ASME 31.3 Butt Weld Radiographic Interp on LoF and IP

[tonykenna]

Good morning!! First post

31.3 states on table 341.3.2 in regards to acceptance criteria for Girth, Groove and brance connection welds, Lack of Fusion is listed as Zero for Extent of imperfection, which is fine. Incomplete Penetration however, it is stated that there is a acceptible limit of the lesser between ≤1mm ann ≤0.2Tw

My issue is with how these 2 discontinuities are being called by RT techs. We have been getting several RT rejections based on a call of Lack of Fusion when these issues are obvious incomplete penetration (Missed Edge) on one side of the root face only due to one side of the root face (Land) being thicker than the other. Incomplete penetration does not need to be limited to one side of the root face for it to be called Incomplete Penetration imo.

It makes absolutly no sense to me why there is an allowable limit to incomplete Penetration on both sides of the root and zero allowable on one side in the case of lack of fusion on one side of a root face due to excessive thickness of one land preperation. I can understand the severity of the limits for lack of sidewall fusion, but that must be different than missing an edge on a root face.

Thougts?

 

[Duwe6]

Welcome to the world of RT techs trying to evaluate welds. E-mail their L/III - Level III - and demand that the film be read per Code and "good industry practice". As you have observed, IP is not LF, and IP is a VERY easy 'call' on RT film. Part of this industry-wide problem is that when the techs are taught film interpretation, they get taught to ASME. Per ASME, there are 'linear indications' and 'rounded indications'. In almost all of the ASME construction codes, all Linears are rejectable. In B31.3, you can have 1.5" of IP in any 6-inches of weld. So in B31.3 work, the tech HAS to be taught the differences [or he is a rather expensive detriment to your weld shop].

LF is close to imposssible to see on RT film shot in the field using a Gamma source, unless there is something [slag] trapped in that area. It is slightly easier to see on film actually X-rayed [in the NDT lab's shop] The grains of silver on the RT film are larger than the gap between the weld and the basemetal in "cold-wall" LF on a groove weld. Plus, all an RT image is, is shadows on a sheet of film, and they can be misleading. If you RT'd a phonebook, the RT would look like the world's finest scaffold plank -- a slab of dense wood with no knots or other flaws. But that RT will never, cannot, disclose the hundreds of LF's between the pages.

If you are getting erroneous LF calls, demand that the L/III come out to the weld and reread the film. If he upholds the LF 'call' and you still think the weld is sound, air-arc out the 'defect'. Have the L/III mark on the weld where the LF is. Give him a welding hood, and have him watch the 'unwelding' of the purported flaw by gently air-arcing out the 'flaw', layer by small layer. One of you will be wrong.

If the RT firm is wrong, either fire them and hope for better techs from the next one, OR pay for UT - shearwave ultrasound - of any purported LF 'called' by an RT tech. If UT confirms the LF, you pay for the UT. If the UT reveals that it was just an anamolous shadow on the film, the RT firm pays the UT tech, and you get to yell at the tech reading your film.

Over the years, I've found that most purported sidewall LF is a figment of the RT tech's imagination. Like all shadows, if you stare at RT film long and hard enough, you can see about anything. It's just human nature. Film should be read expeditiously, prefferably twice, and quit. If defects are there, they will have been seen.