Here is a result that runs against what most people assume.
A thin-walled pipe is not harder to inspect than a thick one. It is easier, and you can find smaller defects in it.
The reason is a single number you can calculate before you leave the office.
Guided wave testing works in the T(0,1) mode because it is non-dispersive:
its velocity is the same at every frequency, so a pulse holds its shape however far it travels.
But T(0,1) is not the only torsional mode. Above a certain frequency, the T(0,2) mode also appears.
At even higher frequencies, additional modes, such as T(0,3), T(0,4), and so on, emerge.
These modes (T(0,2), T(0,3), T(0,4), etc.) are dispersive.
Its velocity varies with frequency, so it spreads as it goes and arrives as a smeared signal on top of your clean one.
So that frequency is the cut-off where T(0,2) starts to appear, and it is set by the wall thickness.
Cut-off frequency = shear wave velocity divided by twice the wall thickness.
For steel that gives:
- 1.0 inch wall — 64 kHz
- 0.75 inch wall — 85 kHz
- 0.5 inch wall — 128 kHz
- 0.25 inch wall — 254 kHz
Operate below it and you generate only pure T(0,1). Operate above it and you get two or more modes from one pulse.
Now look at what that means in the field.
On a 1-inch wall you are capped around 45 kHz. Wavelength about 2.8 inches, quarter wavelength about 0.7 inch.
On a 0.25-inch wall you can run 180 kHz. Wavelength about 0.7 inch, quarter wavelength about 0.18 inch.
Why does the quarter wavelength matter? Because the reflection from a defect is greatest when its axial length is about a quarter of the wavelength.
So the thin-walled line lets you reach defects roughly four times shorter than the thick-walled one.
There is a second benefit at high frequency, and it matters on plant piping.
Higher frequency interacts far less with the things bolted and welded to the outside of the pipe — clamps, supports, welded attachments.
That is exactly what you need for corrosion under insulation (CUI) inspection, where the line is covered in supports and the defects are small.
So the practical rule is simple. Measure the wall thickness before you choose a frequency.
Not the nominal thickness from the drawing — the measured thickness at the probe position.
The wall thickness decides your frequency ceiling. The frequency decides your wavelength. The wavelength decides the smallest defect you can find.
Skip the measurement and you are either generating a dispersive mode you cannot read, or leaving sensitivity on the table that the pipe would have given you.