Five files we're allowed to talk about.

Matter

County DOT requested root-cause analysis on three stiffener plates showing toe cracks at the fillet welds, discovered during a biennial inspection. The bridge had been in service eleven years. Original fabrication records were available but incomplete for the stiffener welds specifically.

Approach

Sectioned transverse through the crack plane on all three plates. Macro-etch with 10% ammonium persulfate. Hardness traverse across HAZ, 0.5mm step. Optical metallography at 200× and 500×.

Finding

Weld metal was sound. Base plate showed banded microstructure with coarse pearlite colonies aligned parallel to the rolling direction. Cracks initiated at weld toes where the banding intersected the HAZ. Not a weld defect — a base-plate susceptibility amplified by joint geometry. Recommended retrofit stiffeners be cut from normalized plate.

Matter

A pressure-vessel fabricator needed PQR for GTAW root / GMAW fill on 2205 duplex, heavy wall. Their first attempt failed ferrite-balance requirements — phase count off by nine points. They needed to know why, fast.

Approach

Re-etched their existing coupon with modified Beraha's, ran ASTM E562 point count at 400× over twenty fields. Cross-checked against the WPS heat-input numbers. Verified interpass temperature against shop log.

Finding

Interpass temperature had drifted to 205°C on the last three passes — well above the 150°C the WPS called out. Excess heat input was dissolving ferrite into austenite on cooling. The procedure was fine on paper. The shop floor wasn't following it.

Matter

A forged crane hook separated at the eye during a lift of a steel coil. No injuries. Insurance carrier for the yard requested independent analysis. Nominal rating of hook was well above the lift weight.

Approach

As-received photography. Fracture-surface examination under stereo microscope. Sectioning through the eye radius. Hardness traverse from bore outward. Metallography at the initiation site. Reviewed prior inspection records.

Finding

Fatigue crack, origin at the bore radius, propagated approximately 70% through the section before final overload. Hardness and microstructure consistent with spec — the steel was fine. Prior inspection records showed the hook had never been MT-checked at the bore, only the saddle. A twelve-year-old defect nobody had looked at.

Matter

Structural fabricator received a shipment of plate stamped A572-50 but with MTRs that didn't match the heat numbers on the plate. Supplier insisted the MTRs were correct. Fabricator insisted something was wrong. Both were right.

Approach

OES chemical composition from three plates. Tensile coupons per ASTM E8. Hardness. Microstructure review.

Finding

Chemistry matched A572-50 on two plates; the third came in on the low side of manganese and high on silicon — closer to A529. Tensile properties still met A572-50 minimums but the third plate was clearly from a different heat. Mill had mis-stamped at the shipping dock. Paperwork was technically right for the heats they meant to send.

Matter

Four of thirty-six anchor bolts on a precast erection failed within eighteen months of install. Bolts were ASTM F1554 Gr. 55, hot-dip galvanized. Install happened in late January during a cold snap.

Approach

Fracture-face examination. Hardness traverse from head to thread. Hydrogen-content measurement on one bolt. Review of mill certs and galvanizing certs.

Finding

Classic delayed hydrogen embrittlement. Bolts tested at HRC 35 — upper edge of the grade, where hydrogen susceptibility climbs. Galvanizing had introduced the hydrogen; cold install and sustained tension had provided the driving force. Grade 55 was within spec but marginally so. A lower-hardness grade or post-galvanizing bake would have prevented it.