IEC 60068-2-64:2008/AMD1:2019

Amendment 1 - Environmental testing - Part 2-64: Tests - Test Fh: Vibration, broadband random and guidance

Amendement 1 - Essais d'environnement - Partie 2-64: Essais - Essai Fh: Vibrations aléatoires à large bande et guide

General Information

• Status: Published
• Publication Date: 08-Oct-2019
• ICS: 01 - GENERALITIES. TERMINOLOGY. STANDARDIZATION. DOCUMENTATION
• ICS: 19.040 - Environmental testing
• Technical Committee: TC 104 - Environmental conditions, classification and methods of test
• Current Stage: PPUB - Publication issued
• Completion Date: 09-Oct-2019

Standards Content (sample)

IEC 60068-2-64
Edition 2.0 2019-10
INTERNATIONAL
STANDARD
NORME
INTERNATIONALE
colour
inside
A MENDMENT 1
AM ENDEMENT 1
Environmental testing –
Part 2-64: Tests – Test Fh: Vibration, broadband random and guidance
Essais d’environnement –
Partie 2-64: Essais – Essai Fh: Vibrations aléatoires à large bande et guide
IEC 60068-2-64:2008-04/AMD1:2019-10(en-fr)
---------------------- Page: 1 ----------------------
THIS PUBLICATION IS COPYRIGHT PROTECTED
Copyright © 2019 IEC, Geneva, Switzerland

All rights reserved. Unless otherwise specified, no part of this publication may be reproduced or utilized in any form

or by any means, electronic or mechanical, including photocopying and microfilm, without permission in writing from

either IEC or IEC's member National Committee in the country of the requester. If you have any questions about IEC

copyright or have an enquiry about obtaining additional rights to this publication, please contact the address below or

your local IEC member National Committee for further information.

Droits de reproduction réservés. Sauf indication contraire, aucune partie de cette publication ne peut être reproduite

ni utilisée sous quelque forme que ce soit et par aucun procédé, électronique ou mécanique, y compris la photocopie

et les microfilms, sans l'accord écrit de l'IEC ou du Comité national de l'IEC du pays du demandeur. Si vous avez des

questions sur le copyright de l'IEC ou si vous désirez obtenir des droits supplémentaires sur cette publication, utilisez

les coordonnées ci-après ou contactez le Comité national de l'IEC de votre pays de résidence.

IEC Central Office Tel.: [phone]
3, rue de Varembé [e-mail]
CH-1211 Geneva 20 www.iec.ch
Switzerland
About the IEC

The International Electrotechnical Commission (IEC) is the leading global organization that prepares and publishes

International Standards for all electrical, electronic and related technologies.
About IEC publications

The technical content of IEC publications is kept under constant review by the IEC. Please make sure that you have the

latest edition, a corrigendum or an amendment might have been published.

IEC publications search - webstore.iec.ch/advsearchform Electropedia - www.electropedia.org

The advanced search enables to find IEC publications by a The world's leading online dictionary on electrotechnology,

variety of criteria (reference number, text, technical containing more than 22 000 terminological entries in English

committee,…). It also gives information on projects, replaced and French, with equivalent terms in 16 additional languages.

and withdrawn publications. Also known as the International Electrotechnical Vocabulary

(IEV) online.
IEC Just Published - webstore.iec.ch/justpublished

Stay up to date on all new IEC publications. Just Published IEC Glossary - std.iec.ch/glossary

details all new publications released. Available online and 67 000 electrotechnical terminology entries in English and

once a month by email. French extracted from the Terms and Definitions clause of

IEC publications issued since 2002. Some entries have been

IEC Customer Service Centre - webstore.iec.ch/csc collected from earlier publications of IEC TC 37, 77, 86 and

If you wish to give us your feedback on this publication or CISPR.
need further assistance, please contact the Customer Service
Centre: [e-mail].
A propos de l'IEC

La Commission Electrotechnique Internationale (IEC) est la première organisation mondiale qui élabore et publie des

Normes internationales pour tout ce qui a trait à l'électricité, à l'électronique et aux technologies apparentées.

A propos des publications IEC

Le contenu technique des publications IEC est constamment revu. Veuillez vous assurer que vous possédez l’édition la

plus récente, un corrigendum ou amendement peut avoir été publié.
Recherche de publications IEC - Electropedia - www.electropedia.org

webstore.iec.ch/advsearchform Le premier dictionnaire d'électrotechnologie en ligne au

La recherche avancée permet de trouver des publications IEC monde, avec plus de 22 000 articles terminologiques en

en utilisant différents critères (numéro de référence, texte, anglais et en français, ainsi que les termes équivalents dans

comité d’études,…). Elle donne aussi des informations sur les 16 langues additionnelles. Egalement appelé Vocabulaire

projets et les publications remplacées ou retirées. Electrotechnique International (IEV) en ligne.

IEC Just Published - webstore.iec.ch/justpublished Glossaire IEC - std.iec.ch/glossary

Restez informé sur les nouvelles publications IEC. Just 67 000 entrées terminologiques électrotechniques, en anglais

Published détaille les nouvelles publications parues. et en français, extraites des articles Termes et Définitions des

Disponible en ligne et une fois par mois par email. publications IEC parues depuis 2002. Plus certaines entrées

antérieures extraites des publications des CE 37, 77, 86 et
Service Clients - webstore.iec.ch/csc CISPR de l'IEC.
Si vous désirez nous donner des commentaires sur cette
publication ou si vous avez des questions contactez-nous:
[e-mail].
---------------------- Page: 2 ----------------------
IEC 60068-2-64
Edition 2.0 2019-10
INTERNATIONAL
STANDARD
NORME
INTERNATIONALE
colour
inside
A MENDMENT 1
AM ENDEMENT 1
Environmental testing –
Part 2-64: Tests – Test Fh: Vibration, broadband random and guidance
Essais d’environnement –
Partie 2-64: Essais – Essai Fh: Vibrations aléatoires à large bande et guide
INTERNATIONAL
ELECTROTECHNICAL
COMMISSION
COMMISSION
ELECTROTECHNIQUE
INTERNATIONALE
ICS 19.040 ISBN 978-2-8322-7458-3

Warning! Make sure that you obtained this publication from an authorized distributor.

Attention! Veuillez vous assurer que vous avez obtenu cette publication via un distributeur agréé.

® Registered trademark of the International Electrotechnical Commission
Marque déposée de la Commission Electrotechnique Internationale
---------------------- Page: 3 ----------------------
– 2 – IEC 60068-2-64:2008/AMD1:2019
© IEC 2019
FOREWORD
This amendment has been prepared by IEC technical committee 104: Environmental
conditions, classification and methods of test.
The text of this amendment is based on the following documents:
FDIS Report on voting
104/848/FDIS 104/855/RVD

Full information on the voting for the approval of this amendment can be found in the report

on voting indicated in the above table.

The committee has decided that the contents of this amendment and the base publication will

remain unchanged until the stability date indicated on the IEC website under

"http://webstore.iec.ch" in the data related to the specific publication. At this date, the

publication will be
• reconfirmed,
• withdrawn,
• replaced by a revised edition, or
• amended.

IMPORTANT – The 'colour inside' logo on the cover page of this publication indicates

that it contains colours which are considered to be useful for the correct

understanding of its contents. Users should therefore print this document using a

colour printer.
___________
INTRODUCTION
Add, after the fourth paragraph, the following new paragraph:

The traditional general purpose broad-band random vibration test utilizes waveforms with a

Gaussian distribution of amplitudes. However, when so specified, this test procedure can also

be utilized with random vibration tests with a non-Gaussian distribution of amplitudes. Such

tests are sometimes alternatively known as high kurtosis tests.
Add, after the last paragraph, the following new paragraph:

Annex C is an informative annex giving information on non-Gaussian distribution/high kurtosis

tests.
---------------------- Page: 4 ----------------------
IEC 60068-2-64:2008/AMD1:2019 – 3 –
© IEC 2019
3 Definitions
Add the following new terminological entries:
3.39
kurtosis

4 statistical moment, which provides a measure of the shape of an amplitude distribution

Note 1 to entry: Typically a waveform with Gaussian distribution will have a kurtosis of 3, if considered over an

infinite period.
Note 2 to entry: Kurtosis is given by:
kurtosis xx−.
( )
∑ i
N σ
i=1
where:
σ is the standard deviation of the N values which describe the waveform;
x are individual values representing the waveform described by N such values;
𝑥𝑥̅ is the mean value of the N values which describe the waveform.
3.40
skewness
3 statistical moment, which provides a measure of non-symmetry of an amplitude
distribution

Note 1 to entry: Typically a waveform with Gaussian distribution will have a skewness of 0, if considered over an

infinite period.
Note 2 to entry: Skewness is given by:
skewness xx−.
( )
∑ i
N σ
i=1
where:
σ is the standard deviation of the N values which describe the waveform;
x are individual values representing the waveform described by N such values;
𝑥𝑥̅ is the mean value of the N values which describe the waveform.
3.41
beta distribution

family of continuous probability distributions defined on the interval [0, 1] parametrized by two

positive shape parameters, denoted by α and β, that appear as exponents of the random

variable and control the shape of the distribution
SEE: Figure 4.
---------------------- Page: 5 ----------------------
– 4 – IEC 60068-2-64:2008/AMD1:2019
© IEC 2019
Figure 4 – Examples of the beta distribution with different α and β values
4 Requirements for test apparatus
4.1 General
Add, at the end of 4.1, the following new paragraph:

For non-Gaussian testing, the test apparatus shall be able to produce a signal with a specified

probability distribution and crest factor. Generally, non-Gaussian random vibration testing

requires shaker and amplifier systems that are designed for Gaussian random vibrations but

with increased crest factor capabilities.
4.6.2 Distribution
Add, after Figure 2, the following new paragraph:

For non-Gaussian tests, the time history shall be recorded and the statistical characteristics of

crest factor, skewness, kurtosis and amplitude probability distribution established, see

Clause C.3. If required by the test specification, additional analysis of the time history shall be

undertaken. The measurement time for kurtosis, skewness and amplitude probability

distribution should be long enough to obtain statistically acceptable results.
5 Severities
Replace, in the second paragraph, the first two sentences with the following:
Each parameter shall be specified by the relevant specification. They shall be:

Add, at the end of the second paragraph, after list item d), the following new text:

For non-Gaussian vibration testing the test severity is determined by the same parameters as

for broad-band Gaussian vibration testing but with the addition of:
– the type of non-Gaussian testing to be undertaken (see Annex C),

– the required probability distribution or kurtosis (and skewness if applicable),

---------------------- Page: 6 ----------------------
IEC 60068-2-64:2008/AMD1:2019 – 5 –
© IEC 2019
– the required crest factor.
8.4.1 General
Add, at the end of 8.4.1, the following new text and Figure 5:

For non-Gaussian vibration testing, the time history shall be recorded and the kurtosis,

skewness (if applicable) and amplitude probability density shall be established as required by

the relevant specification (see also Figure 5).
Key
Kurtosis = 4,5 (see 3.39)
Skewness = 0 (see 3.40)
Figure 5 – Time history of non-Gaussian excitation –
Probability density function compared with Gaussian (normal) distribution
11 Information to be given in the relevant specification
Replace the existing list item h) with the following new list item h):

h) Crest factor* / amplitude distribution, kurtosis and skewness (if applicable)/drive signal

clipping amplitude
---------------------- Page: 7 ----------------------
– 6 – IEC 60068-2-64:2008/AMD1:2019
© IEC 2019
Replace, in list item h) the clause number with "4.6.2 and 5.3"
Add, at the end of Annex B, the following new Annex C:
---------------------- Page: 8 ----------------------
IEC 60068-2-64:2008/AMD1:2019 – 7 –
© IEC 2019
Annex C
(informative)
Guidance on non-Gaussian distribution/high kurtosis tests
C.1 Non-Gaussian random vibration

Random vibration testing has traditionally utilized a nominally Gaussian distribution of

amplitudes, but with the crest factor of the waveform set so as to truncate the amplitude

distribution at around three standard deviations. In recent years, several different techniques

have become available, which allow the required acceleration spectral density to be achieved,

but with a waveform distribution typically modified to permit higher amplitudes to occur with a

greater probability. This commonly results in a distribution with a 4 moment of statistics

(kurtosis), which is greater than that of a normal or Gaussian distribution. Hence, the

approach is sometimes referred to as high kurtosis or non-Gaussian vibration testing. With

some available techniques it may also be possible to use a waveform with a kurtosis less than

that of a Gaussian distribution and/or with a non-zero skewness.

High kurtosis or non-Gaussian vibration testing can be advantageous in a number of

circumstances. A typical application arises when both vibration and shock conditions occur

together. Such conditions commonly occur when equipment is transported on or installed in

land vehicles. In such cases it may be required to incorporate moderately higher amplitude

components to the waveform either randomly or pseudo-periodically. High kurtosis or non-

Gaussian vibration testing may also be used to replicate the effects of repeated impacts of

moderate amplitude such as occurs when equipment is loose or experiences “rattling”. In this

case a waveform with a high kurtosis and non-zero skewness could be used.

The traditional general purpose broad-band random vibration test procedure of this document

can also be used with high kurtosis or non-Gaussian vibration testing with only minor

modifications.

High kurtosis or non-Gaussian vibration testing should not be used in place of a traditional

broad-band Gaussian vibration test, except were explicitly specified in the relevant

specification.
C.2 Methods to generate non-Gaussian random vibration
C.2.1 General

High kurtosis or non-Gaussian vibration testing mostly utilizes the same control strategy as

used with traditional broad-band Gaussian vibration tests. Moreover, the underlying

techniques used by vibration controllers to establish a non-Gaussian waveform, are

essentially the same as for Gaussian vibration testing, but typically with an additional step.

Currently, there are several different available techniques used to modify a Gaussian

waveform into a non-Gaussian one. The different techniques may produce distinctly different

waveforms, simply specifying the acceleration spectral density, skewness, and kurtosis is not

sufficient to produce a waveform with identical characteristics. As a consequence, the

techniques are not necessarily interchangeable and different failure modes may be

stimulated, by the different techniques, even if the test severity is identical. Therefore, it is

essential to record and to interpret the time history characteristics during the test.

Set out below is information on three methods for modifying a Gaussian waveform into a non-

Gaussian one. It should be emphasized that these three different methods are only included

here as guidance for the specifier who may have no prior knowledge. Available approaches

are not limited to those described and indeed many other methods and variants exist which

can also be used to effectively implement the non-Gaussian vibration test of this document.

---------------------- Page: 9 ----------------------
– 8 – IEC 60068-2-64:2008/AMD1:2019
© IEC 2019
The three methods described in this annex are:
– amplitude modulation technique,
– phase modification technique,
– non-uniform phase technique.

The selection of the most appropriate techniques for generating a non-Gaussian waveform

will depend upon the application and characteristics of the waveform required. This document

makes no recommendations as to the most appropriate technique for a particular application.

Any choice to be made is left to the relevant specification.
C.2.2 Amplitude modulation technique

Commonly, general purpose broad-band Gaussian digital random vibration test controllers

generate the required waveform by randomizing the phase components of a frequency

spectrum, using a uniformly distributed random variable. This frequency spectrum is

converted to a waveform by means of an inverse Fourier transform. The random waveform is

typically multiplied by a window function, overlapped and then added together to form a nearly

stationary input to the vibration excitation system. The multiplication by the window function

accomplishes two objectives: the spectral lines are spread and the window limits the leakage

into nearby frequencies. If the window and overlap are chosen appropriately, the result can be

truly stationary. If the overlap is not sufficient for a given window, the data can be slightly non-

stationary, with a period equal to the overlap duration.

The amplitude modulation approach to generate a non-Gaussian waveform is similar to the

one described above, except the amplitude of the window is a random variable. The

distribution of this random variable can arise from any distribution, although if a beta

distribution is used, it effectively limits the kurtosis.
The amplitude modulation technique:
a) Produces waveforms with a well-defined spectrum.
b) Produces waveforms with zero skewness.

c) Produces waveforms which have relatively long “bursts” of high amplitudes. Hence, long

realizations need to be generated to accurately estimate the kurtosis. The kurtosis of short

segments will vary considerably.

d) The kurtosis of the waveform can be controlled using a defined distribution such as with

the parameters of a beta distribution. Increasing the standard deviation of the beta

distribution monotonically increases the kurtosis. This makes an iterative procedure to

pick the parameters for a specified kurtosis relatively straightforward.
C.2.3 Phase modification technique

Another way to generate non-Gaussian vibrations is to initially set the phase of the frequency

spectrum from a uniform distribution, as in the previous technique. However, the phase is

subsequently modified using an optimization procedure to vary the phase in an attempt to

minimize the difference between the target kurtosis and skewness with that of the generated

waveform.
The phase modification technique:
a) can produce waveforms with kurtosis both less than 3 and greater than 3;
b) can produce waveforms with a skewness other than zero;
c) is computationally intensive.
---------------------- Page: 10 ----------------------
IEC 60068-2-64:2008/AMD1:2019 – 9 –
© IEC 2019
C.2.4 Non-uniform phase technique

If the values chosen for the frequency spectrum are selected from a distribution other than a

non-uniform one, the resultant waveform will have non-Gaussian kurtosis and skewness

statistics.

If, for example, the phase values are selected from a beta distribution, the span of values will

be in the range 0 to 2π and the mean will have a value of π. If the two parameters of the beta

distribution (α and β) are equal and positive, the distribution will be symmetrical about the

mean. If, α = β = 1, a uniform distribution results. As the values of α and β become large, the

distribution starts to approximate to a Gaussian distribution whose variance decreases as the

values of α and β increase.

When all the phase angles of the frequency spectrum are 0, the peak amplitude of the

waveform will be the maximum possible. If the phase distribution moves from uniformly

distributed random toward an impulse at π, a waveform with an increasing peak amplitude

and hence an increasing kurtosis, will be produced. This can be achieved using a beta

distribution with a decreasing variance (an increasing α and β).

It is even possible to make the beta parameters (α = β) a function of frequency and hence the

kurtosis will be a function of frequency. This can for example be used to generate a waveform

which is approximately Gaussian at low frequencies but has a large kurtosis at higher

frequencies.

Variation to the technique can be achieved by modifying the way that the resultant waveforms

are windowed, overlapped and added together. These can be used to control where in the

extended waveform the peaks occur. In this way the peaks can occur randomly or pseudo-

periodically.
The non-uniform phase technique:

a) Can only generate waveforms with a values of kurtosis greater than or equal to 3.

b) Can only generate waveforms with negative and positive skewness.

c) Produces waveforms with many short duration excursions greater than would be expected

from a Gaussian distribution. The frequency of these excursions can be controlled in

either a random or quasi-periodic manner.
d) Is computationally efficient.
C.3 Additional analysis

The relevant specification may specify, when undertaking non-Gaussian or high kurtosis

testing, that additional parameters of the waveform are derived. These additional parameters

are not used to control the test waveform, but rather to establish additional characteristics of

the waveform which can be used for comparison with perceived damage criterion. Different

additional waveform parameters may be used, depending upon the perceived damage

criterion of concern. Several or all of the following parameters may be considered.

a) cycle counting,
b) Rainflow cycle fatigue counting,
c) level crossing counts,
d) maximum response spectrum (MRS),
e) fatigue damage spectrum (FDS),
f) temporal moments.
---------------------- Page: 11 ----------------------
– 10 – IEC 60068-2-64:2008/AMD1:2019
© IEC 2019

The probability density as well as other methods can yield significant different results if their

setup and boundary definitions are not commonly defined for reasons of comparison (see

DIN 45667:1969-10).
C.4 Frequency range

The relevant specification may state a frequency range in which the non-Gaussian distribution

has a significant influence. The upper frequency of this frequency range may be particularly

relevant to non-Gaussian waveforms containing high amplitude events at higher frequencies.

This is because to adequately define such events, the upper frequency of the test frequency

range may need to be greater than would normally be required for an equivalent conventional

Gaussian vibration test. Also some of the techniques used to generate a non-Gaussian

random vibration, inherently impose restrictions on the lower frequency of the usable

frequency range to prevent excessive shocks characterized by low frequencies.
Bibliography
Add the following new references:
DIN 45667:1969-10, Classification methods for evaluation of random vibrations
LALANNE, C. Mechanical Vibration and Shock Analysis. Volume 5: Specification
Development, Second Edition, ISTE – Wiley, 2009
MCNEILL Scot I. Implementing the Fatigue Damage Spectrum and Fatigue Damage

Equivalent Vibration Testing. 79th Shock and Vibration Symposium, October 26-30, 2008,

Orlando Florida, USA.
___________
---------------------- Page: 12 ----------------------
– 12 – IEC 60068-2-64:2008/AMD1:2019
© IEC 2019
AVANT-PROPOS

Le présent amendement a été établi par le comité d'études 104 de l'IEC: Conditions,

classification et essais d’environnement.
Le texte de cet amendement est issu des documents suivants:
FDIS Rapport de vote
104/848/FDIS 104/855/RVD

Le rapport de vote indiqué dans le tableau ci-dessus donne toute information sur le vote ayant

abouti à l'approbation de cet amendement.

Le comité a décidé que le contenu de cet amendement et de la publication de base ne sera

pas modifié avant la date de stabilité indiquée sur le site web de l'IEC sous

"http://webstore.iec.ch" dans les données relatives à la publication recherchée. À cette date,

la publication sera:
• reconduite,
• supprimée,
• remplacée par une édition révisée, ou
• amendée.

IMPORTANT – Le logo "colour inside" qui se trouve sur la page de couverture de cette

publication indique qu'elle contient des couleurs qui sont considérées comme utiles à

une bonne compréhension de son contenu. Les utilisateurs devraient, par conséquent,

imprimer cette publication en utilisant une imprimante couleur.
___________
INTRODUCTION
Ajouter le nouvel alinéa suivant après le quatrième alinéa:

L'essai de vibrations aléatoires à large bande habituel pour un usage général utilise des

formes d’ondes qui présentent une distribution gaussienne des amplitudes. Toutefois, lorsque

cela est spécifié, cette procédure d’essai peut également être utilisée pour des essais de

vibrations aléatoires avec une distribution non gaussienne des amplitudes. De tels essais

sont parfois aussi connus sous la dénomination d'essais à coefficient d'aplatissement élevé.

Ajouter le nouvel alinéa suivant après le dernier alinéa:

L'Annexe C est une annexe informative qui donne des informations sur les essais avec

distribution non gaussienne / à coefficient d'aplatissement élevé.
---------------------- Page: 13 ----------------------
IEC 60068-2-64:2008/AMD1:2019 – 13 –
© IEC 2019
3 Définitions
Ajouter les nouvelles entrées terminologiques suivantes:
3.39
coefficient d’aplatissement

moment statistique d'ordre 4, qui fournit une mesure de la forme d'une distribution d'amplitude

Note 1 à l’article: Normalement, une forme d’onde à distribution gaussienne aura un coefficient d'aplatissement

de 3 si elle est étudiée sur une période illimitée.
Note 2 à l’article: Le coefficient d'aplatissement est donné par
114
coefficient d'aplatissement x−x.
( )
∑ i
N σ
i=1
où:
σ est l'écart-type des N valeurs qui décrivent la forme d’onde;

x sont les valeurs individuelles qui représentent la forme d’onde décrite par N valeurs de ce type;

𝑥𝑥̅ est la valeur moyenne des N valeurs qui décrivent la forme d’onde.
3.40
coefficient d'asymétrie

moment statistique d'ordre 3, qui fournit une mesure de l'asymétrie d'une distribution

d'amplitude

Note 1 à l'article: Normalement, une forme d’onde à distribution gaussienne aura un coefficient d'asymétrie de 0 si

elle est étudiée sur une période illimitée.
Note 2 à l'article: Le coefficient d'asymétrie est donné par:
113
coefficient d'asymétrie xx− .
( )
∑ i
N σ
i=1
où:
σ est l'écart-type des N valeurs qui décrivent la forme d’onde;

x sont les valeurs individuelles qui représentent la forme d’onde décrite par N valeurs de ce type;

𝑥𝑥̅ est la valeur moyenne des N valeurs qui décrivent la forme d’onde.
3.41
distribution bêta

famille de distributions de probabilités continues définies sur l'intervalle [0, 1] paramétrées

par deux paramètres de forme positifs, notés α et β, qui apparaissent comme des exposants

de la variable aléatoire et qui contrôlent la forme de la distribution
VOIR: Figure 4.
---------------------- Page: 14 ----------------------
– 14 – IEC 60068-2-64:2008/AMD1:2019
© IEC 2019
Figure 4 – Exemples de distributions bêta avec différentes valeurs de α et de β
4 Exigences pour l'appareillage d'essai
4.1 Généralités
Ajouter le nouvel alinéa suivant à la fin de 4.1:

Pour les essais non gaussiens, l'appareillage d'essai doit être capable de produire un signal

avec une loi de probabilité et un facteur de crête spécifiés. Généralement, les essais de

vibrations aléatoires à distribution non gaussienne nécessitent des systèmes vibrants et

amplificateurs qui sont conçus pour les vibrations aléatoires à distribution gaussienne mais

avec des capacités de facteur de crête accrues.
4.6.2 Distribution
Ajouter le nouvel alinéa suivant après la Figure 2:

Pour les essais non gaussiens, la variation temporelle doit être consignée et les

caractéristiques statistiques du facteur de crête, du coefficient d'asymétrie, du coefficient

d'aplatissement et de la loi de probabilité de l'amplitude doivent être établies, voir

l'Article C.3. Si cela est exigé par la spécification d'essai, une analyse supplémentaire de la

variation temporelle doit être effectuée. Il convient que la durée de mesure du coefficient

d'aplatissement, du coefficient d'asymétrie et de la loi de probabilité de l'amplitude soit

suffisamment longue pour obtenir des résultats statistiquement acceptables.
5 Sévérités
Remplacer les deux premières phrases du deuxième alinéa par ce qui suit:

Les valeurs de chaque paramètre doivent être stipulées par la spécification applicable. Elles

doivent être:
Ajouter le nouveau texte suivant à la fin du deuxième alinéa après le point d):
---------------------- Page: 15 ----------------------
IEC 60068-2-64:2008/AMD1:2019 – 15 –
© IEC 2019

Pour l'essai de vibrations à distribution non gaussienne, la sévérité d’essai est déterminée par

les mêmes paramètres que pour l'essai de vibrations à distribution gaussienne à large bande

mais en ajoutant:
– le type d'essai non gaussien à réaliser (voir l'Annexe C),

– la loi de probabilité exigée ou le coefficient d'aplatissement (et le coefficient d'asymétrie

s'il est applicable),
– le facteur de crête exigé.
8.4.1 Généralités
Ajouter le nouveau texte suivant et la Figure 5 à la fin de 8.4.1:

Pour l'essai de vibrations à distribution non gaussienne, la variation temporelle doit êt

...