I am going through some of our gauges and trying to follow the internal standard that is based on the AIAG MSA manual, but I am uncertain on a few things.
Capability analysis of measuring systems:
Under measurement standard (Ucal), it says the uncertainty of the measurement standard has to be indicated and it should be ≤ 5% of the smallest tolerance to be inspected. In our case, the lowest tolerance is 0.3 for runout, so 0.015 or less, and usually ± 1mm (2mm total) for length, so 0.10mm? My question for this one is...when you get the uncertainty on a measurement standard, what all is involved? The measurement standard itself is just a shaft so I would imagine that any uncertainty on the standard would be from the measuring device used to certify the standard. Unless you dive deep into other modes such as temperature changes, etc, but I have no way to control that in our facility.
Measuring systems with calibration masters - The difference between the master and measurement standard has to be ≥ 66% of the tolerance. If I understand this right, we need to have: Ex. one shaft at 700.00mm and another at least 701.32 and above or 698.68. I guess I am getting confused at this because just below this section is linearity deviation and it talks about the largest length to be measured has to be > 66% of the operating range, but a minimum of 5 lengths are measured at least 2 times within the measuring range. So for clarification, I need to just have a calibration master at x value (in our case, 846.699) , another shaft atleast 1,188 for 1800mm scale, and then 3 more shafts in between there that have a difference of atleast 1.32mm?
I ran the Type 1 study on a reference shaft in Q-DAS thinking it would calculate uncertainty to some extent and also ran a Type 2 study (10 pcs, 3 ops, 2 trials) and then ran a VDA 5 study combining the two to calculate the QMS/QMP of the measuring system and process and neither actually provided the uncertainty based on all that data. I am trying to find out how it is expected to get that uncertainty data through Q-DAS. Previously we would just measure a master part 20 times and the spreadsheet calculates the stdev and standard uncertainty, k=2, and k=3. You would get the standard uncertainty of the measurement standard and then any Type B data from calibration certs and previous information could also be added and calculated. Now we are supposed to do all this through Q-DAS and this is the document I have to go by. Thoughts?
Capability analysis of measuring systems:
Under measurement standard (Ucal), it says the uncertainty of the measurement standard has to be indicated and it should be ≤ 5% of the smallest tolerance to be inspected. In our case, the lowest tolerance is 0.3 for runout, so 0.015 or less, and usually ± 1mm (2mm total) for length, so 0.10mm? My question for this one is...when you get the uncertainty on a measurement standard, what all is involved? The measurement standard itself is just a shaft so I would imagine that any uncertainty on the standard would be from the measuring device used to certify the standard. Unless you dive deep into other modes such as temperature changes, etc, but I have no way to control that in our facility.
Measuring systems with calibration masters - The difference between the master and measurement standard has to be ≥ 66% of the tolerance. If I understand this right, we need to have: Ex. one shaft at 700.00mm and another at least 701.32 and above or 698.68. I guess I am getting confused at this because just below this section is linearity deviation and it talks about the largest length to be measured has to be > 66% of the operating range, but a minimum of 5 lengths are measured at least 2 times within the measuring range. So for clarification, I need to just have a calibration master at x value (in our case, 846.699) , another shaft atleast 1,188 for 1800mm scale, and then 3 more shafts in between there that have a difference of atleast 1.32mm?
I ran the Type 1 study on a reference shaft in Q-DAS thinking it would calculate uncertainty to some extent and also ran a Type 2 study (10 pcs, 3 ops, 2 trials) and then ran a VDA 5 study combining the two to calculate the QMS/QMP of the measuring system and process and neither actually provided the uncertainty based on all that data. I am trying to find out how it is expected to get that uncertainty data through Q-DAS. Previously we would just measure a master part 20 times and the spreadsheet calculates the stdev and standard uncertainty, k=2, and k=3. You would get the standard uncertainty of the measurement standard and then any Type B data from calibration certs and previous information could also be added and calculated. Now we are supposed to do all this through Q-DAS and this is the document I have to go by. Thoughts?
