How is the interaction physically happening while performing ANOVA GR&R?

  • Thread starter luis.armendariz
  • Start date
L

luis.armendariz

How is the interaction physically happening while performing a GR&R?
What are the causes of it?

I know repeatibility is due to the variation of measuring the same part which is expected to be the same result and reproducibility is the variation when 2 or more appraisers measure the same part which is also expected to be the same result, but what about interaction?
What kind of variation is? how it is present?

Another question is why the 50% or more of the points on the Xbar chart must fall outside the control limits? How those limits are calculated?
I'm assuming they are not calculated as usual in a Xbar Chart; since it would be expected that all fall within the control limits.
 

Miner

Forum Moderator
Leader
Admin
Re: Interaction on GR&R ANOVA method

Operator by part interactions are usually the result of a difference in operator technique combined with part to part differences that are either exposed or hidden by that technique.

Lets use a simple example of measuring the diameter of a metal shaft. Some of the shafts have a taper near the end while other shafts do not. One operator always takes measurements from the middle of the shaft while another operator always takes measurements near the end. This will show up as a significant operator by part interaction because the operators will obtain similar results for those shafts with no taper, but will obtain different results for those shafts with taper.

Regarding your second question, the control limits are calculated based on 100% measurement error. Consequently, the control limits define a blackout zone inside which the measurement device cannot see anything. The device can only see a measurement when the measurement exceeds this blackout zone. Therefore, you would want as many points as possible to exceed this zone.
 
Top Bottom