Ok, The first think I would say is that it's not applying a more stringent criteria. Assuming as you state that the exam was done properly, after you removed the cap and re-examined the reading at that time was +5. Going behind it and blasting and therefore changing the surface did not make the +5 go away, it only changed the surface so that you lost more energy in the transfer 7 db by the numbers you've quoted. I am not trying to be a smart ass here, but this would be akin to taking an RT shot where the weld surface was to course to properly read, grinding the surface, and seeing a lack of fusion in the subsequent shot, then shooting it again at a slightly different angle which will half the time make a lack of fusion image disappear due to simple orientation of the energy source. Did the LOF go away? no, did the image go away? yes. Does the lack of an LOF image without repair in the third and final shot make the weld good? no. The ethical thing to do is reject it if the worst case findings are reject-able by the referencing code.
For a UT example, it would be like finding a flaw prior to galvanizing, then checking it again after. before it was out, after it was in. Did the flaw go away? no. did the galv create an attenuation issue? yes.
The best course of action is to make sure there are no surface irregularities or geometries that can interfere with the exam before you begin. If there are geometric concerns, work the math and sketch it out before you begin. Scan per code and procedure requirements, and stick to the results of that scan.
S10.6 Weld Joint Type and Groove Design. The weld
joint type and groove design are important factors affecting
the capabilities of UT for detecting discontinuities.
The following are design factors which can cause problems
and should be considered for their possible affects:
(1) Backings
(2) Bevel angles
(3) Joint member angles of intercept
(4) PJP welds
(5) Tee welds
(6) Tubular members
(7) Weld surface roughness and contour.
Point number seven is where you have run into a problem. While appendix S is informative, it does specifically mention weld surface roughness as a concern in the design when it comes to UT. Bottom line is, you went from a rough surface, to a smooth surface, and back to a rough surface. It tells me the surface your working with is partially masking indications through ultrasonic attenuation. Suggest checking with transfer correction check per figure S4 in appendix S.
Regards,
Gerald