Ultrasonic Thickness Measurement on Site: Five Steps from Probe to Reading

Jan 25, 2026 Thickness Gauges 5 min read

Same gauge, different operator, readings that differ by 0.3 mm. The difference is probe choice, coupling and how the number is read. These five steps are the standard routine.

1. Choose the probe first

A single-crystal longitudinal probe uses one crystal to transmit and receive; it suits flat, smooth surfaces. A dual-crystal probe has two crystals separated by an acoustic barrier, one transmitting and one receiving, so it hugs rough and curved surfaces with less noise.

Thin plate wants high frequency – weaker penetration, better resolution. Thick castings want low frequency – resolution traded for penetration. Small-diameter tube wants a 6 mm small-diameter probe; a flat probe on a curved surface touches only at a point or along a line, and the energy never gets in.

2. Calibrate velocity and zero

Measure velocity and set zero on a reference block of known thickness. Changing material means resetting velocity: using the velocity of steel on aluminium introduces a systematic error that no amount of careful measurement downstream will remove.

3. Couple properly

Couplant exists to displace the air between probe and workpiece. Low viscosity for smooth surfaces; high viscosity for rough, vertical and overhead surfaces; high-temperature couplant on hot work, applied to the probe rather than the part. Thin and even – not a blob.

4. Measure and read

Hold the probe perpendicular, wait for the reading to settle, then record. Single-point mode shows the thickness at that spot. Scan mode moves along the surface and displays the minimum found – for corrosion survey, scan mode finds the thin spot far faster than point-by-point work.

5. Record and interpret

Write down the location, probe type, velocity and couplant along with the number. Only then is the data comparable when the same asset is re-tested next year.

When a reading looks wrong, do not blame the instrument first. Surface rust, back-wall pitting and internal laminations all shift the value. Work through surface, coupling, probe and velocity in that order before suspecting the electronics.

Summary

Most thickness errors come from a wrong probe, an unset velocity or poor coupling – not from the gauge. Turn those three into fixed habits and the numbers become stable on their own.