authors acknowledge the support provided by the West Virginia Transportation
Division of Highways (WVDOH) and FHWA for Research Project WVDOH RP#312. Special
thanks are extended to our project monitors, Mike Mance, Donald Williams, and Ryan Arnold of
WVDOH.

19
References

Alfano, M., and Pagnotta, L. (2007). “A non-destructive technique for the elastic characterization of thin
isotropic plates.” NDT & E International, 40(2), 112–120.
https://doi.org/10.1016/j.ndteint.2006.10.002
Carneiro, V. H., Lopes, D., Puga, H., and Meireles, J. (2021). “Numerical inverse engineering as a route to
determine the dynamic mechanical properties of metallic cellular solids.” Materials Science and
Engineering: A, 800, 140428. https://doi.org/10.1016/j.msea.2020.140428
ASTM C09 Committee. (2019). Test Method for Fundamental Transverse, Longitudinal, and Torsional
Resonant Frequencies of Concrete Specimens. ASTM International, West Conshohocken, PA.
ASTM E28 Committee. (2015). Test Method for Dynamic Youngs Modulus, Shear Modulus, and Poisson
Ratio by Impulse Excitation of Vibration. ASTM International, West Conshohocken, PA.
Bosomworth P., "Improved Frequency Equations for Calculating the Young’s Modulus of Bars of
Rectangular or Circular Cross Section from Their Flexural Resonant Frequencies," (2010) Journal
of ASTM International 7, No. 8: 1-15. https://doi-org.wvu.idm.oclc.org/10.1520/JAI102953
Brancheriau, L. "An alternative solution for the determination of elastic parameters in free–free flexural
vibration of a Timoshenko beam." (2014) Wood Sci Technol 48, 1269–1279. https://doi-
org.wvu.idm.oclc.org/10.1007/s00226-014-0672-x
Chen, H.-L. (Roger), and Kiriakidis, A. C. (2005). “Nondestructive Evaluation of Ceramic Candle Filter with
Various Boundary Conditions.” Journal of Nondestructive Evaluation, 24(2), 67–81.
https://doi.org/10.1007/s10921-005-3483-z
Chen, H.-L. (Roger), and Leon, G. (2019). “Direct Determination of Dynamic Elastic Modulus and Poisson’s
Ratio of Rectangular Timoshenko Prisms.” Journal of Engineering Mechanics, 145(9), 04019071.
https://doi.org/10.1061/(ASCE)EM.1943-7889.0001643
Cowper, G. R. (1966). “The Shear Coefficient in Timoshenko’s Beam Theory.” Journal of Applied
Mechanics, 33(2), 335–340.
Díaz-de-Anda, A., Flores, J., Gutiérrez, L., Méndez-Sánchez, R. A., Monsivais, G., and Morales, A. (2012).
“Experimental study of the Timoshenko beam theory predictions.” Journal of Sound and Vibration,
331(26), 5732–5744. https://doi.org/10.1016/j.jsv.2012.07.041
Goens, E. (1931). “Uber die Bestimmung des Elastizitätsmoduls von Stäben mit Hilfe von
Biegungsschwingungen.” Annalen der Physik, 403(6), 649–678.
Huang, T. C. (1961). “The Effect of Rotatory Inertia and of Shear Deformation on the Frequency and Normal
Mode Equations of Uniform Beams With Simple End Conditions.” Journal of Applied Mechanics,
28(4), 579–584. https://doi.org/10.1115/1.3641787
Hutchinson, J. R. (2000). “Shear Coefficients for Timoshenko Beam Theory.” Journal of Applied Mechanics,
68(1), 87–92. https://doi.org/10.1115/1.1349417
Ip, K. H., and Tse, P. C. (2001). “Determination of Dynamic Flexural and Shear Moduli of Thick Composite
Beams Using Natural Frequencies:” Journal of Composite Materials, Sage PublicationsSage CA:
Thousand Oaks, CA. https://doi.org/10.1106/U0FU-9BR5-JNTG-B57R
Kolluru, S., Popovics, J., and Shah, S. (2000). “Determining Elastic Properties of Concrete Using Vibrational
Resonance Frequencies of Standard Test Cylinders.” Determining Elastic Properties of Concrete
Using Vibrational Resonance Frequencies of Standard Test Cylinders.
Larsson, P.-O. (1991). “Determination of young’s and shear moduli from flexural vibrations of beams.”
Journal of Sound and Vibration, 146(1), 111–123. https://doi.org/10.1016/0022-460X(91)90525-O
Lee, B. J., Kee, S.-H., Oh, T., and Kim, Y.-Y. (2017). “Evaluating the Dynamic Elastic Modulus of Concrete
Using Shear-Wave Velocity Measurements.” Advances in Materials Science and Engineering,
2017, 13. https://doi.org/10.1155/2017/1651753
Leon, G., and Chen, H.-L. (Roger). (2019). “Direct Determination of Dynamic Elastic Modulus and Poisson’s
Ratio of Timoshenko Rods.” Vibration, 2(1). https://doi.org/10.3390/vibration2010010
Medina, R., and Bayón, A. (2010). “Elastic constants of a plate from impact-echo resonance and Rayleigh
wave velocity.” Journal of Sound and Vibration, 329(11), 2114–2126.
https://doi.org/10.1016/j.jsv.2009.12.026
Park JY, Sim S-H, Yoon YG, Oh TK (2020). "Prediction of Static Modulus and Compressive Strength of
Concrete from Dynamic Modulus Associated with Wave Velocity and Resonance Frequency Using
Machine Learning Techniques." Materials. 13(13):2886. https://doi.org/10.3390/ma13132886

20
Quaglio O.A., Da Silva J.M., Rodovalho E.C. and Costa L.V. (2020). " Determination of Young's Modulus
by Specific Vibration of Basalt and Diabase." Advances in Materials Science and Engineering
https://doi.org/10.1155/2020/4706384
Rossit C.A., Bambill D. V. and Gilardi G. J. (2018). "Timoshenko theory effect on the vibration of axially
functionally graded cantilever beams carrying concentrated masses". Structural Engineering and
Mechanics. 66(6), 703-711. http://dx.doi.org/10.12989/sem.2018.66.6.703
Safari M.., Mohammadimehr M. and Ashrafi H. (2021). "Free vibration of electro-magneto-thermo
sandwich Timoshenko beam made of porous core and GPLRC". Advances in Nano Research.
10(2), 115-128. http://dx.doi.org/10.12989/anr.2021.10.2.115
Timoshenko, S. (1937). Vibration problems in engineering. D. Van Nostrand Company, New York.
Wang, J.-J., Chang, T.-P., Chen, B.-T., and Wang, H. (2012). “Determination of Poisson’s ratio of solid
circular rods by impact-echo method.” Journal of Sound and Vibration, 331(5), 1059–1067.
https://doi.org/10.1016/j.jsv.2011.10.030

21

View publication stats

You might also like