Source-labelled market dataSource, observation time and freshness are shown with every quoteData statusAll pages
GoldObserve

INTERIOR / SOUND / COVERAGE / POROSITY / INTERPRETATION

Ultrasonic Testing for Gold Bars

Understand back-wall signals, internal reflections, material velocity, scan coverage and benign porosity without mistaking ultrasound for a gold fineness assay.

THE SHORT ANSWER

Ultrasound looks inside a bar, but it does not assay the gold

A probe sends sound into a bar and records returning signals. A predictable back-wall reflection and material-consistent travel time can support an interior-consistency assessment; unexpected reflections or signal loss can indicate an interface, insert, pore or other discontinuity. LBMA is explicit that ultrasound provides no fineness information. Coupling, coverage, bar geometry, operator technique and benign casting porosity can all change the scan.

METHOD BOUNDARY

Separate the observation from the conclusion

QuestionDefensible answerUnsafe shortcut
What it measuresSound travel time, back-wall response and reflections from acoustically distinct internal interfaces within the covered scan path.Calling the observation proof of authenticity.
What it can missFineness, fully alloyed compositional changes with low acoustic contrast, uncovered regions and indications outside the method's sensitivity.Treating an unobserved defect as an impossible defect.
Best roleNon-destructive interior-consistency screening of suitable bars by trained operators using documented calibration and coverage.Replacing identity, provenance, recourse and other tests.

SCAN RECORD

An image without setup data is not a reproducible test

Bar geometryMass, length, width, thickness map, faces and surface condition.
EquipmentInstrument, probe, frequency, software, calibration and reference block.
CouplingCouplant, contact quality, surface access and any poor-contact regions.
CoverageScan grid, probe path, inaccessible areas and orientation.
Raw responseTravel time, back-wall signal, amplitude and stored images or data.
InterpretationAcceptance basis, operator competence, ambiguity and required secondary method.

POROSITY PROBLEM

A discontinuity is not automatically fraud

Cast gold is not a perfect optical or acoustic monolith. LBMA's phased-array work reports that microscopic air pockets and porosity can form during solidification without changing purity, mass or market acceptability. Those features can strongly reflect or scatter ultrasound, reduce the back-wall response and create local depth variation.

The correct response to an ambiguous scan is not to declare the bar fake or ignore the signal. Preserve the data, compare it with the documented acceptance procedure and use a qualified secondary method. Conductance or eddy current can add evidence, but LBMA notes that secondary methods may also fail to provide an unambiguous answer.

PRIMARY AND STANDARDS SOURCES

Methods are only as strong as their scope and controls

Sources reviewed July 31, 2026. Instrument procedures, safety data, standards and acceptance criteria can change. Use the current method documentation for the exact item, instrument, laboratory and decision.

FREQUENTLY ASKED QUESTIONS

Ultrasonic gold-bar questions

Can ultrasound tell how pure a gold bar is?

No. LBMA says ultrasound provides no information on fineness. It can identify reflections or travel behavior associated with interfaces, inserts, pores or other internal features.

Can ultrasound detect tungsten in gold?

It can be useful because sound velocity differs between gold and tungsten and an interface may reflect sound, but coverage, orientation, geometry, coupling and interpretation remain method-specific.

Why can a genuine cast bar fail an ultrasonic test?

LBMA describes benign porosity and small voids from casting that can scatter or reflect sound, weaken the back-wall signal and resemble a problematic indication.

Is phased-array ultrasound better than a single probe?

It can provide greater coverage, focusing and recordable data, but LBMA also identifies interpretation, sensitivity and standardisation challenges. More data does not eliminate method limits.