1 / 35
文档名称:

six sigma for elec- design and mfg.pdf

格式:pdf   页数:35
下载后只包含 1 个 PDF 格式的文档,没有任何的图纸或源代码,查看文件列表

如果您已付费下载过本站文档,您可以点这里二次下载

six sigma for elec- design and mfg.pdf

上传人:管理资源吧 2012/2/21 文件大小:0 KB

下载得到文件列表

six sigma for elec- design and mfg.pdf

文档介绍

文档介绍:Chapter
2
The Elements of Six Sigma
and Their Determination
In this chapter, the concepts needed to define six sigma quality in de-
sign and manufacturing are differentiated from each other. Several
techniques are developed for analyzing individual parts, as well as
higher orders plexity such as assemblies, modules, systems,
and product designs. In addition, techniques for measuring manufac-
turing line performance are also developed for use in the six sigma
concept. The following topics are discussed in this chapter:
1. The quality measurement techniques: SQC, six sigma, Cp and
Cpk. This section is a review of the different methods used to de-
sign for quality as well as to control quality. Several techniques are
outlined and the differences between the methods are contrasted.
2. The Cpk approach versus six sigma. In this section, the concept of
Cpk is analyzed pared to six sigma. The Cpk approach re-
duces some of the ambiguities of the ␴ shift of the process aver-
age used in the traditional Six Sigma calculations. Cpk calcula-
tions, including negative Cpk, are analyzed, and the effects of
average shifts on Cpk are also shown.
3. Calculating defects using normal distribution. In this section, de-
fect calculations are shown for variable and attribute processes
and designs. Many examples are shown for different conditions of
average shift and process variability.
4. Are manufacturing processes and supply parts always normally
distributed? Assuming normality of manufacturing process distri-
33
34 Six Sigma for Electronics Design and Manufacturing
bution is an important part of calculating defects, yields, and per-
forming other statistical analyses of six sigma. In this section, the
requirements for assuming normal distribution of manufacturing
processes are examined, as well as tests that can be made to re-
view normality of data. In addition, methods for handling nonnor-
mal distribution of data for six