W
Hi all,
we are in charge of develping methods to measure the dialing torque of (plastic) insulin pens. The production is at the very beginning so we don't have any information of the process variation.
We do have upper and lower specification limits. We don't have a reference part or normal.
The measurement itself is non destructive.
We started with 10 parts 3 Appraisers 3 trials - and perfectly failed, even all single values are within the spec limits.
My question is threefold:
1) Is there a way to perform somewhat like approach 1. E.g.
- Measureing a part which is at LSL 25 times
- Measureing a part which is close to the spec, use it as reference?
- Measureing a part 25 times which is at the USL
to get c_g-values at both ends of the distribution?
2) Is a GRR in principle possible if we just have the tolerance and a not stable process?
3) Is a reduction of a single part which is close to the spec useful as supposed by the project manager and use its mean wrt. the spec value for a calculation of c_g?
Thank you very much in advance
Waeller
we are in charge of develping methods to measure the dialing torque of (plastic) insulin pens. The production is at the very beginning so we don't have any information of the process variation.
We do have upper and lower specification limits. We don't have a reference part or normal.
The measurement itself is non destructive.
We started with 10 parts 3 Appraisers 3 trials - and perfectly failed, even all single values are within the spec limits.
My question is threefold:
1) Is there a way to perform somewhat like approach 1. E.g.
- Measureing a part which is at LSL 25 times
- Measureing a part which is close to the spec, use it as reference?
- Measureing a part 25 times which is at the USL
to get c_g-values at both ends of the distribution?
2) Is a GRR in principle possible if we just have the tolerance and a not stable process?
3) Is a reduction of a single part which is close to the spec useful as supposed by the project manager and use its mean wrt. the spec value for a calculation of c_g?
Thank you very much in advance
Waeller