Calibration Merry Go Round - Torque Wrenches and Testers

[josephs6801]

I'm on a merry go round with my auditor and I don't know how to get off. We calibrate our torque wrenches and torque testers twice a year. Every great once in a while one of the wrenches fails calibration. We have a very detailed procedure how to remove the wrenches and they may not get back into the calibration until they have been re-calibrated and certified by an outside source. Now here is where I find it complicated, I need to backtrack all repairs made with said failed wrench, but when did it fail? It was found in tolernace, it left in tolerance. During the calibration process it fails, how do I decide which of the repairs are "valid" and which ones are not?
Any advise \ assistance words of wisdom are welcome!

 

[Michael_M]

This is also a requirement of AS9100. I will use thread gages as my example (since this is what we have to do this with).

We calibrate gages once a year. I have to track what part number the gages are used on (or job number could work as well). If the gage comes back as 'failed' or 'adjusted' I have to back track the part numbers that were run and verify the product is still good since we do not know when the gages lost calibration. I have a form that gets filled out and if we potentially shipped parts, we have to notify our customer. This is a pain but a requirement so..........

If this becomes a huge burden on you, could you do a 'calibrate prior to use' against a good standard instead of doing it once a year? Or calibrate prior to use and send out once a year for full calibration? I am just tossing out an example: You have a device that the torque wrench fits into and you perform the torque and compare it to what the device says the torque actually is. This device would need to be calibrated of coarse and if it fails then you have a potential issue but it's an idea.

 

[John Broomfield]

I'm on a merry go round with my auditor and I don't know how to get off. We calibrate our torque wrenches and torque testers twice a year. Every great once in a while one of the wrenches fails calibration. We have a very detailed procedure how to remove the wrenches and they may not get back into the calibration until they have been re-calibrated and certified by an outside source. Now here is where I find it complicated, I need to backtrack all repairs made with said failed wrench, but when did it fail? It was found in tolernace, it left in tolerance. During the calibration process it fails, how do I decide which of the repairs are "valid" and which ones are not?
Any advise \ assistance words of wisdom are welcome!

josephs6801,

You don't say what business you are in.

Trucking companies, for example, invest in their own equipment to verify the calibration of their torque wrenches at any time to manage the work at risk.

Your procedure of using a wrench for six months before sending it out for recalibration means that you are not supporting the users with the means to check a wrench for any reason.

...and over or under stretched bolts could be dangerous.

If your business is low risk then an option could be to have a back-up wrench that can be used at any time the user suspects a damaged or faulty wrench.

John

 

[PaulJSmith]

You'll have to go back to the last known confirmation that the wrench was in calibration. Everything between that time and the time you found it out of calibration should be considered suspect until verified otherwise.

Not pretty, but that's the best way to ensure you have good product.

 

[josephs6801]

Service Industry. Truck repair to be exact. We have both company owned and employeed owned wrenches. Around 90 at various locations. We do test internally and send the testers to a lab for calibration.
Thanks and have a good day.

 

[John Broomfield]

Service Industry. Truck repair to be exact. We have both company owned and employeed owned wrenches. Around 90 at various locations. We do test internally and send the testers to a lab for calibration.
Thanks and have a good day.

josephs6801,

Then, perhaps, you need one or more of these:

http://www.mountztorque.com/products/torque-analyzers-sensors

Engaging top management in the risk assessment may be necessary to loosen the purse strings.

John

 

[AndyN]

Service Industry. Truck repair to be exact. We have both company owned and employeed owned wrenches. Around 90 at various locations. We do test internally and send the testers to a lab for calibration.
Thanks and have a good day.

Joe:

Firstly, you'll get a lot of benefit from NOT calling for torque wrenches to be "calibrated". Trust me on this, they are NOT measuring equipment!

Here's what you do. Torque is a process applied to fasteners. It's achieving the lowest required torque value you want to show. The way you do this is to either study the torque wrench types (by style and brand) and then do a study on each type for the process variation. You'll find some are really accurate - Snap on for example - and some (especially indicator type) are rubbish. You really want the breaking beam type. Throw out the cheap ones and buy robust accurate ones. Buy a torque analyzer and do a periodic analysis, plot the results and show your auditor - data speaks for itself. Auditors don't always know what they're talking about...

Added in edit: I've had this done at a major tractor manufacturer where all they do, all day long is bolt stuff together. Torque is their life and this is what they do.

 

[josephs6801]

Great to know. I am new to the truck industry coming from 22 years in manufacturing. I have been here just a year and find everything has two meanings.

 

[dgriffith]

... Now here is where I find it complicated, I need to backtrack all repairs made with said failed wrench, but when did it fail? It was found in tolernace, it left in tolerance. During the calibration process it fails, how do I decide which of the repairs are "valid" and which ones are not?
Any advise \ assistance words of wisdom are welcome!

This is called either traceability: Finding the tool or instrument used on a product when the product fails, or in your case reverse traceability: finding the product when a tool or instrument fails.
Unless the wallet opens to purchase sophisticated torque monitoring in real time, the only way to narrow down the failure impact time frame is to test the wrench more frequently.
The use of an electronic torque tester similar to this http://buy1.snapon.com/catalog/item...roup_ID=739762&store=snapon-store&dir=catalog can be used before each shift when the tool is checked out to an individual. If you know all assemblies that individual worked on during a shift, the scope of your problem is narrowed. It can also serve as evidence the user can properly use the wrench (there are many who think they can, but it can be demonstrated they cannot). If you maintain the link between assemblies, technician, and tool in a small time frame it may help you.

(My best example is a failed torque sensor on our calibrator! Had to find all wrenches that sensor was used to calibrate--with appologies to Andy....--and then find all aircraft where those wrenches were used!
Fortunately, it was a high sensor, so the wrenches and uses were few. We had recently and satisfactorily completed a proficiency test and were able to show that the sensor was good up to a specific date. So all wrenches cal'd after that date were recalled, and those wrenches lead to jobs on 5 aircraft. The affected bolts were retorqued to satisfy the impact investigation.)

 

[josephs6801]

Wow, learning alot here. Problem is we have up to 600 ft-lb and they are at 5 different locations. I believe I'm going to have to bite the bullet and calibrate every quarter. Does anyone know of any labs that do in-house calibrations on torque wrenches?