Introduction to Statistical Quality Control, 5th edition：统计质量控制介绍第五版.ppt

《Introduction to Statistical Quality Control, 5th edition：统计质量控制介绍第五版.ppt》由会员分享，可在线阅读，更多相关《Introduction to Statistical Quality Control, 5th edition：统计质量控制介绍第五版.ppt（111页珍藏版）》请在淘文阁 - 分享文档赚钱的网站上搜索。

1、Chapter 71Chapter 72Learning ObjectivesChapter 73P Control ChartThe most versatile and widely used attributes control chart -used when the subgroup size is not constantUsed to evaluate fraction defectiveControl limits are based on Binomial DistributionChapter 74P Control ChartUCL PERCENT908070605040

2、3020100 2 4 6 8 10 12 14 16 18 20 22 24 26P CONTROL CHART ALCAS PEFLOW SOLDER MACHINEPERCENT DEFECTIVESUBGROUP NUMBERChapter 75P ChartCompute the control limitsThe UCL and LCL are not straight lines,they rise and fall with respect to the subgroup sizeChapter 76The P ChartWhen to use:Only when variab

3、le data cannot be obtained.When charting fraction rejected as nonconforming from a varying sample size.When screening multiple characteristics for potential monitoring on variable control charts.When tracking the quality level of a process before any rework is performed.Chapter 77The P ChartHowBy co

4、unting the number of defective items from a sample and then plotting the percent that are defective.ConditionsIn order to be of help,there should be some rejects in each observed sample.The higher the quality level,the larger the sample size must be to contain rejects.For example,if 20%of a product

5、is rejectable,a sample size of 5 will be needed.However,a sample of 1,000 will give an average of only one reject per sample if 0.1%of the product is rejectable.Chapter 78The P ChartChartControl limitsCenterline Plot point Sample sizepVaryingTable 1 Formulas for the P ChartChapter 79The P ChartTo ca

6、lculate Plot PointsThe p plot point is the fraction defective in a sample.The centerline is the average fraction defective in series of samples.Figure 1 is a cross section showing countersunk holes for rivet installation.Chapter 710The P ChartTo calculate Plot PointsIn a sheet metal assembly shop a

7、common process is bucking rivets.Because of the combined variation in the rivets,the drilled holes,and the bucking process,there are quality problems.After the rivets in an assembly have been bucked into place they are checked for nonconformity.A P chart is used to track the first-time-through fract

8、ion defective.Chapter 711Figure 1Cross section of sheet metal plates with countersunk holes for rivet installationChapter 712P ChartChapter 713P ChartTOTALSChapter 714X RelayChapter 715Fluorocarbon Leak Test76543211 2 3 4 5 6 8 9 14 15 16 17 18 19 20 21 22 23 24 25 28 29UCLFLUOROCARB.LEAK TESTP Char

9、t(%)Chapter 716NP Control ChartNumber of defectivesUsed when subgroup size is constant(n)The actual number of defects is represented by pn(or np)Chapter 717NP Control ChartNP Control Chart plating defects of assembled parts 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30151413121110987654321LOT NUMBERUCLCh

10、apter 718NP Control ChartsComputationsControl limits are based on the Binomial Distribution1)Central line:2)3)Chapter 719NP ChartsChapter 7204 Steps in control chat for Fraction RejectedI Preparatory DecisionsII Starting the control chartIII Continuing the control chartIV Reports and Action based on

11、 control chartChapter 721I.Preparatory DecisionsPurposeSelect Quality CharacteristicSelection of subgroup P or NP chartControl limit calculatedChapter 722II.Starting the control chartRecording data ,CalculationTrial limitsPlotting pointsChapter 723III Continuing the control chartSelection of P0Calcu

12、lation of control limits Plotting the points/limitsInterpretation of lack of controlPeriodic Review/Revision of PiIf chart in control Chapter 724IV Reports and Action based on control chartAction to bring process into control-Pareto,high spot/low spot.Review of Design and specificationInformation to

13、 Management(Quality level)Sensitivity of p chart:0.1%requires 1000sChapter 725Example of Attribute control chartLOTNumber inspectingNumber of defectivesP1500270.054250120.2403800120.0154100140.1405150150.100Total160080Chapter 726Example of Attribute Control ChartSubgroupSubgroup Size n#of Defectives

14、Percent Defective P(%)UCL(%)LCL11151513.017.71.2220188.215.43.32102310.915.63.42202210.015.43.5255187.015.04.6365154.14.72554415.04.8300134.34.92803311.714.8103304212.714.3Total2550243Chapter 727Purpose of maintaining attribute charts is to continuously improve the processes for defect free producti

15、on by highlighting the key problems.How to work on getting desirable patternRecommend working on 3 top problems for the day or week of month as the time permits and solve by a systematic problem solving method namely(1)Define problems(2)Find key causes(3)Solution to cure the key causes demonstrated

16、by statisticsChapter 728Chapter 729Chapter 730Chapter 731Chapter 732Chapter 733Chapter 734Chapter 735Chapter 736Chapter 737Chapter 738Chapter 739Chapter 740Design of the Fraction Nonconforming ChartThree parameters must be specified1.The sample size2.The frequency of sampling3.The width of the contr

17、ol limitsCommon to base chart on 100%inspection of all process output over timeRational subgroups may also play role in determining sampling frequencyChapter 741Chapter 742Chapter 743Chapter 744Chapter 745Chapter 746Average sample size approachChapter 747Chapter 748Chapter 749Chapter 750Chapter 751C

18、hapter 752Chapter 753Chapter 754Chapter 755Charts for NonconformitiesOften times there is interest in knowing how many defects an item has -C charts:total number of nonconformities in subgroups of fixed size(defects per square yard of cloth)-U charts:average number of nonconformities per unit(defect

19、s per TV set)C cart and U chartChapter 756Charts for NonconformitiesUsed during inspection of complex assembliesControl limits based on Poisson DistributionChapter 757C Control ChartUCL and LCL are straight lines because of fixed sample sizes CL=UCL=LCL=Where is the central line and equal to Chapter

20、 758The data in table 2 represent the types of defects found on the first two boxes.Box 1 has nine defects and box 2 has twelve.Notice that the types and quantity of defects are different for the two boxes;nevertheless,the total number of defects is plotted on the C chart.Chapter 759Note:A C chart(a

21、s well as any other attribute chart)should only be used when there is absolutely no way to obtain variable data from the characteristic in which measurable data is available.To get started,this data can be analyzed on an attribute chart to get ideas(using the Pareto analysis)on the characteristics.C

22、hapter 760U Control ChartAs the subgroup size varies,UCL and LCL varies CL=UCL=LCL=where is the central line and equal to Chapter 761The U ChartWhen to use:Only when variable data cannot be obtained.When plotting the average number of defects found per unit.When screening multiple characteristics fo

23、r potential monitoring on variable control charts.How:Each unit is examined and the average number of defects found are plotted.Chapter 762Conditions:Constant unit size,but any convenient number of units per plot point.Unit size is different from samples size.For example,one unit could be defined as

24、 1 square-foot of material.On a particular day,12.3 square-feet of material is inspected.The plot point would represent the average number of defects per unit,but the sample size would be 12.3.The U ChartChapter 763The U ChartThere are potentially several different types of defects per unit,but none

25、 of which would necessarily render the part a defective.For example:paint blemishes on a skin panel,or various electrical faults on a circuit board.A unit can be single part,an assembly of several parts,an area of material,or any rational grouping in which the likelihood of defect(s)is constant from

26、 unit to unit.Chapter 764In order for this type of analysis to be of help,there should be some defects in each observed unit.Chapter 765The U ChartTo Calculate Plot Points:The u plot point is the average number of defects per unit in a sample of n units.The centerline is the average of all the plot

27、points on the chart.Figure 6-17 shows a roll of composite material/dyed cloth with potential multiple defects.Chapter 766The U ChartComposite materials are generally made up of several layers of materials that are bonded to form the desired characteristics.Prior to lay up,are used each day,the numbe

28、r of rolls inspected for defects.Since different amounts of the material are used each day,the number of rolls inspected also changes daily.Chapter 767Attribute Control ChartsChart typeNames/remarksCentral lineControl LimitsFraction defectiveNumber defectivesNumber of defects per subgroupNumber of d

29、efects per inspection unitChapter 768AttributedataN50N not constantFractiondefectiveN50N is constantFractiondefectiven1n not constantDefectsper unitn1Is constantDefects per unitc chartVariable dataP chartnp chartu chartn6n25n=1IndividualchartX-bar chartR chartX-bar chartS chartX-bar chartS2 chartCha

30、pter 769Chapter 770Chapter 771Chapter 772Chapter 773Chapter 774Chapter 775Chapter 776Chapter 777Chapter 778Chapter 779Chapter 780Chapter 781Chapter 782Chapter 783Chapter 784Chapter 785Chapter 786Chapter 787Chapter 788Chapter 789Chapter 790Chapter 791Chapter 792Chapter 793Chapter 794Chapter 795Chapter 796Low Defect LevelsChapter 797Chapter 798Chapter 799Chapter 7100Chapter 7101Chapter 7102Chapter 7103Chapter 7104Chapter 7105Chapter 7106Chapter 7107Actions taken to improve a processChapter 7108Chapter 7109Chapter 7110Learning ObjectivesChapter 7111