Jody at Welding Tips and Tricks recently released a video comparing MIG welds on hot rolled steel. For one he did not remove the mill scale, and for the other he did. The welds look more or less identical. Guess which one failed with barely an effort?A strange experience:
I experienced a weld - from a factory process, with speed controlled and jig held MIG torch, so gas, feeds and speeds, etc. all calibrated and set, and batch recorded.... but of thousands of welds joining structural end fittings to steel tubes, one single end casting had some "contamination" so the weld that looked as perfect as the rest had zero penetration into the casting. Sent to the Welding Institute for a report on "Why? and what went wrong?" they were completely bamboozled. This was not a pressure weld, but a tension member: When the load was applied as the structure was assembled, a ring of unwelded steel casting appeared as the end piece pulled-out of the tube. Micrograph sections showed perfect penetration and welding to the tube and a non-welded friction joint between the weld and the end casting. The report suggested the end casting may have been simply VERY cold, or had some insulating pollution (Oxide?) on the surface that prevented the arc from "hitting" real metal so the feed metal "spray" simply sat on the surface of the end casting. Further metallurgical lab studies had failed to find any contamination, so the Welding Institute report had "no conclusion" for the failure. - Had this been a pressure vessel shell, pipe-joint or whatever, the pressure test would have definitely exposed the faulty weld. The Company had stringent checks in place, so all welded and bolted joints were examined for movement as the structure was built and during loading up to completion. These checks had identified the failure part way through construction of the building. No other faults were found.
Conclusion. A visually externally perfect weld can have a total failure within. But if it "holds the load without failure", then it can be deemed OK. - Except the "Factor of Safety" may be compromised.
For Boilers in service, the equivalent loading "to ensure safety" is the periodic Hydraulic test at an excess well above the Safety valve release pressure. (ASME and UK regs require different pressures, but both have "Excess" pressure applied hydraulically).
I personally feel that there can be many boilers "in service" where joint integrity is "Imperfect", even by design, such that an over-pressure of the boiler could fail a joint at well below the "design expectation". The Safety valve must therefore be suitably respected and checked against the gauge at every steaming to be sure that boilers cannot experience adverse conditions. ALL Boiler operators MUST also be well trained and "compos mentis" when operating boilers to prevent operation outside the correct operating limits.
(compos mentis: = (law) Of sound mind, sane; thus criminally responsible for one's eventual wrongdoing).
K2
My sense is that the "visual test" is far more applicable/useful for TIG and SMAW than for MIG - but I'm no expert by any stretch, so perhaps someone else can correct or confirm ...