GR&R for Torque application

[DavidJ.BlackII]

Hello all,

I am looking for advice on what kind of GRR to perform on a assembly line process. In our production plant we have audit stations to compare to our atlas copco transducers on the line. These audit stations have clicker-style calibration certified wrenches at each, and are used on housing bolts to determine if they were tightened to print specifications, or if they will breakaway before the minimum value is reached.

I believe either an attribute study or a cap study is the correct format, but am unable to determine between the two. The test seems more destructive as once the bolt is loosened you can’t breakaway on the same bolt again, so would it be an attribute study? But (if I’m correct) an attribute study wouldn’t give me a Cpk value to make a final decision, and I wouldn’t be able to validate whether or not the gage is capable. Any help is appreciated!

 

[Miner]

Have you ever detected a problem with your Atlas-Copco system? Using a clicker style torque wrench to evaluate an Atlas-Copco system strikes me as similar to using a tape measure to inspect after a micrometer, and non-value-added. The Atlas-Copco system has its own built-in measurement system. Plus, as you mentioned, breakaway torque is not measuring the same characteristic as the run-down torque.

 

[DavidJ.BlackII]

Have you ever detected a problem with your Atlas-Copco system? Using a clicker style torque wrench to evaluate an Atlas-Copco system strikes me as similar to using a tape measure to inspect after a micrometer, and non-value-added. The Atlas-Copco system has its own built-in measurement system. Plus, as you mentioned, breakaway torque is not measuring the same characteristic as the run-down torque.

It feels counterintuitive but we were given a non-conformance for not having a R&R for our audit torque wrench stations. It seems that specifically we have to calibrate all of our measuring equipment, and then perform R&Rs on each equipment type for each type of inspection they do to determine if said equipment is capable of accurately measuring that feature. The redundancy is palpable. We’ll have a vendor come in and service our Promess presses, and then the engineers have to do studies to ensure that what the software is relaying to us is accurate (using a calibrated strain gage and load cell kit).

 

[Miner]

You may find my blog post on Non-replicable measurement systems helpful. In your application, I would recommend the "consecutive specimens" taken across multiple setups approach.

You did not answer my question. If you have not seen an issue with the Atlas-Copco equipment, eliminate the audit.

 

[Matt's Quality Handle]

You did not answer my question. If you have not seen an issue with the Atlas-Copco equipment, eliminate the audit.

Not OP, but perhaps the audit is useful in finding bolts that were skipped instead of bolts that were undertorqued by the Atlas-Copco equipment. The equipment is more than capable, but the audit could be targeting a "manpower" or "method" problem, not a "measurement" issue.

It does seem counterproductive to verify the values of a higher accuracy method with a lower accuracy method.

 

[Miner]

Not OP, but perhaps the audit is useful in finding bolts that were skipped instead of bolts that were undertorqued by the Atlas-Copco equipment. The equipment is more than capable, but the audit could be targeting a "manpower" or "method" problem, not a "measurement" issue.

It does seem counterproductive to verify the values of a higher accuracy method with a lower accuracy method.

The Atlas-Copco equipment that we use also keeps track of how many bolts are supposed to be torqued vs. how many were torqued.

 

[Matt's Quality Handle]

The Atlas-Copco equipment that we use also keeps track of how many bolts are supposed to be torqued vs. how many were torqued.

Fair point. I guess another question for OP would be has the audit ever found anything? Were the findings RC'ed?