CHINESE JOURNAL OF SHIP RESEARCH，VOL.14，NO.

3，JUN 2019 21

To cite this article：Wang X X, Li X T, Ma Xiaohui, et al. Numerical analysis on spectrum dip characteristics of shock response
spectrum of submarine equipment[J/OL]. Chinese Journal of Ship Research, 2019, 14(3). http://www.
ship-research.com/CN/Y2019/V14/I3/31.
DOI： 10.19693/j.issn.1673-3185. 01206

Numerical analysis on spectrum dip

characteristics of shock response
spectrum of submarine equipment
Wang Xiaoxin*1，2，3，Li Xiaotian1，2，3，Ma Xiaohui4，Du Guowei1，2，3
1 Institute of Nuclear and New Energy Technology，Tsinghua University，Beijing 100084，China
2 Collaborative Innovation Center of Advanced Nuclear Energy Technology，Tsinghua University，
Beijing 100084，China
3 Key Laboratory of Advanced Reactor Engineering and Safety of Ministry of Education，Beijing 100084，China
4 Tongfang Industrial Co.，Ltd，Beijing 100083，China

［Objectives］ Design shock spectra are used as crucial inputs for shock response analysis of equipment
Abstract：
mounted on submarines. In order to study the spectrum dip characteristics of installed large-mass shipboard
equipment，［Methods］through finite element numerical simulation of the submarine cabin structure，this paper，
using a submarine as an example，investigates the variation of shock spectrum with mass and stiffness of shipboard
equipment in different connection forms. The spectra obtained are then compared to that of the widely used national
military standards GJB1060.1-91.［Results］The results show that for single-DOF systems，the spectra specified in
GJB1060.1-91 standards are more conservative in low and high frequency ranges (i.e. displacement and acceleration
zone respectively)，while in the medium frequency ranges (velocity zone)，the shock response spectrum derived from
the current research is greater than that given in GJB standard. For multi-DOF systems，ignoring the interaction effect
between various orders of modes following the national military standard leads to conservative calculation results. In
addition， the connection between equipment and cabin have a strong effect on response of the equipment.
［Conclusions］The current study has obtained the influence of various factors on the spectrum dip effect，which may
be significant in the anti-shock design of massive equipment mounted on submarines.
Key words：submarine；shock response analysis；shock spectrum；spectrum dip
CLC number: U661.4

ment due to its small calculation amount and accessi⁃

0 Introduction bility. Design shock spectrum is the input condition
The impact of explosion shock loading is an impor⁃ of the shock resistance analysis based on the method
tant factor leading to the failure of shipboard equip⁃ of response spectrum, which determines the calcula⁃
ment, and the analysis and evaluation of equipment tion accuracy of shock resistance[2]. In the 1960s, the
shock resistance is an important part of the design[1]. envelope spectrum similar in earthquake engineering
The numerical calculation method of shock resis⁃ is used by researchers to calculate the shock load of
tance based on shock response spectrum is extensively the shipboard equipment, which is enough to cause
applied in the design analysis of shipboard equip⁃ damage to the submarine structure, but the equip⁃

Received：2018 - 03 - 06
Supported by: National Ministries Foundation
Author(s): Wang Xiaoxin, female, born in 1984, Ph.D., assistant research fellow. Research interests: structural mechanics in reac⁃
tor technology, shock dynamics. E-mail: [email protected]
Li Xiaotian, female, born in 1970, Ph.D., professor. Research interests: structural mechanics in reactor technology,
shock dynamics. E-mail: [email protected]
Ma Xiaohui, female, born in 1988, master , engineer. Research interests: structural mechanics in reactor technology,
shock dynamics. E-mail: [email protected]
Du Guowei, male, born in 1991, Ph.D. candidate. Research interest: shock dynamics. E-mail: [email protected]
*Corresponding author：Wang Xiaoxin
downloaded from www.ship-research.com
 22 CHINESE JOURNAL OF SHIP RESEARCH，VOL.14，NO.3，JUN 2019

ment remains intact in the on-board test. It can be hull frame. In the figure, M and K are the mass and
seen that there is a large difference between calcula⁃ support stiffness of the equipment respectively, and
tion results and experimental results, which can pro⁃ A is the connection point between the equipment
mote the researchers to delve into the shock re⁃ and the hull.
sponse spectrum of submarine. O'Hara et al. [3-4] pro⁃
posed the concept of spectrum dip, pointing out that
the interaction between shipboard equipment and
the submarine can reduce the response spectrum at
the natural frequency of the equipment by about one M

order of magnitude. Based on this theory, Remmers K

et al. [5] proposed the DDAM (Dynamic Design Analy⁃ A

sis Method) spectrum analysis method for submarine

Fig.1 Illustration of numerical model for spectrum dip effect
design, which is still widely used in the field of ship⁃ analysis
board equipment design. The corresponding research
on the spectrum dip effect has been carried out all The three-dimensional model of the submarine
over the world [6-8]. The shock design response spec⁃ cabin structure is established by ABAQUS finite ele⁃
trum considering the spectrum dip effect is given in ment analysis software, and the numerical calcula⁃
the DDAM method, and the spectrum value is relat⁃ tion of the shock dynamic response is carried out.
ed to the equipment mass. The current code for de⁃ Fig. 2(a) shows the finite element model of the sub⁃
sign of submarine equipment shock follows the na⁃ marine cabin structure, and Fig. 2(b) shows the con⁃
tional military standard GJB 1060.1-91[9] , and its de⁃ nection method of the single-degree-of-freedom
sign response spectrum refers to the response spec⁃ (SDOF) equipment and the hull. In the figures, the
trum in DDAM. hull and the frame are respectively simulated by the
Although scholars in China have carried out re⁃ shell element and the beam element. The equipment
searches on the shock resistance design spectrum [10-11] , is simulated by concentrated mass, and the equip⁃
the derivation process, the source of parameters and ment and the hull are connected by a spring having a
the application scope (e.g. , submarine mass and
form, equipment connection) of the response spec⁃
trum formula are not well understood. Therefore, it is
necessary to conduct an in-depth study on the gener⁃
RP-2
ation law of submarine spectrum dip effect. For sub⁃
RP-1
marine with specific hull form, this paper will study
the interaction between the submarine with specific
hull form and the equipment with different mass and
frequencies under different connection modes, thus y

to obtain the design response spectrum considering z x

the spectrum dip effect, and to discuss the applicabil⁃ （a）Hull model
ity of the DDAM method and the response spectrum
of national military standard.

1 Calculation model
1.1 Finite element model
RP-1
This paper considers the interaction between the
submarine equipment and the hull, and analyzes the
shock dynamic response of the typical submarine
cabin structure. Fig. 1 shows the simplified mass
y
spring structure used to study the "spectrum dip ef⁃
fect", which is adopted to simulate the equipment z x

connected to the hull and the influence of its sup⁃ （b）Connection mode of SDOF equipment

port, wherein the equipment is connected with the Fig.2 Finite element model for spectrum dip effect analysis
downloaded from www.ship-research.com
 Wang X X, et al. Numerical analysis on spectrum dip characteristics of shock response spectrum of submarine equipment 23

degree of freedom in the vertical direction. The natu⁃ 200

ral frequency of the hull at the support position of

the equipment is about 102 Hz. The shock accelera⁃ 150

tion time history acts on both ends of the cabin mod⁃

Acceleration/g
el along the vertical direction (y). The acceleration 100

response time history of point A in 0.5 s is obtained

and the shock response spectrum is generated by cal⁃ 50
culating the dynamic response of the model in the
time domain. 0
0 0.01 0.02 0.03 0.04
1.2 Shock input Time/s

The shock design spectrum of submarine cabin （b）Positive and negative triangular wave
acceleration time history curve
adopts a typical trilinear spectrum, and the speed
Fig.3 Inputs for shock response analysis of the submarine cabin
zone and the acceleration zone adopt the shock spec⁃
trum in the national military standard. Specifically,
2 Spectrum dip law of SDOF system
the shock acceleration design value A 0 of the sub⁃
marine equipment and the shock speed design value 2.1 Influence of equipment mass and
V 0 are respectively calculated by the following for⁃ frequency on spectrum dip
mulas:
217.73 + M a The finite element model in Section 1.2 is used to
A 0 = 102.02 ´
9.07 + M a
（1） calculate the influence of additional SDOF systems
217.73 + M a of different mass M and natural frequencies f e on
V 0 = 0.51 ´ （2） the shock dynamic response of submarine cabin
45.36 + M a
where M a is the modal effective mass of the addi⁃ structure. For SDOF systems, M a = M in Eq. (1)
tional equipment of the cabin. With M a = 0, A 0 and and Eq. (2), so the SDOF mass is recorded as M a .
V 0 should be the shock design values without consid⁃ In the calculation, M a is taken as 1, 5, 10, 50, 100,
ering the mass of the additional equipment, which 150, 200, 250 and 300 t respectively. With the
are used as the shock design values of the cabin, spring stiffness adjusted, the natural frequency of the
namely, A 0 = 2 450 m/s2, V 0 = 2.45 m/s. The dis⁃ SDOF system is 2.5, 5, 10, 25, 50, 100, 120, 200,
500 and 1 000 Hz respectively. Based on the above
placement spectrum value U 0 of the cabin shock de⁃
values, the equipment support system with different
sign spectrum is taken as 0.04 m. Fig. 3(a) shows the
stiffness on the hull is simulated in a total of 90
shock design spectrum of the cabin.
kinds of mass-frequency combinations. SDOF sys⁃
The positive and negative triangular wave acceler⁃
tem is connected with a frame between the midpoints
ation time history curves shown in Fig. 3(b) corre⁃
of the bow and stern of the hull. Fig. 4 shows the cal⁃
sponding to the trilinear spectrum in Fig. 3(a) are ob⁃
culation results of the shock response spectra at the
tained on the basis of the BV 043/85 Shock Resis⁃
connection point A between the equipment and the
tance Specification for Bundeswehr Ships[12].
cabin under a typical natural frequency of the equip⁃
101 1
00 ment.
cm cm
0g
g cm
10 1 0.
1 As the mass of the equipment increases, the differ⁃
ence between the calculation result of the response
100 10 spectrum and the response spectrum without equip⁃
Velocity/（m·s-1）

0g
ment increases in various frequency ranges, and the
response spectrum value at the natural frequency f e
10-1 10
decreases gradually. The response spectrum corre⁃
g
sponding to the frequency is the design acceleration
value used to analyze the stress of the equipment by
the response spectrum method. Therefore, if the re⁃
10-2
100 101 102 103
sponse spectrum without equipment is adopted with⁃
Frequency/Hz out considering the interaction between the equip⁃
（a）Shock design spectrum ment and the submarine cabin, the acceleration will
downloaded from www.ship-research.com
 24 CHINESE JOURNAL OF SHIP RESEARCH，VOL.14，NO.3，JUN 2019

102 101 1
00
cm cm cm cm cm 0g cm
0 10 1 10 0.
1 cm
10 10 1 0.
1
00
0g
101
100 10
Velocity/（m·s-1）

Velocity/（m·s-1）
10 0g
000
00
gg
100

10 Ma=1 t
0g Ma=5 t
f=50 Hz 10-1 Ma=10 t 10
Ma=50 t g
10-1 Ma=1 t Ma=100 t
Ma=10 t Ma=150 t
Ma=50 t 10 Ma=200 t
Ma=100 t g Ma=250 t
Ma=300 t Ma=300 t
10-2 10-2 0
100 101 102 103 104 10 101 102 103
Frequency/Hz Frequency/Hz
（a） f e = 50 Hz
Fig.5 Calculation results of the shock spectrum value for
102 equipment with different combinations of mass and
cm cm cm cm
0 1 frequency
10 10 1 10 0.
00
0g
101 ually increased; the spectrum values of the response
spectrum at each frequency are gradually reduced;
Velocity/（m·s-1）

1
00
0g
10 0
1 and the spectrum dip effect is significant.
00
10
0g It can be seen from Fig. 5 that in the response
0g
f=200 Hz spectrum considering the spectrum dip effect, the
10-1 Ma=1 t
Ma=10 t low frequency ranges is an obvious equal-displace⁃
Ma=50 t 10
Ma=100 t g ment zone, and in the high frequency ranges, the ac⁃
Ma=300 t
10 -2
celeration does not substantially change with the fre⁃
100 101 102 103 104
Frequency/Hz quency, namely that it is the equal-acceleration
（b） f e = 200 Hz zone. Although there is an equal-velocity zone be⁃
Fig.4 Calculation results of the shock response spectra tween the two, the corresponding interval is relative⁃
ly small. Fig. 6(a) shows the variation of acceleration
be too large, resulting in a more conservative stress
spectrum value of the equal-acceleration zone with
calculation result. This shows that the influence of
the mass of equipment. In the figure, the dotted line
the interaction between the hull and the equipment,
is the linearly fitted curve of the acceleration in loga⁃
namely, the "spectrum dip effect", should be consid⁃
rithmic coordinates; A0 = 286.92 M a -0.575; and the lin⁃
ered in the analysis of the stress of the submarine
equipment using the response spectrum method. Af⁃ ear degree of fitting is r2 = 0.992 8. In addition to the
ter the submarine is installed with large-scale equip⁃ case of small equipment mass ( M a < 2 t), the calcu⁃
ment such as reactors, the interaction between the lated acceleration spectrum value is smaller than the
hull and the installed equipment will be very strong. spectrum value given by the national military stan⁃
If the spectrum dip effect is ignored in the calcula⁃ dard. When M a = 5-300 t, the relative deviation
tion, the result may be completely inconsistent with between the acceleration spectrum value given by
the true response and stress level. the national military standard and the acceleration
Fig. 5 shows the spectrum value at f e of the re⁃ spectrum value of the fitted spectrum is 30%-40% .
sponse spectrum curve of the connection point A be⁃ Fig. 6(b) shows the variation of the velocity spectrum
tween the submarine cabin and the equipment with value in the equal-velocity zone with the mass of the
different mass-frequency combinations. For different equipment. In the figure, the dotted line is the linear⁃
equipment mass, the calculated results at each fre⁃ ly fitted curve of the speed in logarithmic coordi⁃
quency can constitute a response spectrum curve. nates; V0 = 3.928 7 M a -0.219; and the linear degree of
When M a = 1 t, the influence of installed equipment fitting is r2 = 0.959 3. For equipment with various
on the submarine cabin structure is small, and the re⁃ mass, the calculated velocity spectrum value is great⁃
sponse spectrum curve is basically coincident with er than that given by the national military standard,
the design response spectrum without considering and the maximum relative deviation is about 60%.
the influence of the equipment. With the increase of It can be found from the analysis results that there
M a , the influence on the installed equipment is grad⁃ is no obvious influence of the spectrum dip effect in
downloaded from www.ship-research.com
 Wang X X, et al. Numerical analysis on spectrum dip characteristics of shock response spectrum of submarine equipment 25

103 101 11
00
Calculated results cm cm
0g cm
10
g 1
Linear fitting 1 0.
GJB 1060.1-91

102 100 10

Velocity/（m·s-1）
0g
Acceleration/g

101 10-1 10
g

Response spectrum in this paper

GJB 1060.1-91
100 0 10 -2

10 101 102 100 101 102 103

M a /t Frequency/Hz
（a）Acceleration spectrum （a） M a =5 t
101
101 1
Calculated results 00
cm cm 0g cm
Linear fitting 10 1 0.
1
GJB 1060.1-91
Velocity/（m·s-1）

100 10
Velocity/（m·s-1）
0g
10 0

10-1 10
g

Response spectrum in this paper

10-1 0 GJB 1060.1-91
10 10 1
10 2 10-2 0
10 101 102 103
M a /t
Frequency/Hz
（b）Speed spectrum （b） M a =50 t
Fig.6 Variation of the calculated spectrum values with Fig.7 Comparison between the current calculated spectrum
equipment's mass values and the GJB 1060.1-91

the equal-displacement zone, and the shock re⁃ shock design spectrum, while the shock design spec⁃
sponse spectrum of equipment with different mass trum given by the national military standard adopts
has little change. Therefore, the influence of the the equal-velocity assumption in the medium and
spectrum dip effect is ignored in the equal-displace⁃ low frequency ranges. Therefore, the response spec⁃
mentzone, and the displacement spectrum value U 0 trum calculated in the low frequency ranges is small⁃
is conservatively taken as 0.04 m. er than that given by the national military standard.
In summary, for the hull form of the submarine The response spectrum in this paper is smaller in
studied in this paper, the shock acceleration design the equal-acceleration zone and larger in the
value A0, the shock speed design value V0 and the dis⁃ equal-velocity zone than that given by the national
placement spectrum value U 0 can be obtained con⁃ military standard, so it can be seen from the spec⁃
trum type that the velocity zone of response spectrum
sidering the influence of the spectrum dip effect as
given in this paper is shorter, and the interval of the
follows: A0 = 286.92 M a -0.575, V0 = 3.928 7 M a -0.219,
equal-velocity zone is further reduced as the mass of
and U 0 = 0.04 m, respectively.
the equipment increases.
2.2 Comparison with shock design spec- According to the above analysis, for the submarine
trum of national military standard with specific shape studied in this paper, the design
response spectrum given by the national military
Fig. 7 shows the comparison between the shock de⁃ standard is relatively large in the high frequency
sign spectrum calculated by this paper and the shock ranges and small in the medium frequency ranges.
design spectrum given by the national military stan⁃ Considering that the frequencies of the shipboard
dard when M a = 5 and 50 t. equipment and its support are likely to be in the me⁃
The equal-displacement zone is considered for dium frequency ranges, it is not necessarily com⁃
the low frequency ranges in the calculation of the pletely conservative to use the design response spec⁃
downloaded from www.ship-research.com
 26 CHINESE JOURNAL OF SHIP RESEARCH，VOL.14，NO.3，JUN 2019

trum given by the national military standard for re⁃ According to the DDAM method [5] , the response
sponse analysis of the equipment, so it is necessary spectrum value of the multi-DOF system consider⁃
to carry out targeted analysis for submarines of a spe⁃ ing the spectrum dip effect is determined by the natu⁃
cific shape. ral frequency f ei and modal effective mass M ai cor⁃
responding to each order of modal. The spectrum val⁃
3 Spectrum dip law of multi-DOF
ue can be considered as the response spectrum value
system obtained when the SDOF system corresponding to
In this section, the calculation results of SDOF each modal acts independently on the submarine
system are extended to multi-DOF systems. Accord⁃ cabin structure. When the equipment is supported
ing to the modal analysis theory, the decoupled mo⁃ on a completely rigid foundation, the result obtained
tion equation of multi-DOF system can be trans⁃ by this method is accurate. But for the shipboard
formed into motion equations of several independent equipment, its interaction with the hull will cause de⁃
SDOF systems. The right side of the diagram in Fig. 8 formation of the hull, so a certain deviation will oc⁃
is the equivalent model of multi-DOF system [5]. In cur if this calculation method is adopted.
the figure, M ai is the effective mass of the ith-order The red, green and blue curves shown in Fig. 9
modal of the left system, and the natural frequency of are the response spectra obtained from the indepen⁃
the spring mass system composed of M ai and Kfi is dent connection of submarine cabin structure with
the natural frequency f ei of the ith order of the left the SDOF system with the natural frequency of fei
and the mass of M ai (i = 1, 2, 3). The response spec⁃
system. The effect of the two systems on the founda⁃
tion is equivalent. Table 1 shows the mass and stiff⁃ trum value at f ei is the design response spectrum
ness characteristics of the two typical equipment con⁃ value calculated for DDAM spectral analysis and is
sidered in this section (equipment CA of medium and marked with the corresponding colored dot, which
low frequency ranges and equipment CB of high fre⁃ can be calculated by Eq. (1). The black solid line is
quency ranges). And Table 2 shows the natural fre⁃
101 1
quencies and modal effective mass of the multi-DOF cm
00
cm
cm 0g
1
model, so that the equivalent SDOF system in Fig. 8 10 18 Hz 1 0.
29 Hz
can be constructed. 41 Hz
100 10
Velocity/（m·s-1）

0g
M3

K3

M2 M a1 M a2 M a3
10-1 10
g
K2 Kf1 Kf 2 Kf 3
Multi-DOF
M1 Single DOF M a1
Single DOF M a2
K1 Single DOF M a3
10-2 0
10 101 102 103
Frequency/Hz
Fig.8 Schematic diagram of decoupling of a multi-DOF system
（a）Equipment CA
Table 1 Mass and spring stiffness of the multi-DOF model 101 1
87 Hz 00
cm cm 140 Hz 0g cm
10 1
Equipment Equipment mass M/t Spring stiffness K/（MN·m ） 1 0.
-1

197 Hz
model M1 M2 M3 K1 K2 K3
CA 50 10 5 1 603.8 370.1 123.4 100 10
Velocity/（m·s-1）

0g
CB 50 10 5 36 952 8 527.3 2 842.4

Table 2 Natural frequencies and modal effective mass of

the multi-DOF model 10-1 10
g
Equipment model Order i f ei /Hz M ai /kg
Multi-DOF
1 18.132 32 948 Single DOF M a1
Single DOF M a2
CA 2 29.282 29 273 Single DOF M a3
10-2 0
3 41.095 2 779 10 101 102 103
Frequency/Hz
1 87.034 32 947
（b）Equipment CB
CB 2 140.55 29 274
Fig.9 Comparison between the calculation results of multi-DOF
3 197.26 2 779 models and equivalent single DOF models
downloaded from www.ship-research.com
 Wang X X, et al. Numerical analysis on spectrum dip characteristics of shock response spectrum of submarine equipment 27

the response spectrum of the connection point A cal⁃ sults. In some cases, the deviation of the DDAM
culated and generated from the direct time-history method is relatively large, and the calculation is too
response of multi-DOF system in the left side of conservative. In order to obtain more accurate calcu⁃
Fig. 8, which can be considered as the exact solution lation results, it is necessary to establish an overall
of the response spectrum input by the response spec⁃ model for the cabin structure and equipment to ob⁃
trum method of multi-DOF system. tain the true response spectrum at the support posi⁃
It can be seen from the comparison of the black tion of the equipment for subsequent calculations.
and colored dots in Fig. 9 that there is a certain devi⁃
ation in the response spectrum values between the
4 Influence of connection mode
multi-DOF system and the equivalent SDOF system, between equipment and cabin
so the two systems are not completely equivalent. structure
This is because the deformation of the hull under the
In the above study on the interaction of the SDOF
simultaneous action of multi-DOF is different from
system and the multi-DOF system with the cabin
that of SDOF, and the acceleration response of the
structure, the connection mode between the equip⁃
equipment is also different. In this case, the modals
ment and the hull is simplified into single-point con⁃
of the multi-DOF system have mutual influence and
nection. In actual situations, the equipment support
cannot be simply decoupled. Moreover, the current
and the hull are connected in different methods. For
DDAM method also does not consider the effect of
example, there are large connection areas between
multi-modal interaction caused by deformation of
large equipment such as reactors and cabin struc⁃
the hull on the response spectrum.
tures, and such connection methods are bound to
From the calculation results of the low-frequency
have a certain impact on the dynamic response of the
and high-frequency equipment in Fig. 9, it can be
cabin. This section considers the influence of differ⁃
found that the acceleration response of the SDOF
ent connection areas between the equipment and the
model may be too large or too small, which is related
hull on the shock response spectrum through four
to the natural frequency and the modal effective
models (Model 1-Model 4). The equipment in these
mass of the equipment, so the influence law is more
models is two-DOF system, in which the equipment
complicated. Because the low-order modal effective
mass and the natural frequencies are M a1 = = 50 t,
mass is relatively large, the low-order modal has a
greater influence on the high-order modal, and the f e1 = 30 Hz and M a2 = 15 t, f e2 = 50 Hz, respective⁃
high-order modal has larger deviations in spectrum ly. The equipment in Model 1 is connected to the
value. Under the influence of the low-order modal, hull by means of point connection, while the equip⁃
the spectrum dip effect at high frequency is more ob⁃ ment in Model 2-Model 4 adopts skirt support. The
vious, so that the response spectrum value is re⁃ connection radii R to the hull are 0.5, 1.0 and 1.5 m,
duced. It can be seen from Fig. 9 that the maximum respectively.
deviation is about 5 times when the interaction be⁃ Fig. 10 shows the calculation results of response
tween modals in high-order modal is considered and spectrum at the equipment support position of the
not considered. And because the high-order modal four models with different connections. It can be
effective mass is small, the effect on the low-order seen from the figure that as the connection radius be⁃
modal is relatively small. After the calculation re⁃ tween the equipment and the hull increases, the stiff⁃
sults of all orders of the modal are effectively com⁃ ness of the joint is enhanced, and the first-order fre⁃
bined, the accelerations of the mass points of the sys⁃ quency (the frequency corresponding to the first
tem on the left side of Fig. 8 can be obtained. Com⁃ peak of the response spectrum curve) of the system
pared with the calculation result of the multi-DOF constituted by hull and equipment is increased. The
model, the calculation result of the SDOF decoupling values of the response spectrum curve at the natural
model is too large, and the maximum deviation is frequencies f e1 = 30 Hz and f e2 = 50 Hz are the de⁃
about 2 times for the equipment calculated in this sign values of the shock acceleration of the equip⁃
section. ment. As the connection radius between the equip⁃
According to the above results, the DDAM method ment and the hull increases, the response spectrum
neglecting the interaction between modals causes a values at the above-mentioned frequencies increase.
certain deviation from the real response spectrum, re⁃ When the connection radius increases from 0.5 to
sulting in conservative dynamic response analysis re⁃ 1.5 m, the acceleration response spectrum increases
downloaded from www.ship-research.com
 28 CHINESE JOURNAL OF SHIP RESEARCH，VOL.14，NO.3，JUN 2019

101 1
00 servative and can meet the engineering require⁃
cm 30 Hz cm 0g cm
10 1 0 .1 ments. However, in some special cases, ignoring the
50 Hz
interaction between modals may result in too conser⁃
100 10 vative equipment design and too large cost.
Velocity/（m·s-1）

0g
3) For the influence of the equipment support form
on the response spectrum, the interaction between
10-1
the equipment and the hull will change under differ⁃
10
g ent connection modes, and the local stiffness of the
Model 1（R=0 m） hull at the support will change as well. The connec⁃
Model 2（R=0.5 m）
Model 3（R=1.0 m）
Model 4（R=1.5 m） tion radius also has a significant effect on the re⁃
10-2
101 101 102 103 sponse spectrum of the equipment. As the connec⁃
Frequency/Hz tion area increases, the spectrum dip effect is weak⁃
Fig.10 Calculation results of the models with different ened.
connections
In summary, the response spectrum of equipment
by 1.5 times, because both the influence of equip⁃ is affected by many factors. In engineering design, it
ment mass on the hull and the spectrum dip effect is necessary to carry out targeted analysis in the se⁃
are weakened after the increase in the stiffness of lection of the shock design response spectrum to ob⁃
connection between the hull and the equipment. tain a reliable shock response for the submarine of
According to the above analysis results, the con⁃ different shapes. Due to the limitations of the DDAM
nection mode between the equipment and the hull method and its design response spectrum, for
has a significant impact on the design response spec⁃ large-scale equipment with high precision require⁃
trum. The design spectrum of the DDAM method can⁃ ments, during the dynamic response analysis, it is
not cover the influence of various support modes and necessary to perform coupling analysis on the overall
the connection area. For large equipment with a model of the hull and equipment to fully consider the
large connection area, it is necessary to establish a mutual interaction between the hull and the equip⁃
cabin-equipment coupling model to fully consider ment.
the interaction between the equipment support and The simplified finite element model established in
the hull, thus to obtain accurate and reliable calcula⁃ this paper is only the cabin model, and the shock
tion results of dynamic response. load is input in the form of acceleration, which may
be different from the real response of the whole sub⁃
5 Conclusions marine under the impact of underwater explosion
In this paper, the dynamic response characteris⁃ load.
tics of submarine after the installation of large-mass In the next step, the fluid-solid coupling of the
equipment are analyzed; a simplified finite element submarine-equipment model under the impact of un⁃
model of cabin-equipment is established; and the in⁃ derwater explosion load will be precisely calculated
fluence of different factors on spectrum dip charac⁃ to more accurately simulate the coupling between the
teristics is studied and analyzed through numerical submarine and the hull after the installation of
calculation, obtaining the following conclusions: large-mass equipment, and the influence law and
1) For the hull and SDOF equipment model, the functional mechanism of spectrum dip characteris⁃
shock spectrum given by the national military stan⁃ tics will be further investigated on this basis.
dard is more conservative in low frequency ranges
(displacement zone) and high frequency ranges
References
(acceleration zone), and the response spectrum given ［1］ Liu J H，Zhou X T，Pan J Q，et al. The state analysis
by the national military standard is small in the medi⁃ and technical development routes for the anti-explo⁃
um frequency zone (velocity zone). sion and shock technology of naval ships［J］. Chinese
Journal of Ship Research，2016，11（1）：46-56，71
2) For the hull and multi-DOF equipment model,
（in Chinese）.
the currently used response spectrum method ig⁃
［2］ Edward A J. Shock response spectrum［J］. Sound and
nores the interaction between modals and has a cer⁃ Vibration，2009，43（6）：6-15.
tain deviation from the true spectrum dip characteris⁃ ［3］ O'Hara G J. Shock spectra and design shock spectra：
tics of multi-DOF system. In most cases, the process⁃ Report No 5386 ［R］. Washington DC： Naval Re⁃
ing method of national military standard is more con⁃ search Laboratory，1959.
downloaded from www.ship-research.com
 Wang X X, et al. Numerical analysis on spectrum dip characteristics of shock response spectrum of submarine equipment 29

［4］ O'Hara G J，Cunniff P F. The shock spectrum dip ef⁃ ［9］ Industry Committee of National Defense Science and
fect［J］. Journal of Sound and Vibration，1985，103 Technology. General requirement for environmental
（3）：311-321. conditions of naval ships-mechanical environments：
［5］ Remmers G M，O'Hara G J，Cunniff P F. Dynamic GJB1060.1-91 ［S］. Beijing：China Standard Press，
design analysis method DDAM［J］. Shock and Vibra⁃ 1991（in Chinese）.
tion，1996，3（6）：461-476. ［10］ Zhang X Y，Liu J H，Pan J Q，et al. Review on an⁃
［6］ Scavuzzo R J，Hill G D，Saxe P W. The spectrum dip ti-shock criteria for equipments in some primary navy
of deck mounted systems［J］. Shock and Vibration， countries［J］. Journal of Ship Mechanics，2011，15
2010，17（1）：55-69. （11）：1322-1334（in Chinese）.
［7］ Scavuzzo R J，Hill G D，Saxe P. The "spectrum dip"： ［11］ Huang R B，Du J Y，Wang G X. Study on shock crite⁃
dynamic interaction of system components［J］. Journal ria for naval vessels of USA［J］. Ship Science and
of Pressure Vessel Technology，2015，137（4）：044701. Technology，2007，29（2）：138-142（in Chinese）.
［8］ Yu B Y，Yan M，Zhang L，et al. Influence of spec⁃ ［12］ German Federal Office for Military Technology and
trum dip effect on the equipment damage under Procurement. Shock resistance specification for
non-contact underwater explosion［J］. Noise and Vi⁃ Bundeswehr Ships：BV 043/85［S］.［S.1.： s.n.］，
1985.
bration Control，2016，36（5）：6-10（in Chinese）.

潜艇设备冲击响应谱谱跌特性数值分析
王晓欣 *1，2，3，李笑天 1，2，3，马笑辉 4，杜国伟 1，2，3
1 清华大学 核能与新能源技术研究院，北京 100084
2 清华大学 先进核能技术协同创新中心，北京 100084
3 清华大学 先进反应堆工程与安全教育部重点实验室，北京 100084
4 同方工业有限公司，北京 100083

摘 要：
［目的
目的］设计冲击谱是舰载设备响应谱抗冲击分析的关键输入条件。为了研究大质量舰载设备安装后
的谱跌特性，
［方法
方法］以潜艇为例，通过对舱段结构的有限元数值模拟，研究舰载设备在不同连接方式下冲击谱
随质量及刚度变化的规律，并与目前广泛采用的国军标 GJB 1060.1-91 给出的冲击设计谱进行对比。
［结果
结果］结
果表明：对于单自由度系统，在低频段（位移段）和高频段（加速度段），国军标给出的冲击谱更保守；在中频段
（速度段），研究得到的响应谱大于国军标给出的响应谱。对于多自由度系统，依据国军标给出的方法忽略模态
间的相互作用将导致计算偏于保守。此外，设备与舱段的连接方式对设备响应谱存在明显影响。
［结论
结论］研究得
到的各种因素对谱跌特性的影响规律，对潜艇大型设备的抗冲击设计具有借鉴意义。
关键词：潜艇；冲击响应分析；冲击谱；谱跌

downloaded from www.ship-research.com