For ultrasonic testing of piping welds, ASME Sec V Art. 4 allows sensitivity calibration on notches using the standard block for piping (Fig. T-434.3-1) or side-drilled holes using the alternate block (Fig. T-434.3-2). They are both acceptable reference targets but behave quite differently due to a couple of super fun reasons.
For a comparison of echo-dynamics, simulation was performed with CIVA on a 20.5 mm curved block using both notch and side-drilled hole reflectors. A 16-element, 5MHz transducer was modeled (you know which one, check out the fancy render in the title block). No sensitivity normalization was performed... what you see is what you get. The gentle descent of the SDH signal is due to attenuation with increasing sound path.
Note: multiple reflections of the hole were made, and intermediate data points interpolated to produce the same sound path as the notch (because you can't drill a hole at the ID, right?).
Well would you look at that? Just look at that! Notches are really loud, aren’t they? And wobbly, check out that dip at 60 degrees!
For one, notches are relatively massive reflectors compared to side drilled holes, with sound reflecting primarily via the corner trap. Imagine playing pool and missing a corner pocket shot. You can miss by a little, or by a mile (1.61 kilometers, go metric system!), and the ball will still come back at you.
Side drilled holes are basically cylinders with sound reflecting off only the narrow strip perpendicular to the sound beam. Imagine the line of reflected light you’d see when shining a flashlight on a chromed cylinder rod.
Notches are asymmetric reflectors. That’s a fancy way of saying the profile and reflection characteristics continuously change with angle of incidence. At a refracted shear wave angle of 45°, the signal is massive as the sound reflects efficiently from the corner trap. With the beam nearing 60°, mode conversion occurs which reduces the reflected shear waves and produces the dip in the chart above. Beyond 60° it picks back up a little but never fully recovers the oomph it had at lower angles.
Meanwhile, side-drilled holes are symmetric reflectors, producing the same reflection profile at all angles.
So what? What are you trying to get at?
A sensitivity calibration (e.g. DAC/TCG) created with notches will result in generally lower sensitivity than side-drilled holes. The gain must be turned way down in order to set the peak of the notches at 80% FSH. There is a 10 to 12 dB difference in the scenario shown in the chart above.
In addition, a sensitivity calibration made on notches will produce very different amplitudes depending on angle. We already deal with this in UT anyways... the amplitude always changes depending on what angle you hit the flaw with (unless your weld flaw is a perfect cylinder or a sphere, which it probably isn't). But calibrating on notches amplifies this effect by amplifying the signal at around 60°.
I would like to thank Ed Ginzel for his ongoing, seemingly bottomless wealth of expertise and for running all the CIVA analyses.