Not logged inAmerican Welding Society Forum
Forum AWS Website Help Search Login
Up Topic American Welding Society Services / Technical Standards & Publications / AWS D1.1: 4.9.5. Retest
- - By Wouter De Neve Date 03-18-2015 12:34
Section 4.9.5 'Retest' of AWS D1.1 states: "If any one specimen of all those fails to meet the test requirements, two retest for that particular type of test specimen may be performed with specimens cut from the same WPS qualification material. The results of both test specimens shall meet the test requirements."

Does this mean that only 1 retest is allowed?
Or can a failed retest be retested again (double of course)?

I'm currently in this situation with a welding procedure.
Position is 6GR and I had a double in the root bends.
After the second retest, the root bends were succesfull (I even had 4 specimens retested instead of 2)

The failure was due to a welder performance defect over a portion of the whole weld length.
I am sure of this since I have about 150 PQR's in our database with these welding consumables, on the same base material and with the same parameters.
Furthermore, A similar PQR was welded at the same time by the same welder.
The only difference: 6G.
The base material was of the same heat number as the smaller wall thickness pipe of the 6GR test.
This PQR passed immediately

I know that in this case, rewelding the procedure brings no added value.

I'm curious to hear other opinions.
Parent - - By Lawrence (*****) Date 03-18-2015 17:47
Rewelding the procedure may not bring value... But it will bring compliance.

The root cause analysis you went through shows that the welder caused the defect on the entire joint.

You mention similar... But the difference between 6G and 6GR is a greater challenge to the welder (as you know) and that is the part that failed.

A pesky auditor (I'm sure we have both met some of those) might make you very miserable, seeing your planned actions and their justifications.

There is value in avoiding misery.... Recommend retest  :(

And Welcome to the Forum Wouter !
Parent - - By Wouter De Neve Date 03-18-2015 18:55
Thanks for the welcome!

The defect wasn't even on the entire joint.
There were even 3 assemblies welded to have enough test material.
Only one showed this failure (partly).
And the succefull bend tests came from another.

Furthermore, this welder has a 6GR certificate (one of our standard qualifications, since we weld a lot of jackets)
Plus, he already welded several 6GR PQR's in the past.
We only made new PQR's because the client ordered us to.

Therefore, I'm still stuck with the question: "Why reweld something in the same manner, which gives you no more, nor less information since it has not failed due to the essential variables?"
Parent - By welderbrent (*****) Date 03-18-2015 19:41
Wouter,

WELCOME TO THE AWS WELDING FORUM!!

Since you didn't tell us, could it be the original PQR's were over a couple of years old?  Customer may just have wanted to make sure someone on staff could still do it, or that a change in shop equipment did not affect your ability to do it, or even if the electrode manufacturer was the same a change of lot may have been a consideration.

Bottom line: we don't have all the information and possibly neither do you.  Customers pull the strings and can request you go beyond the code.  So can the engineer.  He who has the gold makes the rules.  You want the job, you do what he wants.

The worst part of this, you can qualify a PQR and/or a person one day to the exact process and all parameters and the next day a different customer wants the same test taken fresh.  No, other one won't do, take it again.  Maybe their rep watches, maybe not.  No matter. 

Now, as to your exact situation, easy to fail a 6GR once in a while.  Depending upon my project I could see making the contractor retest to prove he could.  Not just lucky the first time.  They may have had a bad previous experience with a different contractor or heard a rumor about you and wanted proof.  But, once a test has been requested, it must meet all applicable criteria totally on it's own merit with no consideration of previous tests and results. 

Retests during a test to qualify a PQR are different for a PQR than for a WPQR, I'm sure you know that, just sayin, 4.9 vs 4.33. 
'Retests', if that is what one wants to call it because 'we have already done that and have the results on file', at the request of a customer aren't really 'retests'.  They are contract conditions. 

He Is In Control, Have a Great Day,  Brent
Up Topic American Welding Society Services / Technical Standards & Publications / AWS D1.1: 4.9.5. Retest

Powered by mwForum 2.29.2 © 1999-2013 Markus Wichitill