I am reading through some UT procedures that are based on the EPRI scan. It is for performing angle beam inspection on power piping. An ASME basic block is used to set the calibration. However, only 3 points (1/4T hole, 1/2T hole, and 3/4T hole) are used to set the TCG. This scan is includes the second leg. Am I reading something wrong, or can a full V-Path scan be performed if the only points set in the TCG are in the first leg?
You can include many points in your DAC curve (some equipments lets you use 10 points or more), just place the transducer in the appropriate position to achieve reflections related to the sound path of 5/4 T for example, as you can see in the attached image. You can make the side drill holes in positions that let you easily find the 3/2T and 7/4T reflections and so on
05:10 Jan-17-2010 Nigel Armstrong Engineering, - Specialist services United Kingdom, Joined Oct 2000 1094
Re: Why only a 3-point calibration? 3 vertical in-line SDH's is most convenient for T/4, T/2 and 3T/4 on half-skip - 5T/4, 3T/2 and 7T/4 on full-skip. You could manage with holes at T/4 and T/2 by flipping the block.
16:27 Jan-17-2010 Ed Ginzel R & D, - Materials Research Institute, Canada, Joined Nov 1998 1199
Re: Why only a 3-point calibration?In Reply to Nigel Armstrong at 05:10 Jan-17-2010 .
David...Perhaps we are missing some information in this posting. You indicated the concern arises from an EPRI "scan". By this I guess you mean there is a document (technique) that describes how to go about a particular setup for a specific weld. Power Piping examined to ASME is generally setup up on the piping calibration block with ID and OD notches so the 3 holes you refer to deviate from the rules (although this is OK if demonstrated that the SDHs are more sensitive). ASME itself is lacking in its description of how to set up a DAC or TCG but we must assume it means over the range of inspection (EN 583-2 is more specific as it states "The time base is first calibrated to accommodate the maximum sound path length to be used. If the technique for the weld in question does include the "second leg" this is from the half-skip to the full skip region.
As noted, the description you give is from an EPRI document that clearly deviates from ASME Code. If you were following ASME you would use the three points made by the notches, i.e. 0.5, 1 and 1.5 skip, which easily addresses the working range. Otherwise, for 3 SDHs you should be using the Appendix B instructions (where you see the 3 vertical holes Nigel referenced). Our normal practice was to make a DAC that included the 9/8 node point since we had to extend the DAC past the full skip point where we could expect toe cracking. A 3 point DAC is therefore a MINIMUM when using the SDHs and the 3 points will need to cover the maximum range inspected.
A further deviation from ASME should be noted here...ASME allows TCG for non-piping but it oddlly does not allow it for piping welds. Specific wording is made in T-463.1.3 to allow TCG...T-464.1.2 in piping weld calibrations by contrast makes no reference to TCG and only a DAC is permitted. Perhaps you could ask the Level 3 in EPRI how this document was intended to be used to meet the intent of ASME rules if it contains the deviations you noted.
Re: Why only a 3-point calibration?In Reply to David Bunch at 00:32 Jan-17-2010 (Opening).
The block used for calibration is a modified ASME basic block, with 1/16" side drilled holes. Here is what the procedure says for calibration:
Sensitivity Adjustment: A DAC shall be constructed by obtaining an 80% from the 3/4T hole of the pertinent Calibration Block and then recording the amplitudes from the 1/4T and 1/2T holes. If warranted, the recording of the 1-1/4T shall be done. If your machine is equipped, you can use a TCG and set all reflections at 80%.
Would this be code compliant to inspect in the second and even a bit of the third leg?
This sounds better...a 3 point DAC is not actually stipulated here. It indicates the depths of the 3 SDHs normally seen in ASME blocks. You can use these to constract a DAC for the distance pertinent to the test. The 1-1/4T corresponds to what we used to call the 9/8th node (the full skip was divided into 8 "nodes" the 1/4T SDH was then seen at the 1/8, 7/8 and 9/8 nodes). Since there is no DAC target at the full skip the next one is the 1-1/4T (9/8th node) which would be necessary for proper coverage where the calibration block was of the same thickness as the test piece.
EPRI work usually requires that a technique/procedure was demonstarted to be effective on a component and therefore conforms to the Article 1 T-150 requirements. Use of different targets and modifying the techniques to use a TCG are therefore within the Code provided it is suitably documented.
Thanks for your customarily detailed and helpful references to code. I would argue with you on one point: you said that ASME does not allow TCG for piping welds [para T-464.1.2]. I agree that the piping code does not specifically allow TCG, but I would argue that it does not specifically disallow it, either. The paragraph simply requires the technician to obtain the responses necessary to generate a three-point distance-amplitude correction (DAC) curve. But it does not go on to specify whether the data is to be displayed with a distance amplitude response curve or in TCG format, both of which are types of DAC. Referencing sections such as ASME Sec I, Sec VIII, B31.1, and B31.3 all set reject criteria based partly on the reference level, but, again, they do not state if the reference level needs to be displayed as a curve or in TCG format. All that matters is that the data is compared to reference.
20:06 Jan-20-2010 Ed Ginzel R & D, - Materials Research Institute, Canada, Joined Nov 1998 1199
Re: Why only a 3-point calibration?In Reply to David Mackintosh at 19:35 Jan-20-2010 .
I agree with you that the "intent" is present. This was just my way of pointing out that ASME requies some tidying up. If it is explicitly stated as allowed in non-piping it SHOULD be treated the same way for piping. In fact, in non-piping they go out of the way to include reference levels down to 40% which is a good option for computerised systems.