Ultrasonic thickness mapping on corroded girders

I’m using a dual-element, 5 MHz, 0.5 in transducer with glycerin couplant to map section loss on a 1978 plate girder, and I’m debating 25 mm vs 50 mm grid spacing to capture the true minimum remaining thickness for a conservative capacity check. For those prioritizing fracture resilience and buckling resistance, how are you setting field calibration (IIW V1 vs step wedge) and scan density to balance time on steel with confidence in the critical path?

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍‌‌‌‍⁠‍‌‍‌⁠‌‍‍‌‌‍⁠‍‌‍‌‌‌‍‌‌‌⁠​‍‌‍‍‌‌‍⁠‍‌‍‌⁠​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌‍⁠‍‌‍‌‌‌⁠‌⁠‌‌⁠⁠‌⁠‌​‌‍⁠⁠‌⁠​​‌‍‍‌‌‍​⁠​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​‍​‍‌‍⁠‍‌‍‌‌‌⁠‌⁠​‍​‍​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‌​⁠​‌​⁠​‍​⁠​‌​⁠‍​​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌​‌‌​⁠​​​‍⁠‌‌​⁠‌‌‍​‍‌‌​⁠​⁠‌‍‌​‌​‌‍⁠‍​⁠‌‍‌​​⁠‌​‍​‌⁠‌​​⁠‍​‌⁠‌‍‌‍⁠‌​‍​‍‌⁠⁠‌

On similar 70s girders I calibrate with a steel step wedge bracketing the expected thickness; IIW V1 stays for angle/velocity checks, but the wedge keeps thickness linearity on rough, dual‑element reads. I run 25 mm only in hot zones — weld toes, near stiffeners, obvious pits — and 50 mm elsewhere, then a quick zig‑zag sweep to “chase the min,” like finding the deepest pothole before your tire does. If the couplant struggles, swap to a tackier gel and, on the worst pitting, drop to 3.5 MHz for a cleaner backwall.

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍‌‌‌‍⁠‍‌‍‌⁠‌‍‍‌‌‍⁠‍‌‍‌‌‌‍‌‌‌⁠​‍‌‍‍‌‌‍⁠‍‌‍‌⁠​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠​‌​⁠‌‍​⁠‍‌​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‌​⁠​‌​⁠​‍​⁠​‍​⁠​‍​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠‍​​⁠‍‌‌‍‌​‌⁠​⁠‌‍‌​‌‌‌‌​⁠​⁠​⁠‌‍‌⁠‌⁠​⁠‌⁠​⁠‌‍‌⁠‌​‌​⁠⁠‌‍‌⁠‌‌‌‍​‍​‍‌⁠⁠‌

‘5 MHz, 0.5 in’: 25 mm at weld toes/drip lines if patchy, 50 mm elsewhere; step wedge, IIW V1 spot-check.

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍‌‌‌‍⁠‍‌‍‌⁠‌‍‍‌‌‍⁠‍‌‍‌‌‌‍‌‌‌⁠​‍‌‍‍‌‌‍⁠‍‌‍‌⁠​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠​‌​⁠‌‍​⁠‍‌​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‌​⁠​‌​⁠​‍​⁠​‍​⁠‌​​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌​⁠‌‌⁠​⁠‌‍‌‌‌‍​‌‌​⁠‍‌‍‌‌‌⁠​​‌⁠​‌‌‌‌​‌​‍​‌‌​⁠‌‌‍‌‌⁠​‍‌‌⁠⁠‌⁠‍‍​⁠‌​​‍​‍‌⁠⁠‌