MSA实例讨论,期待高手

客户要新产品的MSA分析,具体测量方法,取10个样本,3个测量员,每人每件测量3次,量具公差0.6%.以下是我们的测量结果
序号 测量值 测量人测量次数116.22411216.2211316.21511416.22211516.21811616.21611716.21511816.2211916.218111016.21811116.22412216.2212316.21512416.22212516.21812616.21612716.21612816.2212916.219121016.21812116.22413216.22113316.21513416.22213516.21913616.21613716.21513816.2213916.218131016.21813116.22321216.2221316.21521416.22221516.21821616.21621716.21521816.22121916.218211016.21921116.22422216.2222316.21522416.22222516.21822616.21522716.21522816.2222916.218221016.21822116.22423216.2223316.21523416.22223516.21823616.21623716.21523816.2223916.218231016.21823116.22431216.22131316.21531416.22331516.21831616.21631716.21531816.2231916.218311016.21831116.22432216.2232316.21532416.22232516.21832616.21632716.21632816.22132916.218321016.21932116.22433216.2233316.21533416.22233516.21833616.21633716.21533816.2233916.218331016.21833
我首先做正态分析,P<0.05
然后,使用R&R(交叉)分析,分析方法"Xbar R",结果如下:
来源 标准差(SD) (6 * SD) 异 (%SV)
合计量具 R&R 0.0002664 0.0015981 9.49
重复性 0.0002560 0.0015357 9.12
再现性 0.0000737 0.0004422 2.62
部件间 0.0027952 0.0167715 99.55
合计变异 0.0028079 0.0168475 100.00
重复和再现差异都很小,测量系统OK.但是从分析的图形来看总是觉得有些怪异,如下:

大家可以看下左中\下的两张控制图,中间的一章波动很怪异仿佛能看到某种规律,而最下面的很多点都不在控制限内,带着这些疑惑,我又做了直方图,结果发现分布呈现多个波峰.

至此我完全迷惘了:从分析的数据来看一切正常,但是从各种图表分析又非常怪异,到底我的分析漏掉了那里?