1、CANICGSB-4.2 No./N 1407 i ! TEXTILE TEST METHODS MTHODES POUR PREUVES TEXTILES PRECISION AND ACCURACY PRCISION ET EXACTITUDE OF MEASUREMENTS DES MESURES Nat i onal Standard of Canzda i .* Norme nationale du Canada Canada Copyright Canadian General Standards Board Provided by IHS under license with C
2、GSBNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-CGSB CANKGSB-4.2 NO.L-fl87 Xt = 1874650 0010b51 038 CANlCGSB-4.2 No.lN* 1407 NATIONAL NORME STANDARD NATIONALE OF CANADA DU CANADA TEXTILE TEST METHODS MTHODES POUR PREUVES TEXTILES PRECISION AND ACCURACY PRCISION
3、ET EXACTITUDE OF MEASUREMENTS DES MESURES Prepared byPrepar6e par Canadian General Standards Board lOffice des normes generales du Canada Published December 1987 by the Canadian General Standards Board Ottawa, Canada K1 A 1 G6 Winister of Supply end Services Canada - 1987 No part of Ihn publication
4、may be reproduced in any form without the pior permission of the publisher. Approved byApprouv6e par le Standards Council of Canada mm Conseil canadien des normes Publie, dcembre 1987, par lOffice des normes generales du Canada Ottawa, Canada K1A 1G6 Ministre des Approvisionnements et Secvices Canad
5、a - 1987 Aucune partie de Ite publicstion ne peut tre reproduite daune maniere sans la permission praiable de Idiiteur. C Copyright Canadian General Standards Board Provided by IHS under license with CGSBNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-CGSB CAN/CGSB
6、-4-2 NO.L-MB7 * m LB74650 0030652 T74 D The CANADIAN GENERAL STANDARDS BOARD (CGSB), under whose auspices this standard has been developed is a govemment agency within the Federal Deparmient of Supply and SeMces. The CGSB is engaged in the production of voluntary standards in a wide range of subject
7、 areas through the media of standards committees and the consensus process. The standards committees are composed of representatives of relevant interests including producers, consumers and other users, retailers, gmments, educational instiaitions, technical, professional and trade societies, and re
8、search and testing organizations. Any given standard is developed on the consensus of views expressed by such representatives. The Ministers Advisory Council on CGSB reviews the results of the consensus process. The CGSB has been accredited by the Standards Council of Canada as a national standards-
9、writing organization. The standards that it develops and offers as National Standards of Canada conform to the criteria and procedures established for this purpose by the Standards Council of Canada In addDon to standards it publishes as national standards, the CGSB produces standards to meet parbcu
10、lar needs, in response to requests bom a variety of soums in both the public and private sectors. Both CGSB standards and national standards developed by the CGSB are developed in conformance with the policies described in the Policy Manual for the Development and Maintenance of Standards by CGSB. C
11、GSB standamls are subject to review and revision at any time, so as to ensure that ihey keep abreast of technological progress. Suggestions for their improvement, which are always welcome, should be brought to the notice of the standards committees concerned. Changes to standards are issued either a
12、s separate amendment sheets or in new editions of standards. An up-to-date listing of CGSB standards, including details on latest issues and amendments, and ordering instructions, will be found in the Catalogue of Standards and Qualified Products Lists which is published annually and is available wi
13、thout charge upon request. Although the intended primary application of this standard is stated in its Scope, it is important to note that it remains the responsibility of the users of the standard to judge its suitability for their parbcular purpose. Many tests required by CGSB standards are inhere
14、ntiy hazardous. The CGSB neither assumes nor accepts any responsibility for any injury or damage that may occur during or as the result of tests, wherever performed. The CGSB takes no position respecting the validity of any patent rights asserted with any item connected with this standard. Users of
15、this standard are expressly advised that determination of the validity of any such paient rights are entirely their own respansibility. Further infarmation on the CGSB and its seMces and standards may be obtained from: La prsente norme, a , this remains the continuing responsibility of the accredite
16、d standards-writing organization. a Those who have a need to appiy standards are encouraged to use Nabonal Standards of Canada whenever practicable. These standards are subject to penodic revim therefore, users are cauboned to obtain the latest edition from the organizaban preparing the standard. Th
17、e mspodbility for approving National Standards of Canada resrswiththe: standards Council of canada, 350 sparks street, Ottawa, Ontario. KIR 7S8 Le CONSEIL CANADIEN DES NORMES est lorganisme de coordination du Systme de normes nationales, une fdration dorganismes indpendants et autonomes qui travaill
18、ent au dveloppement et lamlioration de la normalisation vobntaire dans Iintrt national. Les pnncipaux buts du Consetl sont dencourager et de promwvar la normalisation volontaire comme moyen dam cet aspect demeure ia responsabilit de lorganisme rdacteur de normes accredit. II est recommand8 aux perso
19、nnes qui ont besoin de nom de se servir des Nonnes nabonaies du Canada lorsque la chose est possibie. Ces nonnes font lobjet dexamens priodiques; cest pourquoi il est recommand aux utilisateurs de se procurer ldition la plus rcente de la norme auprs de lorganisme qui la par it has been designated as
20、 CANICGSB-4.2 No. 1-M simply for identification and to conform with the general layout adopted. I 2. GENERAL 2.1 The observed value of any measurement may differ from the true value of the quantity being measured for a number of reasons. For example: 1. the value have recorded incorrectly 2. the mea
21、suring instrument may have an incorrectly located zero point 3. the observer using an instrument may be unable to repeat determinations that agree exactly regardless of how hard he tries to keep conditions uniform, etc. 2.2 It is sometimes said that whenever a measurement differs from the true value
22、, the difference represents error. It is suggested here, however, that to the extent that reason 1. is responsible, mistake, blunder, or some such work is more appropriate than error. Similarly, the type of error described in 2. is well described by the word bias. Sometimes this is referred to as a
23、systematic, constant, or determinate error. The practice is to reserve the word “error“ for what is left after mistakes and biases have been accounted for. It OBJET La prsente mthode: a) formule un expos sur la prcision et lexactitude des mesures et nonce une faon de procder afin dtablir un interval
24、le de confiance pour une moyenne calcule et b) dcrit les modes opratoires pour dterminer le nombre de spcimens a slectionner dans le cas dun essai particulier. La forme de la prsente mthode diffre des autres mthodes de la norme CANICGSB-4.2, Mthodes pour preuves textiles, en raison de la nature du s
25、ujet trait. Elle est dsigne CANCGSB-4.2 No 1-M simplement a des fins de rfrence et pour se conformer la disposition gnrale adopte. GNRALITS La valeur observe pour chacune des mesures peut diffrer de la valeur exacte de la quantit mesure pour de nombreuses raisons. Par exemple: 1. la possibilit dune
26、erreur denregistrement 2. le point zro de linstrument de mesure peut precision implies grouping about some value that need not be the true one. DEFINITIONS The following definitions are applicable to this method. Accuracy: the closeness of agreement between an observed value and an accepted referenc
27、e or true value. Bias: a systematic error that contributes to the difference between a population mean of the measurements or test results and an accepted reference or true value. Precision: the closeness of agreement between randomly selected individual measurements or test results. Sample: a porti
28、on of a lot of material which is taken for testing or for record purposes. Specimen: a specific portion of a material or laboratory sample upon which a test is performed or which is taken for that purpose. APPLICABLE PUBLICATIONS The following publications are applicable to this method: ASTM E 122 -
29、 Choice of Sample Size to Estimate the Average Quality of Lot or Process. sont prises en compte. Lexprience dmontre que les erreurs ainsi dfinies forment une distribution de frquence stable ayant tendance sannuler. Ce nest toutefois pas le cas pour les raisons 1. et 2. susmentionnes. Dans la prsente
30、 norme, le terme “erreur” dsignera toujours les discordances entre les mesures qui ne sont pas imputables aux fautes ou aux erreurs de justesse, c.-d., celles ayant des distributions de frquences qui leur sont associes. Les mthodes statistiques dcrites dans le prsent document ont t labores afin de m
31、esurer et de dmontrer leffet de ces erreurs. Elles ne visent pas a liminer les effets des erreurs de justesse mais peuvent servir les dceler. La prsente mthode est juge comme tant dune prcision acceptable si lerreur associe la mthode de mesure est ngligeable. Lorsquon parle de prcision, cela ne sign
32、ifie pas que les mesures se rapprochent ncessairement de la valeur exacte mais plutt quelles prsentent un accord suffisant entre elles. Lorsquune mthode est de prcision acceptable et que lerreur de justesse est inexistante ou ngligeable, la mthode est alors dune exactitude acceptable. Par exactitude
33、, on entend le regroupement des mesures autour de leur valeur exacte et par prcision, on entend le regroupement des mesures autour dune certaine valeur qui nest pas ncessairement la valeur exacte. DFINITIONS Les dfinitions suivantes sappliquent la prsente mthode. Exactitude: troitesse de laccord ent
34、re une valeur observe et une valeur de rfrence accepte ou sa valeur exacte. Erreur de justesse: erreur systmatique contribuant la discordance entre la moyenne dune population de mesures ou de rsultats dessai et une valeur de rfrence accepte ou la valeur exacte. Prcision: troitesse de laccord entre d
35、es mesures individuelles ou des rsultats dessai choisis au hasard. chantillon: une partie dun lot de matriau prleve des fins dessai ou de rapports. Specimen: une partie spcifique dun matriau ou dun chantillon de laboratoire soumis un essai ou prlev des fins dessai. PUBLICATIONS APPLICABLES Les publi
36、cations suivantes sappliquent la prsente mthode: ASTM E 122 - Choice of Sample Size to Estimate the Average Quality of Lot or Process. 2 N0.W 1-M87 Copyright Canadian General Standards Board Provided by IHS under license with CGSBNot for ResaleNo reproduction or networking permitted without license
37、from IHS-,-,-CGSB CAN/CGSB-4.2 NO.L-M7 * D 1874650 0030657 556 D 4.1.2 MIL-STD-105 - Sampling Procedures and Tables for Inspection by Attributes. 4.1.3 Statistical Tables for Biological, Agricultural and Medical Research, by Fisher and Yates. 4.2 Reference to the above publications is to the latest
38、issues, unless otherwise specified by the authority applying this method. Sources for these publications are shown in the Notes section. 5. CHARACTERIZATION OF A FREQUENCY DISTRI- BUTION 5.1 Experience shows that errors of measurement are usually distributed symmetrically, and that they are densely
39、clustered about the central region and sparsely represented elsewhere. From this experience, and with the aid of a few reasonable assumptions and some mathematical development, the notion of the normal distribution of errors arises. (Figure 1 ) MIL-STD-105 - Sampling Procedures and Tables for Ins- S
40、tatistical Tables for Biological, Agricultural and Medical Sauf indication contraire de lautorit appliquant la prsente mthode, ces publications sentendent de ldition la plus rcente. La source de diffusion est indique dans la section intitule Remarques. pection by Attributes. Research, by Fisher and
41、Yates. CARACTRISATION DUNE DISTRIBUTION DE mi- QUENCES Lexprience dmontre que les erreurs de mesure sont ordinairement distribues symtriquement et fortement groupes dans la rgion centrale et clairsemes ailleurs. A partir de ces observations et avec laide dhypothses raisonnables et de certaines formu
42、les mathmatiques, il est possible dexpliquer la notion de la distribution normale des erreurs. (Figure 1) O FIGURE l 5.2 A normal distribution is fully specified by two numbers that are called the mean, X, and the standard deviation, u (Note 1). The mean gives the location of the center of a normal
43、distribution and the standard deviation specifies the extent to which the distribution spreads around the mean. 5.3 The standard deviation therefore describes the precision of the method of measurement, and the mean, provided no bias is present, represents the true value of the quantity measured. No
44、te I: The prime (7 on a symbol denotes true or population values. Symbols without primes represent estimates calculated from actual measurements. X Une distribution normale est entirement dfinie par deux valeurs, soit la moyenne, X, et lcart-type, u (remarque 1). La moyenne indique la position centr
45、ale dans une distribution normale, tandis que lcart-type exprime la dispersion des rsultats autour de la moyenne. Lcart-type dfinit donc la prcision de la mthode de mesure et la moyenne, pourvu quil ny ait pas derreur de justesse, reprsente la valeur exacte de la quantit mesure. Remarque 7: le signe
46、 prime (I) sur un symbole dnote une valeur exacte ou une valeur de population. Les symboles sans signe prime reprsentent des evaluations calcules daprs des mesures reelles. NO./No 1487 3 Copyright Canadian General Standards Board Provided by IHS under license with CGSBNot for ResaleNo reproduction o
47、r networking permitted without license from IHS-,-,-CGSB CANICGSB-4.2 NO.L-Pl87 * L874b.50 OOLOb58 492 D 5.4 The distribution is introduced to describe all the measurements of the property under consideration that might theoretically be made by the specified method or, in statistical language, the p
48、opulation. Sections 5 and 6 are concerned with the situation in which both numbers (the mean and the standard deviation) that specify the population, are unknown. The problem is, then, to arrive at estimates of these two numbers from a set of measurements or, in more general terms, from a sample drawn from the population. Since the object here is to obtain a faithful description of distribution, any samples drawn for this purpose must be taken in such a manner that the frequencies in the distribution are not persisten
copyright@ 2008-2019 麦多课文库（www.mydoc123.com）网站版权所有
备案/许可证编号：苏ICP备17064731号-1