D
Dawn
Can someone tell me where I can find more information on gage correlations?
D
Dawn
I have been inside and out of the MSA book, but what I'm looking for is actual instructions on how to compare a gage to another to determine how far off the readings are from each other.
I would recommend looking at NCSL (National Conference of Standards Laboratories) RP (Recommended Practice) 15. That is interlaboratory comparison practices.
If you are looking for information on gage R&R studies, a copy of MSA (Measurement Systems Analysis) book from the QS9000 series.
------------------
If you are looking for information on gage R&R studies, a copy of MSA (Measurement Systems Analysis) book from the QS9000 series.
------------------
I was unclear as to what you were specifically requesting. I still do not know exactly what your need is. If you could share what specific type of gauge, I have calibrated many, many types and can share some ideas.
If you are, for example speaking of pressure gauges, you need either a lab standard, or a Tee connector and a variable pressure source. Once you connect to the tee, adjust the pressure to a desired reading range, and compare the two.
From a metrology perspective, I feel it important to point out that you cannot ever expect two gauges to be exactly the same. If they are rated at +/-2%FS (full scale)with a 100 PSIG full scale, for example, they can be legitimately 4 PSI away from each other at virtually any point on the scale, and still considered to be as accurate as required by their specifications. And that is only if you have had them calibrated at necessary intervals.
If you want a method to correlate pressure gages installed in a system on a regular basis, you can install tee connectors with cutoff valves adjacent to the installed gages, and periodically install your correlation gage as a check.
But bear in mind with the +/- accuracy qualifiers in my earlier paragraph (above), you may still not know which one of the two is right, if indeed either of them is.
------------------
If you are, for example speaking of pressure gauges, you need either a lab standard, or a Tee connector and a variable pressure source. Once you connect to the tee, adjust the pressure to a desired reading range, and compare the two.
From a metrology perspective, I feel it important to point out that you cannot ever expect two gauges to be exactly the same. If they are rated at +/-2%FS (full scale)with a 100 PSIG full scale, for example, they can be legitimately 4 PSI away from each other at virtually any point on the scale, and still considered to be as accurate as required by their specifications. And that is only if you have had them calibrated at necessary intervals.
If you want a method to correlate pressure gages installed in a system on a regular basis, you can install tee connectors with cutoff valves adjacent to the installed gages, and periodically install your correlation gage as a check.
But bear in mind with the +/- accuracy qualifiers in my earlier paragraph (above), you may still not know which one of the two is right, if indeed either of them is.
------------------
D
Dawn
I'm sorry I wasn't clear in my request. I am trying to determine how much difference in a drop indicator on the floor and a (micrometer, depth mic, 1" mic.)We are finding differences up to .004 and we need to do a gage correlation to see how big a difference betwen the two. Also, once I find the difference, what do I do? Thanks for your help, new at this.
There are a number of possible directions you could go with an issue such as this. I'm trying to think of a simple way to answer...
One thing you will need is something stable to measure (I'll call it a DUT (Device Under Test). Make multiple measurements of the DUT with one measurement device. Determine how much variability you are getting measuring the DUT with one measurement instrument. Do this for each measurement instrument. If your minimum resolution on your measuring instruments is .001 and you have less than .001 variation in the multiple measurements done with one measurement instrument (for each of them), that says for starters that you have good repeatability. Next compare the mean value between the measuring instruments. Compare that delta between them also against their operating specifications. To do this, you have to add the specs together of the measuring instruments. If they each have specs of
+/-.002 for example, and you have two measuring instruments, they could then legitimately be as far off from each other as +/-.004.
Compare that total delta against your performance specs for the DUT. If the performance specs of the DUT are less than this, then your measuring instruments are not adequate for the job.
Whatever the DUT spec is, your measurement instrument must have sufficiently tighter specs to assure that measurement instrument accuracy, variability, etc. stays well within the specs of the DUT. A 10 to 1 ratio is a good rule of thumb in many cases.
If you need further elaboration, please feel free to post further. Hope I have been of help.
------------------
One thing you will need is something stable to measure (I'll call it a DUT (Device Under Test). Make multiple measurements of the DUT with one measurement device. Determine how much variability you are getting measuring the DUT with one measurement instrument. Do this for each measurement instrument. If your minimum resolution on your measuring instruments is .001 and you have less than .001 variation in the multiple measurements done with one measurement instrument (for each of them), that says for starters that you have good repeatability. Next compare the mean value between the measuring instruments. Compare that delta between them also against their operating specifications. To do this, you have to add the specs together of the measuring instruments. If they each have specs of
+/-.002 for example, and you have two measuring instruments, they could then legitimately be as far off from each other as +/-.004.
Compare that total delta against your performance specs for the DUT. If the performance specs of the DUT are less than this, then your measuring instruments are not adequate for the job.
Whatever the DUT spec is, your measurement instrument must have sufficiently tighter specs to assure that measurement instrument accuracy, variability, etc. stays well within the specs of the DUT. A 10 to 1 ratio is a good rule of thumb in many cases.
If you need further elaboration, please feel free to post further. Hope I have been of help.
------------------
Similar threads
