Weld failure in HAZ

By Shane Feder (****)

Date 08-10-2011 08:44

Dave,
Not enough information.
Thickness and material are the two main requirements needed.
Regards,
Shane

By 803056

Date 08-10-2011 15:05

It is amazing how many times I've encountered situations where the wrong diameter bending mandrel was used.

Shane is correct in his comment that it is essential to know the thickness of the test specimen and the material being bent.

ASME, AWS, and B2.1 used for aerospace and military work tie the bend radius (or diameter if you like it better) to the ductility of the base metal or the filler metal used for the test.

It isn't a case where one bending mandrel is suitable for all materials and all thicknesses. Every welding standard will provide different bend radii for different base metals and thicknesses. In the case of ASME and AWS B2.1, there is an equation that can be used to determine the correct bend radius.

I agree with Shane's position that not enough information is provided, but if the weld is sound and the failure occurs in the HAZ there is a good chance the wrong bending apparatus was used.

Best regards - Al

By MBSims

Date 08-10-2011 17:21

A failure in the base metal HAZ is unusual for side bends on carbon steel pipe. If the failure was in the base metal, and not along the fusion line, then I would suspect a problem with the bend radius or base metal problem. Being a Common Arc test and knowing the attention they give to both test coupon materials and bend testing fixtures, the likelihood of either is very low. I would want to see the bend samples to confirm the failure location.

By 99205

Date 08-10-2011 17:37

A picture of the failed samples, would tell if the radius was incorrect or there was a problem with fusion.

By DaveBoyer (*****)

Date 08-10-2011 22:37

Thanks to everyone who posted. I was not there, and I don't think He got to keep the straps, and I don't think He has any pictures of them.

He said something about the pipe being 2 3/4XX, I guess the XX relates to wall thickness, but I don't know for sure.

I am aware of the basic concept of material thickness/bend radius/ductility relationship, but I don't know the standards used in testing.

My friend is in the early stages of apprenticeship with the Boilermakes in eastern New Jersey. He has been to the union school in western Pensylvania and welded and bent a lot of straps there, He never saw any fail in the HAZ. He has passed a test with a contractor.

One would expect that when taking a test at a recognised facility with 15-20 others testing, CWIs giving the test and perhaps 20 contractors representatives there to take part in the process, that everything would be right, proper test material, proper mandrel in the jig etc. He was suprised when given rust pitted material to test on.

For what it is worth 1/4 to 1/3 of the guys passed, including His brother in law, and some of the others that failed had failures in the HAZ.

I too, would like to have seen the straps to be sure what I think He was describing to Me on the phone is actually what they looked like.

By 803056

Date 08-10-2011 22:49

The information I have at hand indicates transverse face and root bends would be in order for any 2 1/2 inch pipe with a wall thickness less than schedule 160. I don't have any information handy for 2 3/4 inch pipe. The ductility would be better than 20% if the pipe was a P1 base metals. There should be little issue with passing the bend test if the proper bend radius was used and the pipe was prepared properly, i.e. the face reinforcement and root reinforcement removed flush with the base metal. The HAZ should not be adversely affected by welding with 3/32 inch diameter electrode even if there was no maximum interpass temperature listed. I would still question whether the proper bend radius was used given the weld was sound, i.e., it didn't fail through the weld.

Assuming the bend radius was appropriate for a P1 base metal, the next question I would ask is, "What was the base metal?"

Then I would ask, "Was the pipe properly identified?" since it was stated it was corroded, marking could have been easily lost. There is an outside chance it was not a P1 base metal, but something more in line with a P4 or P5 such as a SA 213 or SA335 P21 or P22 which could fail in the HAZ if proper welding procedures were not followed.

Good luck finding the real answer. Even a photograph may be of little help in solving this question.

Best regards - Al

By DaveBoyer (*****)

Date 08-11-2011 03:16 Edited 08-11-2011 03:18

Your comments are pretty much in line with what I told him could cause the problem.

If the pipe was mild steel, the rust pits should not be great enough stress risers to initiate a fracture.

He questioned the higher speed that the air powered bender operates at compaired to the hydraulic one at the school, I don't think this is an issue.

I had suggested that it may have been bent with too small a mandrel for the material and thicknes or that if the wrong material was used, pre heat, interpas temperature requirements and posibly post heat might have been required for proper welding of that [wrong] material.

I don't think He has any recourse to challenge the test process, He is out the money, time and agravation.

The exact cause may never be known.

If there is a process where the test can be reviewed, let me know and I will send it along to Him.

He said the contractors representatives seemed to think that something [not His fault] was wrong due to the location of the failure, but the CWI busted Him out [and others] because the straps failed, failure location not making any difference.

As of this point, He plans to test for jobs as they become available.

By MBSims

Date 08-11-2011 03:24

For the Common Arc testing, each contractor representative inspects the test and has the ability to accept or reject at any point. A rejection by one does not mean all the others will reject. Any of the representatives has the authority to request a retest if they felt there was a problem with the test coupons or bender that was not attributable to the welder.

By 803056

Date 08-11-2011 03:44

The CWI was correct in failing the tests if they broke within the HAZ or the weld.

That being said, if there were several cases where the guided bend samples failed in the HAZ there should have been ample reason to question the material or the bend radius.

There are those CWI's that pass their examinations with flying colors and there are those that squeak by with the minimum test score. There are CWI's that are on top of their game and there are those that never crack a book once they receive their stamp. Who knows what the circumstances were given the nature of the information provided. I wasn't there when the testing was done, so it isn't right to condemn the CWI's actions based on the limited information we have.

In general, I return the guided bend samples to the client, in this case the welder if he paid for the test. There is nothing that says the samples have to be returned to the paying customer, I simply find it good practice.

Best regards - Al

By DaveBoyer (*****)

Date 08-11-2011 04:02

I wish I knew more of what happened there. I am hearing this from a rather unhappy friend who is still fairly new to this game, and probably didn't try to get His straps before He left.

I was not condemning the inspector, because I don't know the inspection criteria, But My friend was not happy with Him/them.

I understand Your point that not all CWIs are equally knolegable, and that some may be more inclined to get to the bottom of a problem if they thought there was one.

The cost of this one is on Him, and with all that was involved, He isn't interested in doing it again right away.

The conversation ended with My suggestion that He put this test behind Him, and on the next test, concentrate on the job at hand and not fret over this one.

By DaveBoyer (*****)

Date 08-11-2011 03:45

I guess they all busted Him out then. I know little about Common Arc, or how it works.

By 803056

Date 08-11-2011 14:57 Edited 08-12-2011 03:08

Most welders are upset when they fail a test. I can understand why he is upset in this case, but there are certain things a welder needs to know when taking a performance test. As the old adage says, "Buyer beware." is as true now as it always was.

It doesn't matter where the welder takes his performance test or who administers the test. As a minimum, the welder should be provided with a copy of the applicable WPS. The welder should also be informed of the acceptance criteria so there are no surprises as to what is expected. The welder should also be informed of what he can and can't do as far as the use of power tools for cleaning, etc. The test administrator (CWI) is responsible to ensure the proper base metals are provided and the testing machines are appropriate for the testing that is required. In a perfect world everything goes according to plan, but in the real world, that isn't always the case.

It is easy to envision that practice pieces could get mixed in with test pieces if there isn't a mechanism to ensure the material identification is properly maintained. If the testing facility is also used for the purpose of training welders, it isn't too difficult to imagine some pipe nipples composed of P21 or P22 getting mixed up with A53 or A106. They all look the same once they sit around for a while and they are not protected from corrosion. The fact that they would be using corroded samples where the identification is questionable is troubling. Then again, we were not there to verify what the actual situation was.

Not all testing programs are the same and a lot of it has to do with the individuals in charge of the testing program. It doesn't matter if the test is taken at an ATF, with an individual CWI, or a program like Common-Arc, the program is only as "strong" as the individual administering it.

It is easy to sit back and play the role of the Monday Quarterback, but unless we were there to see how the program was administered, we don't know what actually transpired. The story sometimes gets twisted by the time we hear the facts secondhand.

Best regards - Al