ISO 3497 — coating thickness measurement by X-ray spectrometric methods

Scope — and what it does not cover

ISO 3497 specifies how metallic coating thickness is measured with X-ray fluorescence (XRF): calibration, reference standards, spot size, counting time and the sources of error. It is the practical, non-destructive way to check zinc thickness on small plated parts. It does NOT define how thick your coating must be — ISO 2081, ISO 4042, ASTM B633 and F1941 do that — and it does not replace a microsection (ISO 1463) or a coulometric test (ISO 2177), which remain the referee methods when the geometry is difficult or the result is disputed.

Standards are copyrighted documents. We explain scope, intent and practical application in our own words; obtain the current edition from ISO, ASTM, DIN or your national standards body.

What actually bites in production

• 1XRF reports an average over the measured spot. On a curved or threaded surface the spot sees several geometries at once and the number drifts.
• 2Calibration standards must match the substrate, the coating and, as far as possible, the geometry. A flat-standard calibration read on a thread crest is not a valid measurement.
• 3Small parts need a defined flat significant surface for routine measurement: head face, shank or washer face. Thread thickness, when it truly has to be known, is a microsection question.
• 4Counting time trades against uncertainty. A fast reading on a thin coating carries a wider spread than the display suggests.
• 5Passivation films and sealers sit on top of the zinc and are effectively invisible to the measurement. The number is the metal, not the whole system.
• 6Report the method with the result. "8.4 µm" means nothing without the method, the spot location and the calibration.

What to write on the drawing or PO

• The thickness requirement as a minimum local value, and which surfaces are significant.
• The measurement method — XRF to ISO 3497, microsection to ISO 1463, or coulometric to ISO 2177 — and which one governs in a dispute.
• The number of measurement points per part and parts per lot.
• Whether a thickness report is issued with the delivery.

What goes wrong when it is cited loosely

• Comparing an XRF number from one laboratory against a microsection from another and calling the difference a defect.
• Measuring on a thread and rejecting a lot against a requirement written for a flat significant surface.
• Assuming the reading includes the passivate and the sealer.
• Specifying an average thickness when the standard cited requires a local minimum.