Download as pdf or txt
Download as pdf or txt
You are on page 1of 4

Evaluation of kp from disk samples

ABSTRACT
In text books kp is determined either by k31 (Eq. 1) or by measuring the resonance and anti-resonance
frequencies of a disk sample (of D/t>10, Eq. 2). The latter is much easier and more directly, but
calculation from Eq. 2 is not simple. Some approximations were proposed (Eq. 3-5) for a fast estimate,
while for precise evaluation iterations have to be conducted. Here the approximations are numerically re-
examined for practically important piezoelectric ceramics. Some related factors are refined for a better
approximation.

INTRODUCTION
From the theory of piezoelectricity, the electromechanical coupling factor (radial vibration mode) kp can
be evaluated by
2 2d 312 Y0E
k p2 = k 2
= (1) or
1−σ E (1 − σ E )ε 33
31 T

k p2 (1 − σ E ) J1[η1 (1 + ∆f / f s ] − η1 (1 + ∆f / f s ) J 0 [η1 (1 + ∆f / f s ]
= (2)
1 − k p2 η1 (1 + ∆f / f s ) J1[η1 (1 + ∆f / f s ]
Where σE is the Poison’s ratio, ∆f=fp-fs the difference between the anti-resonance and resonance
frequencies, η1 the lowest positive root of (1-σE)J1(η)=ηJ0(η), and J0, J1 are zero and first order of first
kind Bessel functions, respectively [1].
For σE between 0.27 and 0.35, the value of kp can be determined from the curve shown in Fig. 1 with a
precision better than 1% for a thickness to diameter ratio t/D<0.1 [1].

Fig. 1 Planar coupling factor of thin disks as a function of (fp-fs)/fs

During passed decades some approximations have been used for various situations:
f p2 − f s2
k p = 2 .3 (3)
f p2
2
f p − fs  f − fs 
k p = 2.51 − p  (4)
fp  f 
 p 
1
kp = (5)
fs
a +b
f p − fs
Where a=0.389 and b=0.563 for a Poison’s ratio of 0.38 (specifically for KNN ceramics).
Equation (3) is used in PCM-office, a program for piezoelectric measurements of EPCOS CMRD.
Equation (4) is used by APC International Ltd. USA, a company of piezoelectric ceramics [2].
Equation (5) was proposed in Ref. [3] for KNN ceramics.
Approximations of Eq. (3-5) omit the influence of Poison’s ratio as its variation about 0.3 brings about
little change of kp. Practically important piezoelectric ceramics have Poison’s ratio from 0.2 to 0.4,
particularly from 0.3 to 0.4. In the following sections kp will be calculated as a function of Poison’s ratio
and (fp-fs)/fs, and the accuracy of the 3 approaches will be compared.

EVALUATION OF η1
For a given value of σE, η1 can be evaluated by solving the transcendental equation (1-σE)J1(η)=ηJ0(η).
In Tab. I the numerical values are given and fitted with a linear equation for σE from 0.2 to 0.4.
η1=1.861+0.624σE (6)
2,12
Table I Corresponding values of σE and η1
2,10

σE η1 2,08

0.20 1.98441 2,06

0.25 2.01717 2,04


η1

0.30 2.04885 2,02

0.35 2.07951 2,00

0.40 2.10920 1,98

0,20 0,25 0,30 0,35 0,40

η1=1.861+0.624σ E E
σ

Fig. 2 Linear fitting between σE and η1

CALCULATION OF kp
Values of kp are calculated with Eq. (2) for different σE as a function of ∆f/fs, as are shown in Fig. 3 (a).
kp show weak dependence on the Poison’s ratio. Deviation of the kp values from the average is around
1% for σE=0.20 and 0.40, and is 0.5% for σE=0.25 and 0.35 [see Fig. 3(b)]. These data hint that kp can be
estimated by the curve of σE=0.30 with an uncertainty of ≤1% for materials having σE from 0.20 to 0.40.
0,7 E
E 0,010 σ =0.40
σ =0.20
E
0,6 σ =0.25
E
σ =0.30 E
E
σ =0.35
[kp(σ )-kp(0.30)]/kp(0.30)

0,5 σ =0.35 0,005


E
σ =0.40
0,4 E
σ =0.30
0,000
kp

0,3
E
σ =0.25
0,2 -0,005
E

0,1 E
σ =0.20
-0,010
0,0
0,00 0,05 0,10 0,15 0,20 0,00 0,05 0,10 0,15 0,20

∆f/fs ∆f/f s

(a) (b)
Fig. 3 kp as a function of ∆f/fs and σE (a), and its deviation from kp(σE=0.30) (b).
RE-EXAMING OF THE APPROXIMATIONS
Approximations in forms of Eq. (3), (4) and (5) can be readily checked by comparing respectively
derived values with the directly calculated ones shown in Fig. 3(a). As a function of ∆f/fs, these
approximations should be reformed as:
 
 1 
k p = 2.31 −  (3a)
 (1 + ∆f ) 2 
 fs 
2
∆f
 ∆f 
fs  fs 
k p = 2.51 −  (4a)
1 + ∆f  1 + ∆f 
fs  fs 
1
kp = (5a)
f
a s +b
∆f

In Fig. 4 the comparison is given in panel (a) and (b). Equation (3) shows a large deviation of 35%,
which is unacceptable. The deviation of Eq. (4) is around -0.2%, while that of Eq. (5) is around 0.9%.
The deviation of Eq. (5) is reasonable as the Poison’s ratio has been assumed to be 0.38.
0,35
0,9
E 0,30
0,8 σ =0.30
Eq. (3) Eq. (3)
0,7 Eq. (4) 0,25 Eq. (4)
Eq. (5) Eq. (5)
0,6 0,20
∆kp/kp(σ =0.30)

0,5
0,15
E
kp

0,4
0,10
0,3

0,2 0,05

0,1 0,00
0,0
0,00 0,05 0,10 0,15 0,20 0,00 0,05 0,10 0,15 0,20

∆f/fs ∆f/fs

(a) (b)
Fig. 4 Comparison between 3 approximations and directly calculated values (a) and respective deviations
from the latter (b).

In order to use Eq. (3) for a reasonable estimation, the directly calculated curve is fitted with Eq. (3a), by
varying the factor of 2.3. The fitted values coincide with the original data within ±0.3%, as is shown in
Fig. 5. The good fitting gives a very easy way to estimate kp from keff:
f p2 − f s2
k p = 1.2725 (3b)
f p2
or kp=1.128keff (3c)

Similar fitting has been done with Eq. 5 by varying a and b. A new set of factors lead to much higher
precision (~0.05%):
1
kp = (5b)
fs
0.3951 + 0.5828
f p − fs
0,010
0,010
0,6
0,6

0,5 0,005
0,5 0,005

0,4 0,4

∆kp/kp(σ =0.30)
∆kp/kp(σ =0.30)
0,000 0,000
0,3

kp
0,3
kp

E
E
0,2 0,2
-0,005 -0,005
0,1 0,1

0,0 -0,010 0,0 -0,010


0,00 0,05 0,10 0,15 0,20 0,00 0,05 0,10 0,15 0,20
∆f/fs ∆f/fs

Fig. 5 Nonlinear fitting of kp(σE=0.30) with Eq. (3b) (a), and Eq. (5b) and their deviations (b).

CONCLUSIONS
1) kp can be precisely estimated by measuring resonance and anti-resonance frequencies, fp and fs.
For σE between 0.20 and 0.4, the mathematical uncertainty is below ±1% about the values
calculated for σE=0.30.
2) Some available approximations are refined to get a better fitting, as are listed below:

Equations Expressions Uncertainty


Eq. (3b) ±1.3%
f −f 2 2

k p = 1.2725 p s
= 1.128keff
f p2
Eq. (4) 2 ±1.2%
f p − fs  f − fs 
k p = 2.51 − p 
fp  f 
 p 
Eq. (5b) ±1.0%
1
kp =
fs
0.3951 + 0.5828
f p − fs

REFERENCES
[1] B. Jaffe, W. Cook, and H. Jaffe, Piezoelectric Ceramics, Academic Press, London and New York,
1971, pp. 290-3.
[2] APC International, Ltd., Technical document,
http://www.americanpiezo.com/piezo_theory/index.html
[3] M Matsubara, T. Yamaguchi, K. Kikuta, and S. Hirano, Jpn. J. Appl. Phys., 44[8], 2005, pp 6136-42.

You might also like