In discussions with quality managers we keep meeting the same misconception: "Acoustic testing is the same as ultrasound, right?" No. Both use sound waves, but physics, expressiveness and economics differ fundamentally.

The physics compared

Ultrasonic testing (UT)

A piezoelectric probe sends high-frequency waves (typ. 1–25 MHz) into the part. Reflections at material boundaries, cracks or inclusions are detected. Localisation via time of flight.

Strength: high spatial resolution. A 0.3 mm crack is detectable when the probe is close.

Weakness: point inspection. Full inspection of a complex casting needs many probes or repositioning – costly.

Acoustic resonance analysis (ART)

The part is briefly excited and rings freely. The entire set of natural frequencies is captured – depending on geometry, mass, material and integral defect state. Comparison vs. master or ML model decides OK/NOK.

Strength: holistic assessment in < 1 s.

Weakness: no localisation.

Comparison matrix on eight criteria

CriterionUltrasonic testingAcoustic resonance analysis
Inspection speed5–60 s per part0.2–2 s per part
100 % capabilityonly with effortstandard case
Detection capabilitylocal, ≥ 0.3 mm crackintegral, ~1 % stiffness change
Localisationmillimetre-precisenot possible
Material varietymetals, some plasticsmetal, sinter, ceramic, composite
Part geometryflat/tubular preferrednearly arbitrary
Investment50–500 k€ (std.) – 0.5–2 M€ (PA)30–200 k€
Inline integrationhigh (couplant, robot)low (sensor + excitation)

When to pick which

Ultrasound when defect location matters, sampling-only is OK and geometries are simple.

Resonance analysis when 100 % inline matters, geometries are complex, the question is "does it work?" rather than "where exactly is the flaw?", or material mix-ups have to be detected (UT often misses them).

Hybrid strategy

Both methods complement each other. RTE often recommends: 100 % ART as inline filter, UT sample on parts flagged "suspicious" by ART. Lowest cost per defect found.