Immersion water path

In immersion testing the sound crosses a water column before it reaches the part. Water carries compression waves at about 1480 m/s at 20 °C, roughly four times slower than steel, so 1 mm of water path takes as long as about 4 mm of steel — the exact ratio is c_material / c_water, 3.99 for compression in ferritic steel. That single ratio governs every water-path decision.

The water column produces its own multiple reflections between the probe face and the part surface, and they march across the time base at intervals of one interface round trip. If the water path is too short, the second interface echo lands inside the part's thickness window and is easily mistaken for a flaw. The condition for keeping it clear is water path > thickness × c_water / c_material — for steel, a water path of at least a quarter of the wall thickness, and in practice comfortably more.

Focusing works the other way round. A focused immersion probe is specified by its focal length in water. When the converging beam crosses into the faster material, refraction at the surface steepens the convergence, so every millimetre of material depth uses up about four millimetres of water focal length — the same ratio c_material / c_water again. To place the focus at a depth d inside the part the water path must be F_water − d · c_material / c_water. Get this wrong and the focus sits in the water or well past the far wall.

Water velocity is temperature sensitive — it rises by roughly 3 m/s per °C around room temperature — so a tank that warms up during a shift will shift the interface echo position and the focal depth. Degas or at least settle the water: bubbles on the part surface are the most common cause of spurious immersion indications.

Ratio = c_material / c_water
Water path equivalent in material = water path × ratio
Minimum water path = thickness ÷ ratio
Water path to focus at depth d = F_water − d × ratio
Water round-trip time [µs] = 2000 × water path [mm] / c_water