DFT Outlier and a Below-Spec Pull-Off Test on a Bridge Repaint
On a bridge repaint, the specification requires a 6–9 mil DFT epoxy midcoat, measured per SSPC-PA 2, and 500 psi minimum pull-off adhesion per ASTM D4541. The inspector takes five spot measurements in a 100 m² area: 8.4, 8.0, 7.6, 8.1, and 4.5 mil. The average is 7.3 mil, within range. The contractor's gauge was adjusted on the blasted steel using a certified shim on a smooth plate, but not on the actual rough surface. A pull-off test nearby yields 410 psi, with 70% cohesive failure in the epoxy and 30% adhesive failure at the primer/midcoat interface. The contractor argues the average DFT passes and the adhesion test is invalid because the dolly was applied over a curved weld toe.
1. Should the DFT area be accepted? Explain the acceptance logic.
Read the worked answer
No. Under SSPC-PA 2, acceptance is not decided by the average alone. The average must meet the minimum, no individual spot may fall below 80% of the minimum, and no spot may exceed the maximum. With a 6.0 mil minimum, 80% is 4.8 mil. The 4.5 mil spot fails.
Gauge adjustment on a smooth plate is also questionable. Type 2 gauges are affected by surface roughness; adjustment should account for the actual surface or be verified on the roughened steel. Roughness usually causes high readings, so a low reading on a rough surface is more concerning, not less. Reject the area, map the low readings, take additional measurements, and repair the low film before accepting the midcoat.
2. How should the 410 psi pull-off result be interpreted?
Read the worked answer
The result is below the 500 psi minimum and cannot be averaged away. The mixed failure mode is informative: 70% cohesive failure within the epoxy suggests the epoxy itself may be the weak link, while 30% adhesive failure at the primer/midcoat interface suggests contamination, an exceeded recoat window, or inadequate surface preparation. Either way, the system did not demonstrate the specified adhesion.
The contractor's objection about the curved weld toe has some merit: a dolly on a curved or irregular surface can produce eccentric loading and invalid results. But that justifies retesting on a properly prepared flat area, not ignoring the failure. Conduct additional pull-off tests on representative flat areas, review primer age and recoat records, and if failures repeat, remove and recoat the affected area.
3. What quality-control actions should follow these results?
Read the worked answer
Issue a nonconformance report and place a hold on the area. Do not permit the topcoat until both DFT and adhesion are resolved.
Actions:
- Map and repair low-film areas; re-measure after cure.
- Retest adhesion on properly prepared flat locations, with documented dolly preparation and failure modes.
- Review primer application and recoat records for contamination or exceeded windows.
- If retests pass, document the resolution; if they fail, remove the midcoat, prepare the primer, and reapply.
Record all readings, gauge identification, calibration, and test locations. The goal is a defensible acceptance decision, not a passing average.