How much porosity?

A problem arose at a structural steel fabricator's works, where plates used for fabricating box girders were coated with an iron oxide epoxy primer. Although the primer was applied by mechanised plant, overlapping passes of the spray gun caused localised excess coating thickness which on a number of occasions had caused porosity in fillet welds during procedure testing.

Before the transverse stiffeners were welded into box sections, therefore, the primer was removed from the corners by grinding. This enabled rutile-flux-cored wires to be used without further removing the primer - which increased productivity compared with manual metal arc (MMA) welding.

To meet the production schedule on the last girder, the grinding operation was omitted, but when this was noticed by a visiting inspector, he asked for all the welds to be ground out and replaced. To avoid this time-consuming operation the fabricator persuaded the inspector to accept radiography of the welds in question, and the quality requirement was agreed to be that laid down in the contract, viz Class Blue No 5 of the IIW collection of reference radiographs.

The radiographic technique was to direct the beam of X-rays towards each fillet weld in turn at an angle of 40° to the stiffener. The film cassette was placed on the opposite side of the T joint.

Difficulties arose in interpreting the radiographs when comparing them with the specified IIW reference radiographs and agreement could not be reached on the acceptability of the welds. A TWI welding engineer made the following comments:

The IIW reference radiographs are of butt welds, not fillet welds. Fillet welds on primed plate tend to have any porosity concentrated in the root region of the weld, and the pores tend to be elongated in the form of wormholes with their major axes aligned with the weld throat. IIW reference radiograph Blue No 5 shows distributed porosity in a butt weld and comparison with aligned porosity would be virtually impossible.
The IIW reference radiographs were at one time used, quite erroneously, as acceptance Standards and this caused great difficulties, as in the present case. The latest edition of the IIW collection states that it is not considered to be an acceptable Standard for welds. It shows typical weld defects of different degrees of severity but the sizes, extent and distribution of defects are not numerically defined, which would be essential for an acceptance Standard.
The fillet welds do not transmit the main loads in the girder, and TWI and others demonstrated many years ago that any reduction of cross sectional area caused by porosity in butt or fillet welds has to be well over 5% before there is any reduction in tensile strength. Work at TWI has also shown that, provided the weld size is adequate, fatigue failure occurs from the toe of the weld and propagates through the parent plate, the presence of porosity in the weld having no effect.

The client's inspector accepted that, in the present case, an increase in the weld sizes from 6mm to 8mm leg length would more that compensate for the root porosity in the fillet welds. In fact, there was no real need to increase the sizes of the welds and they were already overdesigned - not an unusual occurrence!