7钻

null引子引子在企业中，决大多数人都会遇到质量问 题 快递公司问题件快递公司问题件货款处理关于圆的周长面积重点题型关于解方程组的题及答案关于南海问题 5M（人、机、料、法、环） 面对质量问题，你是如何分析处理？ 测试，有多少人关注过？（追问超过个3Why） 按什么顺序来分析呢？希望与大家一块探讨！七颗钻石流程七颗钻石流程 1内容内容7钻的定义 应用目的 第1、2、3钻七颗钻石的流程是什么？七颗钻石的流程是什么？七颗钻石流程是可以用来解决质量问题的工具之一。 它可以用来调查以下方面发现的差异 Global Customer Audit System (GCAS) 全球顾客评审 生产线上巡检 检验工位 2 7颗钻石流程的目的7颗钻石流程的目的 3 7颗钻石流程 是一个构建好的 评价 LEC评价法下载LEC评价法下载评价量规免费下载学院评价表文档下载学院评价表文档下载 我们的制造流程是否满足了顾客所要求的质量标准的一个方法。 7 钻流程7 钻流程 4 Responsibility 责任QET PST (Lead) ¨5b SMT RGM (Lead) ¨6统计工程(红-X方法) ¨7工艺更改能解决问题吗产品更改需要吗？非常复杂的问题SQA - Supplier Quality Assurance QET - Quality Engineering Team PST - Plant Spport Team SMT - Systems Management Team RGM - Reliability Growth Management 职责 岗位职责下载项目部各岗位职责下载项目部各岗位职责下载建筑公司岗位职责下载社工督导职责.docx 回顾职责回顾 5 钻石1-4 生产&可靠性 5a 过滤 工艺控制负责人 钻石5b-7 产品工程7 钻流程的各步骤7 钻流程的各步骤 7 1、2步是用来确定生产部门运行的装配流程满足设计内容。 最初的调查(第一步)应在发现缺陷的地方开始 第2步只能当第一步的调查,确定问题是来源其他 方面时,才可开始 如果工艺没有满足设计内容，当需要工程方面的协助前，必须进行纠正并要有效。 当制造过程满足设计内容并且问题依然存在时，第3步开始。null 8 1 - 正确的过程是否有正确的操作指导文件？员工知道吗？ 操作能执行吗？是否执行操作文件？ 是否满足关键操作点控制图 是否强制的操作序列号已识别，是否执行了 操作者是否正确培训过？（员工是否知道、并理解自己的工作内容及相应的职责？）员工是否能独立操作？ 是否为固定操作工？是否此操作有很多轮换人员 员工的工作情绪是否稳定？ 员工是否擅自离岗？需离岗时，是否知道如何处理？ 生产现场是否有生产指挥人员，生产指挥人员是否知道、并理解自己的职责？ 员工是否理解“三不原则”？并按次执行？ 操作者是否知道他/她的岗位的产品的输出质量 操作者是否理解产品标准 能够发现质量问题吗？理解产生质量问题的原因吗？ 操作者是否知道当他/她的岗位出现质量问题时如何传递沟通信息？ 以下的问题有可能产生问题吗？null 9 是否使用正确的工具/夹具(是否符合工艺要求 ？） 工具是否设置成规定的扭矩 现场是否有可用有效的工具？能否识别工具的状态？ 是否进行适当的校准 工具是否规定了维护保养及频次？ 日常的工具维护是否被理解并执行？ 常见的因工具原因造成的缺陷是否知道、并被理解？ 是否工具被旁路 是否工具/夹具有损毁保护 工位是否有防错设施 是否工位的布置适合操作者有效的工作 是否进行了预防性维修(检查表) 工具功能是否正确2 - 正确的工具以下的问题有可能产生问题吗？null 10 零件的路径是否正确？ 物料参数是否符合工艺要求？ 现场物料是否可用，并且能被识别状态？ 是否在料架上标识有正确的零件号 物料出现异常，是否知道如何处理？ 因物料的原因引起的缺陷是否知道并理解？ 在料架上摆放位置是否正确，包装是否正确？ 包装盒上零件号与料架上是否一致 拉动卡是否有效 是否需要防错措施 是否已有的防错设备运行正确3 - 正确的零件小结小结质量问题分析解决的有效工具----7钻 定义和意义 第1、2、3钻谢谢大家！谢谢大家！ 7钻的掌握需要经过不断的实践！ 更需要大家养成一种良好的习惯！ 希望我们一道通过运用7钻，更快更好提升我们专用车产品的质量。共勉null 11 可靠性工程师确定零件是否更改过并对产品全部质量负责。如果由于零件质量(超差)是根本原因，那么应：填写PR&R通知供应商： 同供应商共同确认纠正措施 采取筛选措施，制造应支持在工位上的物料筛选 同工段长共同确认对策和遏制措施 如果确认零件在规范内，那么RE应： 同供应商共同工作，SQA将零件返回到规范内需要的范围 同质量工程师合作，确定是否需要工程更改这种不能消除的纠正措施Stages 1&2Production ¨1Correct Process? 正确的工艺流程Rel. Engineer/ SQA/Supplier零件质量Production ¨2Correct Tool? 正确的工具Correct Part? 正确的零件Production ¨3¨4 - Part Quality零件质量null 12 Diamonds 1-4 in Review 钻石1-4步回顾Diamonds 1-4 are for evaluating the stability of the process. 1-4步用来评价工艺过程的稳定性 Once a problem has been identified, the automatic response must be to immediately step thru diamonds 1-4. 当问题被定义后，1-4步的负责人员就自动确定下来 It must become inherent in our problem solving thought process. 这些步骤是问题解决过程中固定的流程null Responsibility职责 Process Control Managers 过程控制负责人 ¨5a 13 ¨5a - Engineering Filter 5a-工程过滤Is Engineering support required? 需要工程方面的支持吗5a is a step to filter out and resolve less-complex issues before seeking help from Quality Engineering. 5a是一步过滤出并将不复杂的问题在寻求质量工程帮助前解决 The Process Control Manager is responsible for conducting initial studies prior to escalating the issue. 过程控制负责人负责在提升问题前指导初始研究 These studies may include but are not limited to: 这些研究应包括但不局限在: Problem Definition Tree 问题定义树 5 Whys 5个为什么 Statistical Process Control 统计过程控制 Design of Experiment (DOE) 试验设计 Shainin Techniques (quick & dirty) 技术(快速& 脏) If after the initial studies the cause of variation has not been determined, the issue is referred to Product Engineering. The problem definition tree is the minimum exit criteria to go from Diamond 5a to 5b. 如果经过初始研究，变差的产生原因还没有确定，就需要产品工程的支持。问题定义树是从5a-5b的基本方法。nullProject title: Poor adjustment of RR door panel 805-830 Leader: Pedro Aranda Alfaro Sponsor:Juan Carlos Licea Area: VAET Coach: Leonardo Chávez salcedo Initial date: oct-04-2002 Final date: oct-29-2002 Problem description: This month it was reported by the shipping lot in USA a problem of quality existed in the 805/830 vehicles. The problem is the poor adjustment in RR door panel. Projects statement: Find and improve the Red X that produce the poor adjustment in the RR door panel for the model 805/830 in Silao Assembly plant. Strategy: We initiated this project using the model to model strategy, because there was a good contrast and good leverage based on multi-vari see pag 6. Measurement system: As measurement system a Force gage was used. This measurement system passed the isoplot with DR=13.6, DM=1.0 lbs, and DP=13.6 lbs (see page 4) Clue generation: Stage 1 shows that the Red X doesn’t live in the process. The stage 2 shows that the Red X lives in the door panel. The pair comparison shows that the door molding is the part with the best contrast. Statistical confirmation: Once we detected the Red X candidate, I decided to do a B vs C six pack trial, in order to confirm the door molding is the influence to make the customer satisfaction. The six pack passed with 95% of confidence and end count of 6. Irreversible corrective action: To modified the flange 1.5 mm in length to keep a better position and avoid that the door panel take out of position. Improvements: Reduce warranty cost to repair the door panel caused by poor fit (loose) and eliminate the condition in 100% in vehicles 805/830 models. The amount of money from jul 01 to jul 02 was 11,084.77 USD. Lessons learned: We learned that is necessary looking other alternatives using other strategy when is not possible invest money but always focus to guarantee the customer enthusiasm, therefore this project is an example of this.Executive Summarynull 执行总结项目题目：右后门板调整不好(805/830车型)。 领导； Pedro Adranda Alfaro 发起人： Juan Carlos Licea 区域： VAET 教练： Leonardo Chavez salcedo 起使日： 2002年10月4日 截止日：2002年10月29日 问题描述： 本月收到报告，在美国有关于门饰板的质量问题， 右后门饰板调整不好。(805/830型) 项目综述： 针对Lilao装配厂805/830车型的右后门饰板，发现并 解决其配合不好的问题。 策略： 我们起初应用车型对车型策略，因为多车之间有很好的对比度和平衡关系。(见见第6页) 测量系统： 用测力计作为测量系统。这种测量系统通过等量图。 (DR=13.6, M=1.0lbs, P=13.6lbs见第4页 线索产生： 阶段1表明红X不在过程中。阶段2表明红X产生在门饰板处。成对比较表明门型是成对比较点。 统计确认： 一旦我们查到了候选的红X，我们决定做B vs C六组数据比较，这是为了确认门型影响客户满意度。用6对数据进行比较具有95%的可信度和最大累加值为6. 不可撤销的纠正措施： 为使门边缘加长1.5mm以保证正确的位置不偏离。 改进措施： 减少保修费用来修复由于配合不好产生的门饰板问题，并且100%消除803/805车型的缺陷。 学到的经验： 我们学到了 当投资不会太多的情况下，用其它策略以寻求可选择的方法解决问题。但始终要关注顾客热忱。此项目便是一个这样的例子。 Project Definition Tree/项目定义树Project Definition Tree/项目定义树Problem characterization 问题描述Warranty condition/保修条件The problem is presented only in the rear door 问题只是在后侧车门出现BASED: DISTRIBUTON LOT CAMPAING CURRENT: 100 % door panel 805830 models loose 目前状态：805/830型车门饰板100%松动。 GOAL: eliminate 100% the poor fit. 目标：100%消除不良的配合Project statement 项目综述Need energy to release the trim door panel 需要用力才能卸下门饰板Customer is calling the problem in both sides 客户现在对两侧都在抱怨Effort to release (lbs)/释放力(磅)Frequency频率CURRENT现有DESIRED希望nullSolution tree项目解决树TARGET: Eliminate condition 100% 目标： 100%消除现有问题PASSED ISOPLOT DR=15.2 SEE PAG 4 通过等量图验证 DR=15.2 参见第4页PASSED STAGE 1 SO THE RED X DOESN’T LIVE IN THE PROCESS SEE PAG # 7. 通过第一阶段验证，因此红X不在装配过程中，参见第7页。 THE COMPONENT SEARCH STAGE 2 SHOWS THAT THE RED X DOES NOT LIVE IN THE TRIM DOOR PANEL SO LIVES IN THE REST OF THE DOOR PANEL. SEE PAG # 7./零部件分析阶段2显示红X不是由车门内饰板造成的,所以应是由车门板的其余部分引起的,见7页.PAIR COMPARISON SHOWS THAT THE DOOR PANEL MOLDING HAS THE CHARACTERISTIC THAT PRODUCE THE DIFFERENCE SEE PAG 9/ 配对对比显示车门饰板饰件有造成差异的特征，参见第9页THIS IS THE BEST STRATEGY BASED ON MULTIVARI SEE PAG 5.这是基于多重变换比较得出的最佳策略，参见第5页。PASSED B vs C WHERE IT SHOWS THAT THE DOOR MOLDING KEEP IN POSITION THE DOOR PANEL SEE PAG 10, BUT WE CHOICE OTHER CHARACTERISTIC THAT KEEP IN POSITION THE DOOR PANEL AND WITOUT INVEST MONEY SEE PAG 11 & 13. 通过BC对比试验，显示车门饰板饰件能够使车门饰板保持正确位置，见第10页。但是我们选择能够保证车门饰板位置并且不用增加投资的其它特性，参见第11和13页。Frequency频率GREEN Y RUN CHART SEE PAG 17/绿Y字流程图参见第17页。DESIRED希望CURRENT现有BASED: DISTRIBUTON LOT CAMPAINGEffort to release (lbs)释放力(磅)IsoplotIsoplotDM = 1.0 lbs DP = 13.6 lbs DR = 13.6 Passed Isoplot A Force gage was used as the measurement system拉力计作为测量工具通过ISOPLOT评估nullPOINT TO POINTThere are no contrast in this strategy there are no difference.There is a good contrast but the leverage is not enough because we don’t have tools to measure it.This strategy provide good contrast and the best leverage based on multi-vari chart see pag 6There are no contrast through the timeStrategy diagram PLANT TO PLANTMODEL TO MODELTRUCK TO TRUCKSIDE TO SIDESAME PANELSAME PLANTSAME MODELSAME TRUCKThere is a contrast between truck to truck and good leverage but the contrast is not enough based on multi-vari see page 6.EVENT TO EVENTSAME POINTThere are no contrast in this strategy there are no difference event to event.策略图事件 对 事件同一个点点 对 点同一块扳同一侧 对 同一侧同一台车车 对 车同一型号型号 对 型号同一个工厂工厂 对 工厂在这个对比中,无差异.在这个对比中,无差异.在这个对比中,有差异.但无测量工具.在车与车的对比中,有差异.但不能满足 MULTI-VARI的要求..在这个对比中,有差异.且能满足 MULTI-VARI的要求..在这个对比中,无差异.nullMulti - variNOTE: This graphic shows that the best contrast is model to model.SideTruck1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31EFFORT TO TAKE OFF (lbs)R L R L R L R L R L R L R L R L R L R L R L R L R L R L R L R L R L R L R L R L R L R L R L R L R L R L R L R L R L R L R L Customer satisfaction limitBOBWOWModelC C C C A A A A A A A A A A A S S S S S S S S S S S S S S S SNOTE: S: SUBURBAN 830 MODEL A: AVALANCHE 805 MODEL C: CADILLAC 806 MODEL注:S\A\C三种车型最好最坏顾客的满意底线注:此图显示出在型号与型号的对比中有最大的差异.松脱力(磅)nullComponent searchNOTE: This graphic shows that the Red X suspect does not live in the process because it passed the stage #1. The stage 2 shows that the Red X suspect does not live in the seal window. So the Red X suspect lives in the molding door panel.Effort (lbs)BOBWOWSTAGE 1STAGE 2第一阶段第二阶段最好最坏原状态第一次拆装第二次拆装第三次拆装门窗密封条门饰板门饰板饰条铰链注:这个图显示因为通过了第一阶段的试验,所以RED X不是由操作造成. 第二阶段的试验显示门窗密封条不是RED X,门饰板饰条是RED X要素查找null规则Rule 1. There are complete separation between the BOB and WOW values. 将最好和最坏的值完全分开 规则Rule 2. DY>V1R 设Where: R= RBOB + RWOW 2 那么then RBOB=1.9 RWOW=1.2 R= 1.9 + 1.2 = 3.1 = 1.55. 2 2 DY = YBOB - YWOW = 46.8 - 8.1 = 38.7 如果if DY > V1R 那么then 38.7 > (0.89) (1.55) 所以so 38.7 > 1.38 所以满足规则1和规则2的要求so passed Rule 1 and Rule 2 通过问题为每个单元设定判定极限 Set decision limits for each unit by using the questions: DLBOB= YBOB + V2R DLWOW= YWOW + V2R 那么Then udl = 48.83 DLBOB = 46.8 + 1.31 (1.55) = ldl = 44.77 udl = 10.13 DLWOW = 8.1 + 1.31 (1.55) = ldl = 6.07 结果Results~~~~~~~nullPaired Comparison备注NOTE: We consider that the door panel molding is the Red X candidate in the pair comparison because is the only difference between panel to panel. In 805/830 vs 806 models.我们认为门板装饰件是可能的Red X, 因为在车型805/830 和 车型806对比后只有门板有差异。WOWBOBDOOR MOLDINGRED X CANDIDATEnullPassed B vs C with 95% of confidence that the molding door panel is the Red X with complete separation and end count of 6.在 完全分离，End count 为5的情况下， 以95%的置信度通过B v S 表明门板装饰件是Red XB vs C Six packB = 有门饰件的车Vehicles with door molding. C = 没有饰件的车Vehicles without door molding. Req: 置信度confidence = 95%Once confirmed that the RED X is living in the door panel molding, we decided to change our strategy from “model to model” to “truck to truck” because the door panel molding is unique for Cadillac models. In other words, find one characteristic in the same door panel 805/830 model that makes the difference, because there are some door panels that require almost 20 lbs to pulled out of it position from the door and there are others with less than 10 lbs to pulled out of it position; The customer enthusiasm limit is 18 lbs.this is enough to keep the door panel in position at low cost.一旦确定Red X 为门外板装饰件，我们决定改变策略， 从车型到车型变到卡车到卡车，因为门外板装饰件对Cadillac车型来说是唯一的。换句话说，，同样的门板装在805/830 车型上却产生了差异，原因是一些门板从门上拽出需要20磅的力，二另一些只需要小于10磅的力; 顾客的要求是18磅，这已经足够保证门板以较低的成本固定到相应的位置。nullBOBWOWBIG CONTACT AREASMALL CONTACT AREA大间隙BIGGAP小间隙SMALLGAP门边缘DOOR FLANGE门边缘 DOOR FLANGEThe main characteristic that cause difference between the door panels 805/830 model is the poor contact of the flange on the door due that there are panels that present some deformation on the top edge and cause the poor contact.引起805/830 车型门外板差异的主要特征饰门边缘连接不良，导致门板上边缘变形BOB vs WOW door panel最好、最坏门外板对比nullDOOR FLANGE CURRENT USEDDOOR FLANGE PROPOSEDThis picture shows that if we change the door flange 1.5 mm lenght we guarantee that all door panels require almost 18 lbs to pull out it position because the contact area will be bigger against door flange current DOOR FLANGE CURRENT USEDDOOR FLANGE PROPOSED这张图片表明如果我们把门边缘加长1.5mm，我们确保开门力会增大18磅，原因是接触面积比当前增大了。当前使用的门边缘推存使用的门边缘nullB = Vehicles with new flange longer than 8.5 mm. C = Vehicles with current flange smaller than 7.9 mm. Req: confidence = 95%Passed B vs C with 95% of confidence that the new flange length 1.5 mm is the Red X with complete separation and end count of 6. B vs C Six packB=门边缘比当前长8.5mm的汽车 C=门边缘比当前小7.9mm的汽车 可信度为95% 检测数据通过了B vs C ，95%可以相信加长1.5mm的门边缘是完全独立的红X，并且累计数是6。nullCUSTOMER SATISFACTION LIMITCURRENT PROCESSBREAK POINTEffort (lbs)Green Y Run ChartRL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL RL WEEK 40 WEEK 41 WEEK 42 WEEK 43 WEEK 44 WEEK 45 WEEK 46 WEEK 47 DATE绿Y运行表当前进程断点Product Engineering产品工程Product Engineering产品工程 14SQA - Supplier Quality Assurance QET - Quality Engineering Team PST - Plant Support Team SMT - Systems Management Team RGM - Reliability Growth ManagementLine Balance Mandatory SequenceStage 3Engineering ChangeResponsibilityQET PST (Lead) ¨5b SMT RGM (Lead) ¨6Statistical Engineering (Red X Strategies) ¨7Will Process Change Resolve?Product Change Required?Hardy Perennial or Extremely Complex Problem 产品工程 强制的顺序工程更改第三阶段职责统计工程 (红X策略)质量工程SQA-供应商质量保证 QET-质量工程小组 PST-工厂支持小组 SMT-系统管理小组 RGM-可靠性增长小组 工艺更改会解决吗？需要产品更改吗艰难长期或者及其复杂的问题nullnullnullnullnullnullnullnull