You have a lathe operation. You notice that you are getting a 'trail' on your piece. Your inspection shows that although this 'trail' is visible, the part passes your surface roughness requirement. You slow down feed rate 10% from set-up and the problem worsens. The operator increases the feed rate 20% from set-up and the problem goes away. The responsible engineer changes the CNC set-up program to reflect this 20% increase. Is this a process change? I wouldn't classify it as such but you can get an arguement that it is. It really has no effect on the product - it was passing (technically) prior to the change and the change of a feed rate is not (opinion on word choice) significant.
You have to make some decisions on a case by case basis what is 'significant' to your process.
Where operators can make adjustments they are typically 'minor' - you want an engineer involved if a 'major' problem arises. Where the operators are 'tuning', they are not really changing the process. In the lathe example operators are allowed to adjust feed rate +/- 30% from set-up. They do not have to get authorization to do this. However - it is expected that the responsible engineer (or other 'authority') be (in general) monitoring what is going on so that s/he knows if every operator on every shift is always setting the feed rate up 20%. The operators are not, however, required to record the change unless it is in response to an out-of-control condition where it is recorded on the control chart.
However, if you change the set-up instruction it technically becomes a process change, but is it significant? Not in my opinion. I don't believe 4.9.5 is meant to control this type of minor process change. I see 4.9.5 as important in more 'significant' process changes such as a change from a grind to a hone, a large process parameter change (such as mold temperature or hold time in injection molding) or a sequence change.
[This message has been edited by Marc Smith (edited 23 May 2000).]