BRITISH STANDARD BS ISO

16063-1:1998

Methods for the

calibration of
vibration and shock
transducers —
Part 1: Basic concepts

ICS 17.220.20
BS ISO 16063-1:1998

National foreword

This British Standard reproduces verbatim ISO 16063-1:1998 and implements

it as the UK national standard. It supersedes BS 6955-0:1988 which will be
withdrawn in January 1999.
The UK participation in its preparation was entrusted by Technical Committee
GME/21, Mechanical vibration and shock, to Subcommittee GME/21/2,
Vibration and shock measuring instruments and test equipment, which has
the responsibility to:
— aid enquirers to understand the text;
— present to the responsible international/European committee any
enquiries on the interpretation, or proposals for change, and keep the UK
interests informed;
— monitor related international and European developments and
promulgate them in the UK.
A list of organizations represented on this subcommittee can be obtained on
request to its secretary.
Cross-references
The British Standards which implement international or European
publications referred to in this document may be found in the BSI Standards
Catalogue under the section entitled “International Standards Correspondence
Index”, or by using the “Find” facility of the BSI Standards Electronic
Catalogue.
A British Standard does not purport to include all the necessary provisions of
a contract. Users of British Standards are responsible for their correct
application.
Compliance with a British Standard does not of itself confer immunity
from legal obligations.

Summary of pages
This document comprises a front cover, an inside front cover, pages i and ii,
the ISO title page, pages ii to iv, pages 1 to 21 and a back cover.
This standard has been updated (see copyright date) and may have had
amendments incorporated. This will be indicated in the amendment table on
the inside front cover.

This British Standard, having

been prepared under the Amendments issued since publication
direction of the Engineering
Sector Committee, was
published under the Amd. No. Date Comments
authority of the Standards
Committee and comes
into effect on
15 January 1999

© BSI 06-1999

ISBN 0 580 32049 9

 BS ISO 16063-1:1998

Contents

Page
National foreword Inside front cover
Foreword iii
Text of ISO 16063-1 1

© BSI 06-1999 i
ii blank
BS ISO 16063-1:1998

Contents
Page
Foreword iii
1 Scope 1
2 Normative references 1
3 Terms and definitions 2
4 Characteristics to be measured 4
4.1 General 4
4.2 Direct response 4
4.3 Spurious response 4
5 Calibration methods 7
5.1 General 7
5.2 Primary calibration methods 7
5.3 Comparison calibration methods 16
6 Expression of uncertainty of measurement 17
Annex A (informative) Expression of uncertainty of measurement
in calibration 18
Bibliography 20
Figure 1 — Principle of the ideal interferometer 8
Figure 2 — Example of a measuring system using an interferometer 10
Figure 3 — Dual centrifuge 14

Descriptors: Vibration, mechanical shock, test equipment, measuring instruments, transducers,

calibration, generalities.

ii © BSI 06-1999
 BS ISO 16063-1:1998

Foreword
ISO (the International Organization for Standardization) is a worldwide
federation of national standards bodies (ISO member bodies). The work of
preparing International Standards is normally carried out through ISO technical
committees. Each member body interested in a subject for which a technical
committee has been established has the right to be represented on that
committee. International organizations, governmental and non-governmental,
in liaison with ISO, also take part in the work. ISO collaborates closely with the
International Electrotechnical Commission (IEC) on all matters of
electrotechnical standardization.
International Standards are drafted in accordance with the rules given in the
ISO/IEC Directives, Part 3.
Draft International Standards adopted by the technical committees are
circulated to the member bodies for voting. Publication as an International
Standard requires approval by at least 75 % of the member bodies casting a vote.
International Standard ISO 16063-1 was prepared by Technical Committee
ISO/TC 108, Mechanical vibration and shock, Subcommittee SC 3, Use and
calibration of vibration and shock measuring instruments.
This first edition of ISO 16063-1 cancels and replaces ISO 5347-0:1987, of which
it constitutes a minor revision. A new clause 6, new Annex A, and an enlarged
bibliography have been included.
ISO 16063 will consist of the following parts, under the general title Methods for
the calibration of vibration and shock transducers:
— Part 1: Basic concepts;
— Part 2: Primary calibrations;
— Part 3: Secondary calibrations;
— Part 4: Environmental calibrations.
Parts 2 to 4 are under preparation and will consist of a revision of parts 1 to 23
of ISO 5347.
Annex A of this part of ISO 16063 is for information only.

© BSI 06-1999 iii

iv blank
 BS ISO 16063-1:1998

Introduction
The calibration of vibration and shock transducers has become increasingly important as the need has
grown for accurate measurements of the shocks and vibrations to which man and a wide variety of
equipment are subjected in service. Several methods have been used or proposed for these calibrations and
some of them are described in this part of ISO 16063. Clause 5 describes methods which have proved to be
reliable means for the primary calibration of vibration and shock transducers.
Methods of calibration for both vibration and shock transducers are included in this International
Standard because it has proved to be impracticable to make a distinction between transducers used in
measurements of vibrations and those used in measurements of shocks.
This International Standard is limited to the calibration of acceleration, velocity and displacement
transducers. It does not deal with transducers used for measurements of force, pressure or strain, even
though some of these may be calibrated using similar methods. Furthermore, transducers used to measure
rotational vibratory motion are also excluded because, at present, they are few in number and the
calibration hardware and methods are somewhat different from those for the rectilinear transducers
covered by this International Standard.
This part of ISO 16063 contains definitions and describes basic primary calibration. In addition, it
describes, in general terms, various methods for the calibration of vibration and shock transducers as well
as methods for measuring characteristics other than sensitivity. In order to be able to carry out a
calibration with known accuracy, detailed specifications for instruments and procedures have to be laid
down. Information of this kind for each method of calibration will be specified in subsequent parts of
ISO 16063 (i.e. revisions of parts 1 to 23 of the ISO 5347 series).
The transducer may be calibrated as a unit by itself; it may include a cable connection and/or a conditioning
device. The calibration system shall always be properly described.
A bibliography is included and the references are referred to in the text by numbers in square brackets.

1 Scope
This part of ISO 16063 describes methods for the calibration of vibration and shock transducers. It also
includes methods for the measurement of characteristics in addition to the sensitivity.
One primary calibration method has been selected as the preferred method (see 5.2.1). Comparison
calibration methods for vibration and shock are also described (see 5.3). More detailed descriptions are
given in parts 1 to 23 of ISO 5347 (see references [1] to [22]).
This part of ISO 16063 is applicable to continuous-reading rectilinear acceleration, velocity and
displacement transducers and recommends a preferred method which has proved to give reliable and
reproducible results.
It is not applicable to methods for the calibration of rotational transducers.

2 Normative references
The following normative documents contain provisions which, through reference in this text, constitute
provisions of this part of ISO 16063. For dated references, subsequent amendments to, or revisions of, any
of these publications do not apply. However, parties to agreements based on this part of ISO 16063 are
encouraged to investigate the possibility of applying the most recent editions of the normative documents
indicated below. For undated references, the latest edition of the normative document referred to applies.
Members of ISO and IEC maintain registers of currently valid International Standards.
ISO 1101:1983, Technical drawings — Geometrical tolerancing — Tolerances of form, orientation, location
and run-out — Generalities, definitions, symbols, indications on drawings.
ISO 2041:1990, Vibration and shock — Vocabulary.
ISO 2954:1975, Mechanical vibration of rotating and reciprocating machinery — Requirements for
instruments for measuring vibration severity.
GUM:Guide to the Expression of Uncertainty in Measurement. BIPM/IEC/IFCC/ISO/OIML/IUPAC,1995.

© BSI 06-1999 1
BS ISO 16063-1:1998

3 Terms and definitions

For the purposes of this part of ISO 16063, the terms and definitions given in ISO 2041, together with the
following, apply.
3.1 transducer
device for converting the mechanical motion to be measured, for example acceleration in a given direction,
into a quantity which may be conveniently measured or recorded
NOTE A transducer may include auxiliary equipment for amplifying, supplying necessary operating power, providing necessary
circuit elements, indicating or recording its output, etc.
3.1.1
operating range
range of frequency and of amplitude for which the transducer behaves as a linear transducer within
specified limits of tolerance
3.1.2
reciprocal transducer
bilateral electromechanical transducer for which the ratio of the applied current to force produced (when
the transducer is restrained so the velocity is zero) equals the ratio of the applied velocity to the voltage
produced (when the transducer is open-circuited so the current is zero)
EXAMPLES: Electromagnetic and piezo-electric transducers
3.1.3
unilateral transducer
transducer employing strain gauges as sensing elements for which an electrical excitation does not cause
a perceptible mechanical effect in the transducer
3.2
input signal
signal applied to the input of the transducer
EXAMPLE: The acceleration applied to the mounting surface
3.3
output signal
signal generated by the transducer in response to a given input signal
NOTE 1 For single-ended transducers, the acceleration vector is considered positive when directed into the mounting surface of the
transducer. For back-to-back reference accelerometers, the acceleration vector is considered positive when directed from the top
surface into the accelerometer to be calibrated by comparison.
NOTE 2 The phase of the output quantity (e.g. voltage, charge, current, resistance, etc.) should be specified with reference to the
defined positive acceleration vector or the derived quantities (velocity or displacement).
3.4
sensitivity
for a linear transducer, the ratio of the output to input during sinusoidal excitation parallel to a specified
axis of sensitivity at the mounting surface
NOTE 1 In general, the sensitivity includes both amplitude and phase information and is, consequently, a complex quantity which
varies with frequency.
the sinusoidal input motion may be represented by the following equations:
s = ŝ exp [j(Êt + Î1)] = ŝ [cos(Êt + Î1) + jsin(Êt + Î1)] (1)
v = jÊs = v̂ exp[j(Êt + Î1 + ;/2)] = v̂ [cos(Êt + Î1 + ;/2) + jsin(Êt + Î1 + ;/2)] (2)
a = jÊv = â exp[j(Êt + Î1 + ;)] = â[cos(Êt + Î1 + ;) +jsin(Êt + Î1 + ;)] (3)
u = û exp[j(Êt + Î2)] = û[cos(Êt + Î2) + jsin(Êt + Î2)] (4)

where
s is the complex quantity of the displacement;
v is the complex quantity of the velocity;
a is the complex quantity of the acceleration;
u is the complex quantity of the output;

2 © BSI 06-1999
 BS ISO 16063-1:1998

ŝ is the peak amplitude of sinusoidal displacement;

v̂ is the peak amplitude of sinusoidal velocity;
â is the peak amplitude of sinusoidal acceleration;
Ê is the angular frequency;
Î1 and Î2 are the phase angles;
t is the time;
j is the imaginary unit.
the displacement sensitivity, Ss, expressed in the units of the output signal per metre, is

(5)

where

is the magnitude of the displacement sensitivity;

(Î1 — (Î2) is the phase lag.

the velocity sensitivity, Sv, expressed in the units of the output signal per metre per second, is

(6)

where

is the magnitude of the velocity sensitivity;

(Î1 + ;2 — Î2) is the phase lag.

the acceleration sensitivity, Sa, expressed in the units of the output signal per metre per second squared, is

(7)

where

is the magnitude of the acceleration sensitivity;

(Î1 + ; – Î2) is the phase lag.

usually, the displacement sensitivity is determined for a displacement transducer, the velocity sensitivity
for a velocity transducer, and the acceleration sensitivity for an acceleration transducer. In general, the
sensitivity magnitudes and the phase angles are functions of the frequency, f = Ê/2;.
NOTE 2 A displacement, velocity or acceleration transducer in which the corresponding sensitivity does not become zero as the
frequency approaches zero is said to have a zero-frequency response (direct-current response). Sensitivity under constant acceleration
corresponds to Ê = 0 and the phase lag is zero. Examples of transducers with zero-frequency response are acceleration transducers
employing strain gauges, potentiometers, differential transformers, force-balance (servo) or variable reluctance circuits as sensing
elements. Seismic self-generating transducers, such as piezo-electric and electrodynamic transducers, are examples of transducers
without zero-frequency response.
3.5
transverse sensitivity ratio (TSR)
ratio of the output of a transducer, when oriented with its axis of sensitivity transverse to the direction of
the input, to the output when the axis of sensitivity is aligned in the direction of the same input

© BSI 06-1999 3
BS ISO 16063-1:1998

3.6
vibration generator
any device for applying a controlled motion to the mounting surface of a transducer
NOTE Vibration generators are sometimes referred to as exciters or shakers.

4 Characteristics to be measured
4.1 General
The primary object of the calibration of a transducer is to determine its calibration factor (sensitivity) over
the amplitude and frequency range for the degree of freedom for which the transducer is to be used. In
addition, it may be important to know its response to motions in the other five degrees of freedom; for
example, for a rectilinear acceleration transducer, its response should be known to motions at right
angles to the sensitive direction and to rotations. Other important factors include damping, phase lag,
non-linearity or variation in response with amplitude of motion, effect of temperature and pressure
changes, and other extraneous conditions such as motion of the connection cable.
4.2 Direct response
4.2.1 Frequency response and phase response
The sensitivity of a transducer is obtained by placing the transducer with its sensitivity axis parallel to the
direction of motion of the vibration generator, measuring the motion or input applied by the vibration
generator, and measuring the output of the transducer. Both continuous-reading and peak-reading
transducers can also be calibrated with a controlled transient excitation whose amplitude and frequency
components are within the working range of the transducer. To detect any resonances, the output of the
transducer should be observed while varying the vibration generator frequency slowly and continuously
over the frequency range. In general, only information concerning magnitude sensitivity calibration is
given as a function of frequency. However, for the use of a vibration transducer close to its upper or lower
frequency limits, or for special applications, the phase response may be required. This is determined by
measuring the phase lag between the output signal and the mechanical excitation over the frequency range
of interest.
4.2.2 Non-linearity
Deviations from linearity of the output of a transducer (amplitude distortions) are determined by
measuring its output magnitude as the magnitude of the input is increased from the smallest value to the
largest value for which the transducer is designed. When a sinusoidal vibration generator is used, the
measurement should be repeated for several frequencies.
Non-linearity may take several forms. The sensitivity of the transducer may change progressively with
increasing amplitude, there may be a permanent change leading to a displacement of the zero after
subjecting the transducer to vibration or shock, or there may be stops that limit the range of motion
suddenly.
The type and magnitude of the non-linearity of a transducer may be indicated by its amplitude distortion
and by comparing its resonance curve, its phase lag, and its decrement with the corresponding
characteristics for the idealized linear transducer. The permissible deviations from linearity will depend
on the measurements to be made. Non-linearity should be expected at the upper limit of the useful dynamic
range of the transducer.
4.3 Spurious response
4.3.1 Temperature dependency
The sensitivity, damping ratio and resonance frequency of many transducers change as a function of
temperature. Temperature response calibrations are usually performed using a comparison method. The
standard transducer is mounted axially in line with the test transducer. The test transducer is placed
inside a temperature chamber and the standard transducer is located outside the chamber or otherwise
protected from changes in temperature in such a way that its sensitivity remains constant to within 2 %
for the ambient temperatures present during the entire calibration. The vibration generator is used only
at frequencies where it is known that the transverse motion is less than 25 % of the axial motion. The
vibration generator is selected and a fixture designed so that there is negligible relative motion between
the test and standard transducers at frequencies at which the calibration is to be performed.

4 © BSI 06-1999
 BS ISO 16063-1:1998

An alternative procedure for performing temperature response calibrations is to mount the standard and
test transducers on a suitable fixture inside the temperature chamber. This method is limited to
temperature ranges over which the response of the standard transducer is known.
For transducers which respond to static acceleration, the zero unbalance is measured at the maximum and
minimum temperatures.
Transducers with internal damping greater than 10 % of the critical damping should be calibrated at a
minimum of four frequencies at a single vibration amplitude and at each of four temperatures in addition
to room temperature. This method is equally applicable to transducers, such as the electrodynamic types,
which utilize a coil of wire in their operation. The frequencies are selected throughout the frequency range
of intended use.
The internal capacitance and resistance of piezo-electric transducers shall be measured after stabilization
at the maximum calibration temperature.
If the measured resistance of a piezo-electric accelerometer at the maximum calibration temperature is so
low that it affects the low-frequency response of the type of amplifier to be used, a low-frequency response
calibration should be performed at that temperature. A number of frequencies shall be selected to describe
adequately the frequency response. The calibration should be performed on the complete system, using the
amplifier that is used with the accelerometer.
NOTE High temperature may affect the low-frequency response of the accelerometer as well as the noise and stability of the
accelerometer-amplifier combination. Temperature response deviations are computed as the change in calibration factor determined
at the test temperature referred to the room temperature (20 °C) calibration factor (measured at a frequency in the range of
frequencies over which the transducer response is uniform). This change is expressed as a percentage of the room temperature
calibration factor. It is usually desirable to select transducers which have temperature response deviations not exceeding + 15 %
throughout the temperature range of intended use.
4.3.2 Transient temperature sensitivity in piezo-electric transducers
Pyroelectric outputs are generated in all piezo-electric transducers subjected to transient temperatures.
This is especially true for ferroelectric materials. The magnitude of the pyroelectric outputs depends upon
the material constituting the crystal and the design of the transducer. Usually, the predominant frequency
of the pyroelectric output is considerably less than 1 Hz. Also, most of the pyroelectric output from the
transducer is filtered owing to the low-frequency characteristics of most amplifiers.
Accordingly, the pyroelectric output is dependent on the rate of change in temperature and on the
characteristics of the amplifier, together with the characteristics of the transducer. The pyroelectric test is
performed using the type of amplifier normally used with the transducer. The transducer is attached to an
aluminium block by the usual means of attachment. Both are quickly immersed in an ice-water bath or a
bath of other suitable liquid at a temperature which differs by approximately 20 °C from room
temperature. The liquid in the bath should be described. The mass of the block should be
approximately 10 times the mass of the transducer. Precautions are required to ensure that the liquid does
not penetrate the transducer or that electrical leakage resistance is not lowered by the liquid at the
connector, etc. The maximum amplifier output and the time from the start of the transient at which this
maximum output is reached are measured on a direct-current oscilloscope or recorder. If the output
reverses within the first 2 s and reaches a peak of opposite polarity, the magnitude and time of this peak
are also recorded. For an accelerometer, the transient temperature sensitivity is expressed in equivalent
metres per second squared per degree Celsius [(m/s2)/°C] by dividing the maximum transducer output by
the product of the difference between the bath temperature and room temperature and the accelerometer
sensitivity.
For special applications using amplifiers having significantly different low-frequency characteristics, the
pyroelectric test is performed with the specific amplifier to be used. Also, for applications in which the
transient temperature rate differs greatly from that described by the above conditions, the test may be
performed by simulating the particular temperature environment.
4.3.3 Transverse sensitivity ratio
The transverse sensitivity ratio (TSR) is usually determined at a single frequency below 500 Hz. The
frequency used shall be reported. Sinusoidal motion is applied at a frequency at which it is known that the
motion in a plane perpendicular to the sensing axis is at least 100 times the motion in the direction of the
sensing axis. For transverse sensitivity ratios less than 1 %, the requirements for motion are more severe
and extreme care and skill are required to obtain the value of the transverse sensitivity ratio.

© BSI 06-1999 5
BS ISO 16063-1:1998

The transducer is mounted and rotated about its sensing axis through 360°, in increments of 45° or less, to
determine the maximum transverse response.
NOTE Experimental transverse sensitivity measurements on accelerometers indicate no detectable frequency dependence up to
about 2 000 Hz. Only limited data are presently available regarding the transverse response within the frequency range
from 2 000 Hz to 10 000 Hz. Several experimenters have stated that their measurement results usually indicate the high-frequency
transverse response (that is, 2 000 Hz to 10 000 Hz) to be of the same order of magnitude as in a low-frequency determination (that
is, less than 500 Hz). Generally, it is considered that for accelerometers whose axial resonance frequency is greater than 30 kHz,
major transverse resonances will be greater than 10 kHz and, thus, beyond a transducer’s normal operating range. For vibration
transducers of other types, even less information is currently available. If possible, the lowest frequency of transverse resonance
should be determined.
4.3.4 Sensitivity to rotational motion
Certain rectilinear vibration transducers are susceptible to rotational inputs. Examples of these include
flexion-type piezo-electric and piezo-resistive accelerometers, and pendulum force-balance (servo)
accelerometers. Attention is drawn to the existence of rotational sensitivity, and precautions may have to
be taken to preclude a measurement error due to this effect. The rotational sensitivity of rectilinear
vibration transducers can be determined by special methods developed for sensitivity calibrations of
rotational vibration transducers (see reference [36]).
4.3.5 Strain sensitivity
The technique described below is the preferred method to determine the error produced in a transducer
output due to bending of its base.
The transducer is mounted on a simple cantilever beam which produces a radius of curvature of 25 m and
a strain of 250 × 10–6.
A steel cantilever beam is clamped to a rigid support. The beam is 76 mm wide and 12,5 mm thick with a
free length of 1 450 mm.
The natural frequency is very close to 5 Hz. The strain is measured by strain gauges bonded to the beam
near the pickup mounting location about 40 mm from the clamped end. The motion at the mounting
location can be checked by means of a transducer attached using extra isolation against base bending. A
transducer with a calibration factor more than 10 times higher than the units under test is normally
adequate. The outputs from the strain gauges and the transducer under test are recorded. The system is
excited by manually deflecting the free end of the beam. The output of the transducer is recorded at a point
where the strain in the surface of the beam is 250 ×10–6. (This is equivalent to a radius of curvature
of 25 m.) The error is the difference between the motion of the beam at the mounting location and the
motion indicated by the transducer. The strain sensitivity, for a strain of 10–6, is determined by dividing
the above difference by 250.
The strain sensitivity should be tested at various strain amplitudes, in various directions. The maximum
strain sensitivity of some transducers can produce significant errors in certain applications and mounting
conditions. For example, some piezo-electric accelerometers produce error signals of several per cent at
certain frequencies where strains are produced in vibration generators used for calibration purposes.
4.3.6 Magnetic sensitivity
The transducer is placed in a known magnetic field at 50 Hz or 60 Hz, and rotation of the transducer is
started. The maximum electrical output of the transducer is recorded. For accelerometers, metres per
second squared per tesla is recorded as the equivalent based on the sensitivity. For velocity transducers,
metres per second per tesla over the useful frequency range is recorded as the equivalent. Induced
mechanical vibrations and spurious electrical noise shall be eliminated from the test assembly.
4.3.7 Mounting torque sensitivity
The change in calibration factor due to transducer mounting torque is determined by applying torques of
one-half the specified mounting torque, the specified, and twice the maximum specified. This test applies
only to transducers that are mounted by screws, bolts, or other threaded fasteners. If more than one
fastener is used in the normal mounting, the torques should be applied to each fastener.
Care should be taken to ensure that the transducer mounting surface is free from burrs or other surface
defects which would prevent a flat mounting. The test surface to which the transducer is to be mounted
should be flat and smooth and made from steel. The recommended values of flatness and roughness are a
curvature less than 5 4m and an r.m.s. ground finish of 2 4m or better.

6 © BSI 06-1999
 BS ISO 16063-1:1998

The test surface on which the transducer is to be mounted should be drilled and tapped square to the
mounting surface with a perpendicularity of 0,05 mm or better (see ISO 1101). The interface lubrication
normally recommended should be used and stated. The torque should always be applied from an
unmounted condition, that is from zero torque for each of the three test torques. The torque sensitivity is
recorded as the change in transducer calibration factor for one-half and twice the specified torque in
relation to the specified torque. The uncertainty in the applied torque should not exceed ± 15 %.
4.3.8 Special environments
The operation of some transducers may be adversely affected in certain special environments, such as
strong electrostatic, variable magnetic or radio-frequency fields, acoustic fields, in the case of cable effects,
and nuclear irradiation. At present, there are no generally accepted techniques for measuring the effect of
such special environments on a transducer, although special tests have been developed in instances where
adverse effects could be expected (see ISO 2954).

5 Calibration methods
5.1 General
In order to perform a direct calibration of a transducer, it is necessary to use a vibration generator which
applies a controllable and measurable input to the transducer and to provide a means for recording or
measuring the output of the transducer. The transducer shall be attached to the vibration generator (or
placed near it in the case of transducers whose output depends on the relative motion between the
transducer and the vibrating object).
The attachment shall be sufficiently rigid to transmit the motion of the vibration generator to the
transducer over the frequency range of the transducer. This requires that the natural frequency of
the system, consisting of the transducer regarded as the mass and the attachment as the spring of a
single-degree-of-freedom system, be high compared with the highest frequency component of the motion of
the vibration generator. The vibration generator may be a support for tilting the transducer relative to the
pull of gravity, a centrifuge, an electrodynamic vibration generator, or the anvil of a ballistic pendulum.
The tilting support and centrifuge are used for calibration at zero frequency. Rotational calibration is used
for low-frequency calibration for the Earth’s gravitational field. The electrodynamic vibration generator is
normally used for steady-state sinusoidal calibrations. Ballistic pendulums, which apply transient
excitation, may be used as a complementary method to the electrodynamic vibration generator, to bring out
natural frequency response and to permit calibration a high accelerations and velocities. In addition, shock
excitation may be used to verify transducer performance for high accelerations and velocity changes and to
check that auxiliary instrumentation connected to the transducer functions properly under transient
conditions.
A number of calibration methods are described in this part of ISO 16063 and they may be used for special
purposes. However, the use of a laser interferometer is recommended for primary calibration. Whenever
possible, it is recommended that standard transducers be calibrated by this method, and if only one
frequency is used, this should preferably be 160 Hz, 80 Hz, 16 Hz or 8 Hz depending on the application.
Frequency response may be obtained by calibration at discrete frequencies over the frequency range of
interest or as the frequency response relative to the sensitivity at the reference frequency with less
accuracy. Most other calibration needs can be covered by comparison against a standard transducer having
primary calibration. The calibration is always referred to the moving base of the transducer and, for
“back-to-back” calibration standards, to the mounting base for the unknown transducer.
5.2 Primary calibration methods
5.2.1 Calibration by measuring displacement amplitude and frequency
5.2.1.1 General
Many dynamic calibration methods depend on the accurate measurement of the displacement amplitude
of the vibration to which the transducer is subjected. This method is generally used for continuous-reading
transducers. The sinusoidal motion applied by the vibration generator should be along a well-defined
straight line; lateral motions should be negligible.

© BSI 06-1999 7
BS ISO 16063-1:1998

The measured displacements can be used to calculate velocities, v, and accelerations, a, using the
formulae v̂ = 2;f ŝ and â = (2;f)2 ŝ which are derived by single and double differentation, respectively, for
the sinusoidal displacement, s, and frequency, f. These formulae assume that the harmonic and noise
content of the motion remains negligible even after the differentation. They emphasize the need for
minimizing the distortion due to the electrical power sources or due to other causes such as mechanical
resonance. Harmonics are also objectionable since they may excite resonant response in a transducer.
Once the displacement amplitude is known, the transducer sensitivity may be calculated as the ratio of the
measured transducer output to the velocity or the acceleration amplitude. The displacement shall be
measured by laser interferometry. The method is well described in references [23] to [28], [37] and [38].
The methods of sensitivity calculation based on displacement amplitude measurement by laser
interferometry generally give good accuracy from 0,1 Hz to 10 kHz (corresponding to displacement
amplitudes of 0,5 m to 20 nm). Special methods based on interferometric displacement measurement allow
primary phase calibration to be performed in addition to the sensitivity calibration. As an alternative to
laser interferometry based on displacement measurement, good accuracy in absolute sensitivity and phase
calibration of vibration transducers may also be achieved by current-state laser doppler velocimetry [39].
Considerable errors in the measurement of displacement will occur if the reference mirror is perturbed at
the frequency (or a harmonically related frequency) at which the accelerometer is vibrated. Error may also
result from perturbation of the beam splitter. It is advisable to monitor for such pertubation using a very
sensitive accelerometer.
5.2.1.2 Theory for the ideal interferometer
The principle of operation is shown in Figure 1, where E0, E1 and E2 represent the electric field vectors,
and l1 and l2 represent the actual path lengths the beams have to travel after the beamsplitter. The
displacement to be measured is represented by s (mirror 2).
The electric field vectors E1 and E2 can be represented by the formulae

where Æ is the wavelength of the laser light.

Figure 1 — Principle of the ideal interferometer

8 © BSI 06-1999
 BS ISO 16063-1:1998

The intensity of the photodetector I(t) is given by the formula

where
A and B are constants of the system;
L = l2 – l1
From the intensity expression, it can be seen that the maxima will occur when
4;
---------- ( l2 – l 1 + s ) = 2n;
2
and, therefore, the displacement corresponding to the distance between two intensity maxima is given by
%s = Æ/2. The number of maxima, Rf, for one vibration cycle is then
R = 4 ŝ /(2/2) = 8 ŝ /Æ
which is commonly referred to as the “frequency ratio” because it can be calculated by dividing the number
of fringes counted during 1 s by the vibration frequency.
The displacement amplitude, ŝ , is thus given by the formula
ŝ = Rf·2/8
If, in addition to the frequency ratio, the vibration frequency is measured, one can also compute the velocity
and acceleration.
The same system can be used to measure displacement amplitude at frequencies outside the range
recommended earlier for the fringe-counting method. Several other methods can be devised by considering
the frequency spectrum of the intensity I(t). As given in reference [23], the expansion gives

The following two examples adequately illustrate the type of signal processing that is required here.
a) Adjusting the vibration amplitude to a level which makes the nth harmonic component zero, one can
4;
solve the equation J n æ ------- ˆsö = 0, to obtain ŝ .
è Æ ø
b) In cases where it is not possible or practical to calibrate at amplitude levels required by
4;
the J n æ ------- ˆsö = 0 method, one can extract the value of ŝ from the ratio of two harmonic components, for
è Æ ø
example, by solving for ŝ from

where û1 and û3 are the measured magnitudes of the first and third harmonics.

© BSI 06-1999 9
BS ISO 16063-1:1998

5.2.1.3 Measuring system

An example of a measuring system is shown in Figure 2. The transducer is a so-called reference transducer
and the sensitivity shall be determined for the upper surface (reference mounting surface). The laser has
an output power of 1 mW, and the detector is a normal silicon phototransistor. The pulse generator is used
to obtain a well-defined signal for the counter input instead of the internal crystal oscillator. The frequency
analyser is used to select the appropriate frequency when the zero-point method is used. The laser and the
interferometer system and the vibrator system should be mounted on independent heavy vibration
isolation blocks (for example each of mass more than 400 kg) to avoid perturbation of the reference mirror
or the beam splitter by the reaction of the vibrator support structure.
5.2.2 Calibration by reciprocity method (see references [28] to [30])
Primary calibrations may also be carried out using the technique of reciprocity calibration. The reciprocity
theory is applicable to the calibration of vibration standards in the amplitude range where the transducer’s
electrical output is linearly proportional to the motion of the vibration generator on which it is calibrated.
The theory shows a reciprocity relationship for the driver coil of the vibration generator and equates the
ratios of force/current and potential difference/velocity.
When the calibrator is energized with current in the driving coil at a specified frequency, the sensitivity,
Suc is defined as the ratio of the potential difference, u13 in volts, generated by the accelerometer, to the
acceleration, â in metres per second squared, at the mounting surface, that is
u 13
S uc = --------
- (8)
â

Figure 2 — Example of a measuring system using an interferometer

The object of the reciprocity method is to determine the sensitivity Suc that when the potential difference
u13 is measured, the acceleration â may be computed by use of equation (8). Suc is determined from the
following equation:
Suc = S0 + sZZm (9)

where Zm is the mechanical impedance of the transducer, in kilograms per second.

The quantities S0 and sZ are determined by the following two experiments and computational procedure.

10 © BSI 06-1999
 BS ISO 16063-1:1998

Experiment 1
Several weights are attached to the mounting table. For each weight, and with no weight attached, the
transfer admittance Ye in amperes per volt, is measured between the driving coil and the accelerometer,
and is given by the formula
i
Y e = --------- (10)
u 13

where
i is the current, in amperes, in the driving coil;
u13 is the potential difference, in volts, generated by the accelerometer.

Experiment 2
The moving parts of the calibrator are set into sinusoidal vibration by connecting the mounting table of the
calibrator to a vibration generator and then energizing the vibration generator1). The ratio, u13/u15, of the
potential, u13, generated in the accelerometer to the open-circuit potential, u15, generated in the driving
coil is measured.
Computational procedure
Determine the ordinate intercept, J, and the slope, Q, of the function W/(YeW – Ye0) when plotted against
the mass W, of the weight attached to the mounting table in experiment 1, and where YeW is the value of
Ye with a weight of mass W attached, and Ye0, is the value for W = 0. This plot is made by separating
W/(YeW – Ye0) into its real and imaginary parts from which the real and imaginary parts of J and Q are
determined. The quantities S0 and sZ, in equation (9) are then given by

S0 = j v Ju 13 /u 15 (11)

sZ = ( u 13 /u 15 )/jvJ (12)

where
Ê is the angular frequency, in radians per second;
j is the imaginary unit.
5.2.3 Calibration by centrifuge
5.2.3.1 Single centrifuge
A centrifuge consists of a balanced table or arm which can be made to rotate about a vertical axis at a
uniform angular velocity. With this device, an accurately known constant acceleration can be applied to an
acceleration transducer for as long a time as is desired. Centrifuges capable of subjecting transducers to
masses of several kilograms and to accelerations up to 6 × 105 m/s2 have been built and centrifuges rated
at lower accelerations are commercially available.
Only rectilinear acceleration transducers having zero-frequency response can be calibrated on a centrifuge.
To make a calibration, the acceleration transducer shall be mounted on the table or arm of the centrifuge
with its axis of sensitivity carefully aligned on a radius of the circle of rotation. The acceleration acting on
the transducer is
a = Ê2r (13)
where
Ê is the angular frequency, in radians per second, of the centrifuge;
r is the distance from the axis of rotation to the centre of gravity of the mass element of the
transducer.

1)Some electrodynamic vibration generators have been constructed having two drive coils mechanically connected to the
armature and mounting table of the vibration generator. In this instance, it is not necessary to use a separate vibration
generator.

© BSI 06-1999 11
BS ISO 16063-1:1998

It is necessary to mount the transducer at such a distance from the axis of rotation that the deflection of
the mass element of the transducer can be neglected in the determination of r.
Most transducers are so constructed that it is not easy to measure r directly. The value of r may be
determined from readings taken for the transducer mounted at two positions separated by a known
distance, %r, while it is rotated on the centrifuge. It is good practice to adjust the speed so that
approximately the same value of acceleration is applied to the transducer in both positions. The value of r
in the second position, designated r2, is

(14)

where
Ê1 is the angular frequency in the first position, for which r = r1 = r2 – %r;
Ê2 is the angular frequency in the second position, for which r = r2;
u1 is the transducer output due to the angular frequency Ê1;
u2 is the transducer output due to the angular frequency Ê2.

With this value of r2 and the angular frequency Ê2, the acceleration can be determined from equation (13).
The determination of r may be eliminated if the transducer has a linear range which extends down to the
acceleration due to gravity, g. If this is the case, the transducer is first calibrated at ± g by the tilting
support method. The transducer is then placed on the centrifuge and the angular frequency Ê1 at which
the output is that corresponding to g is determined. The applied acceleration at another angular frequency,
Ê, is then
a = gÊ2/Ê21 (15)

The angular frequency, Ê, has to be determined more accurately than the radial distance, r, because the
applied acceleration varies as the square of this quantity. Most centrifuges that are designed for calibration
purposes are equipped with a tachometer which gives a direct indication of the rate of rotation to within
about 2 %. A much more accurate determination of the rate of rotation is possible either by stroboscopic
means or by any one of several devices (for example a device using a photoelectric cell or a magnet) which
produces pulses at a rate proportional to the speed. The pulse rate may be determined with an electronic
counter.
In the calibration of electromechanical transducers on a centrifuge, leads are brought through slip rings
and brushes. Since acceleration transducers with zero-frequency response are relatively low-impedance
devices, shielding from external fields and cable noise presents no particular problem. The electrical noise
from a worn-in slip-ring assembly of good design is negligible under normal circumstances. However,
certain acceleration transducers using strain gauges as the sensing element contain only one or two active
elements, the other resistances of the Wheatstone bridge circuit being added externally. With such a
transducer, the entire bridge should be mounted on the rotating table to avoid false signals which would
otherwise occur as a result of small changes in resistance of the slip-ring assembly. Alternatively, a Kelvin
bridge circuit may be used.
In the calibration of transducers at small accelerations, gravity may have a significant effect if the
transducer is sensitive to transverse accelerations. If possible, the transducer should be placed on the
centrifuge with the axis of maximum transverse sensitivity of the transducer in a horizontal plane.
The sensitivity of an acceleration transducer at zero frequency can be determined with an expanded
uncertainty of measurement of 1 % or better on a good centrifuge. Calibration on a centrifuge will, of
course, give no indication of the usable frequency range.

12 © BSI 06-1999
 BS ISO 16063-1:1998

5.2.3.2 Tilted centrifuge

When the axis of rotation of a centrifuge is not parallel with the Earth’s gravitational vector, the
acceleration applied to a transducer mounted on the centrifuge will have a gravitational component,
g sin Ú sin(Êt), which will be added to the centripetal acceleration, Ê2r, where Ú is the angle between the
rotational axis and the vertical axis. The effect of the gravitational component on a transducer
approximated by a spring-mass system is discussed in reference [31]. Sinusoidal acceleration can be
applied to the transducer when r = 0 and Ú = 90°. In this instance, the acceleration due to gravity can be
determined with minimal uncertainty. The limit is, of course, ± g at whatever frequency is produced by the
rotation. The maximum frequency is generally less than several hundred hertz as limited by the centrifuge
and balancing ability.
5.2.3.3 Dual centrifuge (see reference [32])
The dual centrifuge consists essentially of a small centrifuge eccentrically mounted on a large centrifuge
made to rotate about this vertical axis in accordance with 5.2.3.1 (see Figure 3). The vibration transducer,
A, is attached to the small centrifuge, which is driven independently. When the two centrifuges are driven
at constant angular frequencies, the sensitivity axis of the transducer is alternatively changing its
direction relative to the centre of the large centrifuge at a frequency, in hertz, equal to the angular
frequency, in revolutions per second, of the small centrifuge relative to the large centrifuge. The component
of acceleration, a, applied along the sensitive axis of the transducer at any time t is
a = rcÊ2 cos(Êpt) – r(Ê ± Êp)2 (16)

where
Ê is the angular frequency, in radians per second, of the large centrifuge;
Êp is the angular frequency, in radians per second, of the small centrifuge relative to the large
centrifuge;
rc is the distance between the centres of the two centrifuges;
r is the distance from the centre of gravity of the mass element of the transducer to the centre of the
small centrifuge.
The sign in the last term of equation (16) is plus when Ê and Êp are either both clockwise or both
counterclockwise and minus when they are in opposite directions.
When the term r(Ê ± Êp)2 can be neglected, the applied acceleration along the sensitive axis of the
transducer reduces to the sinusoidal term
a = rcÊ2 cos(Êpt) (17)

Thus, the component of acceleration applied along the sensitive axis of the transducer is sinusoidal. There
is also a component of acceleration applied transverse to the sensitive axis, which renders this method
non-applicable to transducers with high transverse sensitivity ratios. The term r(Ê ± Êp)2 is zero
if Ê and Êp are equal but in opposite directions.
When a dual centrifuge of this type is built in the following manner, equation (17) is exact and the
component of acceleration along the sensitive axis of the transducer should be purely sinusoidal. A pulley
is fixed in space with its centre coincident with the centre of the large centrifuge. A pulley of the same size
is fixed concentrically to the small centrifuge and connected to the other pulley by a belt. The large
centrifuge is driven by a motor and the angular frequency of both centrifuges about their respective centres
will always be equal and opposite in direction.
The dual centrifuge is useful for applying sinusoidal accelerations up to 500 m/s2 in the approximate
frequency range from 0,7 Hz to 10 Hz.

© BSI 06-1999 13
BS ISO 16063-1:1998

Figure 3 — Dual centrifuge

5.2.3.4 Tilting support calibrator
The tilting support calibrator utilizes the Earth’s gravitational field for the calibration of rectilinear
acceleration transducers with zero-frequency response and with negligible transverse sensitivity. It is
useful over the range from – g to + g. The transducer to be calibrated is fastened to a platform at the end
of an arm which indicates the component of acceleration along the arm. The arm may be set at an
angle, Î, relative to the vertical between 0 and 180°. It is furnished with a pointer to read off the angle, Î,
from a divided circle. Care should be taken to level the base to which the transducer is attached in the
position Î = 0. Positioning of the arm to ± 0,1° or better is possible with an accurately divided circle.
The component of acceleration along the arm is given by
a = g cos Î (18)
The change in acceleration corresponding to an angular displacement, %Î, is therefore
%a = g sin Î% Î (19)
The acceleration transducer is subjected to a component of acceleration at right angles, at, to its sensitive
direction equal to
at = g sin Î (20)

Usually, this does not affect the results of the calibration of transducers with negligible transverse
sensitivity.
NOTE Vibration isolation of the test apparatus may be required to achieve satisfactory results. Electronic filtering and averaging
in direct-current voltage-sensing instrumentation may help to reduce the effects of bench-top vibration, etc.

14 © BSI 06-1999
 BS ISO 16063-1:1998

5.2.4 Shock calibration method

To measure the time-varying acceleration, laser interferometry based on fringe counting and time interval
measurement may be applied [40]. From the measured displacement values and the conjugate times, the
acceleration is obtained through an interpolation polynomial and double differentation. A small
measurement uncertainty of the acceleration peak value or of shock sensitivity of high quality
accelerometers has been demonstrated by using a specially developed shock acceleration exciter whose
hammer and anvil are airborne [41]. Most primary shock calibrations are based on the principle of change
in velocity (see references [28] and [33]). This is because velocity is a physical parameter which can be
measured practically. The usual configuration is to mount the transducer to be calibrated on an anvil
suspended by some means in a resting position (see reference [34]). A hammer of some sort is then allowed
to strike the anvil, thus generating a transient motion of the anvil. The impact shall be controlled so that
the velocity change is not so rapid, or so slow, as to excite important frequency components outside of the
response range of the instruments and rigid-body dynamics is a suitably accurate approximation. The
velocity or acceleration transducer to be calibrated shall have a mass which is small compared with the
anvil mass to which it is attached and should have its sensitivity direction carefully aligned with the
direction of the impact force during collision. During impact, the accelerometer output versus time is
recorded. Immediately following impact, the anvil velocity, %v, is measured.
The velocity may be measured by timing the anvil over a known distance. Photoelectric or magnetic
transducers can be used to trigger an electronic timer. The velocity is a direct result of the acceleration
applied during impact:

(21)

where
%v is the velocity increment, in metres per second;
a(t) is the time-varying acceleration, in metres per second squared.
The accelerometer output, ur(t), is then
ur(t) = Sr a(t) (22)

where Sr is the sensitivity of the reference standard, in units of the output signal per metre per second
squared.
Combining equations (21) and (22), and solving for Sr, gives

(23)

Equation (23) makes possible the calibration of a linear acceleration transducer from its recorded output
during the ballistic impact. If the impact is against a linear spring, it will have the shape of a half-sine pulse
of area A = 0,637 hb where h and b are the height and the width respectively of the pulse. The pulse shapes
and durations are usually adjusted by varying mass, impacting mediums and some initial conditions such
as drop height, air pressure, or other physical parameters depending on the nature of the shock generator.
Both techniques (impact against an anvil and impact against a linear spring) are practical to obtain the
value of %v as required for equation (23). The output of the acceleration transducer may be recorded during
impact as a function of time by a high-speed oscillograph or from a storage-type oscilloscope and
photographed. Potential difference scales and times scales can be verified by superimposing a known
potential difference signal, uc and a known time area, yc and xc, respectively. The potential difference, K1,
and time scale, K2, factors are given by
K1 = vc/yc K2 = tc/xc (24)

© BSI 06-1999 15
BS ISO 16063-1:1998

The area under the record of accelerometer output versus time is defined by

(25)

where x1 and x2 denote the beginning and the end of the impact.
Substitution of these quantities into equation (23) results in
K 1 K 2 KA
S r = -----------------------
- (26)
%v

The area, A, can be obtained by graphic integration of the recorded acceleration time history. A planimeter
is useful in measuring the area under the acceleration-time record. Care has to be taken in the
determination of area with regard to zero offsets, overshoot and ringing.
The integral shown in equation (23) may also be determined by electronic integration or by digital
recording and summation techniques. This accelerates the calibration process and reduces the subjective
errors and operator fatigue.
Based on the principle of change in velocity, primary shock calibrations of quality accelerometers can be
performed with uncertainties of less than 5 % over most reasonable ranges of shock amplitudes and
durations. An important assumption made is that the transducer being calibrated has linear frequency
response in the frequency range of interest. If it does not, errors will result which are very difficult to
evaluate. Furthermore, a single value of sensitivity is determined and does not in any way yield any
practical information regarding frequency or phase response.
5.3 Comparison calibration methods
A vibration transducer calibrated by a primary calibration method may be used as a reference standard for
the calibration of other transducers. The method described in 5.2.1 is the preferred method.
In making such a calibration, both the calibrated reference transducer 1 and the transducer 2 to be
calibrated are subjected to the same input motion by suitably mounting the transducers, and their outputs,
u1 and u2, or the ratio of the two outputs, are measured. If the two transducers sense the same vibration
parameter, i.e. both sense velocity or both sense acceleration, and if the responses of both transducers 1
and 2 are linear, the amplitude sensitivity, S2, of transducer 2 is related to the corresponding amplitude
sensitivity, S1, of the calibrated transducer 1 by
u
S 2 = -----2- S 1 (27)
u1

Calibration by this comparison method is limited to the range of frequencies, time duration of pulses, and
amplitudes for which transducer 1 has been calibrated. The complex sensitivity S2 may also be determined
if the phase response of transducer 1 is known and if the phase relationship between u2 and u1 is measured.
If the two transducers sense different vibration parameters, for example if a velocity transducer is
compared to an accelerometer, the sensitivity S2 will be related by some power of jÊ. Best results are
generally obtained when the two transducers are rigidly mounted in a back-to-back configuration with the
sensing axes parallel with the direction of motion. Care has to be taken to ensure that transducers 1 and 2
experience the same motion. If both transducers are rectilinear and are placed on the table of a vibration
generator, the rocking motion of the table has to be negligible. Wave-form distortion is generally not as
critical for comparison calibrations as for primary calibrations, especially if both transducers sense the
same vibration parameter. However, the presence of harmonic components in the motion may increase the
output voltages of the two transducers by different amounts, depending on the ratio of the sensitivities at
the frequencies of the harmonics.
Practical calibration is normally carried out using an electrodynamic vibration generator. A back-to-back
reference accelerometer is mounted on the vibration generator with the calibrated surface opposite the
table. The unknown accelerometer is placed on the back-to-back reference accelerometer. For comparison
at high frequencies, the reference accelerometer shall be calibrated with a mass load of the same order of
magnitude as that of the unknown accelerometer.

16 © BSI 06-1999
 BS ISO 16063-1:1998

The electrical output of the reference accelerometer preamplifier may be conveniently and precisely
compared against the output of the unknown accelerometer using a precision amplifier with attenuator and
comparing the signals by adjusting to the zero indication using a balance comparator meter.
It shall be noted that the sensitivity of the reference transducer may depend on the mass load at the surface
where the unknown transducer is mounted. Therefore the sensitivity values for the reference transducer
shall be known for the mass of the unknown transducer [35].

6 Expression of uncertainty of measurement

The uncertainty of measurement in calibration shall be expressed by the expanded uncertainty U in
accordance with the Guide to the Expression of Uncertainty in Measurement (GUM), based on the approach
recommended by the International Committee for Weights and Measures (CIPM).

© BSI 06-1999 17
BS ISO 16063-1:1998

Annex A (informative)
Expression of uncertainty of measurement in calibration
A.1 General
See clause 6.
The purpose of U is to provide an interval y – U to y + U within which the value of Y, the specific quantity
subject to calibration and estimated by y, can be expected to lie with high probability. To assert confidently
that y – U < Y < y + U, determine the expanded uncertainty U as follows.
A.2 Calculation of expanded uncertainty of measurement
A.2.1 Every effort has to be made to identify each effect that significantly influences the measurement
result and to compensate for such effects by applying the estimated corrections or correction factors.
If an effect influencing the measurement result is appropriately described by a probability distribution
(preferably probability density, cf. A.2.2) having a significant expected value (in particular for an
asymmetrical distribution), treat the latter as systematic error and compensate by correction.
A.2.2 Represent each component of uncertainty that contributes to the uncertainty of the measurement by
2
a standard deviation ui, termed standard uncertainty, equal to the positive square root of the variance u i .
Some standard uncertainties may be obtained as statistically estimated standard deviations by the
statistical analysis of series of observations (referred to as type A evaluation of standard uncertainty in
GUM). Evaluate other standard uncertainties as standard deviation of a probability distribution describing
the scientific judgement of all possible values of the respective quantity (type B evaluation of standard
uncertainty). The judgement is based on all information available about the quantity. In particular, if there
is no specific information about the possible values of a quantity responsible for systematic effects except
that these values are within the bounds b– and b+, a uniform distribution over the interval [b–; b+] may be
used to represent this information. It has a standard uncertainty b/ 3 where b = (b+ –b–)/2. The expected
value (b+ +b–)/2 is to be used for correction in this case.
If an influence quantity can be considered uniformly distributed (rectangular probability density) but is
known to be transformed into the measurement result with a specific nonlinear function (e.g. sinusoidal;
polynomial of second or third order), take this information into account by choosing the associated
distribution model.
EXAMPLE
The sensitivity S of an accelerometer to sinusoidal accelerations in the nominal measurement direction is
calculated from the output, voltage or charge amplitude x̂ , stimulated by a vibration, acceleration
amplitude â, using the formula S = x̂ /â. Among the various disturbing effects influencing the measurement
result in calibration, there may be a significant transverse motion component from the vibration exciter,
acceleration amplitude âT, transformed into an error component c x̂T in the output, in conjunction with the
accelerometer’s transverse sensitivity, ST. It is assumed for this example that the acceleration to be
measured and the transverse acceleration have the same frequency and that there is no phase angle
difference. As the transverse sensitivity is usually sinusoidally dependent on the angle ¶ between the
direction of maximum transverse sensitivity (ST,max) and the direction of a transverse excitation, the error
component can be expressed by
e = ST âT = ST,max âT,maxcos ¶.
x̂T
If the values of the maximum transverse sensitivity (ST,max) and the maximum transverse acceleration
(âT,max) are known while the angle ¶ is not, it is reasonable to assume a rectangular distribution of ¶ within
the interval [–;;;]. Thus, the influence quantity, i.e. transverse acceleration, with rectangularly
distributed angle ¶ leads to a measurement error component e whose probability density is described by
x̂T

(often referred to as arcsin distribution). The associated standard uncertainty is

u( e x̂T ) = b/ 2

The expected value E { e x̂T } is zero in this case. This is the best estimate of the error e .
x̂T

18 © BSI 06-1999
 BS ISO 16063-1:1998

A.2.3 Determine the combined standard uncertainty uc, as the standard uncertainty of the measurement
of Y, by combination of the individual standard uncertainties (and covariances as appropriate) using the
law of propagation of uncertainty. Accordingly, the combined standard uncertainty is obtained from

(A.1)

This equation is based on a first-order Taylor series approximation of

Y = f(X1,X2,...XN) (A.2)

where Y is the measurand determined from N input quantities X1, X2, ..., XN through a functional
relationship f.
An estimate of the measurand Y, denoted by y, is obtained from equation (A.1) using input estimates
x1, x2, ..., xN for the values of the input quantities. Thus the output estimate which is the result of
measurement is given by
y = f(x1, x2,..., xN). (A.3)

In equation (A.1), the symbols ·f/·xi are often referred to as sensitivity coefficients ci. They are equal to the
partial derivatives ·f/·Xi evaluated at Xi = xi. Symbol u(xi,xj) designates the estimated covariance
associated with xi and xj.
For the case where no significant correlations are present, equation (A.1) is reduced to

(A4)

The first-order Taylor series approximation of equation (A.2) resulting in equation (A.1) is only applicable
if the model function f is sufficiently linear with respect to the variation of the input estimates xi within the
ranges characterized by the uncertainties u(xi). This is not the case in the example given in A.2.2 if the
angle ¶ is considered to be an input quantity Xi. To overcome this obstacle which similarly exists with other
influence quantities acting in measurements within calibrations of vibration and shock transducers, an
adequate model has been introduced (cf. reference [42]). To briefly specify this model for the example above,
a factor (1 – e x̂T / x̂ ) with e x̂T / x̂ << 1 is introduced, as an input quantity Xi, into the functional relationship
used for calculating the measurand. Equation (A.2) specially tailored to the example, is reduced to three
input quantities,
Y = f(X1, X2, X3)
where Y designates the measurand (sensitivity S), X1 the accelerometer output (voltage or charge
amplitude x̂ ), X2 the acceleration amplitude and where X3 = (1 – e x̂T / x̂ ). Thus, the relationship
X
Y = ------1- X 3
X2
can be established. The first Taylor series approximation can be used now, leading to the relative combined
standard uncertainty

if there are no significant correlations.

Using the symbols introduced in the example, the latter relationship can be written as follows:

© BSI 06-1999 19
BS ISO 16063-1:1998

where u ( x̂ )/ x̂ is the relative uncertainty of the output voltage amplitude measurement, u(â)/â the relative
uncertainty in acceleration amplitude measurement and u( e x̂T / x̂ )= u( e x̂T / x̂ , with u( e x̂T ) = b/ 2 as
explained in the example. Accordingly, further factors whose deviations from the value 1 are similarly
expressed by the relative error component of the respective quantity (e.g. voltage, acceleration, or the
sensitivity as a whole) might be introduced as input quantities (X4, X5, ...), allowing the variety of
uncertainty sources to be taken into account separately.
A.2.4 Determine the expanded uncertainty U by multiplying uc by a coverage factor k:
U = kuc
where a value of k = 2 is preferably to be used. If it can be assumed that the possible values of the
calibration result are approximately normally distributed with approximate standard deviation uc, the
unknown value can be asserted to lie in the interval defined by U with a level of confidence, or probability,
of approximately 95 %.
A.2.5 When reporting the result of the measurement y, the expanded uncertainty and the value of the
coverage factor k used, if different from k = 2, shall be stated. In addition, the approximate coverage
probability or level of confidence of the interval may be stated.

Bibliography
[1] ISO 5347-1:1993, Methods for the calibration of vibration and shock pick-ups — Part 1: Primary
vibration calibration by laser interferometry.
[2] ISO 5347-2:1993, Methods for the calibration of vibration and shock pick-ups — Part 2: Primary shock
calibration by light cutting.
[3] ISO 5347-3:1993, Methods for the calibration of vibration and shock pick-ups — Part 3: Secondary
vibration calibration.
[4] ISO 5347-4:1993, Methods for the calibration of vibration and shock pick-ups — Part 4: Secondary shock
calibration.
[5] ISO 5347-5:1993, Methods for the calibration of vibration and shock pick-ups — Part 5: Calibration by
Earth’s gravitation.
[6] ISO 5347-6:1993, Methods for the calibration of vibration and shock pick-ups — Part 6: Primary
vibration calibration at low frequencies.
[7] ISO 5347-7:1993, Methods for the calibration of vibration and shock pick-ups — Part 7: Primary
calibration by centrifuge.
[8] ISO 5347-8:1993, Methods for the calibration of vibration and shock pick-ups — Part 8: Primary
calibration by dual centrifuge.
[9] ISO 5347-9:1993, Methods for the calibration of vibration and shock pick-ups — Part 9: Secondary
vibration calibration by comparison of phase angles.
[10] ISO 5347-10:1993, Methods for the calibration of vibration and shock pick-ups — Part 10: Primary
calibration by high impact shocks.
[11] ISO 5347-11:1993, Methods for the calibration of vibration and shock pick-ups — Part 11: Testing of
transverse vibration sensitivity.
[12] ISO 5347-12:1993, Methods for the calibration of vibration and shock pick-ups — Part 12: Testing of
transverse shock sensitivity.
[13] ISO 5347-13:1993, Methods for the calibration of vibration and shock pick-ups — Part 13: Testing of
base strain sensitivity.
[14] ISO 5347-14:1993, Methods for the calibration of vibration and shock pick-ups — Part 14: Resonance
frequency testing of undamped accelerometers on a steel block.
[15] ISO 5347-15:1993, Methods for the calibration of vibration and shock pick-ups — Part 15: Testing of
acoustic sensitivity.
[16] ISO 5347-16:1993, Methods for the calibration of vibration and shock pick-ups — Part 16: Testing of
mounting torque sensitivity.
[17] ISO 5347-17:1993, Methods for the calibration of vibration and shock pick-ups — Part 17: Testing of
fixed temperature sensitivity.
[18] ISO 5347-18:1993, Methods for the calibration of vibration and shock pick-ups — Part 18: Testing of
transient temperature sensitivity.

20 © BSI 06-1999
 BS ISO 16063-1:1998

[19] ISO 5347-19:1993, Methods for the calibration of vibration and shock pick-ups — Part 19: Testing of
magnetic field sensitivity.
[20] ISO 5347-20:1997, Methods for the calibration of vibration and shock pick-ups — Part 20: Primary
vibration calibration by the reciprocity method.
[21] ISO 5347-22:1997, Methods for the calibration of vibration and shock pick-ups —
Part 22: Accelerometer resonance testing — General methods.
[22] ISO 5347-23:, Methods for the calibration of vibration and shock pick-ups — Part 23: Primary shock
calibration using laser interferometry2).
[23] HOHMANN, P. Schwingungsmessungen mit Laserinterferometern zur Kalibrierung von
Aulnehmern, Acustica, 26, pp. 122–136 (1972).
[24] HOHMANN, P. and MARTIN, R. Kalibrierung eines Schwingungsaufnehmer-Vergleichsnormals,
Physikalisch-Technische Bundesanstalt, Braunschweig (1968), VDI-Berichte Nr. 135 (1969).
[25] DERERRARI, H.A., DARBY, R.A. and ANDREWS, F.A. Vibrational displacement and
mode-shape measurement by a laser interferometer, Journal of the Acoustical Society of America, 52,
No. 5, p. 982 (1967).
[26] ACTLEY, R.A. and LOGUE, S.H. A Laser interferometer and its applications to vibration amplitude
measurement, Proceedings, Institute of Environmental Sciences Annual Meeting, pp. 235–240 (1967).
[27] LICHT, T., BRUEL and KJAER. Technical Review, No. 1 (1971).
[28] HARRIS, C.M. and CREDE, C.E. Shock and Vibration Handbook, 3rd edn., McGraw-Hill (1987).
[29] LEVY, S. and BOUCHE, R.R. Calibration of vibration pick-ups by the reciprocity method, J. Research
NBS, 57, No. 4, pp. 227–243 (1956).
[30] BOUCHE, R.R. and ENSOR, L.C. Use of reciprocity calibrated accelerometer standards for performing
routine laboratory comparison calibrations, Shock and Vibration Bulletin (Ap 440-502 Bull. 34, Part 4)
(February 1965).
[31] WILDHACK, W.A. and SMITH, R.O. A basic method of determining the dynamic characteristics of
accelerometers by rotation, Proceedings of the Instrument Society of America, 9, Part V (1995), p. 4
(Paper 54-40-3).
[32] SMITH, R.O., WILLIS, E.A. and HILTEN, J.S. A dual centrifuge for generating low-frequency
sinusoidal accelerations, NBS Journal of Research, 66d, No. 4 (1962).
[33] WlTTKOWSKI, U. Stosspendelverfahren zum Kalibrieren und Prüfen von
Beschleunigungsmessgeräten, Technisches Messen tm, No. 9, pp. 323–328 (1979).
[34] BOUCHE, R.R. The absolute calibration of pickups on a drop-ball shock machine of the ballistic type,
Endevco Corporation TP 206 (April 1961).
[35] LAUER, G. Interferometrische Bestimmung der Belastungsabhängigkeit von
Beschleunigungsaufnehmer-Vergleichsnormalen, Fortschritte der Akustik — FASE/DAGA 82, Göttingen.
[36] v. MARTENS, H.-J. Dynamic calibration of rotational transducers by diffraction grating
interferometry. PTB-Mitteilungen 103(1993)1, pp. 19–27.
[37] CLARK, N.H. An improved method for calibrating reference standard accelerometers.
Metrologia, 20 (1983), pp. 103-107.
[38] v. MARTENS, H.-J.: Interferometric counting methods for measuring displacements in the
range 10–9 m to 1 m. Metrologia, 24 (1987), pp. 163–170.
[39] LEWIN, A.C., ROTHE, V., SELBACH, H. Laser doppler interferometer for applications in the
automotive industry. Proc. of the 25th ISATA Symposium, Florence 1992.
[40] v. MARTENS, H.-J., TÄUBNER, A. Interferometric measurement of translational and rotational
motion quantities. Recent advances in experimental mechanics. S. Gomez et al. (eds), Balkema,
Rotterdam, 1994.
[41] v. MARTENS, H.-J., SCHLAAK H.-J., TÄUBNER, A. Interferometrische Kalibrierung von
Translations- und Rotationsaufnehmern mit Sinus- und Stossanregung. Fortschritte der
Akustik — DAGA 94, Teil B, pp. 681–684.
[42] v. MARTENS, H.-J. and ROGAZEWSKI, P. Representation and transfer of the units of vibration
quantities in the GDR. Bulletin OIML No. 108, Sept. 1987.

2)
To be published.

© BSI 06-1999 21
BS ISO
16063-1:1998
BSI — British Standards Institution
BSI is the independent national body responsible for preparing
British Standards. It presents the UK view on standards in Europe and at the
international level. It is incorporated by Royal Charter.

Revisions

British Standards are updated by amendment or revision. Users of

British Standards should make sure that they possess the latest amendments or
editions.

It is the constant aim of BSI to improve the quality of our products and services.
We would be grateful if anyone finding an inaccuracy or ambiguity while using
this British Standard would inform the Secretary of the technical committee
responsible, the identity of which can be found on the inside front cover.
Tel: 020 8996 9000. Fax: 020 8996 7400.

BSI offers members an individual updating service called PLUS which ensures
that subscribers automatically receive the latest editions of standards.

Buying standards

Orders for all BSI, international and foreign standards publications should be
addressed to Customer Services. Tel: 020 8996 9001. Fax: 020 8996 7001.

In response to orders for international standards, it is BSI policy to supply the

BSI implementation of those that have been published as British Standards,
unless otherwise requested.

Information on standards

BSI provides a wide range of information on national, European and

international standards through its Library and its Technical Help to Exporters
Service. Various BSI electronic information services are also available which give
details on all its products and services. Contact the Information Centre.
Tel: 020 8996 7111. Fax: 020 8996 7048.

Subscribing members of BSI are kept up to date with standards developments

and receive substantial discounts on the purchase price of standards. For details
of these and other benefits contact Membership Administration.
Tel: 020 8996 7002. Fax: 020 8996 7001.

Copyright

Copyright subsists in all BSI publications. BSI also holds the copyright, in the
UK, of the publications of the internationalstandardization bodies. Except as
permitted under the Copyright, Designs and Patents Act 1988 no extract may be
reproduced, stored in a retrieval system or transmitted in any form or by any
means – electronic, photocopying, recording or otherwise – without prior written
permission from BSI.

This does not preclude the free use, in the course of implementing the standard,
of necessary details such as symbols, and size, type or grade designations. If these
details are to be used for any other purpose than implementation then the prior
written permission of BSI must be obtained.

If permission is granted, the terms may include royalty payments or a licensing

agreement. Details and advice can be obtained from the Copyright Manager.
BSI Tel: 020 8996 7070.
389 Chiswick High Road
London
W4 4AL