Mechanical Systems and Signal Processing 85 (2017) 193–203

Contents lists available at ScienceDirect

Mechanical Systems and Signal Processing

journal homepage: www.elsevier.com/locate/ymssp

Analytically optimal parameters of dynamic vibration absorber

with negative stiffness
Yongjun Shen a,n, Haibo Peng b, Xianghong Li c, Shaopu Yang a
a
Department of Mechanical Engineering, Shijiazhuang Tiedao University, 050043, China
b
Departmentof Engineering Mechanics, Shijiazhuang Tiedao University, 050043, China
c
Departmentof Mathematics and Physics, Shijiazhuang Tiedao University, 050043, China

a r t i c l e i n f o abstract

Article history: In this paper the optimal parameters of a dynamic vibration absorber (DVA) with negative
Received 12 October 2015 stiffness is analytically studied. The analytical solution is obtained by Laplace transform
Received in revised form method when the primary system is subjected to harmonic excitation. The research shows
1 August 2016
there are still two fixed points independent of the absorber damping in the amplitude-
Accepted 11 August 2016
Available online 19 August 2016
frequency curve of the primary system when the system contains negative stiffness. Then
the optimum frequency ratio and optimum damping ratio are respectively obtained based
Keywords: on the fixed-point theory. A new strategy is proposed to obtain the optimum negative
Dynamic vibration absorber stiffness ratio and make the system remain stable at the same time. At last the control
Negative stiffness
performance of the presented DVA is compared with those of three existing typical DVAs,
Fixed-point theory
which were presented by Den Hartog, Ren and Sims respectively. The comparison results
Parameters optimization
in harmonic and random excitation show that the presented DVA in this paper could not
only reduce the peak value of the amplitude-frequency curve of the primary system
significantly, but also broaden the efficient frequency range of vibration mitigation.
& 2016 Elsevier Ltd. All rights reserved.

1. Introduction

Vibration control has been important for several decades and many efficient devices have been presented. One of the
common devices for vibration control is dynamic vibration absorber (DVA), which is widely used for its properties such as
efficiency, reliability, and low cost. Researches on DVA have been developed for more than 100 years since the first DVA
without damping was invented by Frahm [1] in 1909. In 1928, Den Hartog and Ormondroyd [2] found that a DVA with
damping element could suppress the amplitude of the primary system in a broader frequency range, which had been
recognized as the typical Voigt type DVA. Den Hartog and Ormondroyd also found that the amplitude-frequency curve of
the damped DVA would pass through two fixed points independent of the absorber damping, and they proposed an op-
timization criterion to design the Voigt type DVA. Hahnkamm [3] derived the optimum natural frequency ratio in 1932
according to this criterion. Later, the optimum damping ratio was obtained by Brock [4]. In 1956, Den Hartog gave a detailed
introduction on this optimization criterion in his famous monograph [5], and he called this method as the fixed-point
theory. In 2002, Nishihara and Asami [6] derived the exact series solutions for the optimum frequency and damping ratios of
the Voigt type DVA, and compared the control performance with that by Den Hartog. In 2007, Sims [7] introduced a new

n
Corresponding author.
E-mail address: shenyongjun@126.com (Y. Shen).

http://dx.doi.org/10.1016/j.ymssp.2016.08.018
0888-3270/& 2016 Elsevier Ltd. All rights reserved.
194 Y. Shen et al. / Mechanical Systems and Signal Processing 85 (2017) 193–203

analytical solution based on the criterion of minimizing either the positive real part or negative real part of the frequency
response function, so as to efficiently control the chattering phenomenon in machining tool. In 2001, Ren [8] presented a
DVA where the damping element was not connected to the primary system, but to the earth or the base structure. The result
indicated it could present better control performance than Voigt type DVA under the same parameters condition. In 2005,
Liu [9] got the same result using another method. However it was hard to get the analytical optimal solution when the
damping of the primary system could not be ignored, so that some numerical methods were proposed to get the optimal
solution [10–12]. In addition to the above passive DVAs, Shen et al. [13,14] studied the approximately analytical solutions for
four types of semi-active DVAs, and the effect of the time delay on the control performance was also studied.
The positive stiffness means that the deformation is in the same direction as the applied exterior force, which is com-
monly encountered in practical engineering. On the contrary, the negative stiffness means the relationship between the
exterior force and the displacement in deformed objects is reverse. The researches about the properties and stability
condition of the negative stiffness system had been reported in references [15–21]. There are some generation mechanisms
for negative stiffness, where the pre-compressed member or inverted pendulum are typical devices. When the springs with
positive and negative stiffness are put in parallel and connected to different masses, the system may show nonlinear
characteristic near the equilibrium position, such as instability or bifurcation. The load-bearing capacity of the system with
negative stiffness is better than that of the system with only positive stiffness, and at the same time the natural frequency of
the system will be reduced. The system with negative stiffness may have a better vibration control performance if it remains
stable. Thus, the introduction of the negative stiffness to the vibration control system is necessary and meaningful. Platus
et al. [22] produced negative stiffness around equilibrium using the buckling of beams under axial load, and they got an
isolation system by combining it with a linear positive spring. Trimboli [23] proposed the application of the negative
stiffness in the mechanical vibration isolator, where the negative stiffness and positive stiffness were arranged in parallel.
Park [24] studied the active control vibration isolator with negative stiffness, and the basic system characteristics were
experimentally verified. Mizuno et al. [25–28] studied an active vibration isolation system combining zero-power magnetic
suspension analytically and experimentally. They found that a zero-power system behaved like negative stiffness system,
and could generate infinite stiffness if it was connected with a normal spring in series. Then they proposed a new vibration
isolation system, where the negative stiffness was realized by active control technique [29]. In 2013, Acar et al. [30] studied
an adaptively passive DVA with a negative stiffness mechanism analytically and experimentally, and found that it could
suppress the system amplitude by appropriately adjusting the parameters. Yang et al. [31] also investigated a nonlinear
vibration isolator with a negative stiffness mechanism. Many scholars had studied the application of the negative stiffness in
vibration isolation system for its advantages, but little work has been done about the optimal parameters of the DVA with
negative stiffness until now.
In this paper the effect of the negative stiffness on the amplitude of the primary system is studied by introducing the
negative stiffness which is connected the DVA with the earth. Section 2 presents the optimal parameters based on the fixed-
point theory. In Section 3, the control performance of the presented DVA in this paper is compared with those of the DVAs
by Den Hartog, Ren and Sims when subjected to both sinusoidal and random excitations. The results verify the DVA in this
paper has more significant control performance. At last the conclusions are made in Section 4.

2. Analytical investigation on DVA with negative stiffness

The typical Voigt type DVA proposed by Den Hartog is shown in Fig. 1(a), and the DVA proposed by Ren is shown in Fig. 1
(b). By adding the negative stiffness directly connecting the mass of DVA and the earth in the model by Ren, one could get
the model with negative stiffness shown in Fig. 1(c).
Obviously, the model in Fig. 1(b) is a special case of Fig. 1(c), if the coefficient of the negative stiffness is set to zero.
According to Newton's second law, the motion equation of the system with negative stiffness can be established as
⎧ m1x¨1 + k1x1 + k2 (x1 − x2 ) = F cos (ωt )
⎨ ,
⎩ m2 x¨2 + cx2̇ + (k + k2 ) x2 − k2 x1 = 0 (1)

where m1, m2 , k1 and k2 are the masses, linear stiffness coefficients of the primary system and the DVA respectively. c is the
damping coefficient of the absorber, and k is the negative stiffness coefficient. x1 and x2 are the displacements of the primary
system and the DVA respectively. F and ω are the amplitude and frequency of the force excitation.
Using the following parametric transformation

m2 k1 k2 c k F
μ= , ω1 = , ω2 = , ξ= , α= , f= ,
m1 m1 m2 2m2 ω2 k2 m1

Eq. (1) becomes

⎧ x¨ + (ω 2 + μω 2 ) x − μω 2 x = f cos (ωt )
⎪ 1 1 2 1 2 2
⎨
⎪
.
⎩ x¨2 + 2ω2 ξx2̇ + (α + 1) ω22 x2 − ω22 x1 = 0 (2)
 Y. Shen et al. / Mechanical Systems and Signal Processing 85 (2017) 193–203 195

x2 x2
m2 m2
c
k2 k2
x1 x1
m1 c
m1
f (t ) = F cos(ωt ) f (t ) = F cos(ωt )
k1 k1

(a) The Voigt type DVA (b) The DVA by Ren

x2 m2

k2
x1
m1 c k

f (t ) = F cos(ωt )
k1

(c) DVA with negative stiffness

Fig. 1. The models of dynamic vibration absorbers. (a) The Voigt type DVA (b) The DVA by Ren. (c) DVA with negative stiffness.

2.1. The analytical solution

According to Laplace transform, Eq. (2) can be rewritten as

⎧ 2
⎪ s X1 + (ω
2 2 2 jωt
1 + μω2 ) X1 − μω2 X2 = fe
⎨
⎪ 2
.
2
⎩ s X2 + 2sω2 ξX2 + (α + 1) ω2 X2 − ω2X1 = 0 (3)

Supposing X1 = H ( jω) x1e jωt , X2 = H ( jω) x2 e jωt , and letting s = jω , one could obtain

⎧ A2 f
⎪ H ( jω) x1 =
⎪ A1A2 − μω24
⎨ ,
⎪ ω22 f
⎪ H ( jω) x2 = 4
⎩ A1A2 − μω2 (4)

where

A1 = − ω2 + ω12 + μω22, A2 = − ω2 + 2jωω2 ξ + αω22 + ω22.

Considering the modulus of Eq. (4), and introducing the parameters

ω2 ω
ν= , λ= ,
ω1 ω1

one can get

f ⎡⎣ λ − (1 + α ) ν ⎤⎦ + (2λνξ )
2 2 2 2 2
H ( jω) x1 = 2
,
ω1 Δ (5a)
196 Y. Shen et al. / Mechanical Systems and Signal Processing 85 (2017) 193–203

6
μ=0.1
υ=1.3
5 α=−0.3
ξ=0.4
4 ξ=0.2
ξ=0.1

3
A
P
2 Q

0
0 1 2 3
λ
Fig. 2. The normalized amplitude-frequency curves.

fν 2 1
H ( jω) x2 = ,
ω12 Δ (5b)

where

Δ = ⎡⎣ (1 + μν 2 − λ2)(αν 2 + ν 2 − λ2) − μν 4⎤⎦ + (1 + μν 2 − λ2)2 (2νλξ )2 .

2

Accordingly, the amplitude-frequency equations of the primary system and the DVA can be obtained

X¯1 = H ( jω) x1 , X¯2 = H ( jω) x2 . (6)

2.2. The parameters optimization of the DVA with negative stiffness

Based on the above analysis, the amplitude amplification factor of the primary system should be

X1 (αν 2 + ν 2 − λ2)2 + (2νλξ )2

A= = ,
Xst ⎡⎣ (1 + μν 2 − λ2)(αν 2 + ν 2 − λ2) − μν 4⎤⎦2 + (1 + μν 2 − λ2)2 (2νλξ )2
(7)

where Xst = F /k1 is the static deformation of the primary system.

The normalized amplitude-frequency curves with several different damping ratios are given in Fig. 2 to investigate the
system characteristics. It could be clearly seen that there exist two commonly fixed points P and Q on all the curves, which
are independent of the damping ratio. The optimum natural frequency ratio can be obtained by adjusting the responses at P
and Q to the same level. Then the optimum damping ratio can be obtained by making P and Q as the maximum values of the
amplitude-frequency curve.
Due to the fixed-point theory, there exist two equal values if ξ → ∞ and ξ = 0 in Eq. (7)

αν 2 + ν 2 − λ2 1
A= = ± .
(1 + μν 2 − λ2)(αν 2 + ν 2 − λ2) − μν 4 1 + μν 2 − λ2 (8)

It could be found that there is no meaning when the right part in Eq. (8) is positive.
Accordingly, taking the negative one and simplifying the equation, one can get

2λ 4 − 2 (1 + ν 2 + μν 2 + αν 2) λ2 + 2ν 2 (1 + α ) + μν 4 (1 + 2α ) = 0. (9)

Supposing the roots of Eq. (9) are λP and λQ , one can get the following equation

(λ2 − λP2 )(λ2 − λ Q2 ) = λ 4 + (λP2 + λ Q2 ) λ2 + λP2 λ Q2 = 0. (10)

Comparing Eqs. (9) and (10), one can find

λP2 + λ Q2 = 1 + ν 2 + μν 2 + αν 2. (11)

The values at P and Q should be the same in order to get the optimum natural frequency ratio
 Y. Shen et al. / Mechanical Systems and Signal Processing 85 (2017) 193–203 197

1 1
= − .
1 + μν 2 − λP2 1 + μν 2 − λ Q2 (12)

Simplifying Eq. (12) one can get

λP2 + λ Q2 = 2 + 2μν 2. (13)

Combining Eqs. (13) and (11), and solving the equation one can obtain the optimum natural frequency ratio

1
νopt = .
1+α−μ (14)

The two fixed points can be obtained under this condition

1 ⎛ μ⎞
λP2 = ⎜1 + α − ⎟,
1 + α − μ⎝ 2⎠ (15a)
1 ⎛ μ⎞
λ Q2 = ⎜1 + α + ⎟.
1 + α − μ⎝ 2⎠ (15b)

Based on the optimum natural frequency ratio, the response at P and Q can be obtained

X1 2
A| λp , λQ = = (1 + α − μ) .
Xst λp , λQ μ (16)

Next step one could adjust the damping ratio so as to make the maximum amplitude of the primary system locating at
the two fixed points. The condition can be achieved if the derivatives of the amplitude amplification factor are zero at the
two fixed points

∂A2
| λ , λ = 0.
∂λ2 p Q (17)

Solving Eq. (17) and substituting the optimum natural frequency ratio into the results one can get
3μ
ξp2 = ,
8 (1 + α − μ /2 ) (18a)
3μ
ξQ2 = .
8 (1 + α + μ /2 ) (18b)

Taking an average of ξp2 and ξQ2 , one can get the optimum damping ratio

3μ (1 + α )
ξopt = .
4 { 2 (1 + α )2 − μ} (19)

It is obvious that this result cannot exactly make the points P and Q as the maximum values of the amplitude-frequency
curve. But this approximation can be accepted as a simple design law. However there still exist an adjustable parameter in
the optimum natural frequency ratio and damping ratio, i.e., the optimum negative stiffness ratio α . It is clearly that an
appropriate negative stiffness ratio α will make the two fixed points lower and that means we can get a better vibration
absorption performance. According to the characteristics of the negative stiffness system, it can be achieved only when the
negative stiffness material is applied by preload, which could be realized by pre-compressed member or inverted pendulum.
The preload will cause a pre-displacement of the primary system, so an approximation is taken that the pre-displacement is
selected as the amplitudes at the fixed points. That means the response to zero-frequency excitation is the same as the
response at the fixed points

X1 (1 + α )(1 + α − μ) 2
A= = = (1 + α − μ) .
Xst (1 + α )2 − μ μ (20)

Solving Eq. (20), one can get two possible roots of the negative stiffness ratio

α1 = 2μ − 1, (21a)
μ
α2 = − − 1.
2 (21b)

Considering Eqs. (14) and (19), one can find that taking α2 as the negative stiffness ratio will make the optimum natural
frequency ratio as imaginary number and the optimum damping ratio infinity simultaneously. Obviously this value can not
be used. So we take α1 as the optimum negative stiffness ratio, namely
198 Y. Shen et al. / Mechanical Systems and Signal Processing 85 (2017) 193–203

-0.1 2.8 0.5

-0.2
2.6 0.45

-0.3
2.4 0.4
-0.4

2.2 0.35
-0.5
opt

opt
opt
α

ξ
ν
-0.6 0.3
2

-0.7
1.8 0.25

-0.8

1.6 0.2
-0.9

-1 1.4
0 0.1 0.2 0.3 0.4 0 0.1 0.2 0.3 0.4 0 0.1 0.2 0.3 0.4
μ μ μ
Fig. 3. The relationship between mass ratio and optimum parameters.

αopt = 2μ − 1. (22)

Then the three adjustable parameters of the DVA are analytically obtained. It must be pointed out that the three am-
plitudes on zero-frequency excitation, fixed point P and fixed point Q are not exactly the same, because the approximation in
the optimum damping ratio is adopted.
In the design procedure of DVA, the range of m2 is not arbitrary due to many practical situations, such as installation
space and manufacturing cost. The smaller m2 may cause that the two natural frequencies, i.e., λP and λQ , are too close, which
will be harmful to vibration control. The larger m2 will significantly increase the manufacturing cost and make the in-
stallation difficult. In the existing designing method for DVA, including the methods by Den Hartog [2,5], Ren [8,9] and other
works [3,4,6], m2 are all designated in advance. For example, restriction of μ between 0.05 and 0.5 will have little effect on
the ultimate control results.
For practical purpose, we give the relationships between the optimum parameters in Eqs. (14), (19), and (22) with the
mass ratio in Fig. 3. Based on the figure one can get the optimum parameters easily, and we can find that the negative
stiffness ratio remains a small value for the consideration of the system stability.

3. Comparisons of the control performances

Considering the transformed system Eq. (2) and selecting the mass ratio as μ = 0.1 and μ = 0.05, one could obtain the
optimal parameters of the DVA based on Eqs. (14), (19) and (22). From the optimal parameters one could get the normalized
amplitude-frequency curve, which is shown in Fig. 4.

3.1. The comparison with other DVAs under sinusoidal excitation

In order to illustrate the control performance of the presented DVA with negative stiffness, the normalized amplitude-
frequency curves of other typical DVAs proposed by Den Hartog and Ren are also presented. One can get the analytical
optimal parameters for the Voigt type DVA and the model by Ren shown in Fig. 1(a) and (b) by solving the following
equations and then fulfilling the same procedure as literatures [5,8].
⎧ m1x¨1 + c (x1̇ − x2̇ ) + k2 (x1 − x2 ) + k1x1 = F cos (ωt )
⎨
⎩ m2 x¨2 + c (x2̇ − x1̇ ) + k2 (x2 − x1) = 0 (23)

⎧ m1x¨1 + k1x1 + k2 (x1 − x2 ) = F cos (ωt )

⎨
⎩ m2 x¨2 + cx2̇ + k2 (x2 − x1) = 0 (24)

Here in order to compare the control performances, two different mass ratios μ = 0.1 and μ = 0.05 are selected to get the
 Y. Shen et al. / Mechanical Systems and Signal Processing 85 (2017) 193–203 199

9
The result by Den Hartog
8 The result by Ren
The result by Sim1
7 The result by Sim2
6 The result in this paper

5
μ=0.1
A 4

0
0 0.5 1 1.5 2 2.5 3
λ
(a)

12
The result by Den Hartog
The result by Ren
10 The result by Sim1
The result by Sim2
8 The result in this paper

6 μ=0.05
A

0
0 0.5 1 1.5 2 2.5 3
λ

(b)
Fig. 4. The comparison with other DVAs.

optimal parameters independently by the results Den and Ren presented [5,8]. The two curves are shown in Fig. 4. Fur-
thermore, the results by Sims in reference [7] aiming to get a better control performance in chatter problem are also
presented. In Fig. 4, Sim1 represents the result when the real part of the response is negative, while Sim2 represents the
positive case. One can find a detailed derivation in Section 3 of reference [7].
From the comparison, it could be concluded that the presented DVA in this paper can not only mitigate the amplitude of
the primary system significantly in resonance region, but also broaden the efficient frequency range of vibration absorption.
It should be emphasized is that the presented DVA is not so sensitive to the mass ratio, which means one can get a better
control performance using a smaller absorber mass.

3.2. The comparison with DVAs under random excitation

It is more important and meaningful to investigate the system response under random excitation, especially when DVA is
applied in earthquake and civil engineering. A further study is conducted on the DVA under random excitation in this
subsection. When a single degree-of-freedom system (the primary system) shown in Fig. 5 is subjected to random excitation
with zero mean and power spectral density as S (ω) = S0 , the mean square response of the SDOF system can be gotten as
200 Y. Shen et al. / Mechanical Systems and Signal Processing 85 (2017) 193–203

m1
f (t )
c1
k1

Fig. 5. A single degree system under random excitation.

πS 0
σ2 = .
2ω13 ξ1 (25)

The detailed analytic derivation for Eq. (25) could be found in the relevant literatures [5,32]. It is obvious the mean
square response will become infinity if the damping coefficient approaches zero. So it is necessary to connect a DVA to
reduce the system response.
Considering the primary systems subjected to the aforementioned random excitation, and as a comparison the power
spectral density functions of the primary systems attached with different DVAs become
2 2 2
SV (ω) = HVx1 ( jω) S0, SR (ω) = HRx1 ( jω) S0, SNe (ω) = HNx1 ( jω) S0, (26)

where the subscript V, R and N represent for the Voigt type DVA, the model by Ren, and the DVA in this paper respectively.
Based Eqs. (23) and (24), one can get the transfer functions for the Voigt type DVA and the model by Ren. Then the mean
square responses of the primary systems for different DVAs can be calculated as
∞ ∞
2 πS0 [1 + ν 4 (1 + μ)2 + ν 2 (4μξ 2 + 4ξ 2 − μ − 2)]
σV2 = ∫−∞ SV (ω) dω = S0 ∫−∞ HVx1 ( jω) dω =
2ω13 μξν
,

∞ ∞
2 πS0 [1 + ν 4 + (μ + 4ξ 2 − 2)]
σR2 = ∫−∞ SR (ω) dω = S0 ∫−∞ HRx1 ( jω) dω =
2ω13 μξν 5
,
∞ ∞
σN2 = ∫−∞ SN (ω) dω = S0 ∫−∞ HNx ( jω) 2 dω
1

πS0 { 1 + α + ν 4 [(1 + α )3 − 2αμ (1 + α − 2ξ 2) + αμ2 ] + ν 2 [ − 2 + 4ξ 2 + μ + 2α ( − 2 − α + 2ξ 2 + μ)] }

= .
2ω13 μξν 5 (1 + α + μαν 2) (27)

The DVA model by Sims is based on the Voigt type DVA, so that the results presented by Sims are also presented here. As
the aforementioned designation, Sims1 represents the result when the real part of the response is negative and Sims2
represents the positive case. According to the optimal parameters in literatures [5,7,8] and the results in this paper, the
mean square responses of the primary systems under the respectively optimal parameters can be obtained when μ ¼0.1
6.401πS0 7.659πS0 7.693πS0 5.780πS0 2.736πS0
σV2 = 2
, σSims1
= 2
, σSims 2
= , σR2 = , σN2 = .
ω13 ω13 ω13 ω13 ω13 (28)

The results show that the DVA with negative stiffness has the minimum mean square response. This means the presented
model can get a better performance than other DVAs even under the random excitation. Moreover, it could be found that the
presented model is still superior to other two DVAs if different mass ratios are selected.
In order to get more realistic results, 50 s random excitation is constructed, which is composed of 5000 normalized
random numbers with zero mean value and unit variance. The time history of the random excitation is shown in Fig. 6. Here
we take the primary mass m1 ¼1 kg and the stiffness of the primary system k1 ¼100 N/m. Then the other parameters can be
calculated according to the existing literatures for the models proposed Den Hartog, Ren and Sims. And the parameters for
the DVA proposed in this paper can be calculated according to Eqs. (14), (19) and (22). Based on the fourth-order Runge-
Kutta method the response of the primary systems without DVA and with different DVAs can be obtained. The time histories
of these primary systems are shown in Figs. 7–12. The response variances and decrease ratios of the primary systems for the
different systems are summarized in Table 1.
From Figs. 7–12 and Table 1, it could be concluded that the DVA in this paper could present more satisfactory control
performance than other DVAs, even when the primary system is subjected to random excitation. These results verify that
the presented DVA in this paper could reduce not only the peak value of the response, but also the statistical response in
 Y. Shen et al. / Mechanical Systems and Signal Processing 85 (2017) 193–203 201

Fig. 6. The time history of the random excitation.

0.05
Displacement

-0.05
0 5 10 15 20 25 30 35 40 45 50
Time
Fig. 7. The time history of the primary system without DVA.

0.02

0.01
Displacement

-0.01

-0.02
0 5 10 15 20 25 30 35 40 45 50
Time
Fig. 8. The time history of the primary system with DVA by Den Hartog.

0.02

0.01
Displacement

-0.01

-0.02
0 5 10 15 20 25 30 35 40 45 50
Time
Fig. 9. The time history of the primary system with DVA by Sims1.

0.04

0.02
Displacement

-0.02

-0.04
0 5 10 15 20 25 30 35 40 45 50
Time
Fig. 10. The time history of the primary system with DVA by Sims2.
202 Y. Shen et al. / Mechanical Systems and Signal Processing 85 (2017) 193–203

0.02

0.01

Displacement
0

-0.01

-0.02
0 5 10 15 20 25 30 35 40 45 50
Time
Fig. 11. The time history of the primary system with DVA by Ren.

0.015

0.01
Displacement

0.005

-0.005

-0.01
0 5 10 15 20 25 30 35 40 45 50
Time
Fig. 12. The time history of the primary system with DVA in this paper.

Table 1
The variances and decrease ratios of the displacements of the primary system.

Variances Decrease ratios (%)

Without DVA 2.0083e  04

DVA by Den Hartog 3.1235e 05 84.45
DVA by Sims1 3.3254e  05 83.44
DVA by Sims2 3.3568e  05 83.29
DVA by Ren 2.6979e 05 86.57
The presented model 1.0928e  05 94.56

whole-frequency range excellently.

4. Conclusions

A DVA with negative stiffness is investigated in this paper. The investigation shows that there still exist two fixed points
at the amplitude-frequency curves of the primary system. Based on the fixed-point theory, the optimum natural frequency
ratio and the optimum damping ratio are obtained. A method based on the fixed-point theory to determine the optimum
negative stiffness ratio is proposed, and the optimum negative stiffness ratio is also determined. The comparisons with
other DVAs show that the presented DVA in this paper can largely reduce the resonance amplitude and broaden vibration
frequency range. Even it can make the amplitude of the primary system remain small in the whole-frequency range. The
results provide a different idea to design more effective dynamic vibration absorber. Due to the significant control per-
formance of the devices with negative stiffness, one could develop the vibration control system with negative stiffness. The
study on the construction and dynamical analysis of vibrational system with negative stiffness will be more meaningful.

Acknowledgements

The authors are grateful to the support by National Natural Science Foundation of China (No. 11372198), the Cultivation
plan for Innovation team and leading talent in Colleges and Universities of Hebei Province (LJRC018), the Program for
advanced talent in the universities of Hebei Province (GCC2014053), and the Program for Advanced Talent in Hebei Province
(A201401001).
 Y. Shen et al. / Mechanical Systems and Signal Processing 85 (2017) 193–203 203

References

[1] H. Frahm, Device for damping vibrations of bodies, U.S. Patent 989,958, 1911.
[2] J. Ormondroyd, J.P. Den Hartog, The theory of the dynamic vibration absorber, ASME J. Appl. Mech. 50 (1928) 9–22.
[3] E. Hahnkamm, The damping of the foundation vibrations at varying excitation frequency, Master Archit. 4 (1932) 192–201.
[4] J.E. Brock, A note on the damped vibration absorber, ASME J. Appl. Mech. 13 (4) (1946) A284.
[5] J.P. Den Hartog, Mechanical Vibrations, 3rd ed. McGraw-Hall Book Company, New York (1947), pp. 112–132.
[6] O. Nishihara, T. Asami, Close-form solutions to the exact optimizations of dynamic vibration absorber (minimizations of the maximum amplitude
magnification factors), ASME J. Vib. Acoust. 124 (2002) 576–582.
[7] N.D. Sims, Vibration absorbers for chatter suppression: a new analytical tuning methodology, J. Sound Vib. 301 (3) (2007) 592–607.
[8] M.Z. Ren, A variant design of the dynamic vibration absorber, J. Sound Vib. 245 (4) (2001) 762–770.
[9] K.F. Liu, J. Liu, The damped dynamic vibration absorbers: revisited and new result, J. Sound Vib. 284 (3) (2005) 1181–1189.
[10] S.E. Randall, D.M. Halsted, D.L. Taylor, Optimum vibration absorbers for linear damped systems, J. Mech. Des. 103 (4) (1981) 908–913.
[11] A. Soom, Optimal design of linear and nonlinear vibration absorbers for damped systems, J. Vib. Acoust. 105 (1) (1983) 112–119.
[12] E. Pennestri, An application of Chebyshev's min–max criterion to the optimal design of a damped dynamic vibration absorber, J. Sound Vib. 217 (4)
(1998) 757–765.
[13] Y.J. Shen, L. Wang, S.P. Yang, G.S. Gao, Nonlinear dynamical analysis and parameters optimization of four semi-active on–off dynamic vibration
absorbers, J. Vib. Control 19 (1) (2013) 143–160.
[14] Y.J. Shen, M. Ahmadian, Nonlinear dynamical analysis on four semi-active dynamic vibration absorbers with time delay, Shock Vib. 20 (4) (2013)
649–663.
[15] R.S. Lakes, Extreme damping in composite materials with a negative stiffness phase, Phys. Rev. Lett. 86 (13) (2001) 2897–2900.
[16] R.S. Lakes, Extreme damping in compliant composites with a negative-stiffness phase, Philos. Mag. Lett. 81 (2) (2001) 95–100.
[17] R.S. Lakes, T. Lee, A. Bersie, Y.C. Wang, Extreme damping in composite materials with negative-stiffness inclusions, Nature 410 (6828) (2001) 565–567.
[18] R.S. Lakes, W.J. Drugan, Dramatically stiffer elastic composite materials due to a negative stiffness phase? J. Mech. Phys. Solids 50 (5) (2002) 979–1009.
[19] Y.C. Wang, R.S. Lakes, Extreme stiffness systems due to negative stiffness elements, Am. J. Phys. 72 (1) (2004) 40–50.
[20] Y.C. Wang, R. Lakes, Negative stiffness-induced extreme viscoelastic mechanical properties: stability and dynamics, Philos. Mag. 84 (35) (2004)
3785–3801.
[21] Y.C. Wang, R. Lakes, Stability of negative stiffness viscoelastic systems, Q. Appl. Math. 63 (1) (2005) 34–55.
[22] D.L. Platus, Negative-stiffness-mechanism vibration isolation systems, Int. Soc. Opt. Photonics (1992) 44–54.
[23] M.S. Trimboli, R. Wimmel, E.J. Breitbach, Quasi-active approach to vibration isolation using magnetic springs, in: Proc. SPIE 2193, Smart Structures and
Materials 1994, Passive Damping, 73, May 1, 1994.
[24] S.T. Park, T.T. Luu, Techniques for optimizing parameters of negative stiffness, Proc. Inst. Mech. Eng. Part C: J. Mech. Eng. Sci. 221 (5) (2007) 505–510.
[25] T. Mizuno, Proposal of a vibration isolation system using zero-power magnetic suspension, in: Proceedings of the Asia-Pacific Vibration Conference,
Hangzhou, China, vol 2, 2001, pp. 423–427.
[26] T. Mizuno, Vibration isolation system using zero-power magnetic suspension, The 15th Triennial World Congress, Barcelona, Spain (2002), pp. 955–
960.
[27] T. Mizuno, Y. Takemori, A transfer-function approach to the analysis and design of zero-power controllers for magnetic suspension systems, Electr.
Eng. Jpn. 141 (2) (2002) 67–75.
[28] T. Mizuno, M. Takasaki, D. Kishita, K. Hirakawa, Vibration isolation system combining zero-power magnetic suspension with springs, Control Eng.
Pract. 15 (2) (2007) 187–196.
[29] T. Mizuno, T. Toumiya, M. Takasaki, Vibration isolation system using negative stiffness, JSME Int. J. Ser. C 46 (3) (2003) 807–812.
[30] M.A. Acar, C. Yilmaz, Design of an adaptive–passive dynamic vibration absorber composed of a string–mass system equipped with negative stiffness
tension adjusting mechanism, J. Sound Vib. 332 (2) (2013) 231–245.
[31] J. Yang, Y.P. Xiong, J.T. Xing, Dynamics and power flow behaviour of a nonlinear vibration isolation system with a negative stiffness mechanism, J.
Sound Vib. 332 (1) (2013) 167–183.
[32] D.E. Newland, An Introduction to Random Vibrations, Spectral & Wavelet Analysis, Courier Corporation, London, 2012.