The conventional single-setup ToFD arrangement places the intersection of the two beam axes at two-thirds of the wall thickness. Substituting d = 2T/3 into PCS = 2*d*tan(theta) gives PCS = (4/3)*T*tan(theta).

The two-thirds rule is a deliberate compromise. Diffracted energy falls away from the beam axis, and the geometry already favours shallow flaws because their paths are shorter and less attenuated. Aiming the crossing point below mid-wall pushes sensitivity down toward the lower third, where the sound path is longest, while still leaving enough energy in the upper wall for the lateral wave and near-surface region to remain usable.

The rule applies to a single setup covering the full wall. Above roughly 50 mm of thickness the examination is split into depth zones with separate probe pairs, and the two-thirds rule is then applied within each zone rather than to the whole thickness.

The lateral wave and backwall times that follow from the resulting geometry are the numbers to check on the instrument. If they do not appear where predicted, the PCS, the thickness or the velocity is wrong, and no depth taken from that setup can be trusted.

Compression wave

Worked example

Wall thickness	30 mm
Refracted beam angle	60 deg
Compression velocity	5900 m/s
Beam crossing depth	20 mm
Half separation	34.64 mm
Probe centre separation	69.28 mm
Lateral wave time in material	11.74 µs
Backwall time in material	15.53 µs
Lateral wave to backwall window	3.79 µs

d_cross = 2 x 30 / 3 = 20 mm. tan(60 deg) = 1.7320508, so S = 20 x 1.7320508 = 34.6410 mm and PCS = 69.2820 mm (69.28 to 2 dp). c = 5.9 mm/us. t_LW = 69.2820 / 5.9 = 11.7427 us, so 11.74 us. sqrt(S^2 + T^2) = sqrt(1200 + 900) = sqrt(2100) = 45.8258 mm, so t_BW = 2 x 45.8258 / 5.9 = 15.5342 us, that is 15.53 us. Window = 15.5342 - 11.7427 = 3.7914 us, so 3.79 us.