GageRR

•ClicktoeditMastertextstyles•Secondlevel•Thirdlevel•Fourthlevel•Fifthlevel6sSIXSIGMABlack&Decker©Black&Decker(US)Inc.,2003-1GageR&RStudy测量系统重复性和再现性研究---FredGu•ClicktoeditMastertextstyles•Secondlevel•Thirdlevel•Fourthlevel•Fifthlevel6sSIXSIGMABlack&Decker©Black&Decker(US)Inc.,2003-2学习目标•定义CTQ.•讨论怎么测量CTQ’s.•GRR理论•可变量GRR•属性量GRR•ClicktoeditMastertextstyles•Secondlevel•Thirdlevel•Fourthlevel•Fifthlevel6sSIXSIGMABlack&Decker©Black&Decker(US)Inc.,2003-3CTQCriticalToQuality质量关键点CTQ是满足顾客要求的产品或服务关键特性CTQ定义•ClicktoeditMastertextstyles•Secondlevel•Thirdlevel•Fourthlevel•Fifthlevel6sSIXSIGMABlack&Decker©Black&Decker(US)Inc.,2003-4QCS(QC)•统计控制•CTQ的能力要大于1.0Cp,尺寸要集中•用可变量GRR衡量测量系统•用质量控制计划和统计过程来保证Cpk=1.33.QCF•功能性•功能性CTQ的潜在影响因素但很少要尺寸集中•功能检具（通止规）或目视，属性量GRR•用质量控制计划控制失效率DPMO1350QCS（QC)与QCF•ClicktoeditMastertextstyles•Secondlevel•Thirdlevel•Fourthlevel•Fifthlevel6sSIXSIGMABlack&Decker©Black&Decker(US)Inc.,2003-5•重复性和再现性–确定检验员和设备产生的测量变异•量具的校准–确定量具的读数和标准值匹配.测量系统概要•ClicktoeditMastertextstyles•Secondlevel•Thirdlevel•Fourthlevel•Fifthlevel6sSIXSIGMABlack&Decker©Black&Decker(US)Inc.,2003-6研究测量系统GRR目的.确认测量系统的精密度.确保测量值的可信度•ClicktoeditMastertextstyles•Secondlevel•Thirdlevel•Fourthlevel•Fifthlevel6sSIXSIGMABlack&Decker©Black&Decker(US)Inc.,2003-7•尺寸–直径,长度,宽度•速度–速率,转速•品质–颜色,擦伤•数量–#完成,#合格,#不合格•测量系统会带来多少的误差???我们要知道.为什么要测量?•ClicktoeditMastertextstyles•Secondlevel•Thirdlevel•Fourthlevel•Fifthlevel6sSIXSIGMABlack&Decker©Black&Decker(US)Inc.,2003-8TheNecessityofTrainingFarmHandsintheFatherlyHandlingofFarmLiveStockisForemostintheEyesofFarmOwners.SincetheForefathersoftheFarmOwnersTrainedtheFarmHandsforFirstClassFarmsintheFatherlyHandlingofFarmLiveStock,theFarmOwnersFeeltheyshouldcarryonwiththeFamilyTraditionofTrainingFarmHandsofFirstClassFarmersintheFatherlyHandlingofFarmLiveStockBecausetheyBelieveitistheBasisofGoodFundamentalFarmManagement下文有多少个“s”?热身练习•ClicktoeditMastertextstyles•Secondlevel•Thirdlevel•Fourthlevel•Fifthlevel6sSIXSIGMABlack&Decker©Black&Decker(US)Inc.,2003-9这个练习的%精确度是否可接受?90%,95%,99%或99%?%精确度=数对的人数总人数%精确度•ClicktoeditMastertextstyles•Secondlevel•Thirdlevel•Fourthlevel•Fifthlevel6sSIXSIGMABlack&Decker©Black&Decker(US)Inc.,2003-10•假设有一零件轴外径为QC尺寸10±0.01mm,需测量并进行质量控制.•你安排测量时会考虑测量系统哪些因素?–量具精度的选择–测量的过程:人员–其他考虑:量具状态，工装如何保证测量的结果是正确的?测量系统•ClicktoeditMastertextstyles•Secondlevel•Thirdlevel•Fourthlevel•Fifthlevel6sSIXSIGMABlack&Decker©Black&Decker(US)Inc.,2003-11平均值观测值变异曲线测量系统变异曲线零件实际变异曲线例11.s2观测=s2实际+s2测量2.s2实际是我们需要的.•ClicktoeditMastertextstyles•Secondlevel•Thirdlevel•Fourthlevel•Fifthlevel6sSIXSIGMABlack&Decker©Black&Decker(US)Inc.,2003-12操作员观测值变异实际变异测量系统变异重复性再现性量具刻度偏移稳定性线性GageRandR分析测量观测值的变异中由于人员和设备引起的变异所占比例.校验保证量具的读数刻度和已知标准相符合.测量系统量具校验的三个主要内容量具的选择•ClicktoeditMastertextstyles•Secondlevel•Thirdlevel•Fourthlevel•Fifthlevel6sSIXSIGMABlack&Decker©Black&Decker(US)Inc.,2003-13量具再现性是指不同的操作员用同一量具测量同一零件的同一尺寸结果间的差异.再现性操作员A操作员B再现性•ClicktoeditMastertextstyles•Secondlevel•Thirdlevel•Fourthlevel•Fifthlevel6sSIXSIGMABlack&Decker©Black&Decker(US)Inc.,2003-14量具重复性是指一个操作员用同一量具测量同一零件的同一尺寸结果间的差异.重复性几次测量的平均值重复性•ClicktoeditMastertextstyles•Secondlevel•Thirdlevel•Fourthlevel•Fifthlevel6sSIXSIGMABlack&Decker©Black&Decker(US)Inc.,2003-15•量具应有足够的精度来分辨所测量的尺寸•要求10倍的精度–量具的最小刻度为尺寸公差的1/10–如:尺寸10+/-0.1mm,则量具的最小刻度要0.02mm量具的选择量具的精度是量具最小刻度•ClicktoeditMastertextstyles•Secondlevel•Thirdlevel•Fourthlevel•Fifthlevel6sSIXSIGMABlack&Decker©Black&Decker(US)Inc.,2003-16•国家计检具局保管标准块•校验后应在量具盒中贴校验标签•量具的校验标签应和量具检验控制记录一致量具的校验校验：量具的输出值与已知的标准块应一致.•ClicktoeditMastertextstyles•Secondlevel•Thirdlevel•Fourthlevel•Fifthlevel6sSIXSIGMABlack&Decker©Black&Decker(US)Inc.,2003-17控制量具误差:刻度偏移误差量具和标准值比较稳定性确保在不同时间段有相同的测量值线性误差确保量具在测量范围内测量值的正确性量具校准目的•ClicktoeditMastertextstyles•Secondlevel•Thirdlevel•Fourthlevel•Fifthlevel6sSIXSIGMABlack&Decker©Black&Decker(US)Inc.,2003-18已知标准值偏移测量均值量具的刻度偏移量具刻度偏移是测量均值和已知标准值（由质检局保存）的差值•ClicktoeditMastertextstyles•Secondlevel•Thirdlevel•Fourthlevel•Fifthlevel6sSIXSIGMABlack&Decker©Black&Decker(US)Inc.,2003-19量具的稳定性时间2时间1量具的稳定性量具的稳定性是同量具同零件同尺寸在不同时间至少两组测量均值间的差异.•ClicktoeditMastertextstyles•Secondlevel•Thirdlevel•Fourthlevel•Fifthlevel6sSIXSIGMABlack&Decker©Black&Decker(US)Inc.,2003-20标准值观测均值(小尺寸)小差异(小尺寸）小尺寸测量大尺寸测量量具的线性误差量具的线性误差是量具测量范围内所有测量均值与标准值差异的线性关系标准值观测均值(大尺寸)大差异(大尺寸）•ClicktoeditMastertextstyles•Secondlevel•Thirdlevel•Fourthlevel•Fifthlevel6sSIXSIGMABlack&Decker©Black&Decker(US)Inc.,2003-21数据分类.可变量数据，可用小数点（如直径，长度等）.属性量数据，只能用整数（如不和格数)•ClicktoeditMastertextstyles•Secondlevel•Thirdlevel•Fourthlevel•Fifthlevel6sSIXSIGMABlack&Decker©Black&Decker(US)Inc.,2003-22GRR分类.可变量数据GRR.属性量数据GRR•ClicktoeditMastertextstyles•Secondlevel•Thirdlevel•Fourthlevel•Fifthlevel6sSIXSIGMABlack&Decker©Black&Decker(US)Inc.,2003-23选择合适的量具---量具测量精度是产品要求精度的10倍.---量具已校验.准备10个零件,3个操作员—10个零件应覆盖80%以上的过程.每个操作员对每个零件测量3次---零件的选取是随机的.根据4项指标的结果判断测量系统是否合格.根据图表确定不合格原因及改善项目.可变量数据GR&R操作步骤•ClicktoeditMastertextstyles•Secondlevel•Thirdlevel•Fourthlevel•Fifthlevel6sSIXSIGMABlack&Decker©Black&Decker(US)Inc.,2003-24测量系统的精密度%公差=---------------------------=公差范围LSLUSLMSs*15.5LSLUSL公差范围测量系统精密度与公差相比的精密度证明系统有能力区