Sei sulla pagina 1di 74

INTERNATIONAL IEC

STANDARD CEI
NORME 60068-3-11
INTERNATIONALE First edition
Premiére éditíon
2007-05

Environmental testing -

Part 3-11:
Supporting documentation and guidance -
Calculation of uncertainty of conditions
in climatic test chambers

Essais d'environnement -
Partie 3-11:
Documentation d'accompagnement et guide -
Calcul de l'incertitude des conditions
en chambres d'essais cllimatiques

Reference numbar
Numéro de rélérence
IEC/CEI 60068-3-11 :2007

Licensed to l+f Co. LTD 2013-07-18. Arry formof reproduction and redistribution are strictly prohibited.
� THIS PUBLICATION IS COPYRIGHT PROTECTED
� Copyright© 2007 IEC, Geneva, Swltzerland

Al nghts reserved. Unless otherwlse sp&elfied, no part of th1s pubficahon may be reproduced or ullllzed In any torm
or by any means, eteetronlc or mechank:al, lncludlng pholocopytng and microfilm. whhout permlsslon In writlng from
either IEC or IEC's member National Committee in 1he coun1,y ol the requester.
lí you have any questlons abOut IEC copyrigh1 or have an enquwy about Obtamlng add1Uonal rights to thls pubhcation,
pt.ease contact the address below or your local IEC member National Commktee fa< further informalion.

Oroits de reproduclion réservés. Sauf indication oontra1re, aucooe partte de celte publication ne peut Aire reproduite
ni utillsée sous queícue forme que ce soit et par aucun pn:,cédé, électronique ou mécaruqoe. y compris la photocople
et les microfilms, sans reccoro écrit de la CEI ou ctu Comité national de la CEI du pays du demandeur.
Si vous avez des questions sur le copyright de la CEI ou si vous désírez obtenir des d.roits supplémenlaires sur cene
publteation. utilisez les coordonnées cl·aprés ou contactez le Comité national de la C.EI de votre pays de résldence.

IEC Central Oftioe


3. ruede Varembé
CH-1211 Geneva. 20
Switzerta.nd
Emall: lnmall@lec.ch
W<.t, wwwiecd>

About the IEC


� lnlernatíon111 Electrotechnlcal Commi11ion (IEC) 11 the luct.ng gtot. o,ganlz.etlon that proparo1 end publl1he1
lnternatlonal Standard& for all oloctriea1. olectronlc and re&ated technologi.s

About IEC publlcatlons


The technk::al content of IEC publleellons Is kepi und&r oon1tan1 revtew by the IEC Ptease make sure thot you have 1he
l.atest ecUuon, a corrigenda oran amendment mlght have been publilhed
• Catak>gue of IEC pt¡bllcatlons: www.iec ch/search.pub
The IEC on-line Catalogue enables you to search by• v3riety of cnteriii (reference number, ren. technicat commit1ee .... ).
ll atso gives lnformation on projects, withdrawn and replaced publicabons.
• IEC Jusi Published: www.iec.ch/online_news/justpub
Stay up to date on au new IEC pubUcaUons. Jusi PLibhshed detalls tW'ice a montl'I au new publlcauons released. Avallable
OJMl.ne and al:so by email.
• Customer Service Centre: www.lec.Ch/webStore/custs-eN
lf you wish to give us your feedback on this publ�tion � need funher assistance, please visit the Customer Service
c.ecitre FAQ or c:onlacl us:
Email: csc:@iec.ch
Tel.: +41229190211
Fax: +4122919 03 00

A propos de la CEI
Le Commis.sion Eleclrotecimique lnternalionale (CEI) est la premiére org.anisatH)f'I mondiale qui élabore et publie des
normés 1nlérnationa1es pour tout e& qui a trait • l'éléClrblé. ti réleetronque él aux léehnok>giés apparentéés.

A propos des publlcations CEI


Le contenu technlque dos publicollon1 do la CEI nt cons.tammtint revu. VeulUez vous assuror quo vous pon•d•z
f'6dWon la ptus r6cente. un cotrlgendum ou amendomen1 peUI e\lOW ••• publl6.
• Catalogue de:t publlcallOf'IS do la CEI: www.lee.ch/11archpubkur_fut·f hCm
Le Catalogue en.flgne de la CEI vous perme1 d'effeciuer del recherehel en u111t1ant d1tft!iren1s crlléres (numt!iro de
fiffflnce, texto, comité d'étude1, •.• ). 11 donne au11I des in1o,mati0n1 sur les projetl et les publica1ion1 reti�e, ou
refflplac6es.
• Just Pubfish&d CEI: www.lec.ch/onllne_new1r,U'$tpub
Restez lnform6 sur tes nouvelles publk:aUons de la CEI. Jusi Publlshed detadle deux fols par mols les nouvelles
pubkaOons parues. Disponible en·ligne 01 aussi par email.
• Service CUents: www.lec.et\/Wébstoreleustservlcostserv_entry-f.htm
Si vous désirez nous donner des commentalres sur cene publaóon ou si vous avez des questions, vlsftez te FAO du
Service clients ou c:ontactez-nous:
Email: csc@jec ch
Tel.: +4 t 22 919 02 11
Fax: +4122919 03 00

licensed 10 H-1 Co. LTQ 2013-07-18. P(o/ formol reprociJclion and redísiribu!ion are slriclly prohibiled.
INTERNATIONAL IEC
STANDARD CEI
NORME 60068-3-11
INTERNATIONALE First edition
Prernlére édition
2007-05

Environmental testing -
Part 3-11:
Supporting documentation and guidance -
Calculation of uncertainty of conditions
in climatic test chambers

Essaís d'environnement -
Partie 3-11:
Documentation d'accompagnement et guíde -
Calcul de l'incertitude des conditions
en chambres d'essais climatiques

Conin1ksion ElkuotiKfl11lqu• lnte,utlona..


PRICE CODE
CODE PRIX V
lin1emnoMI E1Kttor.ct111lc1I Commlull6n
M�,... 3nenporHHH...eaca,i 1<ovi,cc1ot F0t¡:t.'ict,JHCI MIII�
/Jwpdr.WJirCllt*"'"'�

Licensed to 1-H Co. LTO 2013-07-18. l>"J tormot reproclJction arxt redistribution are strictly prohitited.
-2- 60068-3-11 C> IEC:2007

CONTENTS
FOREWORD 4
INTRODUCTION 6

1 Scope 7
2 Normative references 7
3 Terms and definitíons 8
4 Concept of uncertainty 11
4.1 Uncertainty, error and 'true value" 11
4.2 Statements or uncertainty..... .. 12
4.3 Combining uncertainties 13
5 Tolerance 13
6 Humidity and temperature measurement 13
7 Methods lor determining climatic test chamber uncertainties 14
7.1 Empty chamber 16
7.2 Typ,cal load 16
7.3 Measurement of condilions In the chamber durlng the test 17
7.4 Condit,ons to measure............. .. 17
7.5 Measurements required 18
7.6 Sources of uncertainly 19
7. 7 Essentlal contributions of uncertalnty 20
8 Estimahon of uncertainty componenls and lheir comblnalion 24
9 Overall uncertainty of temperature measuremenl 24
9.1 General 24
9.2 Furtner considerations 26
10 Overall uncertainty of relativa humidity measurement 26
10.1 Uncertainty of temperatura measurement at each sensor point. 27
10.2 Uncertainty of the relative humidity measurement 27
11 Anomalous data and presentation of results 30
11.1 Average case analysis 30
11.2 Worst case analysis 30

Annex A (informativa) Measurement data sets - Loaded chamber 32

Bibliography . . 34

Licensed to tff Co. LID. 2013-07-18. Mi formol reproduction and redistribution are strictly prohibited.
60068·3·11 C, IEC:2007 -3-

Figure 1 -Approaches to callbratlon melhod and uncertalnty calculatlon 15


Figure 2 - lllustration of the íluctuation of a lemperature sensor 23

Table 1 - Combinatlon of temperature uncertalnties 24


Table 2 - Combinatlon of temperature uncertainties at each polnt 27
Table 3 - Comblnation er humidlty uncertalnties 28
Table A. 1 - Typical temperature measurement data set and it's analysis and refs 32
Table A.2 - Humidity measurements anatysls based on Table A.1 temperaturas 33

Llcensed to ttf Co. LTQ 2013-07·18. M¡ formof reproduction and redstrib.llon are strlctly prohibited.
-4- 60068·3·11 e IEC:2007

INTERNATIONAL ELECTROTECHNICAL COMMISSION

ENVIRONMENTAL TESTING -

Part 3-11: Supporting documentation and guidance - Calculation


of uncertainty of conditions in climatic test chambers

FOREWORD
t) The lntemalional Electrotechnioal Commission (IEC) is a woftdWM:le organizalion for stendardization comprising
all natlonal erectrotochnlcal comm1ttae1 (IEC National Comm1ttee,). The object ol IEC Is to promete
lntemalional cc-ooeraiton on au queslions conoeming stendardizallon In lhe etect11cal and olectronlc flelds. To
lhis end and in addllion to olher aclivilies, IEC publt1he1 lntemaoonel Stendard.s Technical Specific,ations,
Technical Rapo,11. PubUcly Availabla Spec1rtc.1bon1 {PAS) and G.11de1 (Mreafter refetted to as �1ec
Publlcotlon{1r). Thtlr properallon 11 ontru1t1d to technK:,.¡ eommlltHI. any IEC N111onal Committoe lntortsted
In the aubjec1 deell wllh mey par1lclpa1e In 1hl1 prepara1ory WOf1c ln1e,na1M>nal. governmental and non-
governmen101 orgonlza1lon1 llalsing wllh the IEC at10 par1ic,p11e in thi1 preparalion IEC eoll1borate1 cto1ely
wilh the lnterneUonal Orgentzetlon for Standard1z11ion (ISO) 1n KCOfdance wllh cond1tlon1 determlned by
agreemenl between the two organlzatlon1.
2) The termal declstons or agreements of IEC on teehnical matters express. as nearty as posslble, an i'n1ernatlonal
consenaus of opinlon on the relevant aubj&cta since each 1echnical ccmmlttee has representatlon from au
interested IEC National Commlltee,.
3) IEC Publications have me form of recommend1tlon1 for lntemalional use and are accepted by IEC National
Committees in that sense. While all reasonable efforts are made to ensure thal the technicat contenl of IEC
PublicaUons Is accurate, IEC cannot be hetd ,espons1b!e fOf the way m wt'ltch they are us&d or for any
mlslnterpretation by any end user.
4) In order to promote lnternatlonal unlformlty, IEC National Comm,nees undenake to apply IEC Publlcatlons
transparently to tne maximum extent possible in 1heir national and regional publications. Any divergence
between any IEC Publication and the corresponding nahonal or regional publication shall be clearly indicated in
the latter.
5) IEC provides no matklng procedore to indicate lts approval and cannol be rende,ed responsible fo, any
equipment declared to be In conformity with an IEC Public.ation.
6) AII users snouio ensure that they neve me tatest ed1bon of th1s publication.
7) No llabllity shall attach to IEC or lts dlrectors. employees. servants or age.nts includ1ng lndlvfdual experts and
members of its technical commiUees and IEC National Commntee.s '"' any personal injury, property damage or
other damage of any nalure whatsoever, whether d1rect or indirect, or for cosls (inciuding legal fees) and
expenses ariaing out of tha publication. use ot. or refiance upon. lhis IEC Publtcation o, any other IEC
Publlcotlons.
8) Auentlon Is drown to tho Normative references clted In thls pubhcauon. Use of the referenced publlcatlons Is
indispensable for the coNect epplicalion of this public11ion
9) Attentlon Is drnwn to 1he posslblllly that some ot the olemontt 01 thls IEC PubOcaOon may be the aubfoct of
patent rfgh11. IEC 1hell not be held re1pon1ible lor idenhty,ng any or 111 auch patent rights

lnlernational Standard IEC 60068-3-11 has been preparad by IEC technlcal commlllee 104:
Envlronmental condltions, classiflcatlon and methods of test.

The text of this standard is based on the following documents:

FDIS Repon on voting


1 041409/FOIS 104/415/RVD

Full information on the voting for the approval ol this standard can be found in the report on
voting indicated in the above table.

This publication has been drafted in accordance with the ISOflEC Directives, Part 2.

Llcensed to l+I Co. LTO 2013·07·18. Mf formol reproduction and redstrilulon are strlctly prohibited.
60068-3-11 O IEC:2007 -5-

A list of all parts in the IEC 60068 series. under the general tille Environmental testíng can be
found on the IEC webslte.

The comminee has decided that the contents of this publication will remain unchanged until
the malntenance result date lndlcated on the IEC web slle under "http://webstore.iec.ch" In
the dala related to the speclflc publlcallon. At thls date, the publication will be

reconfirmed,
withdrawn,
replaced by a revisad edition, or
amended.

Licensed to lff Co. LTQ 2013-07-18. Mf formol reproduction and redistrituion are strictly prohibited.
-6- 60068-3-11 e IEC:2007

INTRODUCTION

Thls part of IEC 60068 provides guidance ro, anatysing uncertainties or temperatura and
humldlty In cllmallc test chambers. 11 has been wrltten ror technlclans, englneers and
managers In environmentat testing. and for anyone who needs to understand the results of
envlronmental tests.

The performance of climatic test chambers is a key concern in environmental test engineer-
ing. To comply with any test specification, lhe performance or the charnber needs to be
characterized to decide whether the generated conditions rall within the specilied limits. This
characterization can be a difficull task, and the analysis of uncertainties in chamber
performance is often surrounded by confusion. This publication is lntended lo ease that
process.

In what lollows, the concept of uncertainty of measurement is introduced first and then
the signilicanoe of tolerance discussed. Aspects of humidity and temperature measurement
are considered, followed by methods for determining and combining uncertainties. The
cases of both calibrating an empty chamber and of measuring conditions in a loaded
chamber are consldered. Finally, detailed guidance and worked examples are given for
analysing results to give estimates ot uncertainty in the measured performance.

Licensed to Hi Co. LT!l 2013-07-18. fin¡ formot reproduction and redistribution are strictly prolibited.
60068·3· 11 C, IEC:2007 -7-

ENVIRONMENTAL TESTING -

Part 3-11: Supporting documentation and guidance - Calculation


of uncertainty of conditions in climatic test chambers

Scope

This part of IEC 60068 demonstrates how to estimate the uncartainly of steady-state
temperature and humidity conditions in temperature and humidity chambers. Since this is
inaxtricably linkad to the methods of measurement. thesa are also described.

This standard is equally appllcable to all environmental enclosures. including rooms


or laboratories. The methods used apply both to temperature chambers and
combined temperature and humidity chambers.

Thls standard is meant to help everyone using cllmatic test chambers. Those already familiar
with uncertainty of measurement wlll flnd lt useful for guldance on typlcal sources or
uncartainty and how they should be quantifiad and combinad. lt is also intanded to assist the
flrst-llme or occaslonal user who has llttle or no knowtedge of the subject.

To discuss uncertalnty, it Is importan! lirst to undarstand what is baing measurad or


characterized. The calibration or characterization of the performance of a chamber is
concernec with the humidity and temperatura of the air in the chamber, as experienced by tne
ítem under test, at a given set polnt. This snoutd not be confused with characterizing or
calibrating the chamber sensor. which is a separate matter.

2 Normative references

The following referenced documents are indispensable for the application of this document.
For dated references, only the edition cited applies. For undated references, the latest edition
of the referenc,ed document (including any amendments) applies.

IEC 60068-3-5: Environmental tesling - Part 3-5: Supporting documentalion and guidance -
Confirmalion of /he performance of temperatura chambers

IEC 60068-3-6: Environmental tesling - Part 3-6: Supporting documentation and guidance -
Confirmatlon of the performance of tsmpsraturelhumldity chambsrs

ISO 3534-1:2006, Slalislics - Vocabulery en« symbols - Part 1: General stalistical terms and
terms used in probability

ISO 3534-2:2006, steusucs - Vocabulary snd symbols - Part 2:Applled stelistics

lnternatlonal Vocabulary of basic and general standard terms in metrology. ISO, Ganava,
Swltzertand 1993 {ISBN 92-67-10175-1)-VIM

licensed 10 l+f Co. LTQ 2013-07·18. Arrf f orrnoí reproduction and redistribt.lion are strictly prohibited.
-8- 60068-3-11 e IEC:2007

Guki« to the expression ot vncertainty in meesurement. ISO. Geneva, Switzerland 1993.


(ISBN 92-67-10188-9)- GUM

3 Terms and definltlons

For the purposes ol this document, the lollowing terms and delinitions apply.

3.1
callbratlon authorlty
laboratory or other organlzatlon that performs callbratlons and Is ltsell accredlted by the
appropriate natlonal accreditation body

3.2
cllmatlc test chamber
enclosure
chamber or enclosed space where the interna! temperature or temperatura and humidity can
be controlled within specified limits

3.3
comblned standard uncertalnty
standard uncertainty of the result of a measurement when that result is obtained from
the values of a number ol other quantities, equal to the positive square root ola sum ol terms,
lhe terms belng the variances or covariances ol these other quantities weíghted accordlng to
how the measurement result varíes with changes in these quanlities

See also GUM.

3.4
correctlon
value added algebraically to the result ol a measurement to compensate determinable
systematlc error

See arso VIM.

3.5
confidence level
value ol probability associaled with a conlidence interval
NOTE The confidence level is the likelihood thet the ·1rue verue" lies within the stated range of uncet1ainty usually
expressed as a percentage. e.g. 95 '4.
See also ISO 3534-1.

3.6
coverage factor
numerical lactor to multiply the combined standard uncertainty to obtain an expanded
uncertainty
NOTE A coverage factor of k=2 corres.ponds to a confidétleé 1.evel or approximately 95 % if normally distributed
and if the number of degrees of freedom is sufflciently larga.
See also GUM.

Llcensed to Hi Co. LTll 2013-07-18. l>n¡ formot reprodlctlon and redistrlbutlon are st rt ctt y prohlbited.
60068-3-11 <O IEC:2007 -9-

3.7
dew point
temperature at which the partial pressure of the water vapour is equal to the saturation vapour
pressure over water or ice
NOTE The temperature to which thé a1r wouk1 have to eoot (at constant pressure and constanl water vapour
content) in ercer to reach saturation. A state of saturation exists when the air is holding the máximum amount of
water vapour possible et the existing 1emperature and pressure.

3.8
dlsperslon
spread of repeated measurements of a quantity

3.9
drift
changa in the lndlcation of a measuring system not relatad to a changa in the quanlity being
measured

See also VIM.

NOTE The drlft alnce the last c1llbr11lon can be e1ttma16d end a correctlon applled to meaaured valuea.

3.10
error
difference between result of a measurement and the true value

3.11
expanded uncertainty
quantity deflnlng an lnterval about the result of a measurement that may be expected to
encompass a large fraction of the distribution of values that could reasonably be attributed to
the measurand

See also VIM.

3.12
fluctuation
change (from the mean) in the temperatura or humidity after stabilization from time to time at
a point in space

NOTE lt may be measured by standard deviation Of maximum dev,allon.

3.13
gradlent
maximum difference in mean value, alter stabilization, at any momenl in time belween two
separate points in the working space

3.14
lncldent alr
condltloned alrstream whlch flows lnto the working space

3.15
partlal vapour pressure
contribution of water vapour in a given volume of air at a constant pressure and temperature
of the atmosphere

3.16
reference lnstrument
previously calibrated instrument used to measure the conditions within the enclosure

Licensed to Hi Co. LTQ 2013·07·18. /ln'j formd reprociJction aro redistribution are strictly prohitited.
- 10- 60068-3-11 O IEC:2007

3.17
relatlve humldity
ratio of actual partial vapour pressure to the saturation vapour pressure at any given
temperature and pressure, expressed as a percentage (% RH)

3.18
repeatability
closeness of agreement between independent results obtained in the normal and correct
operation of the same method on idenlical test material, in a short space of time, and under
the same test conditions (such as the same operator, same apparatus, same laboratory)

3.19
resolutlon
smallest changes between lndications of the chamber controller display that can be
meaningfully d1st1nguished

3.20
saturatlon vapour pressure
when a given volume of air, al a constant temperature, has water vapour present and Is
incapable of holding more water vapour lt Is sald to be saturated

3.21
stabilization
aehievement or the srate or temperature/humidity in the chamber when ali mean values in the
working space are constant and have maintained temperature/humidity within a given
toleran ce

3.22
standard deviation
measure of the dispersion of a set of measurements

NOTE The standard deviation, s, is the best estímate of sigma (the population standard deviation).

See also GUM and/or VIM.

3.23
standard uncertainty
uncertainty of the result of a measurement expressed as a standard deviation

See also GUM.

3.24
tolerance
acceptance llmit spec,fied or chosen for a process or product

See etso ISO 3534-2.

3.25
traceability
property of the result of a measurement or the value of a standard whereby it can be related
to stated references, usually national or international standards, through an unbroken chain of
comparisons, ali having stated uncertainties
NOTE The unb<oken chain ot comparisons is canee a traceabillty chaln,
See also ISO 3534-1 and VIM.

Licensed to tff Co. LTQ 2013-07-18. Mf formol reproduction and redistrituion are strictly prohibited.
60068-3-11 e IEC:2007 - 11 -

3.26
true value
value which characterizes a quantity, perfectly defined in the conditions which exist when that
quantity is considered
NOTE The true vatue of a quannty is a tlleoretical ooncept end in generel, cannol be known exac11y but is
eslimated by measurement.
See also ISO 3534-1.

3.27
uncertainty
parameter, assoclated wllh the result of a measurement, whlch characterlzes the dlsperslon
of the values that could reasonably be attributed to it

3.28
uncertalnty budget
list of sources of uncertainty compiled with a voew to evaluating a combined standard
uncertalnty assoclated wilh a measurement result

3.29
uncertalnty contrlbullon
input to an uncertainty budget

3.30
working space
part of the chamber in which the specified conditions can be maintained within the specified
tolerances

4 Concept of uncertainty

4.1 Uncertainty, error and "true value"

In every measurement - no matter how careful - there is always a margin of doubt about the
result. In simple lerms, the uncertainty of a measurement is a quantification of the doubt
about the measurement resull.

Whlle dlscusslng uncertalnty we often also need to consider a related but separata concept,
"error". A measurement "error" is the difference between the measured vaiue and the "true
value" of the lhing being measured.

The "true vatue" of any quantlty Is In principie unknowable. Thls leads to a problem slnce the
·error" Is deflned as the result of a measurement minus the "true value", Sometlmes this
dlfference can be estlmated. Both terms are best avolded as much as posslble and, when
necessary, should be used wllh care. Dlscussion of ·error analysls", whlch used to be
lncluded in many sclentlñc papers, should have been enlilled "anatysís or the probable limits
of error", or more properly, "analysis of uncertainty". In older publications the term "error" was
widely used when 'uncertainty' would nave been the correct term.

Uncertalnty Is not the same as error. 11 the condilions In a test chamber are measured wlth a
calibrated instrument and the result is 75 % RH when the chamber controller says 90 % RH,
that does not mean the uncertainty is 15 % RH. lt is known that the relative humidity is 75 %
RH. One is aware that either the controller reading is wrong or the chamber is operating
incorrectly. lt has an error estimated to be 15 % RH. The uncertainty is a characteristic of the
measurement that gave the answer 75 % RH. Could that be wrong and, if so, by how much?

Licensed 10 H-1 Co. LTD 2013-07-18. Pcy formol reproci.oclion and redistribulion are strictly prohibiled.
- 12 - 60068-3-11 e IEC:2007

When considering "true value", uncertainty, and error, one of the most important sources of
this type ol inlormation lor a measuring instrumenl is its calibration certiñcate. 11 is vital to use
all of the information provided by lhe calibration certificate lo ensure that the best estímate of
the test uncertainties are obtained.

4.2 Statements of uncertainty

4.2.1 General

When reporting lhe resuns of a measuremenl, three numbers are necessary for a
metrologically corree! and complete statement of the result ol each measuremenl point.
For example, the complete slatement could be:

The "true vetue" Is: 39, 1 'C ± 0,3 K with 95 % conlldence:

39, 1 'C Is the best estímate of lhe true value;


±0,3 K is the conlldence lnterval;
95 % Is the confidence íeveí.

An explanalion of these three components follows.

4.2.2 Best estimate of the true value of the measured quantity

Oflen this wíll simply be the reading on lhe calibrated reference ínslrumenl whích, in lhe case
of a climatic test, could be the temperature measurement system and/or hygrometer reading,
or íl lhe chamber has been calibrated it could be the chamber controller display. lf the
calibration shows either lor an instrument or lor a chamber controller lhat an error exists
(which is not an uncertainty), this should be used lo apply a correction. For example, if the
calíbration ol a thermometer shows that it reads 1 K hígh, 1 K should be sublracted from the
reading to obtain lhe best estímate ol lhe true value.

4.2.3 Confidence interval

This is the range ol measured values wíthín which the "true value" líes with a gíven level of
conlidence. In our example this inlerval ís ±0,3 K.

4.2.4 Confidence level

The ·conlidence rever of a measuremenl is a number (e.g. 95 %) expressing lhe degree of


cenneence ín the result. Thls Is tne probabillty that the real •true value" lies In the glven
range. Most seis of dala are normally dlslributed and about 68 % ol lhe values wlll fall wllhln
plus or minus one standard deviation of the mean. About 95 % ol the values can be expected
to fall wlthln plus or mlnus 2 standard devlations (95 % conhdence level). Put another way,
when many such measurements are performed nol more than 1 in 20 wíll lie outside the
stared límlts. Hence mulllplylng the standard devlalion by 2 ís an accepted way er
encompassing 95 % of the ranga ol values. With a 95 % confidence level, we are 95 % sure
that the "true value" lies ín the stated range.

lt is convenllonal lo work at the 95 % confldence level. Higher confidence levels can be used
but the confidence interval will increase.

4.2.5 Statement of uncertainty

In the above example the staternent ol uncerlainly is that lhe temperalure was 39, 1 ºC± 0,3 K
with 95 % confidence. 39, 1 'C was the best estímate ol the temperatura but because of the
uncertainties there is a possibility of it being in the range 38,8 'C to 39,4 'C with a confidence
of 95 %.

licensed to ltf Co. LTU 2013-07-18. Any formol reproduction and redistril:lJtion are strictly prohibited.
60068·3·11 e IEC:2007 - 13 -

4.3 Comblnlng uncertalntles

Uncertainty contrlbutions shatt be expressed in similar terms before they are combined. They
shatt be in the same unlts and at the same leve! of confidence. Ali contributlons should be
converted into standard uncertaintles (i.e. having a confidence level of plus or mlnus one
standard devlation). Thls is dlscussed further in Clauses 9 and 10.

5 Tolerance

When a test item is to be conditíoned one oí the firsl queslions asked is, "Will the charnber
achieve and maintain the required conditions?" This is asked since the test specification will
often set a tolerance for the required condition e.g. :t2 •e and :t5 % RH. In deciding whether a
tolerance is met, the uncertalnty in the measured chamber performance shall be taken into
account.

Tolerances are not the same as uncertainties. Tolerances are acceptance limits which are
chosen for a process or product. Most often the aim of knowing the uncertainty in a chamber's
performance is to decide whether a toterance is met. In deciding this, the deviation from the
required condition, together with the uncertainty, shall be considerad. Using the values cited
in 4.2.5, it is certain to within 95 % that the true temperature is between 38,8 •e and 39,4 •c.
lf the required condition is 40 ·e ± 2 K, then the probability that the true temperature lies
within the tolerance is considerably better than 95 % because the entire confidence interval
lies within the range of the tolerance.

11 the measured humidity Is 81, 7 % RH, and the confidence interval is :t3,6 % RH at a 95 %
confidence level, then it is certain to within 95 % that the true humidity is between 78, 1 % RH
and 85,3 % RH. 11 the requlred condltlon Is 85 :t 5 % RH, even though the measured condltion
is within thls range, tne probablllty tnat the true humidity is within i5 % RH of the set polnt is
significantly less at a 95 % confidence level, because the enlire confidence interval does not
lle within the range of the tolerance. However. from the uncertainty, there are statlstlcal
methods for maklng a good estlmate of how likely this is.

6 Humidity and temperature measurement

When taking humidity measurements there are many ways of approaching the situation. lt is
generally assumed that the water vapour content of the air is uniform throughout the chamber.
This is a reasonable assumption. and people who have performed measurements of humidity
at multiple polnts in a charnber can confirm that this is normally the case. However, this does
not mean that the relative humidity is uniform.

Oew point, being directly related to vapour pressure. can be assumed to be uniform across
the chamber and is not affected by temperature. lt may be that during routine tests, humidity
measurement is only made in one place. However, at sorne point, either during the test or
when the chamber is operating under similar conditions, humidity measurements shall be
made in at least two places so that an uncertainty can be assigned to the assumption that the
vapour content of the air is uniform.

For most environmental tests, the required humidity is specified in terms of relative humidity.
The importance of relalive humidity arises because the behaviour of most organic rnaterlals
depends on this parameter. Factors such as physical expansion of plastics and wood,
biological activity, electrical impedance and corrosion rates are examples of processes that
are affected by the relative humidity.

In a chamber the vapour pressure Is often nearly unlform.

Llcensed to Hi Co. LTQ 2013-07-18. kty formo! reproductlon and redistrlbutlon are st rl ctt y prohlblted.
-14 - 60068-3-11 O I EC:2007

Even when the air Is lhoroughly stírred there are orten temperatura dlfferences from place to
place in chambers and, although the water vapour pressure is oflen nearly uniform,
the temperature differences cause differences In the relativo humldlty. A single
humidity measuremenl at only one locatoon is often suííiclenl lo tell us aboul the vapour
pressure in the rest or tne chamber. The single measurement should be made al a central
polnt or on the lncldent alr slde o! the object under test.

The measurement can be made with any hygrometer, but normally it will be one of three
types:
a dew-point (dp) hygrometer (mirror condensation);
a psychrometer (weUdry); or
a relative humidity probe.

Examples are shown in Annex A.

7 Methods for determinlng climatic test chamber uncertaintles

There are three basic methods for determining conditions in a climatíc chamber. These three
methods reflect the different requirements in diííerent types or testing and there are good
reasons ro, each approach. These methods are illustrated in Figure 1.

Licensed to H-1 Co. LTQ 2013-07-18. Nff formot reproduction and redistribution are strictly protibited.
60068-3-11 C IEC:2007 -15 -

� 1
e� ..
i!! �
e¡ t )Ji
.. - l �
u.
- -g

�1
Eo
njj S"'�
!o
z
i�
-g ;; �'i�
8
ªJ�
.1:- -
§¡ r e
.!!
3
u
..
¡¡
u

¡i¡ '!:

11
;¡: E f

e
·¡¡¡
.
t:
u

1
N_8l ll
-.!:-
º
�!i·.ll
'.g E i
'ª aí
1t
!?
Jh
e
. ..
:,

...o..
��
.. s�.. E
a
6 o--
r ::,

E Í"' tjl¡i l� � .8
.
� i 8i
&�1:¡¡
4' 6 ..
E
�::: =- e
�e: . !!
(/) 8 á ..

.;
u

ITT
. •e
;¡;
.2
..
Ji 1�

- 8'
..e
s:


u

�i ¡U 1
!

h
;j
¡.ll i
l'l

I¡ <>.
<>.
<(
1

i t l:H . ti
!

l!J j e:,
"'
¡¡:
{Bsl
¡t§ifl
.§ �
� 18' 1 :a
� IF l'l

i
º•·e- g� �
.8 ��
o. !Z

1 f:j
1 8. u "' !? e
.... .. 8 � ii
l u.
o ,: g
o.
E

.3

Licensed to Hi Co. LTO 2013-07-18. lv,; to,mot reproduction and redistrituion are strictly prohibited.
- 16 - 60068·3·11 C> IEC:2007

7.1 Empty chamber

7.1.1 Advantages

Advantages are as follows:

a) The entire working space is calibrated.


b) Calibration need be carried out only once or twice a year.
e) Re-calibration is not required when the test sample is changed.
d) The suitability of the chamber can be assessed without subjecting the test sample to
conditioning.
e) Relatively low cost. Only one set of calibrated instruments required for many chambers.

7.1.2 Dlsadvantages

Disadvantages include:

a) The effect of the test sample Is difficult to quantify, although it may be negligible for
samples that are very smau comparad with the chamber. lt is very difficult to assign an
uncertalnty to lhe effect or the load.
b) The effecl of heat-dlsslpatlng test samples Is very hard to quantlfy.
e) Droft, resolution, and repeatabllity of the chamber controller shall be assessed and thelr
contributions to tne uncertalnty catcutatlons shall be lncluded.

7.2 Typical load

Calibration of the chamber wilh a typical load is ideal where very similar tests are repeated.

7.2.1 Advantages

Advantages are as follows:

a) The affect of the load on the control of the chamber can be accurately assessed without
subjecting the test sample to an unknown stress.
b) The smallest suitable chamber that produces satisfactory conditions can be
chosen prior to test.
e) Careful positioning of the sensors can give detaited information about critical parts of
load.

d) Anomaties from dissipating loads can be quantified.


e) Relatively low cost. Only one set of calibrated instruments required for many chambers.

7 .2.2 Dlsadvantages

Dlsadvantages lnclude:

a) Re·callbratlon Is requlred when the test sample Is changed slgnlflcantly.

b) Drift, resolutlon and repeatabillty of the chamber controller shall be assessed and thelr
contributions to the uncertalnty calculations shall be included.

licensed to 1-H Co. LTQ 2013-07-18. Any f orrnoí reproduction and redstribt.lion are strictly prohibited.
60068-3-11 e IEC:2007 - 17 -

7.3 Measurement of conditlons In the chamber durlng the test

7.3.1 Advantages

Advantages are as follows:

a) This method gives the best estimate in the measured value of the conditions
experienced by the item under test. lt is ideal when different kinds of loads and different
tests are being performed.
b) The effect of the load on the control of the chamber can be accurately assessed.
e) History of cnamber calibration drift need not be assessed.
d) Careful positioning of the sensors can give detailed information about critica! parts of
the load.
e) Anomalies from heat dissipating loads can be quantified.
1) This method can be economical because the chamber is not calibrated for conditions
that are not required.

7.3.2 Disadvantages

Disadvantages include:

a) Measurement equipment is required for every test.


b) Uncertainty calculations shall be made for every test.
c) Can be the most expenslve method because measurement equlpment is required all the
time.

7.4 Conditions to measure

lf measurements are made at the time of the test, then an uncertainty can be calcutated for
that condition. Alternatively, a calibration of the chamber could be performed for each
condition íor which the chamber is to be used. However, in practice it is not always necessary
to perform a calibration at every possible conditlon.

11 measurements are not made at the time of the test, the whole of the
measurement sequence and the analysis shall be repeated for a set of conditions that cover at
least a range of use. For evaluation an example is given in IEC 60068-3-6.

For temperature only {í.e. humidífication OFF) this should include sufficient measurement
points to cover:

the highest temperature;


the lowest temperatura;
at least two temperatures w,th the cooling ON;
at least two temperaturas w,th the heating ON.

In addition to the temperature-only measurements above, measurements should be performed


for at least two humldlty values, coverlng the ranga, for any of the above condltlons where
humldlty tests are to be performed.

ll Is necessary to perform so many measurements because each of the humidity and


temperatura control systems can cause the chamber to have different gradients and
fluctuations. The temperatura control is often much worse when the humidity system is on.

Licensed to 1-H Co. LT!l 2013-07-18. />ir¡ formot reproduction and redistribution are strictly prohitited.
- 18- 60068-3-11 e IEC:2007

lf the chamber ís only used ata few specífíc set poínts then only these need to be calíbrated.

When the test is not performed at one of the calibrated levels, it is necessary to interpolate
between two calibrated levels. lnterpolatíon should be used wíth caution and preferably only if

the calíbrated levels are reasonably close to the test level,


the services used for each calíbrated level (refrigeration, dehumidífiers, heaters, etc.) are
the same.

7 .5 Measurements required

The measurements requíred are the same for all the methods; it is simply a matter of when
the measurements are made and how the results are analysed.

7.5.1 Temperature

For temperalure measurement, an array of temperature sensors Is used to measure the


temperatura at points distributed around the chamber. There are other standards
IEC 60068-3-5 that glve measurement methods but they do not address the consequences of
uncertalnty and only refer to empty chambers.

For an empty chamber, eight sensors are normally used at tne corners of the worklng space,
and a nlnth In the centre. For large chambers more sensors may be requlred.

For a typical load, oran ítem under test, eight sensors, one at each comer of the object,
are usually used. For very small test items fewer sensors may be sufflclent, but at teast
tour should be used. For large or unusually shaped objects, or where sorne particular point
on the test ítem is of special interest, extra sensors should be employed as appropriate.

For a heat dissipating test item, the measurement of the incident air temperature is
usually considered to be the condition of interest for the report but the other sensors should
still be used so that the local effects of the heat from the test ítem can be quantifíed.

7.5.2 Humidity

For humidity measurement, a hygrometer is positioned centrally on the incident aír side of the
test ítem or in the centra of an empty chamber. Thís can be any kind of hygrometer, but ís
most likely to be a relative humidity sensor, a psychrometer, ora condensation (chillad mírror)
hygrometer. The vapour pressure can be computed from the humidity and temperature
measurements. The vapour pressure ís assumed to be the same everywhere in the chamber
and the relative humidity is computed from this vapour pressure and the temperatura at each
of the temperature measurement sensors.

For each condítion, a measurement of vapour pressure gradíents shall be made so that an
uncertainty due to this variation can be calculated. This can be done using several
hygrometer probes of any type. However, relative humldlty probes and psychrometers are
also sensitive to temperatura so usually the estímate obtained using these instruments will be
larger than the true value.

Licensed to 1-H Co. LTQ 2013-07-18. />i'f formof reproduction and redistribution are strictly prolitited.
60068-3-11 e, IEC:2007 -19-

Another method is 10 sample from severa! points through tubing routed 10 a single hygrometer
and swltched In alternallon.

The vapour pressure gradlent Is normally small and need only be evaluated occaslonally.

7 .5.3 Recordlng procedure

To ensure that a valid assessment can be made,at least 5, and preferably 20 or more
recordings should be taken from each sensor at each set condition. The recordings should be
taken over a sufficient period of time so that severa! control fluctuations of the chamber can
be recordad. A period of 30 min is normally sufficient.

Recordings are taken from the array of sensors alter the chamber has stabilized al each set
condition.

Measurements should be taken frequently from each sensor throughout the test period. Table
A.1 shows a typical data set. together with some of the analysis.

lt is essential to ensure that the intervals between the measurements do not coincide with the
cycling interval of the chamber.

7 .6 Sources of uncertainty

In any measurement, uncertainties arise from tour basic sources.

7 .6.1 Callbratlon uncertalntles

Callbratlon uncertalntles for the instruments used are listed on the calibratlon cerllflcate.
These are normally at the 95 % confidence level. When lnterpreting uncertalntles stated on a
callbratlon cerllficale, care shall be taken to conslder ali aspects of uncertalnty mentloned,
lncludlng lnslrument resolution and snort-term changas noted during callbratlon, as well as
the measurement uncertainty.

7 .6.2 lnstrument uncertainties

lnstrument uncerlainties include factors such as the resolution, repeatability and drifl of
the instruments used. Repeated measurements can guard againsl gross errors and give
improved confidence in the estimated uncertainties.

7.6.3 Uncertainties arising at the time of the measurement

The dominant uncertainties at the lime of the measurement usually concern the gradients and
fluctuations in the conditions. The measurement method shall identify these gradients and
fluctuations.

7 .6.4 Uncertainty by radiation

Radlallon effects can be large in some chambers. 11 the temperature sensors glve
measurements which are unexpectedly large or vary from test to test or if there is any reason
to suspect lhat radiation could be a problem (e.g. 11 the temperatura or any part or the
chamber whlch can be sensed by the test ítem is slgnificantly dlfferent from the set
temperatura). lhen an extra test should be per1ormed with temperalure sensors which have
dlfferent radlatlon colours (e.g. a shiny one next to a black one). For temperaturas above
+100 'C. the radlatlon errect has more and more influence on the measuring sensor and the
speclmen.

Llcensed to H-1 Co. LTQ 2013-07-18. krf formof reproá.Jction and redstributlon are st rl ctl y prohlbited.
- 20- 60068-3-11 e IEC:2007

7.6.5 Additional uncertainties

Any other factor which may affect the measurement, for example, differences between the
conditions under which calibration takes place and those in which they are used, shall be
taken lnto account.

lmported calibrations may not be fully relevant to the conditions of use. For example. most
calibration authorities calibrate temperatura sensors in baths of oil, salt or solvent. The tips of
the sensors are lmmersed, leaving the cables exposed to room temperatura, whereas in use
both tne sensors and cables may be immersed in the cnamber environment. Consequently, at
callbration there is a heat conduction effect which is not there in use. This effect shall be
considered and the magnitude of the effect evaluated, or estlmated if this is not possible. The
uncertainty In the evaluated (estimated) correction should then be lncluded In the uncertalnty
budget.

7.7 Essential contributions of uncertainty

The following covers the most importan! sources of uncertainty but is not exhaustiva. Creating
a list of sources of uncertainty is the first step in any uncertainty analysis. The list should
identify all sources of uncertainty, however sorne of them may be thought to be insignificant.

7.7.1 Reference instrument calibration uncertainty

This is the uncertainty stated on the calibration certificate, not to be confused with the "error"
or "correctíon" stated on the certificate. Several reference measurements may be involved
with the calibration uncertainty of each contributing to the overall measurement uncertainty.

7.7.2 Reference instrument drift

Most instruments change slowly with time. Conditions of use may affect the rate of change
and with sorne instruments the effect can be significan!. Data from successive calibrations are
essential to estímate the limits of drift in use, but initially manufacturer's data may be used for
an estímate. However, generally manufacturers quote drifl under ideal conditions so usually a
larger figure should be used. The drifl of all instruments used shall be considerad.

7.7.3 Resolution

This refers to the resolution of the reference instruments used for measuring the conditions in
the chamber. The resolution of the chamber controller shall be included in lhe uncertainty
analysis but should be omitted if the measurements are made at the time of the test.

7.7.4 Temperature effects

Temperatura coefflclents can be very slgnoficant In sorne lnstruments. lnstruments belng


operated near a cllmatlc test chamber are often subjected to ambient temperature varlatlons.
Generally. even when temperatura Is kept nominally constant, temperatura effects are worth
careful evaluatlon because correctlons can then often be applled. However, sorne uncertalnty
remalns even alter correctlons have been applied.

Licensed to H-1 Co. LTD 2013-07-18. Pcy formof reprociJction and redistribution are strictly prohibited.
60068-3-11 © IEC:2007 - 21 -

7.7.5 Hysteresis

Hysteresis is the difference between the output of an instrument for a rising oondition and the
output for a falling condition. Hysteresis can be large for certain types of relative humidity
probe and for sorne types of temperature sensor.

7.7.6 Measurement repeatability

Repeatability Is the closeness of the agreement between repeated measurements of the same
property under the same condilions.

7.7.7 Fluctuatlons

Fluctualions are changas in the condition being measured occurring ovar a perlad of lime.
They are present In most measurement situatlons. In envlronmental testlng they are usually
one of the largest contributions to the overall uncertainty.

Humidity fluctuatlons occur for two main reasons. The vapour content of the alr (the vapour
pressure or dew point) may fluctuate. In additlon, the fluctualions in temperature cause
fluctuatlons In relatlve humldlty, even when the vapour content Is relatlvely stable.
Both factors shall be considered and included in the uncertainty analysis.

Temperatura and humidity sensors can respond al difieren! rates and do not necessarily show
what is really happening in a chamber. To determine the extent of fluctuations many
measurements shall be taken with fast-responding instruments. Often chamber condilions
cycle in a regular pattern about the set point and if measurements are taken al the same point
in each cycle (for exam ple, always at the maximum of each cycle) the result will be
misleading. This is called 'aliasinq' and care shall be taken to detect and avoid lhis; for
example, by sampling at least tour limes faster than the chamber cycle or by sampling al
random intervals.

The effect of fluctuations on the calculation of the mean may be reduced by taking a number
of readings and calculaling the uncertainty of the mean. The improvement in the result is
a function or !he square root er the number of readings and 20 readlngs will normally
be sufficient.

A fast responding (small) sensor will capture the peaks and troughs of the cyclic control of an
climatlc chamber. A slow responding (large) sensor may also show the cycllc nature or the
control, but the recordad paaks and troughs may be compressed due to the stow response of
the sensor. Consequently, an uncertalnty componen! may havo to be included to account for
the inabllity or the sensor to detect the peak temperaturas.

Temperatura fluctuatlons are lllustrated in Figure 2 (below).

7. 7 .8 Gradlents

Gradients are the varlatlons of the measured quantity from place to place. In
most environmental tests, temperature gradients are the largest single source of
uncertainty. To estímate the síze of temperature gradients. measurements shall be
taken from different regions around the item under test or around the chamber working space
when calibrating an empty chamber.

Llcensed to Hi Co. LTQ 2013-07-18. Nr¡ formo( reprowctlon and redistrlbution are strlctly prohlbited.
- 22 - 60068·3·11 C IEC:2007

Gradients in the vapour content of the chamber air are usually comparatively small although
thls may not be true when there Is condensatlon anywhere In the chamber. However, as wlth
fluctuatlons, gradients in temperature cause a gradient in the relativa humidity even when the
vapour content Is comparatlvely constant.

7.7.9 Additlonal uncertainties when calibratlng an empty chamber ora


chamber with a typical load

These can include:

- drift of chamber instruments;


- repeatability of chamber instruments;
- resolution of chamber instruments.

A major consideratlon Is the effect of the test sample on chamber conditions. Unless the size
of the test sample is inslgníficant compared to the charnber working space, then consideration
should be glven to uslng one of the other methods to estabüsh an uncertalnty contributlon.

Particular consideration should be glven to heat-dlsslpating loads.

Licensed to Hi Co. LTll 2013-07-18. MJ formol reproduction and redistritulon are strictly prohibited.
60068-3-11 ,O IEC:2007 - 23 -


11
¡¡

f�.
-
f'i
_ L[
li ...
i
1;:
--i
; ':
� ª
o
".e,'
.";;,'
... �.,
'& N

J!1 g Q.
E
:5 �
o"'e
s1;j
� i=
�.,"
¡:, !
..
N
e
e � s
;
.��
1¡¡ N

j 5l

e::,
e
i
>
"'
¡¡

"e:l
;:¡¡
��"'
i:?
...
i
...
5l
...
..,¿ g
,¿
i.:
... ...;;
8

g ...

...., "'
o,

Llcensed to Hi Co. LTll 2013-07-18. MJ formo( reprOCklCtlon and redistrlbutlon are strlctly prohlbited.
- 24 - 60068-3-11 e IEC:2007

8 Estimation of uncertainty components and their combination

The evalualion and combination of ali the uncertainlies present in a measurement is most
convenien\ly done using a spreadsheet. The following is a typical example based on
evaluatlon of the overall uncertainty of temperatura maasurament as presentad in Clause 9.

The table includes examples rrom dillerent types er uncertainty source such as callbration
uncertainties from cerlificates. uncertainlies from random effects in sets of repeated
measurements. and systematic uncertainlies such as biases and drifts which are eslimated or
calculated.

The evalualion of tne overall uncertainty of a measurement or calibratlon Is a three-step


process:

Each uncertalnty source shall be identified. quantified and represented as a standard


uncertalnty.
The combined elfect or the individual components is round as the square root er tne sum
of their squares. This is known as the combinad standard uncertainty.
This, in turn, is then multiplied by a coverage factor of k = 2 to give the expanded
uncertalnty, whlch may be interpreted as a confidence lnterval that wlll lnclude the true
value of the measured quantity with a probability of 95 %.

The process of ldentllylng and quantifying the sources of uncertainty that are lncluded In this
analysis are discussed in Clause 9 for temperature chambers. and in Clause 10 for combined
temperature and humidity chambers.

9 Overall uncertalnty of temperature measurement

9.1 General

Table 1 below illustrates the method of combining uncertainties using the data sets in Table
A.1. Each uncertainty source is described in detail in the text following Table 1. Thls method
can be used for temperature chambers and is also used for combined temperature and
humidity chambers.

Table 1 - Combination of temperatura uncertalntles

Uncertelnty 1ource v,1u, Dl1trlbullon Divisor Stendard Squ1r1d


·e unc,rtalnty
- -
Celibralion 0,100 Nonnel 2.00 o.oso 0.002 soo
¡;
•e
>-- - 0,010 -
1,00
- 0,010
-
o.ooo
-
100
Repeatabilliy
- Nonnel
-- - º·ººº -
.,ª. Hy1u1re1la
- -
1,73
0,010
- Rec1en9Uler
- 0.006
- -
033

Temperaiure ertecll 0,010 Rec11ngular 1,73 0,006 o.oco 033


!! o.ose
� -Orlfl 0,100 Rectangu1ar 1,73 0,003 333
- o.ooo 133-
..
.!! llnearhy 0,020 Rec11n9Uler 1,73 0,012
"' Resolutlon 0,010 Rec1anguler 1,73 0,006 o.ooo 033
-
Temperature 0,469 Normal 1,00 0,469 0,219 961
;¡;
Gradlent
..
D
e
o
,: Temperature fluctuatlons 0,061 N0<mal 1,00 0,061 0,003 721
overen mean 0,026 Normal 1,00 0,026 0,000 676

Licensed to H-1 Co. LTQ 2013-07-18. Pcy formof reprociJction and redistribution are strictly prohibited.
60068·3·11 e IEC:2007 - 25-

Sum of squares 0,230 525


Comblned standard uncertalnty ±0,480 K
Expanded uncertalnty at 95 o/o coníldence ±0,96 K

The reference lnstrument calibratlon uncertainty, ±0, 1 K. Is the uncerta,nty stated on the
calibration certlficate of the thermometers. Calibratlon unoertalnty can be assumed to be a
normal distributlon because lt Is the sum of many components from the calibralion chain. The
uncertainty is quoted al the 95 o/o confidence level so that the divisor 2 is needed to obtain the
standard uncertainty.

Repeatabílity should be measured by a prior experiment if possible. 11 the value is calculated


from a set of repeat measurements, in this case ±0,01 K. the repeatability of the set of
readings will form a normal distribution characterlzed by a mean value and an estimated
standard deviation. The standard uncertainty is equal to the standard deviation (i.e. the
divisor has a value of 1 ). Alternatively, the value may be taken from the thermometer
manufacturer's specification and treated as a rectangular probability distribution as described
in the nex:t paragraph.

Hysteresis is the difference between the output for rísing and falling temperaturas. The value
of ±0,01 K, is taken from a callbratlon certificate. The first time the thermometers are
calibrated, rising and falling measurements are requested in the mid range to determine this
value. A rectangular distribution is assumed and the limits are divided by the square root of 3
to obtain the standard uncertainty.

Temperature errects arlse from the temperature coefficient of the temperature measuring
instrument electronics and the temperature variations they experiences during use. From the
manufacturer's specification (0,001 K/K) and the conditions of operation (20 t10) •e it is
calculated that the effect cannot exceed ±0,01 K. There is no way ol knowing how likely it is
to be operating at the centre or at the limits of this range, so it is assumed to be a rectangular
probabllity dlstrlbution. The (equlvalent) standard deviation is obtained from the rectangular
distribution's llmlts by dlvldlng by the square root of 3. The standard uncertalnty Is equal to
the standard devialion (the divisor has an overall value of square root of 3).

Dr,ft of the reference lnstrument can be estlmated by comparing the current and prev,ous
calibration certlflcates. In the case clted above a change of 0, 1 K had occurred between
callbrations. lt Is assumed that any subsequent changa could be in either direclion and of the
same magnitude. A rectangular dlstrlbutlon is assumed and the lim,ts are dovided by the
square root of 3 to obtaln the standard uncertalnty.

Non ·linearity is the magnitude of devlatlon of the sensor or instrument output from a straight
line catcutated wllh a linear regresslon method. 1t is assumed that the deviation could be in either
direction and equally probable wlthln the establlshed non-linearity value. Therefore, a
rectangular distribution should be assumed with a value of ±0,02 K and the limits are divided
by the square root of 3 to obtain the standard uncertainty.

The resolution, 0,01 K in this case, is an uncertainty associated with digital displays. lt is not
possible to tell whether the display is just about to change up or down to the next digit simply
by observation. The true reading is equally likely between the upper and lower limits and so a
rectangular distribution is assumed and then the limits are divided by the square root of 3 to
obtain the standard uncertainty. With sorne instruments it can be argued that hall of this figure
can be taken (i.e. t0,005 K).

licensed to l+I Co. LTO 2013·07·18. M¡ formol reproduction and redlstrlwion are strlctly prohibited.
- 26- 60068-3-11 e IEC:2007

The temperatura gradient can be derived from the data in Table A.1. At each measurement,
the average and standard deviation of the eight sensors has been calculated. The standard
deviation is a measure of the temperature gradients. The safest assumption is to take the
largest value of standard deviation, namely t0,469 K. The input to the uncertainty budget is a
standard devlation from repeated measurements and so can be assumed to be normal and
the divisor is 1. lf there is a large variation between the standard deviations it may
be approprtate to take the largest difference and treat the range obtained as a
rectangular dlstrlbution.

Fluctuations are also derived from the <lata in Table A.1. For each sensor the average an<I
stan<lard devlatlon over the perio<I of the test has been calculate<I. The stan<lard devlatlon Is a
measure of the lluctuations. The safest assumption Is to take the largest value of standard
devlation, namely t0.061 K. Again, it is a standard deviation so it can be assumed to be
normal and the divisor Is 1. As wlth gradlents, lf there os a large varlatlon between the
standard deviatlons it may be approprlate to take the targest difference and treat the range
obtalned as a rectangular <listribution.

Finally, the overall mean of all measurement data in Table A.1, 39,8 •c. is statistically basad
and an allowance shall be made for the possibility that it may not represen! the true average.
The value input into the uncertainty budget, 0,026 K, is actually the overall standard deviation
given in Table A.1, t0,397 K. divided by the square root of the number of measurements,
(240). This component, reíerred to as the "standard deviation of the mean", may appear
negligible but is included for completeness. lf only a few measurements were made, it could
become significan!.

The result of this analysis for measurements taken with calibrated reference equipment during
the test is that the following statement can be made:

The mean temperature in this example was 39,8 t 0,96 C at a level of confidence of 95 %.

When performing an uncertainty analysis there are always assumptions to be ma<le, as


indicated In the prevlous clauses, and these should always be clearly stated. In general, small
residual errors in these assumptions will not make a significan! difference to the result, unless
they concern one of the two or three largest uncertainty components.

9.2 Further considerations

II the above example relates to measurements made al the time of test, the uncertainty in the
temperatura is t0,96 K.

II the measurements are made wlth a typlcal load, then lt Is also necessary to take lnto
account the chamber resolutlon and the chamber drift. These need to be combinad (by
summlng In quadrature) wlth the measurement uncertalnty to obtain the overall uncertainty of
the condltlons experlenced by the ltem under test.

11 tne measurements are made during calibration of an empty chamber, a further addition to
the extra uncertainty contributions mentioned above is required to take account for the
possible effect of the load.

10 Overall uncertalnty of relative humldity measurement

Before evaluating the uncertainty in the measurement of relative humidity at the time of
measurement, it is necessary to evaluate the uncertainty in the measured temperature al the
sensor points, as these temperatures are used in the calculation of relative humidity at those
points.

Licensed to H-1 Co. LTD 2013-07-18. />n¡ formal reprod.oction and recistribution are strictly prohibited.
60068-3-11 e IEC:2007 - 27 -

10.1 Uncertainty of temperature measurement at each sensor point

Table 2 covers the evaluation of the temperature uncertainty al points in the chamber which
will then be included in the evaluation of the humidity uncertainty al !hose points. The
uncertainty components involved and the analysis are very similar to those in the example
given in Clause 9, except that the uncertainty contributions for the effect of gradients and
the uncertainty in the overall mean are omitted. The combination of the standard
uncertainties is most conveniently carried out using a spreadsheet.

Table 2 - Comblnatlon ol temperature uncertalntles at each polnt


-
Uncertainty aource V•lue Ol1tribu1ion Divisor Standard Squared
1( uncertalnty
Calibratlon 0.100 Normal 2.00 0.050 0.002 500
"s
$
E
Ropeatablhty 0.010 Normal 1,00 0.010 º·ººº 100

.5
Hy1ter11l1 0.010 R.c.t1ngul1r 1,73 0.006 º·ººº 033
8e
Temperatura 1ff&cl1 0.010 R.c.t1ngul1r 1,73 0.006 º·ººº 033
� Ortfl 0.100 R.c.t1ngut1r 1,73 0.058 0.003 333
.!! Llnearlly 0.020 R.ctangular 1.73 0.012 o.ooo 133
$

°' Resolution 0.010 R.ctangular 1,73 0.006 o.oco 033


Chamber tempereture 0.061 Nonn1I 1,00 0.061 0.003 721
fluctuatlons

Sum of squares 0,009 853


Combined standard uncertainty t0,099 K
Expanded uncertainty al 95 % confidence t0,20 K

10.2 Uncertainty of the relative humidity measurement

In this example, each relative humidity in Table A.2 has been calculated from the dew point
{left hand column of Table A.2) using a single humidity sensor and the temperature of the
corresponding individual temperature sensor in Table A.1.

This analysis considers the uncertainty budget for relative humidity measurement. Generally
the assumptions made in identifying the uncertainty sources and the way the uncertainty
figures are obtalned is similar to the first example, but there Is one lmportant dllference: many
of the values included in the uncertainty budget come from uncertainties in temperature or
dew point. AII of the uncertainty values and standard uncertainties shall be converted into
relativa humidity units before they can be comb,ned.

In our example, the conditlons are nominally 40 •e and 85 % RH. Al these condltlons a
change In temperatura of t0, 1 K corresponds to a changa of t0,45 % RH. A changa in dew
polnt temperature of t0, 1 K has almost the same effect. So, to change from an uncertalnty In
terms of dew polnt or temperatura to an uncertainty in terms of relativa humldity, it Is
necessary to multiply by 4,5 %/K. This factor Is sometimos callad the sensltivlty and should be
ca lculated for ea ch condition.

To cetcutate the sensltlvlty, celcutate the relativa humidity from the dew point ano temperatura
and then repeat the calculation changing the dew point or temperature by 0, 1 K. The result
will give the sensitivity.

Licensed to l+I Co. LTD. 2013-07-18. ktf formot reprociJction and redistrirution are strictly prohibited.
- 28 - 60068·3·11 C> IEC:2007

Table 3 - Comblnatlon ol humldlty uneertalntles

%RH uncert1lnty
-
No,rmal
>----
2,00
>----
0,450
-
0,202 500
lnslrument catlbratlon 0.900
¡; - - - ..-- -- -

Repea1ablll1y 0.225 1.00 0,225 o.oso 625
.i NOfmal
Hysteresls 0.045 R&etangular 1.73 0,026
º·0,016
ººº 675
.
8
.!
Temperature effect
Hyg,ometer drift
0.225
0.450
Rectangular
Rectangular
1.73
1.73
0,130
0,260
900
0,067 600

• Unearlty 0,225 Rectangular 1.73 0,130 0.016 900


"' Resolutlon of hygrometer 0.450 Rectangular 1.73 0,260 0.067 600
Humldlty fluctuatlons 0.755 Normal 1.00 0,755 0.570 025
Humldity gradlents dueto 2.130 Normal 1.00 2.130 4,536 900
.; temperature
o
E Vapour pressure gradient 0,900 Rectangular 1,73 0,520 0,270 400
.:!:
(J Temperature uncertainty
effect on humidily
0,896 Normal 2,00
º·"ª 0,200 704

Overall mean D.124 Normal 1,00 0,124 0,015 376

Sum of squares 6,016 205


Combinad standard uncertainty t2.453% RH
Expanded uncertainty at 95 % conlidence t4,9 % RH

The rererence lnstrument eallbratlon unoertainty, t0,2 K dew point temperature, Is the
uncertainty stated on the calibration certilicate ol the hygrometer. 11 is multiplied by the
sensltivity ractor (4,5) to convert it to a relativa hum,d,ty er 0,900 % RH. Calibration
uncertalnty can be assumed to be a normal distributlon because lt Is the sum er many
components írom the calibration chain at 95 % confidence level. The divisor is 2 to obtaln one
standard uncertalnty.

Repeatabillty should be measured by a prior experiment 11 posslble. 11 the value Is calculated


from a set o! repeat measurements, In this case ±0,05 K dew point temperature, the
repeatability will be normal and it is multiplied by 4,5 to convert it to relativa humidity. The
divisor is 1 to obtain one standard uncertainty. Alternatively, the value may be taken lrom the
lnstrument manulacturer's specification and treated as a rectangular distribution.

Hysteresis is normally negligible ror oondensation hygrometers, but lor relativa humidity
probas it can be several per cent. lt is determinad by specifying rising and lalting rneasure-
ments al the mid range point during the first calibration. In this case the calibration shows no
significan! hysteresis but ±0,01 K dew point temperatura has been assumed as a precaution.
lt is multiplied by 4,5 to convert it to relativa humidity and it is assumed to be a rectangular
distribution and so is divided by the square root ol 3 to give one standard uncertainty.

Temperatura etrect is the temperature coelficient ot the hygrometer electronics. From the
manulacturer's specification (0,005 K dew point temperature/K), and the temperatura
variations it experiences during use (20 ±10) ·c. it is calculated that the errect cannot exceed
t0,05 K dew point temperatura. lt is multiplied by 4,5 to convert it to relative humidity. and it is
assumed to be a rectangular distribution, with the limits divided by the square root ol 3 to give
one standard uncertainty.

Ucensed to Hi Co. LTQ 2013-07-18. Mi formol reproductlon and redstrllulon are strlctly prohlblted.
60068-3-11 e IEC:2007 - 29 -

The drift of the reference hygrometer is estimated by comparing the current and previous
calibration cerlificates. In this case a change ol 0, 1 K dew point temperatura had occurred
between calibrations. lt is multiplied by 4,5 to convert it to relat,ve humidity and is assumed to
be a rectangular dlstributlon. The llmits ol the distrlbution are dlvlded by the square root ol 3
to obtain the standard uncertalnty.

Linearoty Is estlmated by looklng at the callbration cernñcate data or the hygrometer In the
same way as the thermal sensor. The non-llnearlty ol :1:0,05 K dew point temperatura is
multiplied by 4,5 to convert II to relatlve humldity. 11 Is assumed to be a rectangular
distribution and so is dlvided by the square root of 3 to obtain the standard uncertalnty.

The resolutlon or the hygrometer Is :1:0.1 K dew point temperatura. lt Is multiplled by 4,5 anda
rectangular distrlbutlon Is assumed. The limlts ol the dlstributlon are divlded by the square
root ol 3 to obtaln the standard uncertalnty.

Each relativa humldlty value In Table A.2 is catculated lrom the dew polnt In the second
column and the correspondlng temperatura measurement In Table A.1 so the relativo humldity
values include the errect ol temperatura.

Relative humidity lluctuations are derivad from lhe data in Table A.2 in the same way as lor
temperatura. For each temperatura sensor, the average and standard deviation of the relativa
humidity over the perlod ol the test has been calculated. The salest assumption Is to take
the largest value of standard deviation, :t0,755 % RH. lt is a standard deviation so it can
be assumed to be normal and the divisor is 1 for one standard uncertainty.

The humldity gradients are also derivad from the results in Table A.2. As in Table A.1, the
average and standard deviation of the relativa humidity al each of the eight temperatura
sensors has been calculated at each measurement time. The safest assumption is to take the
largest value of standard deviation, ±2, 130 % RH. lt is a standard deviation so it can be
assumed to be normal and the divisor is 1 for one standard uncertainty.

The vapour pressure gradient may be measured by sampling from different points in the
chamber and measuring the dew polnt of the sample. Measuríng vapour pressure gradients
may also be done with several relativa humidily sensors, but if il is done like that the results
are more uncertain (and usually largar) because of the effects of temperatura variations, and
probe calibratlon uncertalntles. The test showed that the gradlent was less than ±0,2 K dew
polnt temperature. lt is multlplied by 4,5 to convert lt to relative humidity and is assumed to be
a rectangular distribution. The limlts ol the dlstribution are dlvided by the square root of 3 to
obtain the standard uncertalnty.

The temperatura measurement uncertalnty al each polnt also contributes to the uncertainty in
relat,ve humldity. This value is avallable from the prior calculation (Table A.2), as evaluated
above. The expanded uncertainty, ±0,20 K, Is multlplied by 4,5 to convert ll to relativa
humidlty and Is assumed to be a normal dlstribution. The value of the combinad standard
uncertainty Is glven In Table 3.

licensed to ltl Co. LTQ 2013-07-18. Any formol reproduc:tion and recistrirution are strictly prohibited.
-30- 60068-3-11 e IEC:2007

Finally, the overall mean of all the measurements, 84,9 % RH, is based on a statistical
analysis of all lhe measurements and an allowance shall be made for the possibility thal il
may not represen! the true average. The value. which appears as 0, 124, is the overall
standard deviation, ±1,924 % RH, divided by the square root or the number or measurements,
(240), to give the standard uncertainty.

The result of lhis analysis for measurements taken with calibrated reference equipmenl during
the test is that we can make the rollowing statement:

The average humldlty experlenced by the ltem under test was 84.9 ± 4,9 % RH wlth
95 % confidence.

lf the measurements were performed to cahbrate a chamber, with or without a load, addltlonal
uncertalntles would have to be included, as mentioned in 8.1.

11 Anomalous data and presentation of results

There are two approaches to calculate the uncertainty in lhe measuremenl of the conditions
achieved wilhin a chamber. The ñrst is to base the calculation on the average condilions
within the chamber and the second is to base the calculation on the worst condition . The
choice of approach depends on the type oí equipment being tested and the requirements of
the customer.

11.1 Average case analysis

The average case analysis leads to a statement of the form: the measured condition during
the test was condition ± uncertainty with 95 % confidence. For example, 39,8 'C ± 0,96 K with
95 % confidence. In this case the condition is the average measured condition, and it may or
may not be what was required by the test.

The 'average case' approach lndicates what the actual condition was as accurately as
possible as well as giving an uncertainty. In humidity tests it will orten show that the test fails
to meet typical specifications. bul will al least indocate whal did happen and with
whal uncertalnty.

11.2 Worst case analysls

The data sncurc always be inspected for anomalies. lf, on inspection, lt Is round that one
particular sensor's output or the average output over a short period was more than three
standard devlatlons from the average, it might be appropriate to make a statement about the
worst case sensor.

The 'worst case' analysis uses only data from the sensor with the largest mean deviation from
the set point. The largest mean deviation of this sensor is added to twice the standard
deviation of its fluctuations, plus the combined expanded uncertainty of all other uncertainties.
This leads to a statement of the form: no point on the item under test was outside the range
set-polnt ± uncertainty with 95 % confidence. In this example, using dala from Tables A.1 and
A.2, this gives 40,0 'C ±1,08 K with 95 % confidence. For temperature. it is often possible to
use this type of analysis to state that the test was within lhe required tolerance, e.g. 40,0 'C
± 2,0 K, with 95 % confidence. For humidity it would rarely be possible to make this type of
stalement.

Llcensed to Hi Co. LTll 2013-07-18. IVt¡ lormot reprod,ctlon and redstributlon are strlctly prohlbited.
60068-3-11 e IEC:2007 - 31 -

Whichever approach is used, the data required is the same and the process of developing the
uncertainty budget can be a very uselul indicator or the major sources or uncertainty. 11 may
show that the temperature gradients in test chambers and the cyclical control that often
occurs are the main sources of uncerlainly. For humidity, the temperature gradients can
cause very large changes in the relatlve humidity lrom place to place within the chamber.
Near saturation at 20 'C. a 1 'C temperature difference causes a 6 % difference in relative
humidity. lndeed thls one uncertainty is often so large that the others could be ignored.

Licensed to Hi Co. LT!l 2013-07-18. /In¡ fa-mot reprociJction and redistribution are strictly prohibitecl.
-32 - 60068-3-11 e IEC:2007

Annex A
(informative)

Measurement data sets - Loaded chamber

Table A.1 - Typlcal temperatura measurement data set and it's analysls and referencos

Temperature 1ensor mea1uremen11 •e At eoch measuremenl


Time 3 5 a 7 8 Avorego SO

09:48 39,15 39.90 39,62 40,06 39,36 40,31 40,53 39,68 39,825 0,469
09:49 39.13 39.88 39.60 39.98 39.30 40,21 40,43 39.68 39.771 0,441
09:50 39,13 39,86 39,56 40,00 39,28 40,23 40,47 39,66 39,773 0,460
09:51 39,13 39,84 39,58 39,96 39,28 40,19 40,37 39,64 39,748 0,429
09:52 39,05 39,74 39,50 39,86 39,20 40,10 40,27 39,60 39,664 0,419
09:53 39,19 39,86 39,64 40,02 39,36 40,23 40,37 39,66 39,790 0,411

09:54 39, 15 39,82 39,62 39,96 39,32 40,17 40,31 39,66 39.751 0,400
09:55 39,17 39,86 39,62 40,02 39,32 40,25 40,47 39,68 39,798 0,445
09:56 39,13 39.80 39.56 39.94 39.28 40,19 40,41 39,66 39,746 0.436
09:57 39, 17 39,82 39.58 39.94 39,28 40.21 40.45 39,68 39.766 0.438
09:58 39,19 39,82 39,58 39,98 39,32 40.23 40,47 39,70 39,786 0,437
09:59 39, 15 39.84 39.62 39.98 39.32 40.21 40,39 39,70 39,776 0.423
10:00 39.17 39.86 39.6" 40.00 39.36 40.25 40,47 39.72 39.808 0.436
10:01 39, 17 39,86 39,6" 40,00 39.36 40,23 40,41 39,72 39,798 0,420
10:02 39.19 39.86 39.62 39.98 39.36 40.23 40,45 39,74 39.803 0.424
10:03 39.19 39.86 39.62 40.00 39.36 40.21 40.39 39.74 39,795 0.409
10:04 39,28 39,94 39.70 40,12 39.44 40,35 40,55 39,80 39,897 0,434
10:05 39.24 39.86 39.68 39.98 39.38 40.21 40,43 39.76 39.815 0.398
10:08 39.23 39.88 39.68 40,00 39.36 40,17 40,37 39.74 39,798 0.389
10:07 39,28 39.94 39.70 40,08 39.42 40,29 40,49 39.78 39.872 0,414
10:08 39.28 39.92 39.88 40,04 39.38 40,25 40,45 39.78 39.848 0.404
10:09 39,21 39.86 39.62 39.96 39.34 40,17 40,39 39,72 39,783 0,400
10:10 39.13 39,78 39.58 39.86 39.28 40,10 40,33 39.70 39.718 0.397
10:11 39,19 39.82 39.82 39.94 39.32 40.17 40.37 39.72 39,768 0,401
10:12 39,21 39,88 39.68 40,04 39,38 40,25 40,49 39,76 39,833 0,429
10:13 39, 19 39,86 39,6" 40,00 39,36 40,21 40,45 39.78 39,810 0,420
10:14 39,21 39,88 39,68 40,02 39.38 40,23 40,45 39,78 39,825 0,416
10:15 39,19 39,86 39,6" 39,98 39.38 40,19 40,39 39,76 39.798 0,401
10:16 39, 19 39,86 39,66 40,00 39,38 40,21 40,41 39,76 39,808 0,408
10:17 39,21 39.86 39,66 40,02 39.38 40,21 40.39 39,78 39,813 0,400

For eacn sensor


Mean 39,180 39,852 39,623 39,987 39,342 40,219 40,424 39,715
SDs 0.052 0,041 0,044 0.053 0.049 0.052 0,061 0.051
Overall mean 39,793
Overall SO 0,397

Li censed to Hi Co. LT!l 2013-07-18. /In¡ fa-mot reprociJction and redistribution are strictly prohibitecl.
60068·3·11 C IEC:2007 - 33-

Table A.2 - Humldlty measurements analysls based on Table A.1 temperaturas

Measured Calculated % RH ot eacn temperature sensors Al e&el'I measurtmtnt

Time Oew po¡nt 1 2 3 4 5 6 8 Average SO

09:48 36.85 88.32 84.84 86. 11 84. 13 87.31 82.98 82.02 85.84 85.192 2.130
09:49 36.74 87.89 84.51 85,69 83,98 87,06 82,92 81.96 85,41 84,927 2,002
09:50 36,73 87.84 84,47 85,82 83,84 87,11 82,78 81.74 85.37 84,871 2,081
09:51 36.62 87,31 84,05 85,22 83,52 86,59 82,46 81.68 84,95 84,471 1.936
09:52 36,63 87,73 84,54 85,63 84,01 87,00 82,95 82,16 85,17 84,898 1,901
09:53 36.72 87,51 84.42 85,41 83,71 86.69 82,74 82,13 85.32 84.741 1,861
09:54 37,02 89.14 85.99 86,91 85,36 88,31 84,37 83.75 86.73 86.321 1.847
09:55 36,60 87,03 83,87 84,94 83, 16 86,31 82,11 81,16 84,67 84,157 2,000
09:56 36.63 87,36 84,27 85.35 83.65 86.63 82.51 81.55 84.90 84.529 1.965
09:57 36,85 88.23 85.20 86,29 84,66 87,68 83,42 82,37 85.84 85.460 1.996
09:58 36,98 88,76 85,81 86,91 85,08 88.12 83,92 82,86 86,36 85,977 2,003
09.59 36.63 87,27 84,09 85.08 83,47 86,45 82,42 81.64 84.72 84.394 1,907
10:00 36.63 88,13 84,93 85,93 64,30 87,21 83,15 82.19 85.58 65, 175 1,980
10:01 36.59 86,98 63.62 64,81 83,20 86.08 82,15 81.38 84.45 84,108 1,887
10;02 36.84 87,13 84,05 85.13 83.62 86.31 82.38 81.43 84.59 84.317 1,908
10:03 36,84 88,08 84,97 88,06 84,35 87,28 83,37 82,58 85,52 85,275 1,865
1004 36.80 88,48 83.51 84,58 82.72 85.78 81.88 80.82 84.14 83.711 1,937
10.05 36.85 87.85 85.02 85.93 84.48 87.22 83.42 82.45 85.47 85.230 1,811

10·06 37,02 88,76 85,81 86,73 65, 18 88,12 84.37 83,48 88.36 86.103 1,789
10.07 36.57 88.33 83,38 84,44 82,78 85.71 81.80 80.94 84.09 83.883 1,848
10:08 36,80 88,48 83.60 84,67 83.07 86,03 82,11 81,25 84.22 83.930 1,808
10:09 36,59 86,80 83,82 84,90 83,38 86,17 82.41 81,46 84,45 84, 173 1,799
10:10 36.84 88,37 85,34 88,25 84,97 87.63 83.90 82.84 85.70 85.825 1,819
10:11 36,74 87,60 84,69 85.59 84, 15 86,97 83,09 82,22 85.14 84,934 1,820
10:12 36,67 87.18 84,10 85,09 83,39 86.36 82.43 81.39 84.84 84,321 1,932
10:13 36.62 87,03 83,96 84,95 83,34 86,22 82.38 81.34 84,32 84,191 1.889
10:14 36,84 87,99 84,88 85,88 84,26 87,17 83,28 82,32 85.34 85,140 1,892
10:15 37.00 88.86 85.72 86,73 85,18 87,93 84,19 83,31 86.18 86.010 1,842
10:16 36.77 87.75 84,65 85,55 84,03 86,84 83,05 82.18 85.10 84.893 1.850
10:17 36,95 88,52 85,49 86,40 84,76 87,69 83,87 83,08 85.85 85.708 1,835

For each sensor


Mean 87,690 84,594 85.632 83,987 86.931 82.954 82.058 85.212
so, 0.746 0,731 0.707 0.739 0.718 0.744 0.755 0.706

Over•II mean 84.88


Overall SO 1.924

Llcensed to lff Co. LTO 2013-07-18. AA¡ formol reproductlon and redistrll:ulon are st rt crl y prohlblted.
-34- 60068-3-11 e IEC:2007

Blblíography

BIPM. IEC. IFCC. ISO, IUPAC, IUPAP, OIML. Guide to the Expression of Uncertainty in
Measurernent. lnternational Organization for Standardization, Geneva, Switzerland. ISBN 92-
67-10188-9, First Edition 1993, corrected and reprinted 1995 (BSI Equivalent: BSI PO 6461:
1995,Vocabulary of Metrology, Part 3. Guide to the Expression of Uncertainty in Measurernent
BSI ISBN O 580 23482 7).

BS 1339-1 :2002 Hurnidily- Parl 1: Terrns, definilions and formulae.

BS 4833:1986 Schedule for Hygrornetric tables ror use in the tesling and operalion of
environrnental enclosures

EA-4/02 Expression of the Uncertainly of Measurement in Calibration. December 1999.


European co-operalion for Accreditalion of Laboratories, Rouerdam

EA-4/02-S 1 Supplement 10 EA-4/02

Measuremenl Good Practico Guide No 11, A Beginner's Guida to Uncertainty of


rneasurernent, Stephanie Bell, NPL, London, 1999

Measurernent Good Practice Guide No 36, Estimating Uncertalnlies in Testing, Keith Birch,
NPL. London, 2001

NIST Technical Note 1297, 1994 Edition, "Guidelines for Evalualing and Expressing the
Uncertainty of NIST Measurement Resutts•

Pragnell R F. An Excel spreadsheet containing the commonly required hygrometric formulae


may be downloaded from the SEE web site. www.environrnental.org.uk

Pragnell R F, The relationship between hurnidity calibration standards and standards for
environmental testing, Environmental Engineering, September 1988

UKAS publication M 3003 The Expression of Uncertainty and Confidence in Measurernent


Edition 1, Decernber 1997

UKAS publlcatlon Lab12 The Expression or Uncertalnty In Tesling, October 2000

Licensed to H-1 CD. LT!l 2013-07-18. hff formot reproduc:tion and redistribution are strictly protitited.
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 37
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 38
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 39
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 40
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 41
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 42
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 43
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 44
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 45
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 46
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 47
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 48
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 49
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 50
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 51
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 52
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 53
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 54
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 55
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 56
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 57
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

下载 收藏 72 / 72

http://www.doc88.com/p-8592318720350.html 58
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 59
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 60
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 61
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 62
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 63
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 64
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 65
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 66
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 67
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 68
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 69
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 70
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

http://www.doc88.com/p-8592318720350.html 71
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

全文阅读已结束,如果下载本文需要使用
1500 积分

下载此文档 文档不适合?发起文档求助吧~

分享至

http://www.doc88.com/p-8592318720350.html 72
/73
31/10/2017 IEC 60068-3-11-2007 - 产业标准 - 道客巴巴

该用户还上传了这些文档

8 p. 8 p. 27 p. 36 p. 34

AHAM TC-1 AHAM SC-1 AHAM SA-1 AHAM RAC-1 AHAM RAC-1
2012.01.01 2007.01.01 (R 2011) 2014.01.01 2015.01.01 1982.01.01 (R

发表评论

验证码: 换一张 匿名评论

提交

关于我们 帮助中心 关注我们


关于道客巴巴 网站声明 会员注册 文 新浪微博

人才招聘 网站地图 档下载

联系我们 APP下载 如何获取积分 关注微信公众号

道客巴巴网站 版权所有 | ©2008-2017 | 网站备案:京ICP备12004711号 京公网安备1101082111号

http://www.doc88.com/p-8592318720350.html 73
/73

Potrebbero piacerti anche