A weld defect is any flaw in a welded joint that compromises its strength, appearance, or ability to perform as designed. These flaws range from tiny gas pockets invisible to the naked eye to large cracks that can cause a structure to fail under load. Some defects sit on the surface where they can be spotted visually, while others hide deep inside the joint and require specialized equipment to find. Understanding what causes them, and how to catch them, is essential for anyone working with welded structures.
Surface vs. Subsurface Defects
Weld defects fall into two broad categories based on where they occur. Surface defects form on or near the outside of the weld and are often visible during a standard inspection. These include undercut, overlap, spatter, and certain types of cracking. Subsurface defects are buried within the weld metal or at the boundary between the weld and the base material. Porosity, slag inclusions, and incomplete fusion are common examples. Subsurface flaws are particularly dangerous because they weaken a joint without any outward sign of trouble.
Porosity
Porosity shows up as tiny cavities, essentially gas bubbles, scattered throughout the weld. These pockets form when gas gets trapped in the molten weld pool as it solidifies. The most common cause is poor shielding gas coverage. Leaks in the gas line, flow rates set too high or too low, drafts in the work area, and turbulence in the weld pool can all allow air to reach the molten metal. When nitrogen or oxygen from the atmosphere dissolves into the weld, it creates voids as the metal cools.
Contamination is the other major culprit. Moisture, grease, or oil on the metal surface introduces hydrogen into the weld pool, which also forms gas pockets. Electrode coatings that haven’t been dried properly and damp flux materials produce the same result. Preventing porosity starts with clean materials, tight gas connections, and controlled airflow around the welding area.
Cracking
Cracks are the most serious category of weld defect because they can propagate under stress, turning a small flaw into a catastrophic failure. Three factors combine to cause cracking in steel welds: hydrogen introduced during the welding process, a hard and brittle microstructure in the heat-affected zone, and tensile stress acting on the joint.
The primary source of hydrogen is moisture trapped in flux coatings, cored wires, or submerged arc welding flux. Hydrogen can also come from the base material itself if its processing or service history left residual moisture, or simply from a humid atmosphere. Once hydrogen diffuses into the hardened, highly stressed zone near the weld, cracking can occur at or near room temperature, sometimes hours or even days after welding is complete. This delayed nature makes hydrogen cracking especially insidious.
The composition of the base metal plays a major role too. Steels with higher carbon content or higher carbon equivalent values are more prone to forming hard, brittle structures during rapid cooling. Controlling heat input, using preheat when required, and selecting low-hydrogen consumables are the standard defenses against this type of cracking.
Undercut
Undercut appears as a groove or channel melted into the base metal along the edge of the weld bead. It happens when excessive heat melts away material at the weld toe without enough filler metal flowing in to replace it. Welding too fast, running too much current, or using an incorrect arc length can all produce undercut. The result is a thinner cross-section right where stress concentrates, which significantly reduces fatigue strength.
Fixing technique issues is the main prevention strategy. For flux-cored arc welding, maintaining the correct contact-tip-to-work distance is critical. If the wire stickout is too short, the flux doesn’t preheat properly; too long, and it overheats. Matching voltage, wire feed speed, and shielding gas type to the manufacturer’s specifications for the specific electrode matters more than many welders realize. Travel angle also plays a role. A slight drag angle (around 10 degrees) in flat and horizontal positions, and a slight push angle for vertical-up welding, helps keep the arc centered and the weld pool properly controlled.
Slag Inclusions
Slag inclusions are small, elongated pockets of non-metallic material trapped inside the weld. They form when slag from the flux coating isn’t fully removed between welding passes, or when the base metal isn’t cleaned thoroughly before welding. These inclusions are often lumpy and sit near the weld’s surface, but they can be nearly invisible without magnification. They act as stress concentrators and weak points within the joint, reducing its load-carrying capacity.
Incomplete Fusion and Incomplete Penetration
Incomplete fusion occurs when the weld metal fails to properly bond with the base material or with a previous weld pass. The root cause is usually insufficient heat. The short-circuiting transfer mode in gas metal arc welding, for example, produces low penetration and can leave unfused regions at the joint root. If the area closest to the electrode is too far from the root of the joint, heat conduction alone may not raise the temperature enough to achieve a full bond.
Incomplete penetration is a related but distinct problem where the weld doesn’t extend through the full thickness of the joint. This commonly happens when welding a groove from one side only if the root face is too thick, the root opening is too narrow, or the groove angle is too tight. Electrodes that are too large for the joint geometry can bridge across the gap without reaching the root. Using insufficient welding current or moving too fast along the joint produces the same outcome. In both cases, the joint has far less strength than its design requires because the weld metal isn’t carrying load across the full cross-section.
Real-World Consequences
Weld defects aren’t just cosmetic problems. Most brittle fractures in welded structures initiate at or near defects located in regions of high stress concentration. The combination of a flaw (such as lack of penetration, an undercut, or an inclusion) with the geometric stress concentration inherent in a welded joint creates conditions ripe for failure.
The Exxon Valdez grounding investigation revealed fillet weld failures in the damaged area of the hull. During the 1995 Kobe earthquake, steel beam connections cracked starting from weld zones near diaphragm plates. Research at MIT confirmed that fillet welds ending abruptly are especially vulnerable under load, and that even a moderate undercut at a weld toe has large negative effects on fatigue life. These aren’t edge cases. They illustrate a consistent pattern: unaddressed weld defects in load-bearing structures reduce the margin between normal service and failure.
How Defects Are Detected
Visual inspection catches surface-level problems like undercut, overlap, and spatter, but subsurface defects require non-destructive testing (NDT) methods.
- Magnetic particle testing works by magnetizing the welded component. Defects near the surface disrupt the magnetic field, causing fine iron particles spread on the surface to cluster at the flaw location. It’s effective for surface and near-surface cracks in ferromagnetic materials like steel.
- Ultrasonic testing sends high-frequency sound waves into the weld through a transducer placed on the surface. When those waves hit a defect, they bounce back. The return signal reveals the size, location, and type of flaw. Ultrasonic testing detects cracks, lack of fusion, porosity, and even residual stresses within the joint.
- Radiographic testing passes X-rays or gamma rays through the weld onto a film or digital detector on the other side. Defects like porosity, slag inclusions, burn-through, and lack of penetration show up as differences in the image because they absorb radiation differently than solid metal. This method produces a permanent record of the weld’s internal condition.
Each method has strengths for particular defect types and joint configurations. In critical applications like pressure vessels, pipelines, and structural steel, codes typically require one or more of these methods at specified intervals throughout production.
Prevention Starts Before the Arc
Most weld defects trace back to one of four root causes: contamination, incorrect heat input, poor joint preparation, or improper technique. Cleaning the base metal, storing consumables in dry conditions, verifying gas line integrity, and matching machine settings to the electrode manufacturer’s recommendations eliminates the majority of porosity and hydrogen-related cracking. Proper joint design, with adequate root openings, appropriate groove angles, and realistic root face dimensions, prevents incomplete fusion and penetration problems. And controlling travel speed, arc length, and gun angle addresses undercut and overlap before they start.
Welding codes like AWS D1.1 set specific acceptance criteria for how large a discontinuity can be before it’s classified as a rejectable defect. Small flaws that fall within tolerance are acceptable. A corner crack up to 1/4 inch on a bend test specimen may pass, for instance, unless it results from slag inclusion or a fusion-type discontinuity, in which case the limit drops to 1/8 inch. The line between an acceptable discontinuity and a defect that requires repair is defined by the applicable code for each project.