1、BRITISH STANDARD BS6434:1984 Guide to The use of BS6065 “Methods for sampling and acceptance testing of shaped refractory products” UDC 666.74:620.113BS6434:1984 This British Standard, having been prepared under the directionof the Refractory Products Standards Committee,was published underthe autho
2、rity of the BoardofBSI and comes intoeffecton 31January1984 BSI 10-1999 The following BSI references relate to the work on this standard: Committee reference RPE/12 Draft for comment82/78559 DC ISBN 0 580 13582 9 Committees responsible for this British Standard The preparation of this British Standa
3、rd was entrusted by the Refractory Products Standards Committee (RPE/-) to Technical Committee RPE/12 upon which the following bodies were represented: British Ceramic Research Association British Metallurgical Plant Constructors Association British Steel Industry Cement Makers Federation Refractori
4、es Association of Great Britain Amendments issued since publication Amd. No. Date of issue CommentsBS6434:1984 BSI 10-1999 i Contents Page Committees responsible Inside front cover Foreword ii Section 1. General 0 Introduction 1 1 Scope 1 2 Definitions 1 Section 2. Technical specifications and accep
5、tance testing 3 Technical specifications 1 4 Terms of acceptance 3 5 Conclusion of acceptance testing 3 Section 3. Application of the scheme 6 Application 3 7 Single plan by attribute: procedure1 4 8 Single plan by measurements with fixed unilateral limit and known: procedure2 4 9 Single plan by mea
6、surements with fixed unilateral limit and unknown: procedure3 5 10 Single plan by measurements with an imposed limiting value Gfor the mean and known:procedure4 6 11 Single plan by measurements with an imposed limiting value Gfor the mean and unknown: procedure5 6 12 Sequential plan by measurements
7、with imposed limiting value Gfor the mean and known: procedure6 7 Section 4. Worked examples 13 Single plan by attribute: procedure1 7 14 Single plan by measurements using fixed unilateral limit and known: procedure2 8 15 Single plan by measurements using fixed unilateral limit and unknown: procedur
8、e3 8 16 Single plan by measurements with an imposed limiting value Gfor the mean and known: procedure4 8 17 Single plan by measurements with an imposed limiting value Gfor the mean and unknown: procedure5 8 18 Sequential plan by measurements with imposed limiting value Gfor the mean and known: proce
9、dure 6 9 Table 1 Application of the scheme 3 Table 2 Summary 4 Table 3 Procedure1 5 Table 4 Procedure2 5 Table 5 Procedure3 6 Table 6 Procedures4 and5 6 Table 7 Procedure6 7 Table 8 Results of testing and acceptance decision 8 Publications referred to Inside back coverBS6434:1984 ii BSI 10-1999 Fore
10、word This standard has been prepared under the direction of the Refractory Products Standards Committee at the request of the Refractories Association of GreatBritain as a guide to the use of BS6065 “Methods for sampling and acceptance testing of shaped refractory products” which is identical with I
11、SO5022:1979 published by the International Organization for Standardization (ISO). A British Standard does not purport to include all the necessary provisions of a contract. Users of British Standards are responsible for their correct application. Compliance with a British Standard does not of itsel
12、f confer immunity from legal obligations. Summary of pages This document comprises a front cover, an inside front cover, pagesi andii, pages1 to10, an inside back cover and a back cover. This standard has been updated (see copyright date) and may have had amendments incorporated. This will be indica
13、ted in the amendment table on the inside front cover.BS6434:1984 BSI 10-1999 1 Section 1. General 0 Introduction This British Standard aims to examine under what conditions technical specifications and acceptance procedures for shaped refractory products should be agreed by the user and the manufact
14、urer. The main aim of any acceptance inspection is to seethat the customer gets the quality required as economically as possible; inevitably the cost of inspection has to be reflected in the final cost of the product. Generally an acceptance scheme comprises: a) technical specifications giving value
15、s for the properties deemed necessary; b) acceptance test terms indicating the procedure for declaring the goods to be in conformity or not. 1 Scope This British Standard guide expands on the instructions and provides a number of examples to clarify the method of sampling inspection described in BS6
16、065. NOTEThe titles of the publications referred to in this standard are listed on the inside back cover. 2 Definitions For the purposes of this British Standard, the definitions, statistical terminology and symbols given in BS6065 apply. Section 2. Technical specifications and acceptance testing 3
17、Technical specifications 3.1 Test methods. Values obtained from a given test piece may depend on the method, apparatus or the size of the test piece used. To avoid any dispute at the time of delivery, it is therefore important that the method used to indicate each property should be accurately defin
18、ed. It is also important that the property should be determined using a British Standard, ISO, or other defined and agreed test method. When agreeing on the specifications, it is important that the chosen standard deviation reflects both the heterogeneity of the refractory product and variation in r
19、esults obtained with the agreed test method when it is correctly applied. 3.2 Choice of properties. When technical conditions for special contracts are drawn up, a certain number of properties will be chosen to ensure consistent quality. As far as possible, these properties should be related to the
20、envisaged conditions in use (e.g.porosity is related to corrosion). Efforts should be made to choose those properties that can be controlled by simple and inexpensive tests so that the necessary number of measurements for an effective control can be made. It should also be remembered that the proper
21、ties of a refractory have to be regarded as a whole. Individual properties depend on the raw materials and the shaping and firing methods. Thus it could be dangerous to ask for, or try to obtain, a better value for a particular property of a given product, as this could have disadvantageous repercus
22、sions on other properties which may prove more important in use than envisaged. For example, in some circumstances a lower apparent porosity than required may adversely affect the thermal shock resistance (which may also have been specified or included). Properties that are too closely related may n
23、ot necessarily be the best choice for control properties as they would yield only limited information about the quality of the batch. For example, bulk density and relative density would not usually be considered appropriate when used together. As many of the properties of refractory materials call
24、for destructive test methods, evaluation of a consignment can only be achieved by using statistical sampling plans to take a number of individual items and considering their operating characteristic curves. An operating characteristic curve (OC curve) is a graph showing what any particular sampling
25、plan can be expected to do in terms of accepting and rejecting batches. A specific sampling plan can be operated without reference to the properties of its OC curve. A clear understanding of the implications of the OC curve is, however, essential to the intelligent application of acceptance sampling
26、 and for such tasks as setting the inspection level (and, if so, what limits) to permitted batch sizes etc. An explanation of the principles etc., of operating characteristic curves is given in clause15 of BS6000:1972. These operating characteristic curves will show whether it is desirable to have t
27、he assurance that a batch complies with the declared value for a certain property by carrying out a large number of tests. It may be seen that control tests may be ineffective if only a few tests on each property are carried out.BS6434:1984 2 BSI 10-1999 It is also necessary to remember that the ope
28、rating characteristic curves of the sampling plans chosen for the acceptance tests of refractory products described in BS6065 are only valid for the evaluation of a single property. If, in this instance, the suppliers risk is limited to5% for one property this risk will exceed30% when seven independ
29、ent properties are checked. This shows that an increase in the number of properties to be guaranteed is often technically and economically not justifiable. In a similar way the consumers risk will increase when the number of control properties is reduced. For example, in a scheme of inspection for a
30、 batch of known standard deviation, with a fixed unilateral limit for the individual value and an acceptable quality level of4%, the determination of three properties will result in a10% probability of acceptance of a batch which contains11% defective items. The number of defective items with a10% p
31、robability of acceptance will increase to21% if only one control property is used (seeFigure 1 of BS6065:1981). Thus it is important to distinguish between the following. a) The control properties that should be subject to strict tests by statistical methods. Failure to comply with the declared valu
32、es can be grounds to declare the goods to be non-conforming. As described in BS6065 the number of properties inspected by destructive tests (excluding chemical analysis) should not be greater than three. b) Other relevant properties. Many properties for various reasons, e.g.lack of standard test met
33、hods, duration and cost of test or non-uniform expression of results, cannot be treated statistically but nevertheless are often important in defining a product for a particular application. With this in mind careful selection of control properties is important since if the relationship between thes
34、e and the other relevant properties is known (even if only approximately) the risk of the other relevant properties deviating significantly will be reduced. BS6065 indicates action to be taken in the use of control properties. If other relevant properties are specified and significant deviations are
35、 noted, it would be necessary to enter into separate negotiations. In the case of newly developed products, where insufficient quality control data are available on which to base agreement on the limits for the properties in question, technical specifications can only be given on a provisional basis
36、 and revised as experience is acquired. Schemes which test batches of materials with properties of unknown standard deviation would be appropriate. 3.3 Presentation of the property values. It has been noted in the past that technical specifications demanded by the user or laid down by the manufactur
37、er have not always clearly stated whether the values are to be considered as mean or limit values. It is necessary that the control properties are given as either: a) a mean value with a standard deviation, remembering that, according to the property under consideration, either too high or too low v
38、alues are cause for complaint; or b) a limit value accompanied by an acceptable quality level, (AQL). It should be noted that, for properties whose values are spread according to a normal distribution, these two methods of presentation are in fact equivalent. Example 1. Consider a plan for strength
39、measurements with a guaranteed mean of1000 (arbitrary units), a known standard deviation of60 and a producers risk of5%. This can be shown to be equivalent (assuming an AQL of4%) to a plan with a fixed lower limit of895 since, for example, with a sample size of10, the sample would be in conformity i
40、f its mean was equal to or greater than: a) for guaranteed mean (using Table 4 of BS6065:1981): G K PRE =10000.52 60=968.8; b) for fixed limit (using Table 9 of BS6065:1981): G + K =895+1.23 60=968.8. Similar agreement is obtainable for other sample sizes and AQL values. Where the standard deviation
41、 has to be calculated from the test sample, plans based on a guaranteed mean are not strictly equivalent to plans based on a limit value. Example2. A plan similar to the above with a guaranteed mean of1000 and a producers risk of5% would again be equivalent (using an AQL of4% and a sample size of6)
42、to a plan with a fixed lower limit of895, provided that the sample standard deviation was60, but if the standard deviation increased to say80, for conformity the sample mean would need to be at least: a) for guaranteed mean (using Table 4 of BS6065:1981): G K PRE S=10000.82 80=934.4; b) for fixed li
43、mit (using Table10 of BS6065:1981): G + KS=895+0.93 80=969.4.BS6434:1984 BSI 10-1999 3 Thus, where the standard deviation has to be calculated from the test sample, a plan based on a fixed limit could include some insurance against excessive product variation. 4 Terms of acceptance 4.1 General. Ther
44、e should be agreement between the parties concerned on the sampling procedures, test methods and subsequent conditions for acceptance or rejection. The implementation of the scheme has to be supervised by a person familiar with the terms and conversant with the statistical possibilities contained in
45、 the standard. 4.2 Sampling plans. These schemes should be chosen from BS6065. In the strict sense, these statistical sampling plans are only applicable to random sampling. Random sampling is sometimes not possible. Therefore the Inspector has to have sufficient practical experience in using samplin
46、g techniques. 4.3 Subdivision into batches. A complete delivery will be subdivided by the manufacturer into batches which will be accepted separately. The subdivision will be made in a way depending on technical and economic conditions. The size of the batches will be fixed by agreement and will nor
47、mally be a function of delivery dates, production capacity, testing capacity and cost of testing. As a general rule, the maximum size of a batch should not exceed500t. 4.4 Time taken for acceptance tests. For a consignment which has to be accepted before delivery, the delivery date will be affected
48、by the following four periods and this has to be borne in mind: a) time taken to be ready for acceptance testing; b) time taken for acceptance testing to be completed; c) time taken to be ready for new acceptance testing in the case of non-conformity; d) time taken for new acceptance tests to be com
49、pleted. 5 Conclusion of acceptance testing The result of the acceptance tests will be the basis of the decision of the conformity or non-conformity of the batch with the agreed technical specifications. In the case of non-conformity, the procedures of subdividing into smaller batches or tightened inspection 1)can be followed and investigations can be made into whether for instance, by sorting or reprocessing it is possible to isolate or eliminate the defective items or improve the properties at fault. If it i
