Weld Defects in Metal Fabrication: Causes and Countermeasures

How Weld Defects Are Classified

Welding joins metal by creating a localised melt that solidifies into a cast structure bounded by partially melted parent metal. Every defect mechanism grows out of that sequence: problems in the liquid pool, problems during solidification, and problems created by the heat that the surrounding metal absorbs. Understanding which of the three stages a defect belongs to is the fastest way to select a countermeasure, because the corrective action for a solidification crack is not the same as the action for a hydrogen crack that appears hours after welding.

Hot Cracking and Cold Cracking

Hot cracking forms while the weld is still solidifying. Low-melting-point films enriched in sulphur, phosphorus or other residual elements remain liquid between growing grains, and the shrinkage stress of a highly restrained joint tears those films apart. It typically appears along the centreline of the bead or at the fusion line where solidification terminates, and it can be visible immediately or only after the joint has cooled and been strained.

Countermeasures target the three ingredients separately:

Reduce restraint. Adjust the weld sequence, allow movement, and use balanced welding on both sides of the joint where the design permits.

Change bead shape. A wider, shallower bead with a slightly lower travel speed and a modest weave reduces the centreline strain concentration; depth-to-width ratios beyond about 1.2 are increasingly prone to centreline cracking.

Control chemistry. Verify that sulphur and phosphorus in the parent plate and filler are within specification, and avoid excessive dilution from a high-sulphur base metal.

Cold cracking is a hydrogen problem and usually a steel problem. Atomic hydrogen dissolved in the weld metal diffuses into the heat affected zone, where a hard martensitic microstructure and high residual stress combine to nucleate a crack. It often appears hours or days after welding, which is why delayed inspection and hold times are specified for thick-section and high-carbon-equivalent steels. The countermeasures are equally well established: low-hydrogen consumables such as E7018 H4 classified to AWS A5.1, correct electrode storage and re-drying, preheat and interpass temperature control, and a post-weld hydrogen release treatment for heavy sections. As carbon equivalent rises above roughly 0.40, preheat becomes mandatory rather than optional practice.

Porosity, Slag Inclusion and Lack of Fusion

Porosity is the most common defect found by radiography and the easiest to eliminate once the source is identified. Moisture on the joint face, oil, paint, mill scale, condensation in the gas line and turbulence that draws air into the arc all produce gas that cannot escape before the pool freezes. Practical controls include drying the joint area, using shielding gas of verified purity with a correct flow rate (excessive flow creates the very turbulence that causes air entrainment), maintaining the correct contact-tip-to-work distance, and eliminating draughts around the arc.

Slag inclusion is specific to flux-bearing processes. It arises when flux is not removed between passes, when the travel angle pushes slag ahead of the pool, or when the arc energy is too low to keep the slag molten and floating. Multi-pass work requires interpass cleaning as a formal step, not an optional one.

Lack of fusion and lack of penetration are geometric errors. They occur when the arc does not reach the root or sidewall: too low a current, travel speed too high, electrode angle misaligned, joint gap too tight, or a bevel preparation that leaves an unmelted shelf. They are also the defects most likely to be missed by visual inspection, which is why ultrasonic or radiographic examination is specified for full-penetration joints.

Detection, Acceptance and Repair

Defect control only works when inspection gives feedback quickly. Surface and near-surface defects are found by visual examination and dye penetrant testing; volumetric defects by radiographic testing or ultrasonic testing, with phased array techniques improving detection of planar defects such as lack of fusion. Acceptance is normally expressed as a quality level rather than a pass or fail judgement: ISO 5817 defines levels B, C and D for steel arc-welded joints, while AWS D1.1 sets acceptance criteria for structural steelwork, and ASME BPVC Section IX and Section V govern procedure qualification and examination for pressure equipment.

Repair itself is a risk. Excavating a defect and re-welding introduces a fourth thermal cycle, additional restraint and a new heat affected zone, so the repair procedure must be qualified and the number of permitted repairs limited. Where defects are extensive, cutting out the joint is usually cheaper than chasing individual indications.

Common Mistakes in Weld Quality Control

Treating weld colour or appearance as the only acceptance criterion in reactive metals such as titanium and aluminium, where contamination is invisible in the early stages.

Storing low-hydrogen electrodes in open quivers and assuming the classification still applies at the arc.

Welding thick sections without a preheat calculation, then finding cold cracks days later during hydrotest.

Using excessive shielding gas flow and creating turbulence that pulls air into the arc while believing purity alone protects the pool.

Skipping interpass cleaning, which converts a routine multipass joint into a slag inclusion report.

Allowing unqualified repair welds, which are statistically the origin of a large share of service failures.

FAQ

Q: What is the most common cause of weld porosity?
Moisture and surface contamination are the leading causes, followed by gas flow problems. Damp electrodes, rusty or oily plate, condensation in the gas line, and turbulent flow that entrains air all produce the same result, so the diagnosis normally starts with joint cleanliness and gas delivery rather than with welding parameters.

Q: How long after welding can a cold crack appear?
Hydrogen-assisted cold cracks may appear minutes to several days after welding, which is why hold times before inspection are specified for susceptible thicknesses and carbon equivalents. Delayed inspection after a defined interval makes little sense if a crack can still initiate after that interval.

Q: Can a weld with porosity still be accepted?
Yes. Acceptance criteria allow a defined amount of dispersed porosity, because round gas pores are far less damaging than planar defects. The quality level in ISO 5817 or the applicable code governs, and the distinction between porosity and lack of fusion matters more than the total count.

Q: What is the difference between lack of fusion and lack of penetration?
Lack of fusion is incomplete bonding between the weld metal and a sidewall or a previous pass. Lack of penetration is a failure to melt through the full joint root. Both are planar defects that concentrate stress, and both are more hazardous than porosity under fatigue or pressure loading.

Q: When is post-weld heat treatment required?
It is specified to reduce residual stress, to allow hydrogen to escape before inspection in susceptible steels, or to restore properties in the heat affected zone. The requirement depends on material, thickness, service and the governing code, and should be written into the welded procedure rather than decided in the shop.

Q: Is preheat needed for all thick-section welds?
Not automatically, but it is required whenever the combination of carbon equivalent, thickness, hydrogen level and restraint creates a risk of cold cracking. Heat input control, clean consumables and residual-stress management all interact with the preheat decision.