量化的equipment可用GR&R,不能量化的呢？
可量化量測的設備一般用GR&R 方式驗證設備可靠性.
非量化,或 GO/ NO-GO設備可以 KAPPA 方式驗證：
Computing way as follows:
a. Select 20 good and 20 bad parts
b. Have each inspector inspect each of the parts and record whether it “Passed” or “Failed”;
capture data in a “Kappa Study Data Table”. Perform the study such that each inspector does not know what the other inspector’s results are.
c. Convert the data into proportions and populate a “Kappa Contingency Table”
d. Compute Kappa = (Pobserved - Pchance)/(1-Pchance)
1. Pobserved is the proportion of units on which both inspections agree
2. Pchance is the proportion of agreements expected by chance
e. Kappa shall be >.7. If Kappa is <.7, the measurement system must be improved before accepting or rejecting parts.
需要Kappa Study Data Table的，apply on demand.
非量化,或 GO/ NO-GO設備可以 KAPPA 方式驗證：
Computing way as follows:
a. Select 20 good and 20 bad parts
b. Have each inspector inspect each of the parts and record whether it “Passed” or “Failed”;
capture data in a “Kappa Study Data Table”. Perform the study such that each inspector does not know what the other inspector’s results are.
c. Convert the data into proportions and populate a “Kappa Contingency Table”
d. Compute Kappa = (Pobserved - Pchance)/(1-Pchance)
1. Pobserved is the proportion of units on which both inspections agree
2. Pchance is the proportion of agreements expected by chance
e. Kappa shall be >.7. If Kappa is <.7, the measurement system must be improved before accepting or rejecting parts.
需要Kappa Study Data Table的，apply on demand.
没有找到相关结果
已邀请:
RSS Feed
1 个回复
jasion_lv (威望:0) (广东 中山)
赞同来自: