Search the Elsmar Cove!
**Search ALL of Elsmar.com** with DuckDuckGo including content not in the forum - Search results with No ads.

Torque Wrench Calibration Frequency (Calibration Interval)

#31
Thank you @dwperron !

Yes the shift leader verifies the torque for the preset wrenches with the torque tester. We have multiple wrenches in use and all of them are verified everyday. However like I said, we did not have the tolerance set up. The torque tester is within the calibration and the frequency is yearly.
We have 3-4 assembly operators (users) using these wrenches. Like you said +/-3.2 in-lb looks very difficult for us based on the historical data.
I would be very concerned about your verification results. It would be highly unusual to have multiple tools that are repeatedly found out of tolerance in both high and low directions, especially when they are being set prior to use.

It sure looks to me like you might have a training issue here. If multiple technicians using multiple tools are getting bad results I would suspect they may not be operating them correctly.
 

dgriffith

Quite Involved in Discussions
#32
Thank you @dwperron !

Yes the shift leader verifies the torque for the preset wrenches with the torque tester. We have multiple wrenches in use and all of them are verified everyday. However like I said, we did not have the tolerance set up. The torque tester is within the calibration and the frequency is yearly.
We have 3-4 assembly operators (users) using these wrenches. Like you said +/-3.2 in-lb looks very difficult for us based on the historical data.
Checking the torque at the end of the day is a very good suggestion.

Can we set out own tolerance based on the product requirement?
Haven't heard anything in a few days. How are your torque measurements going?
There is one other factor that should be considered. Torque testers (and torque calibrators--you have a tester, not a calibrator, right?) have different modes of operation. The wrong mode will easily result in erroneous readings. A clicker/snap type wrench will almost always fail if the mode is not "first peak". Likewise, other wrench types require other modes, like tracking, peak hold, etc. that must be selected and observed correctly when used.
 

IATFuser

Starting to get Involved
#33
Hi @dgriffith! Yes! we have a tester and not a calibrator. You provided a very good information with the type of wrenches. We use Seekonk TA-75 torque tester for pre-set torque wrenches that are clicker type. I am not sure about the mode of operation for the tester. But we still are struggling to be between +/-4% of the limit.
 

IATFuser

Starting to get Involved
#34
@dwperron We don't set the torque wrenches prior to use. They are preset and the shift leader verifies the torque on the tester. Multiple operators use the wrenches however only shift leader does the verification activity.
 
#35
@dwperron We don't set the torque wrenches prior to use. They are preset and the shift leader verifies the torque on the tester. Multiple operators use the wrenches however only shift leader does the verification activity.
A couple of comments.
First, the shift leader should NOT be the one checking the tools. The actual operator using the tool should be the one testing it.
That helps to catch any errors in torque caused by bad technique on the part of the operator. Those can get corrected by training.

The Seekonk TA-75 is the wrong tool for this job!
It is only rated , and calibrated, to 75 in-lb. You say your drivers are set to 80 in-lb, which is beyond the range of the tester. You are getting results up to 90 in-lb which indicates you are substantially over-ranging the tester, which will damage it!
The crude "memory needle" is very subject to producing less than accurate results, especially if the operator is not using proper technique with the wrench.
There are far better digital torque testers out there. Yes, they cost more than the Seekonk, but they are much more accurate and are much better at detecting the actual "click" point on the wrench. They will help you get better, more consistent results.
 

IATFuser

Starting to get Involved
#36
@dwperron That are very valid points you just mentioned. Thank you! I am new to the calibration/instruments subject but it makes sense. Also this has been the practice where I started working so I am also trying to understand. I have a few questions-
1. I checked online and Seekonk has TA-75 and then TA-100 torque tester. And we set our wrenches at 80 in-lb. Do you think we should change it over from 75?
2. Online it says it's a bench mounted torque analyzer. We have installed it on a wall. Does that affect the results?
3. Also do we have to measure more than once for verification of the torque on the wrenches?
4. How do we decide the tolerance on this 80 in-lb? Is there a standard for that or we have to conduct experiments for failure?
 
#38
@dwperron - I apologize. We use TA-150 for 80 in-lb. TA-75 we use for 20 & 17.5 in-lb wrenches. Does that make sense?
OK, you are using the correct range testers. That takes care of one issue. Yes, you can mount that tester to the wall.

Now we move on to torque basics.
Before you test a wrench you should exercise it 5 times (according to ISO 6789).
Alignment of the wrench to the tester is important. Force applied at different angles will give different results. You should be sure the wrench handle is level, ±3° (you can see a 3° error), and you apply the force at the center of the grip and be within 10° of being perpendicular to the handle. This helps eliminate cosine errors from being off angle. You also want to use an absolute minimum of adapters, they add errors.
1.JPG

Typically you would run 3 tests on the wrench. Some places do 5. It is a very good practice to have a log book where the readings for each device are recorded every day. That will show you when a tool starts to drift.

I very strongly recommend that your tool users be the ones who test the wrenches every shift. That way you are accounting for any errors they may add through their technique. It makes no sense to be sure you have an accurate tool when the operator introduces more errors.

The wrench should have a tolerance supplied from the manufacturer. If it doesn't, most click wrenches are ±4% accurate.
You can conduct experiments to determine the accuracy of your wrench if you cannot find the manufacturer specifications.
 

dgriffith

Quite Involved in Discussions
#39
Hi @dgriffith! Yes! we have a tester and not a calibrator. You provided a very good information with the type of wrenches. We use Seekonk TA-75 torque tester for pre-set torque wrenches that are clicker type. I am not sure about the mode of operation for the tester. But we still are struggling to be between +/-4% of the limit.
@dwperron - I apologize. We use TA-150 for 80 in-lb. TA-75 we use for 20 & 17.5 in-lb wrenches. Does that make sense?
DW had excellent advice and references.
The one component not yet mentioned is the wrench. What brand, model, and exact style is it?
Your torque tester--I hesitate to call it an analyzer--looks like it is spec'd to ±2% of reading. If your wrench is ±4%, then the accuracy ratio is only 2:1. It should be 4:1 or better; in this case your tester should be ±1% or ±0.5% accurate. This helps insure the tester does not add its own error to the results.
It looks like the dial has a memory follower. If so, ensure to follow the mfr. instructions for initializing the follower--it can lead to small positive errors, but errors nonetheless.
Depending on how serious the company is about torque (are you the only one who cares...?) they may wish to invest in a quality tester. CDI/Snap-on and Mountz are 2 reputable companies. Log the results in Excel, or some testers have software that autologs.

I would be remiss not to mention that this tester is manual. You really need a quality electronic tester/analyzer. I guarantee you are getting high readings if your wrench is a snap-type, even though it is a fixed value. Almost impossible to accurately test a snap-type and read the first peak manually. It is also almost impossible to prevent a second peak when applying force by hand, resulting in a high reading.


EDIT: You would want to look at the model TT-QC Torque Tool Tester - AWS-QC from the same company. It provides first peak, around $1000.
 
Last edited:
Top Bottom