NDT unit converter hub

Inspection work runs on two unit systems at once. The drawing is in millimetres, the instrument reads velocity in inches per microsecond, the client’s corrosion report is in mils, the hydrotest certificate is in psi and the source certificate is in curies. Most unit mistakes in NDT are not arithmetic errors — they are a number carried into a form without its unit, and the results are expensive: a wall loss judged against the wrong minimum thickness, a hydrotest pressure a factor of ten out, a source activity that puts the barrier in the wrong place.

Three conversions in this list behave differently from the rest and are worth understanding rather than looking up. Temperature is affine, not proportional: 100 °C is not twice 50 °C, and the conversion carries an offset. But a temperature difference converts as a pure ratio — a 1 °C change is 1 K and 1.8 °F, which is what matters when correcting ultrasonic velocity for temperature. Decibels are not a unit at all but a logarithmic ratio between two amplitudes: for the amplitude quantities used in UT the conversion is 20·log₁₀(A₂/A₁), so 6 dB is a factor of two and 20 dB a factor of ten. Using the 10·log₁₀ energy form on a screen-height ratio will halve every gain figure. Leak rate is a throughput — pressure times volume per unit time — so mbar·L/s, Pa·m³/s and atm·cc/s all describe gas flow, and 1 mbar·L/s is 0.1 Pa·m³/s.

Two constants worth committing to memory because they appear on every job: 1 inch per microsecond is exactly 25.4 mm/µs, or 25 400 m/s — so steel compression velocity 5900 m/s is 0.2323 in/µs — and 1 curie is exactly 37 GBq, which makes a nominal 100 Ci iridium-192 source 3.7 TBq. On the dose side 1 Sv = 100 rem, so 1 mSv = 100 mrem, and an older survey meter reading in mR/h is close enough to µSv/h × 0.1 for planning purposes in a gamma field.

Where a number carries a unit into a code check, the unit that matters is the one the code is written in. ASME allowable stresses are tabulated in ksi or MPa depending on the edition and section; converting a psi reading into ksi before comparing avoids the classic factor-of-1000 slip. Density conversions matter mainly for acoustic impedance work, where impedance Z = ρ·c must be in consistent SI units before the reflection coefficient means anything.