無標(biāo)題文章


Q&A: Cpk and SPC Standards (Part 3)

SPC

? Knowledge Sharing

?

We have covered the general concepts of for? Cpk. let’s explore what we expect about Cpk and how to improve Cpk? performance. If you’d like to review the previous sharing sessions, please click? the following links.


我們已經(jīng)了解了Cpk的基礎(chǔ)概念。本次我們來探討一下對(duì)Cpk的期望以及如何提高Cpk的性能。單擊下方鏈接回顧之前的分享丈甸。

?

Q: How do we measure how well a process? is running within the spec limits?

A:?If Cpk is 1.33 or higher, it is commonly? considered that the process is capable.??It is useful to show the real performance? by graphing a trend chart (SPC chart) of process variation against? specification limits.


問:如何衡量制程在規(guī)格限制中的表現(xiàn)?

答:如果Cpk為1.33或更高饶唤,則通常認(rèn)為該制程是有能力的。繪制表現(xiàn)制程變化的趨勢(shì)圖(SPC圖)可以直觀展示制程的實(shí)際性能贯钩。


[if !vml]

[endif]


Q: Why 1.33 of Cpk?

A: 1.33 of Cpk satisfies the SPC? standards of “Key” charts. Based on the formula, we know when Cpk is 1.33, the? distance between spec limits and process mean is at least 4 sigma. At this? time, the specification width is wider enough than the process spread.? Besides, DPPM is only 63.3.



問:為什么Cpk需要達(dá)到1.33募狂?

根據(jù)公式,Cpk為1.33時(shí)魏保,規(guī)格限制和制程均值之間的距離至少是4倍標(biāo)準(zhǔn)差熬尺。此時(shí)摸屠,相比制程分布谓罗,規(guī)格限制足夠?qū)挘褼PPM僅為63.3季二。


Cpk# of SigmaDPPM (Defects parts per Million)

0.331317,311

0.67245,500

1.0032,700

1.33463.3

1.504.53.4

1.6750.57

2.0060.002


[if !vml]

[endif]


Q: Wait, what are “Key” charts? Are they? important?

A: Of course, they are important.? However, “Key” in “Key charts” does not mean “important”. Instead, it is one? of the 4 chart classifications of importance level at DMTM.


Any SPC

? chart at DMTM is classified as one of the? following classifications, ranking from the most to the least important? level.

[if !supportLists]1.???????

? [endif]Program? Control

[if !supportLists]2.????

? [endif]Key

[if !supportLists]3.????

? [endif]Control

[if !supportLists]4.????

? [endif]Engineering


問:等等檩咱,什么是 “Key? 圖表”?它們很重要嗎胯舷?

答:當(dāng)然刻蚯,它們很重要。然而桑嘶,“Key圖表”中的 “Key” 并不代表 “重要”炊汹。實(shí)際上,它是DMTM圖表的4個(gè)重要性級(jí)別之一逃顶。


DMTM的任何SPC圖表都被歸類為下列其中一個(gè)級(jí)別讨便,重要性從高到低排序。

[if !supportLists]1.???????

? [endif]Program? Control

[if !supportLists]2.????

? [endif]Key

[if !supportLists]3.????

? [endif]Control

[if !supportLists]4.????

? [endif]Engineering


Q: What are SPC standards to evaluate? Cpk? Can you specify them?

A: Yes.?For Cpk, we only monitor the two most? important classifications, Program Control and Key charts. They have a direct? relation to one or more output parameters (i.e. yield, quality, reliability,? or performance). Here are the specific SPC rules.


We monitor Cpk based on the latest 28-day? data:

[if !supportLists]·????????

? [endif]For Program? Control charts, the process is manufacturable if Cpk >= 1.67 (5 sigma)

[if !supportLists]·????????

? [endif]For Key? charts, the process is manufacturable if Cpk >= 1.33 (4 sigma)

[if !supportLists]·????????

? [endif]For Control? and Engineering charts, we do not monitor Cpk


問:評(píng)估Cpk的SPC標(biāo)準(zhǔn)是什么以政?能具體說明嗎霸褒?

答:對(duì)于Cpk,我們只監(jiān)測(cè)兩類最重要的圖表盈蛮,Program

? Control 和? Key废菱。它們與一個(gè)或多個(gè)輸出參數(shù)(即產(chǎn)量、質(zhì)量、可靠性或性能)有直接關(guān)系殊轴。以下是具體的SPC規(guī)則衰倦。


我們根據(jù)最新28天的數(shù)據(jù)來監(jiān)測(cè)Cpk。

[if !supportLists]·????????

? [endif]對(duì)于 Program

? Control 圖表梳凛,如果Cpk >=1.67(5倍標(biāo)準(zhǔn)差)耿币,則該制程是可制造的。

[if !supportLists]·????????

? [endif]對(duì)于 Key 圖表韧拒,如果Cpk >= 1.33(4倍標(biāo)準(zhǔn)差)淹接,則該制程是可制造的。

[if !supportLists]·????????

? [endif]對(duì)于 Control 和Engineering圖表叛溢,我們不對(duì)Cpk進(jìn)行監(jiān)控塑悼。


Q: How can we improve the Cpk performance?

A: Technically, Cpk can be improved by

[if !supportLists]1.???????

? [endif]Optimize? process

[if !supportLists]a.?????

? [endif]Centering the? process on target

[if !supportLists]b.????

? [endif]Reducing the? variation in the site-level process

[if !supportLists]2.????

? [endif]Optimize spec? limits


問:我們?nèi)绾翁岣逤pk的性能?

答:從技術(shù)上講楷掉,Cpk可以通過以下方式改進(jìn)

[if !supportLists]1.???????

? [endif]優(yōu)化工藝

[if !supportLists]a.?????

? [endif]將制程以目標(biāo)為中心進(jìn)行

[if !supportLists]b.????

? [endif]減少量測(cè)點(diǎn)級(jí)別制程的變化

[if !supportLists]2.????

? [endif]優(yōu)化規(guī)格限制


References

[if !supportLists]·????????

? [endif]WBT:? INTRODUCTION TO SPC

[if !supportLists]o??

? [endif]Saba ID:

? 00015598

[if !supportLists]·????????

? [endif]ILT: SPC? Concept_Statistics Process Control

[if !supportLists]o?? [endif]Saba ID: 20016444




Question: Data being used in the initial set-up? of a process are assumed to have a normal distribution. If the nominal? (target) is set at the center of the distribution, and the specification? limits are set at ±3s from the center, then the Cpk is equal to:


(A) –0.25

(B) 1.00

(C) 1.33

(D) 1.67



Body text: IntelOne Text, 12pt, black.


When inserting hyperlinks, Word sometimes defaults the link color to

? the old Intel blue or a different color. Make sure to change all hyperlinks

? to use Intel cobalt? blue(RGB 30, 46, 184). Select the hyperlink >? select Font Color > More Colors > Custom tab.


Important: Do not copy hyperlinks from old,? targeted emails. The links may not be valid.


Section? headings, use sentence case, IntelOne Text Bold, 12pt


Use sections to organize key? information.


Other? formatting reminders


Make sure you’re using the built-in? styles up above. These have been modified using the correct fonts, colors and? styles.

List bullets should be square, not? round

Date format above and throughout? comms: months followed by a date should be abbreviated (excluding March,? April, May, June and July) when communicating globally or to the U.S.? Example: Aug. 8, 2020. For other regions local preferences may be followed.

There should only be one space? after sentences, not two

Banner image should be 7” or 700px? x 150px. Click on the banner, select the Picture Format tab > Width and? change size if needed

Correct time format: Monday–Friday? 8 a.m. – 5 p.m.

Refer to the Writing Style Guide for more details.


Tables


Header row text should be IntelOne? Text 11pt, with Intel classic blue fill (RGB: 0, 104, 181) and white text.? Use new add to calendar button. Example:


DateTimeVenueCalendar Link

July? ? 2610–11? ? p.m. PacificWebcast[if !vml]

[endif]

最后編輯于
?著作權(quán)歸作者所有,轉(zhuǎn)載或內(nèi)容合作請(qǐng)聯(lián)系作者
  • 序言:七十年代末厢蒜,一起剝皮案震驚了整個(gè)濱河市,隨后出現(xiàn)的幾起案子烹植,更是在濱河造成了極大的恐慌斑鸦,老刑警劉巖,帶你破解...
    沈念sama閱讀 219,366評(píng)論 6 508
  • 序言:濱河連續(xù)發(fā)生了三起死亡事件草雕,死亡現(xiàn)場(chǎng)離奇詭異巷屿,居然都是意外死亡,警方通過查閱死者的電腦和手機(jī)墩虹,發(fā)現(xiàn)死者居然都...
    沈念sama閱讀 93,521評(píng)論 3 395
  • 文/潘曉璐 我一進(jìn)店門嘱巾,熙熙樓的掌柜王于貴愁眉苦臉地迎上來,“玉大人诫钓,你說我怎么就攤上這事旬昭。” “怎么了菌湃?”我有些...
    開封第一講書人閱讀 165,689評(píng)論 0 356
  • 文/不壞的土叔 我叫張陵问拘,是天一觀的道長。 經(jīng)常有香客問我惧所,道長骤坐,這世上最難降的妖魔是什么? 我笑而不...
    開封第一講書人閱讀 58,925評(píng)論 1 295
  • 正文 為了忘掉前任纯路,我火速辦了婚禮或油,結(jié)果婚禮上,老公的妹妹穿的比我還像新娘驰唬。我一直安慰自己顶岸,他們只是感情好腔彰,可當(dāng)我...
    茶點(diǎn)故事閱讀 67,942評(píng)論 6 392
  • 文/花漫 我一把揭開白布。 她就那樣靜靜地躺著辖佣,像睡著了一般霹抛。 火紅的嫁衣襯著肌膚如雪。 梳的紋絲不亂的頭發(fā)上卷谈,一...
    開封第一講書人閱讀 51,727評(píng)論 1 305
  • 那天杯拐,我揣著相機(jī)與錄音,去河邊找鬼世蔗。 笑死端逼,一個(gè)胖子當(dāng)著我的面吹牛,可吹牛的內(nèi)容都是我干的污淋。 我是一名探鬼主播顶滩,決...
    沈念sama閱讀 40,447評(píng)論 3 420
  • 文/蒼蘭香墨 我猛地睜開眼,長吁一口氣:“原來是場(chǎng)噩夢(mèng)啊……” “哼寸爆!你這毒婦竟也來了礁鲁?” 一聲冷哼從身側(cè)響起,我...
    開封第一講書人閱讀 39,349評(píng)論 0 276
  • 序言:老撾萬榮一對(duì)情侶失蹤赁豆,失蹤者是張志新(化名)和其女友劉穎仅醇,沒想到半個(gè)月后,有當(dāng)?shù)厝嗽跇淞掷锇l(fā)現(xiàn)了一具尸體魔种,經(jīng)...
    沈念sama閱讀 45,820評(píng)論 1 317
  • 正文 獨(dú)居荒郊野嶺守林人離奇死亡析二,尸身上長有42處帶血的膿包…… 初始之章·張勛 以下內(nèi)容為張勛視角 年9月15日...
    茶點(diǎn)故事閱讀 37,990評(píng)論 3 337
  • 正文 我和宋清朗相戀三年,在試婚紗的時(shí)候發(fā)現(xiàn)自己被綠了务嫡。 大學(xué)時(shí)的朋友給我發(fā)了我未婚夫和他白月光在一起吃飯的照片甲抖。...
    茶點(diǎn)故事閱讀 40,127評(píng)論 1 351
  • 序言:一個(gè)原本活蹦亂跳的男人離奇死亡漆改,死狀恐怖心铃,靈堂內(nèi)的尸體忽然破棺而出,到底是詐尸還是另有隱情挫剑,我是刑警寧澤去扣,帶...
    沈念sama閱讀 35,812評(píng)論 5 346
  • 正文 年R本政府宣布,位于F島的核電站樊破,受9級(jí)特大地震影響愉棱,放射性物質(zhì)發(fā)生泄漏。R本人自食惡果不足惜哲戚,卻給世界環(huán)境...
    茶點(diǎn)故事閱讀 41,471評(píng)論 3 331
  • 文/蒙蒙 一奔滑、第九天 我趴在偏房一處隱蔽的房頂上張望。 院中可真熱鬧顺少,春花似錦朋其、人聲如沸王浴。這莊子的主人今日做“春日...
    開封第一講書人閱讀 32,017評(píng)論 0 22
  • 文/蒼蘭香墨 我抬頭看了看天上的太陽氓辣。三九已至,卻和暖如春袱蚓,著一層夾襖步出監(jiān)牢的瞬間钞啸,已是汗流浹背。 一陣腳步聲響...
    開封第一講書人閱讀 33,142評(píng)論 1 272
  • 我被黑心中介騙來泰國打工喇潘, 沒想到剛下飛機(jī)就差點(diǎn)兒被人妖公主榨干…… 1. 我叫王不留体斩,地道東北人。 一個(gè)月前我還...
    沈念sama閱讀 48,388評(píng)論 3 373
  • 正文 我出身青樓颖低,卻偏偏與公主長得像硕勿,于是被迫代替她去往敵國和親。 傳聞我的和親對(duì)象是個(gè)殘疾皇子枫甲,可洞房花燭夜當(dāng)晚...
    茶點(diǎn)故事閱讀 45,066評(píng)論 2 355

推薦閱讀更多精彩內(nèi)容

  • 目錄 第1章 概述 - 4 - 1.1 Hello簡(jiǎn)介 - 4 - 1.2環(huán)境與工具 - 4 - 1.3中間結(jié)果 ...
    輝夜大小姐不可能告白我閱讀 234評(píng)論 0 0
  • Stationarity, cointegration and GARCH model of financialt...
    7813_bbe7閱讀 233評(píng)論 0 0
  • 好的用戶界面-界面設(shè)計(jì)的一些技巧( 原文地:Good UIhttp://goodui.org ) 1 盡量使用單列...
    荷爾蒙閱讀 573評(píng)論 0 1
  • 轉(zhuǎn)至元數(shù)據(jù)結(jié)尾創(chuàng)建: 董瀟偉源武,最新修改于: 十二月 23, 2016 轉(zhuǎn)至元數(shù)據(jù)起始第一章:isa和Class一....
    40c0490e5268閱讀 1,721評(píng)論 0 9
  • # 云計(jì)算實(shí)驗(yàn) 如果說亞馬遜的AWS(Amazon Web Service)是一個(gè)IaaS平臺(tái),為用戶提供計(jì)算服務(wù)...
    喲桑_6481閱讀 609評(píng)論 0 0