Introduction
A fabricator contracts to shop weld 64 mm (2.5 in.) thick plates to 203 mm (8 in.) solid pins. The plates are arrayed like spokes around the outside diameter of the pins and are welded with complete penetration joints (without backing) to the pins. All steel is mild steel. A weld procedure for flux cored arc welding is approved based on the results of complete penetration welds made in one plate. Special conditions include preheat and blanketing of weldments during cooling. No post heat is called for. The assemblies are to become part of a dynamically loaded structure.

The fabricator is required to provide quality control, including a quality control manager and whatever NDT personnel will be necessary to test the welding. The owner is to provide quality assurance personnel to oversee production and testing on behalf of the owner.

 

---

What has happened here is that the cause of a serious redundant welding problem has not been properly investigated.

 

---

 

The Problem
On the first day that my NDT technician arrived at the shop, he called me to tell me of indications in the root area of the plate to pin joint. I pointed out that poor preparation at the root will often result in slag or lack of fusion in the root. I asked him to check all roots visually after back gouging to make sure he had 100 percent sound metal before back welding.

Within the following two weeks of testing, the technician established an ultrasonic rejection rate of 60 percent. This was after intensifying visual testing during root preparation. With rejection rates normally under 5 percent on structural projects, the technician became alarmed. In addition, removal of internal reflectors (slag and cracks) in the root often resulted in the propagation of hairline cracks into the heat affected zone at the pin, requiring further repair and testing.

Suspicions arose on the quality control side that there was a base material problem. Testing of the base materials provided no new information, however. The material appeared to be as specified. I began to wonder if there was a procedural problem resulting from a weld procedure based on test results using 25 mm (1 in.) plate. Quality assurance, on the other hand, took the stance that the fabricator and the technician were incompetent and issued a barrage of noncompliance reports.

Upon visiting the shop as the Level III technician, I noted that the plates were being welded while restrained in a jig to prevent distortion. Joints of this type should not be tightly restrained and I wondered if this might be contributing to the problem. Other than testing the base material and monitoring conditions at the root more closely, no actions were taken by the quality control manager or the quality assurance representative to request review or modification of the procedure, including restraint of members during welding.

During the next week, I received a call from my technician saying that previously accepted welds adjacent to repair welds, when randomly retested, were found to contain rejectable indications not previously noted.

I began talking about delayed cracking. I stated my concerns in writing in my weekly inspection reviews. I then made several calls to the engineer and to the owner's representatives to discuss the potential of delayed cracking. The negative response I received centered around the mild steel base material and the fact that delayed cracking could not be an issue.

I mentioned possible limitations within the approved welding procedure, but met with deaf ears. Concerns over lack of post heat requirements and weldment restraint were addressed as fabrication issues rather than procedural issues. In the end I was assured that all the problems were due to negligence on the part of the fabricator.

The fabricator rapidly fell behind in production and decided to send out a portion of the assemblies to a certified shop. The out of state shop was manned with new quality control and quality assurance personnel. A call to the owner of the newly hired shop resulted in photographs being sent to me by overnight mail. The photographs revealed linear indications in the pins' heat affected zone, discovered during the removal of discontinuities in the weld root area.

The out of state shop suspended fabrication and the assemblies (some of which were successfully welded) were returned to the local shop.

During the time that elapsed, my technician was removed from the project for infractions that included failure to properly calibrate and failure to identify a discontinuity. This is the same technician who identified hundreds of discontinuities and established the initial rejection rate of 60 percent.

Assemblies approved by quality control and quality assurance and shipped to the site were retested in the field and discontinuities were found in previously accepted welds. All assemblies were returned for retesting in the shop.

After weeks of additional repair and testing, the assemblies were delivered to the job site and the issue was brought to closure without additional testing of shop welds in the field.

 

Summary
What are the dynamics of what has happened here?

Ideally, the fabricator could have recognized the welding problem as a redundant problem and worked to establish a new welding procedure, reviewed and approved by the owner and its representatives. Quality assurance could have highlighted the welding problems in written documents to the owner's representatives and sought assistance in recommending a resolution acceptable to all parties. All of this would have, of course, taken up valuable time and led to potential project delay claims. Also, the owners generally avoid recommending means and methods, lest they become liable for the end result.

What appears to have happened here is that, for legal and punitive reasons, sides were drawn which intentionally alienated quality assurance and control, the owner and the fabricator, preventing a spirit of cooperation between the parties in bringing about a satisfactory resolution in the best interest of the end user — in this case, drivers on public highways. What's more, the owner has selected to receive and accept a product in which welds were repaired up to five times before passing ultrasonic and visual testing, satisfied that selected individuals have been admonished, delay claims have been settled and reports of final acceptance have been filed.

What has happened here is that the cause of a serious redundant welding problem has not been properly investigated. Instead, it has been swept under the carpet by testing and retesting until the revolving door is jammed shut with a report of final acceptance.

All of this has caused me to reflect on the ethics standards by which I have gaged myself over the years, and on my first copy of Materials and Processes for NDT Technology purchased from ASNT in 1981. In that book it states, "Decisions to accept or reject following a test result must be based on a thorough knowledge of materials and the properties, processes and their effect on properties, service conditions, and suitable life expectancy." It goes on, "Clearly this much knowledge is seldom located in a single individual, and group decisions or consultations may be necessary. NDT correlation may require the cooperation between test supervisors, designers, metallurgists, manufacturing personnel, customer personnel, and test personnel."

 

REFERENCES
American Society for Nondestructive Testing, Materials and Processes for NDT Technology, Columbus, OH, ASNT, 1981.

 

* Structural Inspection Service; 1617 Acacia Avenue, Torrance, CA 90501; (323) 565-3721; e-mail <csudh@email.msn.com>.

 

Copyright © 2001 by the American Society for Nondestructive Testing, Inc. All rights reserved.

[ Materials Evaluation ]