《M16GaugeRR中英文版PPT优秀课件》由会员分享，可在线阅读，更多相关《M16GaugeRR中英文版PPT优秀课件（24页珍藏版）》请在金锄头文库上搜索。
1、GAUGE R&R ANALYSIS(Repeatability & Reproducibility)(再现性&重复性) 量检具 再现性&重复性 分析1determining how much of a problem is due to inconsistencies in the measurement process.确定问题有多少是由测量过程的不一致造成的。确定问题有多少是由测量过程的不一致造成的。localizing the dominant source of measurement inconsistency in a system so that appropriate cor
2、rective action can be implemented.确定系统中测量不一致的主要来源，以便采取合适的整改行动。确定系统中测量不一致的主要来源，以便采取合适的整改行动。Plan and review before starting开始前进行计划和回顾开始前进行计划和回顾Ensure test samples reflect the full range of the process确保实验样本覆盖所有过程确保实验样本覆盖所有过程Always be present when the tests are run实验进行过程中全程在场实验进行过程中全程在场Ensure samples a
3、re appropriately “blinded” 确保样本是随机抽取的确保样本是随机抽取的Ensure repeat runs with the same person are randomized确保同一个人进行的重复实验是随机的确保同一个人进行的重复实验是随机的2Measurements are critical.测量是关键的If we cant accurately measure something, we really dont know much about it.如果我们不能正确地测量某些东西，我们如果我们不能正确地测量某些东西，我们就对其了解得不多。就对其了解得不多。If
4、we dont know much about it, we cant control it.如果我们对其了解得不多，我们就无法控如果我们对其了解得不多，我们就无法控制它。制它。If we cant control it, we are at the mercy of chance.如果我们无法控制它，我们就任其摆布。如果我们无法控制它，我们就任其摆布。3BreakthroughStrategyCharacterizationPhase 1:阶段阶段1：Measurement测量测量Phase 2:阶段阶段2：Analysis分析分析Optimization优化优化Phase 3:阶段阶段3：
5、Improvement改进改进Phase 4:阶段阶段4：Control控制控制IMPROVEMENT ROADMAP改进路径图Verify the validity of primary (Y) and secondary metric measurement systems.验证首要(Y)和次要公制计量系统的有效性。Six Sigma Uses六西格玛用途六西格玛用途Verify the validity of measurement systems used to measure factors (x).验证用来测量要素(X)的公制计量系统的有效性。Evaluate effectiven
6、ess of strategies to reduce measurement system error评估针对减少测量系统误差的策略的效果突破策略突破策略Strategy特性策略特性策略4WHAT IS AN R&R STUDY?什么是再现性和重复性研究Gage Bias量检具偏差量检具偏差Actual Values真值真值Measurement System Bias测量系统偏差测量系统偏差Measurement System Variance测量系统差异测量系统差异Operator Variation操作者差异操作者差异Gage Variation量检具差异量检具差异Reproducib
7、ility再现性再现性Linearity线性线性Accuracy准确性准确性Repeatability重复性重复性Stability稳定性稳定性Measurement System Output测量系统输出测量系统输出The purpose of R&R studies is to evaluate the variance of the measurement system. The variability is categorized in two areasReproducibility for operator to operator variation.and Repeatabili
8、ty for variability within the remaining parts of the measurement system.再现性和重复性研究的目的是评估测量系统的一致性。差异分为两类再现性和重复性研究的目的是评估测量系统的一致性。差异分为两类再现性是分再现性是分析操作者与操作者的差异析操作者与操作者的差异重复性是分析测量系统中其它因素差异。重复性是分析测量系统中其它因素差异。5WHAT IS VARIANCE ERROR?什么是方差误差Variance Error (s s2) = Accurate but not precise方差误差(s s2) =准确但不精确准确但
9、不精确Bias Error (m m) = Precise but not accurate偏移误差偏移误差(m m) =精确精确但不准确但不准确Little Error = Precise and accurate小误差小误差=精确且准确精确且准确-3-2-10123-6-4-20246Variance is related to the spread of the measurements from the actual value. Average them and you will find they are on target, but the individual measurem
10、ents could be all over the place.方差与测量值在真值周围分布的情况方差与测量值在真值周围分布的情况有关。平均值是正确的情况下，每个有关。平均值是正确的情况下，每个单独的测量值可能散布在所有区间。单独的测量值可能散布在所有区间。6MEASUREMENT SYSTEM ERROR测量系统误差Measurement System Variance测量系统误差测量系统误差Measured Variance = Actual Variance + System Variance测量方差测量方差=真值方差真值方差+系统方差系统方差s s2 2measured = s s2
11、2actual + s s2 2errors s2 2error = s s2 2repeatibility + s s2 2reproducibilityVariance Error Tool = R&R Study方差误差工具方差误差工具=再现性再现性&重复性研究重复性研究-3-2-10123Variance = Spread (s s) Increased方差方差=附加的分布附加的分布Variance error is a function of Reproducibility (Operator influenced variance) and Repeatability (Measu
12、rement system influenced variance)方差误差是由再现性方差误差是由再现性(操作操作者影响方差者影响方差)和重复性和重复性(测测量系统影响方差量系统影响方差)组成的函组成的函数数7R&R STUDY REPRODUCIBILITY再现性和重复性研究-再现性-12-10-8-6-4-2024681012LSLReproducibility is the variation in measurements obtained with one measurement instrument when used by different appraisers while m
13、easuring the identical characteristic on the same part.再现性是指由不同评价者使用同再现性是指由不同评价者使用同一测量工具测量同一物件同一特一测量工具测量同一物件同一特性时得到的值间的差异。性时得到的值间的差异。m Operator操作者 #1Actual Value真值M Operator操作者 #2Operator Variance = Reproducibility操作者差异操作者差异=再现性再现性Reproducibility = Operator to Operator variation.再现性=操作者和操作者间的差异m Ope
14、rator操作者 #38R&R STUDY REPEATABILITY再现性和重复性研究-重复性-6-5-4-3-2-10123456Actual Value真值真值Repeatability is the variation in measurements obtained with one measurement instrument when used several times by one appraiser while measuring the identical characteristic on the same part.重复性是同一评价者使用同一测量工具多次测量同一物件同
15、一特性得到的值的差异Measured Values测量值测量值Repeatability = Remaining measurement system variation.重复性重复性=测量系统中其他因素的差异测量系统中其他因素的差异9TYPES OF R&R STUDIES再现性和重复性研究的类型Defect count or Go/No Go Measurement Systems缺陷计数或者通缺陷计数或者通/止测量系统止测量系统Continuous Data (Variable Data) Measurement Systems连续分布数据连续分布数据(可变数据可变数据)测量系统测量系统
16、R&R STUDY再现性和重复性研究再现性和重复性研究VARIABLE R&R STUDY定量测量再现性和重复性研究定量测量再现性和重复性研究ATTRIBUTE R&R STUDY定性测量再现性和重复性研究定性测量再现性和重复性研究 Examples 举例举例 Visual Defects 目视缺陷目视缺陷Order Errors 指令错误指令错误Typographic Errors 印刷错误印刷错误Inventory Counts 清单计数清单计数 Software Errors 软件错误软件错误Attribute Gages 特性检具特性检具Misplaced Components 零件错
17、装零件错装 Examples 举例举例Dimensional Measures 几何尺寸测量几何尺寸测量Plating Thickness 电镀厚度电镀厚度Turned Diameter 倒角直径倒角直径Material Hardness 材料硬度材料硬度 Voltage Output 输出电压输出电压 Miles per hour 英里每小时英里每小时10Guidelines For Gauge R&R检具再现性和重复性要点 At least two, but preferably three or more operators至少2位，3位或以上操作者为佳 At least 10 unit
18、s to measure (variable)至少测量10件(定量测量) 30 units to measure (attribute)至少测量30件(定性测量)Each unit is measured 3 times by each operator每件由每个操作者测量3次 The equipment must have adequate discriminating power (resolution)设备必须有适当的辨别能力(分辨率) There should by a Gauge R&R strategy for qualifying new operators and new eq
19、uipment需要有针对新的操作者和新的设备的量检具再现性和重复性策略。11R&R STUDY PREPARATION CHECKLISTR&R研究准备工作检查列表Check检查CONDITION条件Determine the scope and breadth of the R&R study.确定R&R研究的范围和深度 Determine the purpose of performing the R&R and develop your plan accordingly.确定R&R研究的目的，并制定相应计划 Consider the cost and time required for
20、the testing in determining the sample size and number of testers etc.考虑所需的成本和时间来确定样本范围和数量。 Complete a calibration study on the measurement system to determine if there are any bias, linearity or stability issues with the system.完成测量系统的校准研究来确认是否有系统偏差，线性或稳定性问题。Ensure that there is little chance that t
21、he part or property being measured is not altered or destroyed by the act of measurement by the system.确定被测物品是否能不因为检测而被系统改变或毁坏。Determine the number of appraisers, the number of sample parts and the number of repeat readings for the study.确定评价者的人数，样件的数量以及重复测量的次数。 Select the appraisers from those who
22、normally operate the measurement system (if possible).从经常使用该测量系统的人中选定评价者。(如果可能的话)。Select the sample parts from the process and ensure that they represent its entire operating range.选择过程中的样件并确认它们能代表整个工作范围。Number each of the parts for later identification.确定所有物件的数量以便日后识别。If a Variable or an Attribute
23、transform R&R is to be conducted, verify the measurement system has a discrimination of 1/10 of the expected process range (ie, if the range is +/- 2 (R=4), then the values reported should be in steps of 4*1/10=.4 or better).如果传导给R&R的一个变量或者一个特性改变了它，验证测量系统的分辨率达到1/10预期测量范围(比如，如果范围是+/- 2 (范围=4)，那报告因该以4
24、*1/10=4作为步长或更小) If an Attribute transform R&R is to be conducted, ensure that the transform has at least 10 levels of response and the values are documented.如果一个特性被传导到R&R,确定这个转换至少有10个层级的相应且测量值被记录在案。 Ensure that the measurement system is operated in the same manner as normally operated.确定测量系统按照惯用的方法进
25、行操作。Document the test procedure in a test plan and review with all participants.把实验过程写入实验计划并让所有的参与者进行审阅。Ensure that measurements are made in a random order and that the appraisers do not have access to previous values. This usually entails giving an appraiser each part to test and having another per
26、son record the reported value on the data sheet.确定测量是以随机的顺序进行的且评价者无法得知上次的测量值。通常指定评价者只进行测量，另指定一个人在数据表上记录测量值。 Ensure that each appraiser uses the same procedure to obtain the readings. In cases where a fixture or placement is involved, the part should be completely removed and replaced for each readin
27、g.确定每个评价者使用相同的流程得到读值。为了防止工装夹具被带入，每次读数工件必须完全从工装夹具上卸下和重新夹紧。 12How Much Variability Is Acceptable?多少的变异是可以接受的？多少的变异是可以接受的？ 30%Not acceptable 无法接受无法接受13Minitab has the capability to do a Variable R&R but not an Attribute R&R. Minitab可以进行变量R&R分析，但不能进行特性R&R分析。Lets walk through how to do a Variable R&R in
28、Minitab.让我们来快速浏览下如何使用Minitab来进行变量R&R分析14Entering the data.输入数据You know from the work we just went through that you will need 10 parts measured by 3 operators and each operator needs to take 3 readings from each part.需要由3操作者来测量10个工件，每个操作者需要对每个工件进行三次读书。Therefore it is reasonable to assume that we will
29、 need a column to put the measurement in (Data), a column to keep track of which operator (Operator) is making the measurement, a column to keep track of which part is being measured (Part) and a column to keep track of which reading (Reading)is being taken.因此有理由假定我们需要一列来输入测量数据，一列输入进行测量的操作者，一列输入被测量的
30、工件，一列输入第几次读书。15Entering the data.输入数据Now the R&R can be run and the data entered.现在可以进行R&R分析，数据可以被输入了。16Performing the Analysis.执行分析Go to Stat Quality Tools. Gage R&R Study.打开Stat菜单Qulity Tools列表Gage R&R Study选项17Select the Part number, operator and data. Then select “OK”.选择Part Number工件数量，Operator操
31、作者和Data数据。然后选择OKPerforming the Analysis.执行分析18Here is our R&R number in the session window. Quite a ways from our less than 10% requirement.这里是对话框里的R&R数字。满足小于10%的要求Interpreting the Results.分析结果19Here is our first indication of a problem. For a “good” R&R you would expect these points to be outside o
32、f the control limits.这是第一个有问题信号。对于好的测量系统来说所有点都不能超出控制极限。Interpreting the Results.分析结果This chart shows no significant problem with the spread of readings for an individual operator.这张表所显示的所有读数对一个操作者来说没有问题。This is usually the first chart you should check. A good R&R will show the largest component of v
33、ariation in the Part-to-Part. This is a very bad R&R.通常这是因该检察的第一张表。好的测量系统因该是工件与工件之间的差异最大。这个测量系统的再现性和重复性很差。20Here is best information so far on the problem. It appears that Operator #3 is significantly biasing the readings.这是体现问题最好的信息。很明显的操作者3读数出现了偏置。Interpreting the Results.分析结果By operator we see fr
34、om the red line (average) confirmation that operator #3 reading are high.通过红线(均值)显示，操作者3的读数偏大。The By Part chart is used to see if a particular part is part of the problem. Here the readings for each part look consistent.按工件分析表格用来分析是否有某个工件是有问题的。由这张表来看读数是连续的。21 CLASS EXERCISE课程练习题Show Files 打开文件lBadr&
35、r.mtwlGoodr&r.mtwBreak out into your teams and conduct an Gage R&R Study with the parts and measuring devices given. 准备好你的团队，用提供的工件和测量装置进行GageR&R研究。22Handling Poor Gage Capability:处理不合格的检具能力 If a dominant source of variation is repeatability (equipment), you need to replace, repair, or otherwise adj
36、ust the equipment. 如果主要的变差来源是重复的如果主要的变差来源是重复的(来自设备的来自设备的)，替换，修复或者调整设备。，替换，修复或者调整设备。 If, in consultation with the equipment vendor or upon searches of industry literature, you find that the Gauge technology that you are using is “state-of-the-art” and it is performing to its specifications, you will
37、have to learn to live with it. 如果，通过和设备供应商协商或者研究工业文献发现所使用的检具技术已经是当前最新并达到了如果，通过和设备供应商协商或者研究工业文献发现所使用的检具技术已经是当前最新并达到了 设计要求，就必须学会适应它。设计要求，就必须学会适应它。 If a dominant source of variation is operator (reproducibility), you must address either training or the quality and use of your standard process or measur
38、ement procedures. You should look for differences between operators to give you some indication as to whether it is a training, skill, and/or procedure problem. 如果主要的变差来自操作者如果主要的变差来自操作者(再现性再现性)，必须确认培训或者标准流程或测量过程的使用和质量。寻，必须确认培训或者标准流程或测量过程的使用和质量。寻 找操作者间培训，技能，操作过程的不同。找操作者间培训，技能，操作过程的不同。 Evaluate the sp
39、ecifications. Are they reasonable? 评价技术要求。是否合理？评价技术要求。是否合理？ If the Gauge capability is marginal (as high as 30% of tolerance) and the process is operating at a high capability (Cpk greater than 2), then the Gauge is probably not hindering you and you can continue to use it. 如果检具能力处于临界状态如果检具能力处于临界状态(超过公差超过公差30%)且过程能力较高且过程能力较高(Cpk大于大于2)，那么，那么 检具可能不会阻碍你，可以继续使用检具可能不会阻碍你，可以继续使用23个人观点供参考，欢迎讨论
