where expertise comes together - since 1996

Web's Largest Portal of Nondestructive Testing (NDT)
Open Access Database (Conference Proceedings, Articles, News), Exhibition, Forum, Network

All Forum Boards
Technical Discussions >
Mapping out ID cracks in plate in pipe
Career Discussions
Job Offers
Job Seeks
Classified Ads
About NDT.net
Articles & News

02:56 Nov-08-2009
David Bunch
Mapping out ID cracks in plate in pipe

When mapping out ID cracks with shearwave ultrasound, does the standard 6dB drop method work? I just read an article that said the main reflector of an ID crack is actually the base of the crack. Well if that is the main reflector, would moving the transducer forward and backward until a 6dB drop was obtained in each direction give the actual throughwall depth of the crack? Or is a more advanced method like crack tip diffraction needed to measure the throughwall depth of the crack?

16:36 Nov-08-2009

Ed Ginzel

R & D, -
Materials Research Institute,
Joined Nov 1998
Re: Mapping out ID cracks in plate in pipe In Reply to David Bunch at 02:56 Nov-08-2009 (Opening).

zoom image[zoom]

Probably even before the PISC trials, the dB drop techniques have been known to be relatively inaccurate. I ran a CIVA model to illustrate the two options. Using a 12.5mm diameter probe with a 45° transverse wave beam I modelled a scan over 3 notches in a 50mm thick plate (middle image). The notches are 1x10, 3x10 and 5x10mm. Amplitude is normalised to the largest response (the 5x10mm notch). The upper image shows the echo dynamic of the 3 notches as you would scan them for the 6dB sizing. Although the amplitudes are different, the probe offsets are essentially identical for the 6dB drops so all would be sized the same vertical extent. In the lower image I magnified the B-scan and added 6dB software-gain so you can see the tip echoes. The 1mm notch ring-time occludes the tip signal but for the 3mm and 5mm notches you can see the precursor (and post-cursor) tip signals separate from the peak corner signal. The notches in the model are separated by 25mm so easier to identify..in a pipe with many closely spaced cracks it may be a bit more difficult to associate the tips and corned signals.
18:34 Nov-08-2009
David Bunch
Re: Mapping out ID cracks in plate in pipe In Reply to Ed Ginzel at 16:36 Nov-08-2009 .

Wow, thanks for such a detailed reply. I agree, the 6 dB method is pretty inaccurate for small flaws. I read that any flaw smaller than the beam spread will be inaccurately measured.

So as far as ID or OD cracks go, the best method is to use a 5mhz 45 degree transducer and use the main beam to flood the base of the crack with sound, and then measure the precursor signal?


© NDT.net - The Web's Largest Portal of Nondestructive Testing (NDT) ISSN 1435-4934

Open Access Database, |Conference Proceedings| |Articles| |News| |Exhibition| |Forum| |Professional Network|