Five Things to Test Before You Buy a Flaw Detector

1. Sensitivity reserve

Using a defined artificial defect on a reference block, see whether the echo stands clear of noise. A set with no reserve forces you to keep raising gain until the display is unreadable.

How to test: with the same block and probe, reduce gain until the echo just disappears, then bring it back. The dB difference is your reserve.

2. Resolution

Can two adjacent defects be separated? Poor resolution merges near-surface defects into the initial pulse and you lose the layering.

How to test: a block with a laminated defect – look for a clear trough between the two echoes.

3. Vertical and horizontal linearity

Vertical linearity determines whether echo height represents defect size; horizontal linearity determines whether position is accurate. Both out of tolerance and quantitative analysis is meaningless.

How to test: holes at different depths in a reference block – check whether amplitude and depth track proportionally.

4. Coupling and signal-to-noise on your own parts

This is the one you cannot do on a demo block. The same probe on your rough surface, your curvature, your wall thickness decides whether the set is usable in production. Demos are always run on polished reference blocks, where everything looks good.

5. Battery life and environmental protection

Site work runs all day. Whether the battery lasts eight hours, what ingress protection the housing offers, and whether you can read the screen in sunlight are all on the datasheet – but must be confirmed in your hand.

A note on "smart" and "phased array"

"Smart" usually means automatic calibration, automatic defect positioning, process recording and report export. Phased array (PAUT) is a different class of instrument using array probes for beam steering and focusing, at several times the price.

They are not the same thing. Do not pay extra for the word "phased array" unless your geometry genuinely needs it – complex weld configurations, for example.