'Cracks' in welds

Are cracks in welds ever acceptable? Is the term 'crack' a misnomer for linear indications, hot tears, lack of fusion, and other weld characteristics, and are these characteristics acceptable?

A recent issue of the "Welding Wire" e-newsletter featured an item that originally appeared in the San Francisco Chronicle. The article, "Questions over welds delay Bay Bridge project," described how inspectors hired by the California Department of Transportation to monitor the fabrication of steel girders being used in the project reported finding cracked welds. The newsletter then posed the questions: Are a few minor cracks in welds for bridge components OK? Can inspectors be too strict?

Welders responded quickly. Here's what they had to say.

Most of the responses in the avalanche that hit my e-mail inbox the day the newsletter went out echoed my own gut reaction when I first read the article: In no way, shape, or form are cracks in welds acceptable in a structure that supports human beings. I wouldn't be caught dead (a real possibility in my mind) driving over a bridge that has known cracks in welds. Who would?

Rich B. wrote, No, inspectors can never be too strict or too thorough when it comes to public safety and longevity of the product. A little crack here, one there, and down comes humpty dumpty."

A welder who is a licensed pilot wrote, "NO! It is not OK! 'Just a few cracks in the wing folks, but I think we can make it!' Famous last words."

A reader who makes custom knives wrote, "Cracked welds OK? Are you crazy? Sounds like we are letting the ridiculous notion and slogan 'No student left behind' infiltrate the work force."

A welding engineer wrote, "In response to the question 'can inspectors be too strict?"—every time you get on a plane, train, boat, or elevator, or work in a high-rise building, say a little prayer that the inspection personnel employed on the fabrication/construction did their job right.

"Inspection personnel are often regarded the same way a police officer is when you are caught speeding. Fabricaters when employed on a contract are out to make money, and any delay/rework is costly.

"If the inspector has done his/her job correctly, everybody that uses the end product benefits. Inspectors have to have a good working knowledge in the methods of fabrication/construction they are employed to inspect and must be able to interpret the relevant codes. When in doubt ask for a second opinion within your inspection company.

"Any cracks in welded joints are undesirable and in most cases can contribute to dangerous/catastrophic results, which are too often measured in human lives. Everybody involved on the job should be qualified and have a good working knowledge of the requirements. Hurry up and that will do does not cut it anymore."

Just when I was thinking that my gut reaction was totally spot on and no one would argue that 'cracks' are OK, an e-mail came in from a welding professional—let's call him Jim—who took me to task for broadcasting the Bay Bridge Project article. Among his comments were: "The newspaper article was more incendiary than technical (using cracks as a term rather than the proper technical term), and I think you have taken the same incendiary tone with your 'Welding Wire' piece."

Realizing that this individual knew far more than I could ever learn about welding, I began an e-mail exchange with him, which taught me things in addition to how easy it is to learn something when you respond cordially to criticism. In the exchange, Jim complemented the inspectors' performance—"they were handed a spec that was excessively difficult and needlessly strict, and the inspectors did what they were told."

Jim went on to say that "the industry accepts cracks every day. NDT [nondestructive testing] misses them, they are too small to fail the acceptance criteria, and fracture mechanics can be used to justify acceptance. It's already in ASME BPV Code. It's in BS 7910. It's called fitness for purpose, and it is a necessary engineering-based approach to quality criteria, which needs to replace the ancient workmanship-based criteria our industry has suffered under for years, and has wasted countless hours of production in making needless fixes to welds that would perform just fine, but didn't look good, or could have been done better.

"There's an enormous body of work in this area. EPRI NP-5380 did a great job on visual weld inspection of structural welds (not vessel welds) in the nuclear industry back in the late 80's, and is still used today. In Europe, FITNET and other systems are in place to make more efficient use of time and materials, as well as assess aging infrastructure. Our codes are stuck in the dark ages when it comes to quality requirements, both visual and for NDT."

Later, I received a response from a senior welding engineer who wrote, "There has never been a perfect weld made. When making a weld, the best we can do is try to match the chemistry, physical properties, and grain size of the base material. There will always be some discontinuity or difference in the weld relative to the parent material.

"Applying this information to the specific application is the tough part. This is where testing and quantifying specific defects come into play. Once this process is completed then guidelines based on the tests are used to inspect the weld.

"The word 'crack' should be outlawed in the welding world. Linear indications, hot tears, lack of fusion are all called cracks by the generalist. The word crack conjures up the idea of a catastrophic failure or something that is going to continue to grow during service until something bad happens. Once quantified and tested, we find that many of these defects do not negatively affect the product because of built-in design factors or location of the defect in the weld or structure.

"If Lehigh University has quantified the allowable defect size and type [as reported in the SFC article] and the inspectors have inspected to this standard, then the bridge will still be standing long after we are both gone. Don't worry, be happy."

The bridge also may be standing long after the San Francisco Chronicle.

Quite honestly, while the comments from Jim and the welding engineer are very reassuring, I'm afraid I still might approach bridges with a little trepidation. Although not linked to welding defects, the Minneapolis I-35W bridge collapse made quite an impression.