Calculation Of a Shock Response Spectrum

Ji Tma

Petr Ko

Faculty of Mechanical Engineering

VSB Technical University of Ostrava
Ostrava Poruba, Czech Republic
[email protected]

Faculty of Mechanical Engineering

VSB Technical University of Ostrava
Ostrava Poruba, Czech Republic
[email protected]

Abstract As it is stated in the ISO 18431-4 Standard, a Shock

Response Spectrum is defined as the response to a given
acceleration acting at a set of mass-damper-spring oscillators,
which are adjusted to the different resonance frequencies while
their resonance gains (Q-factor) are equal to the same value.
The maximum of the absolute value of the calculated responses
as a function of the resonance frequencies compose the shock
response spectrum (SRS). The paper will deal with employing
Signal Analyzer, the software for signal processing, for
calculation of the SRS. The theory is illustrated by examples.
Keywords- shock response spectrum, shock, single degree of
freedom system, measurements, ramp invariant method.

I.

INTRODUCTION

Shock Response Spectrum (SRS) analysis was developed

as a standard data processing method in the early 1960s.
SRS was used by the U.S. Department of Defense. Now this
signal processing method is standardized by ISO 18431-4
Mechanical vibration and shock Signal processing Part
4: Shock response spectrum analysis [1].
Let it be assumed that many small instruments or other
parts as substructures (bodies) are mounted to the base
structure (baseplate) of a mechanical system [2]. The
mounting elements are flexible and could be described by
stiffness and damping parameters. The mass of the
mentioned substructure creates the mechanical oscillator,
which is a single-degree-of-freedom (SDOF) system. Now,
we suppose that a vibration transient is excited on the base
structure as a consequence of either its environment or
normal operation. Many sources of excitation could be
described: Drop impact during handling, explosive bolts on

aerospace structures, bolted joints suddenly opening and

closing with an impact, reciprocating engine fuel explosions
inside cylinders, etc.. For example we could assume that
vibration transient results from the dropping of a mechanical
system onto a concrete surface while producing an impact
load on one corner of the base structure. The impulse force
acting to substructures excites lightly damped oscillations.
The peak value of the substructure acceleration may cause
destruction due to the large inertia force. It is required that
the mentioned substructures, in fact equipments and
instruments, have to survive mechanical shocks. The impulse
force and shock exposure have to be analyzed.
Two main parameters of the SDOF system, which are
specifying free oscillation, are the substructure natural
frequency and damping. There are many measures of
describing damping. This paper prefers the damping ratio
and the quality factor Q = 1 (2 ) , which is approximately
equal to a resonance gain of the lightly damped systems. For
a normalized value of the quality factor Q the substructure
natural frequency f n plays the key role. In this case the
substructure natural frequency determines a peak
acceleration, which is excited by an arbitrary transient
acceleration input, such as shock or impulse force load. The
dependence of the acceleration peak value (maximum or
minimum) on the substructure natural frequency in the form
of a diagram is called a shock response spectrum (SRS) of
the impulse force. SRS of the impulse force or acceleration
forecasts the excited substructure acceleration in dependence
on the substructure resonance frequency.
As it was stated above, SRS assumes that the impulse
force is applied as a common base input (base plate) to a

Natural frequency of the SDOF systems

fn1

< fn2

<

fn3

< fn4

fnK

Response

Resonant gain Q = 10

Input impulse force

Figure 1. Group of single-degree-of-freedom systems

c
978-1-61284-361-2/11/$26.00 2011
IEEE

404

group of independent hypothetical SDOF systems (bodies),

which are only different in the natural frequency, see Fig. 1.
Damping ratio is typically fixed at a constant value, such as
5%, which is equivalent to the quality factor Q, which equals
to 10. An example of the vibration amplitude of the SDOF
system as a response to the input impulse force is
demonstrated in Fig. 1. The peak values of the amplitude
determine SRS of the associated impulse force. In a general
sense SRS allows to compare the dynamic effects of various
shock sources through their effect on the set of the SDOF
systems.
II.

SINGLE-DEGREE- OF-FREEDOM MODEL

The equation of motion, which is describing oscillation

of the single-degree-of-freedom (SDOF) system, shown in
Fig. 2, is written in the form
m x2 = c(x 2 x1 ) k ( x2 x1 ) .

(1)

where x1 is a base structure (baseplate) displacement and x2

is a response displacement of a single-degree-of-freedom
system (body), m, c, k are mass, damping parameter and
stiffness, respectively. The corresponding velocity as a time
function is designated by v1 and v2 and acceleration is
designated by a1 and a2 .
x2
Substructure
m

Alternatively the Laplace transfer function relating the

substructure mass velocity to the base velocity and the
substructure mass acceleration to the base acceleration are as
follows

G (s ) =

V2 (s ) A2 (s )
cs + k
.
=
=
2
V1 (s ) A1 (s ) ms + cs + k

(5)

All the transfer functions are of the same form.

Introducing the SDOF system natural frequency f n , Q
factor (resonance gain) or damping ratio (fraction of
critical damping), we obtain
fn =

1
2

k
, Q=
m

1
km
c
.
=
, =
2Q 2 km
c

(6)

The Laplace transfer function relating the base

acceleration to the substructure mass acceleration could be
written in the form

A2
=
A1

n s
+ n2
R1
R*
Q
=
+ 1 * .
s
s s1 s s1
s 2 + n + n2
Q

(7)

where R1 is a residue, s1 is a pole and R1* , s1* .are

corresponding complex conjugate quantities.
For a given quality factor Q , natural angular frequency
x1
c
k
Base
and base structure acceleration a1 in the form of the time
function, it is theoretically possible to calculate the
structure
acceleration response a2 in the form of the time function as
Figure 2. SDOF, a body (substructure) flexibly mounted on a baseplate
well and consequently to determine the peak value (either
(base structure)
maximum or minimum)) of this time function. The problem
is that the input acceleration is in the form of the sampled
The Laplace transfer function relating the substructure
signal with the sampling interval TS = 1 f S , not in the form
mass displacement to the base displacement is as follows
of a continuous signal. The transfer function (7) has to be
approximated by the Z-transform function, i.e. to be
X 2 (s )
cs + k
G (s ) =
=
.
(2) transformed into a discrete system (digital filter).
X (s ) ms 2 + cs + k
1

III. DISCRETE APPROXIMATION OF A

CONTINUOUS TRANSFER FUNCTION
This part of the paper deals with a mapping of the s-plane
(continuous time) to the z-plane (discrete time) saving the
important properties of the continuous system after
transformation into a discrete system of the same order. In
V1 (s ) = sX 1 (s ), V2 (s ) = sX 2 (s ),
.
(3) the case of the transfer functions (2) and (5), which are
identical, it concerns the resonance frequency and resonance
A1 (s ) = sV1 (s ), A2 (s ) = sV2 (s )
gain. The approximation of the differential equation by the
difference equation can be performed by using the numerical
where
integration of the sampled function using trapezoidal rule.
This method results in the bilinear transform, which distorts
L{x1 (t )} = X 1 (s ) , L{v1 (t )} = V1 (s ) , L{a1 (t )} = A1 (s )
. (4) the frequency due to the fact that the infinity frequency
L{x2 (t )} = X 2 (s ) , L{v2 (t )} = V2 (s ) , L{a2 (t )} = A2 (s )
interval of the continuous system is mapped onto the finite
frequency interval of the discrete system. To avoid the
Assuming the zero initial condition, the relationship
between the Laplace transform of displacement and velocity
and relationship between the same transform of velocity and
acceleration are as follows

2011 12th International Carpathian Control Conference (ICCC)

405

frequency distortion the bilinear transform is extended by a

frequency correction saving the identity of the transfer
function for the resonance frequency
G * (z ) = G (s )

s=

2 f p

1 z 1

f p 1+ z 1

tan

fS

(8)

1
.
s+a

(9)

of the same form as (7) (except a residue) by the z-plane

transfer function
G1 (z ) =

10 + 11 z 1
.
1 + 11 z 1

(10)

The parameter a determines the pole (-a) of transfer

function (9) and parameters 10 , 11 , 11 are coefficients of
the approximating difference equation of the first order. The
impulse response, corresponding to the linear system (9), is
as follows

g1 (t ) = L {G1 (s )} = exp( a t ) .
1

nTS + TS

exp( a(nT

+ TS )) y1 () d =

nTS + TS
nTS

= exp( aTS ) exp( a(nTS )) y1 () d + (...) d

0

nTS

TS

where
f S is the sampling frequency and
f P is the
prewarping frequency, both in Hz. The frequency responses,
before and after mapping, match exactly for the prewarping
frequency. It is therefore desirable to substitute the
prewarping frequency by the resonance frequency.
The approximation, which is preferred by the ISO 184314 Standard, is called the Ramp Invariant Method and was
designed by Ahlin, Magnevall, and Josefsson [3]. This
method is based on the approximation of the s-plane transfer
function
G1 (s ) =

y 2 (nTS + TS ) =

= exp( aTS ) y 2 (nTS ) + exp( aTS ) exp(a u ) y1 (u + nTS ) du

0

(13)

()

The continuous function y1 t in between two adjacent

samples may be approximated either by a constant value or
by a ramp or by any suited function. Undependably of the
approximation of the unknown values of the function in
between samples y1 (nTS ) and y1 (nTS + TS ) , the value of
y2 (nTS + TS ) is computed by using a formula

y2 (nTS + TS ) = exp( aTS ) y2 (nTS ) +

+ 10 y1 (nTS + TS ) + 11 y1 (nTS )

(14)

As it was mentioned before, the approximation by the

ramp is preferred [3]
G1* ( z ) =

10 + 11 z 1
.
1 + exp( aTS ) z 1

(15)

This approximation enables to compute the value of the

parameters 10 and 11 .

y1 (t ) = y1 (nTS ) +

( y1 ((n + 1)TS ) y1 (nTS )) (t n T )

S

(11)

TS

n TS t < (n + 1)TS

. (16)

Now let us return to the kinematic quantities as

displacement, velocity and acceleration, which were
designated by x1 (t ), v1 (t ) and a1 (t ) respectively, for the
base structure and by x2 (t ), v2 (t ) and a2 (t )
for the
substructure. The formula (7) demonstrates the
decomposition of the second order system into two parallel
systems of the first order. The continuous transfer function
(9) can be approximated by the discrete transfer function
t
(15). Therefore, the approximation transfer function of the
y2 (t ) = g1 (t ) y1 () d .
(12) second order system can be composed of two discrete
0
transfer functions (15), but each of them is multiplied by the
corresponding residue. The digital filter corresponding to
The measurement of continuous signals results in a
SDOF system responses is the second order filter of the IIR
sequence of samples in the discrete time t = nTS , where
(Infinite Impulse Response) type, with the general Zn = 0, 1, 2,... For the linear system, which is described by
transform expression
the impulse response (11), the value of the output signal at
the time t = (n + 1)TS could be evaluated by using the
A (z ) 20 + 21 z 1 + 22 z 2
G * ( z) = 2
=
.
(17)
mentioned convolution integral
A1 ( z ) 1 + 21 z 1 + 22 z 2
In order to derive the general formula for a response of
the system with the transfer function (9) the general input
and output signal are introduced. Let y 2 (t ) be an output
signal of a continuous linear system with input y1 (t ) and the
impulse response g1 (t ) can be computed by using the
convolution integral

406

2011 12th International Carpathian Control Conference (ICCC)

where 20 , 21 , 22 , 21 , 22 are parameters of the second

order filter.
The transfer function (17) corresponds to the second
order difference equation relating the samples x1 (nTS ) and
x2 (nTS ) of the base structure and substructure
displacements, respectively

The other transfer functions, which are relating the velocity

or displacement difference to the base structure acceleration,
are shown in table I. The corresponding filter coefficients
could be found in [1].
V.

SOFTWARE TOOLS FOR SRS CALCULATION

There are many software tools for signal analysis, for

example Matlab, which is a high-performance language for
technical computing. The disadvantage is that the high
1
2
frequency sampling is not integrated in the unified

z
+

z
(
)
A
z
21
22
G * ( z) = 2
= 20
.
(18) environment and a user has to import the input data. On the
1
2
A1 (z ) 1 + 21 z + 22 z
other hand there is a software tool, which is specialized only
on signal measurements and recordings without signal
where n is an integer number of the same meaning as in
processing.
formula (13).
Signal Analyzer, the VSB - Technical University of
The displacement samples can be substitute by the
Ostrava indoor software, extended by scripts is an
velocity, v1 (nTS ) and v2 (nTS ) , or acceleration, a1 (nTS )
experiment how to overcome the mentioned disadvantage.
Signal Analyzer is software intended to support laboratory
and a2 (nTS ) , samples.
measurements. The recorded data may be immediately tested
This chapter describes how to approximate the
by an analyzer virtual instrument. The script language is
continuous transfer function by the discrete transfer function.
described in paper [4].
Detailed description of a procedure how to create a formula
for calculation of the filter parameters is omitted. The value
Shock Response Spectrum;
of the digital filter parameters is calculated by formulas
CrLf;
given below, which are tabulated in the ISO 18431-4
ymax=[];
Standard
ff=[];
20 = 1 exp( A) sin( B ) / B .
21 = 2 exp( A)[sin( B) / B cos( B)] .
22 = exp(2 A) exp( A) sin( B ) / B .
21 = 2 exp( A) cos( B ) .

(19)
(20)
(21)
(22)

22 = exp(2 A) .

(23)

where
A=

nTS
1
.
, B = nTS 1
2Q
4Q 2

(24)

IV. TRANSFER FUNCTIONS FOR RELATIVE

VELOCITY AND DISPLACEMENT RESPONSE
TABLE I.

TRANSFER FUNCTIONS

Relative velocity response

spectrum

Relative displacement response

spectrum

V2 V1
ms
=
2
A1
ms + cs + k

X 2 X1
m
=
2
A1
ms + cs + k

X 2 X1
mn
n =
A1
ms 2 + cs + k

mn2
X 2 X1 2
n =
A1
ms 2 + cs + k

For some substructure, the difference of the velocity or

displacement between the base structure and the mentioned
substructure, which is excited by the base structure impulse
load, is more dangerous than the substructure acceleration.

fmin=1;
fmax=1000;
n=90;
qv=(fmax/fmin)^(1/n);
T=1/get(input1,'freq');
Q=10;
for(i=0;i<n;i=i+1)
{
fn=fmin*qv^i;
ff=[ff,fn];
wn=2*pi*fn;
A=wn*T/2/Q;
B=wn*T*sqr(1-1/4/Q/Q);
b0=1-exp(-A)*sin(B)/B;
b1=2*exp(-A)*(sin(B)/B-cos(B));
b2=exp((-2)*A)-exp(-A)*sin(B)/B;
a1=(-2)*exp((-1)*A)*cos(B);
a2=exp((-2)*A);
BB=[b0,b1,b2];
AA=[-a1,-a2];
y=filter(input1,BB,AA);
yy=max(y);
ymax=[ymax,yy];
};
save(ff);
save(ymax);
Figure 3. Signal Analyzer script for calculation of SRS

VI.

HALF-SINE AS AN INPUT FUNCTION

The popular testing function for evaluation of SRS is a

half-sine function as the time history of the input
acceleration. The mentioned testing signal and the result of
calculation of the SDOF response in the form of SRS are
shown in Fig. 4.

2011 12th International Carpathian Control Conference (ICCC)

407

Acceleration response is represented on the y-axis, and

natural frequency of any given hypothetical SDOF is shown
on the x-axis. The y-axis of both the plots is in gs
(acceleration of free fall). Any units of acceleration are
admissible as well. SRS covers frequency over three
decades. 30 responses of the SDOF system in each decade
are calculated for different values of the natural frequency.
The natural frequencies form a geometric sequence with
common ratio 1.080. Therefore SRS in Fig. 4 seems to be a
continuous function. The maximum value of the acceleration
response is reached for the substructure natural frequency
approximately around of 70 Hz. For a given impulse it is
dangerous to tune the resonance frequency of the
substructure, which is attached to the base structure, to the
value in the interval ranging from 30 to 200 Hz due to the
large impact loading if the resonance frequency greater than
200 Hz causes the impact loading to be at the margin of
acceptability.
VII. MEASUREMENT EXAMPLES
An examples of the shock measurements and calculation
SRS is added to demonstrate the described calculation of
SRS. The example shows effect of hammer tip material on
the shock response spectrum. The hammer tip was made
from various materials, such rubber, plastic and aluminum.
The base structure is a steel cylinder of the 1.55 kg mass,
which is hanging on flexible tapes allowing displacement in
axial direction of the cylinder. It is assumed that some
substructure is mounted on the basic structure. An
accelerometer was attached to a cylindrical body, whose
acceleration is measured. Because that the hammer is
intended to excite a mechanical structure for experimental
modal analysis it is equipped with a force transducer. The
hammer hit to the steel cylinder excites both the force and
base structure acceleration.
The time history of the impulse force is shown in Fig. 5
while the corresponding acceleration responses are shown in
Fig. 6. It must be emphasis that the vibration responses are
needed as the input data for computing of the shock response
spectrum in order to assess mechanical loading the
substructures, which are mounted on the base structure,
when a shock occurs. The mechanical impulse in table II, as
an integral of force with respect to time, serves to the
comparison of the mentioned test hit intensity.
TABLE II.

MECHANICAL IMPULSE

Material

rubber

plastic

aluminum

Impulse [Ns]

0.1506

0.1311

0.0661

The shock response spectrum was computed for unified

values of impulse with respect to the reference impulse of the
hammer hit with the rubber tip. According to the impulsemomentum theorem, the resulting momentum of the testing
mass, i.e. velocity of the testing mass, is identical for all the
hammer tip materials. The result of the SRS calculation is
shown in Fig. 7. As it is evident the maximum excitation for
the rubber tip of the hammer is reached at the resonance
frequency of 250 Hz, while the plastic tip excites the

408

maximum vibration response at 2000 Hz and the aluminum

tip is able to excite the maximum response of the mass
vibration at 3500 Hz. The peak at approximately 2 Hz
corresponds to the natural frequency of the mass-spring
system. SRS informs on the effect of the impulse force on
the vibration frequency and can be used to compare
individual hits by the hammer for the experimental modal
analysis.
12

Half-sine impulse (11 ms)

10
8
g 6
4
2
0
0,00 0,01 0,02 0,03 0,04 0,05
Time [s]

Shock Response Spectrum

100,0

10,0
g
1,0

0,1
1

10

100

1000

Frequency [Hz]
Figure 4. Shock Response Spectrum of a half-sine impulse (interval of 11
ms)

The effect of the impact force on the single-degree-offreedom system, which is tuned to a resonance frequency,
describes the properties of the mentioned force better than its
time history of the impact force or acceleration. The reason
for this statement is that SRS draws attention to particularly
sensitive frequency at which can be excited extremely high
amplitude of oscillation
VIII. CONCLUSION
The paper presents the method of calculation of a shock
response spectrum, which is corresponding to an acceleration
exciting the resonance vibration of substructures. SRS
determines the maximum or minimum of the substructure
acceleration response as a function of the natural frequencies
of a set of the single-degree-of-freedom systems modeling
the mentioned substructures.

2011 12th International Carpathian Control Conference (ICCC)

The work was supported by research project GACR No

102/09/0894.
REFERENCES
[1]
[2]

Time [s]
Rubber

Plastic

[3]

Aluminum

Figure 5. Impulse force acting on the base structure

[4]

The vibration (or shock) is recorded by signal analyzer in

the digital form, commonly as acceleration. The transfer
function of the single-degree-of-freedom system is
approximated by an IIR digital filter in such a way that the
filter response to the sampled acceleration may be easily
calculated. This shock response spectrum shows how the
individual components of the impulse signal excite the
mechanical structure to vibrate.

Time : Plastic : Time Acc

350
300
250
200
150
100
50
0
-50
-100
0,00
0,02
0,04

Time : Rubber : Time Acc

50
40

m/s2

30
m/s2

G ISO 18431-4 Mechanical vibration and shock Signal processing

Part 4: Shock response spectrum analysis
Harris Shock and Vibration Handbook, Cyril M. Harris, editor, Allan
G.Piersol, editor, 5th ed., McGraw-Hill 2002. Chapter 23: Concepts
in shock data analysis.
K. Ahlin, M. Magnevall, A. Josefsson, Simulation of forced
response in linear and nonlinear mechanical systems using digital
filters
http://www.bth.se/tek/amt/nonlinear.nsf/(WebFiles)/F75E7709B9C6F
47DC12571D50033078C/$FILE/ISMA_2006_444.pdf
J. Tma, J. Kulhanek, Using scripts in Signal Analyzer. In: 8th
International Scientific Technical Conference PROCESS CONTROL
2008, Editor Taufer, Ivan. 1. edition Pardubice : esk spolenost
prmyslov chemie, June 9-12, 2008, Kouty nad Desnou, Czech
Republic.
.

20
10
0
-10
0,00

0,02

0,04

Time : Aluminum : Time Acc

600
400
m/s2

ACKNOWLEDGMENT

1000
800
600
400
200
0
-200
-400
0,000 0,001 0,002 0,003 0,004 0,005

200
0
-200
-400
0,00

Time [s]

0,02

0,04

Time [s]

Time [s]

Figure 6. Impulse force acting on the base structure

10000,0
1000,0
Rubber

100,0
m/s2

Plastic
10,0

Aluminum

1,0
0,1
1

10

100

1000

10000

Frequency [Hz]
Figure 7. SRS of the shock, which is excited by the hammer

2011 12th International Carpathian Control Conference (ICCC)

409