Dave,
I don't have the latest D1.5, but look at 6.34 and 6.36 of D1.1. Phased array techniques can be used, but it's one of those situations that will require Engineer approval, a qualified procedure, etc. This is basically an exception to the code, so there are no acceptance criteria there to be found - that would have to be developed and approved by the Engineer.
Mankenberg
No published D1.5 criteria specialized to AUT/PA. You *could* evaluate the A-scans to the same criteria as conventional UT. You may not have all the A scans you need, though, because if you want to have the data recorded in position with a tracking wheel, then the angle of the probe with respect to the weld is fixed. No "Movement A" or "Pattern E" per D1.5 Figure 6.7. This is what the article means by "the probes do not oscillate". They could do repeat scans with the probes set at different angles to the weld to make up for this, though--but that would take away some of the time savings of automated PA.
Another way is non-automated PA. I guess that's considered a time savings because instead of coming through with two different angle probes you can use the one with a range of angles, but then you don't get the nice auditable data record.
In some industries they use the C-scan like X-ray film and interpret that to get dimensions of discontinuities. I'm not convinced yet.
The study mentioned in the last paragraph of the article is being done in Florida with involvement of a university, a fabricator, and some AWS people. It'll be interesting to see what they come up with.
We have used PA on a DOT project. We made them supplement the PA that they did normal to the weld with conventional UT per D1.5 clause 6.24.2 to detect transverse discontinuities. But we only trusted the PA enough to tell whether the weld was clear or not; we didn't use it to evaluate defects. If there was an indication, they had to go back to conventional UT+RT to establish acceptability.
What's the rationale given on your project for using PA?
Hg