Welding Defects Guide: Types, Causes and Prevention

Dimensional imperfections

Dimensional imperfections refer to issues in the shape or size of the weld bead, which can impact the weld’s structural integrity and durability. These defects often result from incorrect welding techniques or improper filler metal application, leading to weak joints or excessive material on the weld. 

Discontinuities

Across different welding processes,
discontinuities are imperfections that interrupt the uniformity of the weld, creating areas of weakness that may compromise the welded structure. These defects can result from poor welding practices or environmental factors during the welding process. 

Surface Defects

Surface defects refer to imperfections that appear on the weld's surface, impacting both its appearance and, potentially, its structural integrity. These defects often arise from issues like contamination, rapid cooling, or inconsistent welding techniques. Surface defects can weaken the weld by creating stress concentration points or by allowing external elements to penetrate, but they can be identified through non-destructive visual inspection and advanced NDT methods.

Metallurgical Defects

Imperfections related to the material properties and structural composition of the weld metal. These issues often arise due to improper heat treatment, rapid cooling, or variations in the composition of the weld and base metals. Metallurgical defects can weaken the weld by creating areas with different hardness, brittleness, or strength, leading to potential failure under stress. They may also affect the weld's resistance to corrosion or impact, depending on the environment and the welding process. 

The reduced weld volume directly impacts the joint’s load distribution, making it vulnerable in applications with dynamic or concentrated stress points. Underfill welds often require rework in industrial settings, where precision and safety are paramount. The concave profile also creates a notch effect that can serve as a fatigue crack initiation site under cyclic loading.

Underfill is typically caused by a combination of incorrect welding parameters. The most common causes include: 

•  Low filler metal input, where insufficient filler metal is deposited during welding.

•  Low heat input, which creates a concave, undersized weld bead.

• Inappropriate travel speed, which fails to allow adequate material deposition. In robotic welding, a programming error in the wire feed rate or travel speed can also produce consistent underfill across an entire production run.

• Optimizing the balance between deposition rate and travel speed.

• Increase filler metal deposition to create an adequately sized weld bead and avoid concave profiles.

• Adjust heat input to enable full coverage across the joint and maintain a steady travel speed for uniformity.

• Apply multiple passes if needed to achieve the required thickness for load-bearing joints.

• In automated systems, verifying the programmed parameters against the actual joint geometry before production is essential.

Underfill can occur in any welding process, but it is particularly common in high-speed automated processes like GMAW and SAW if the parameters are not perfectly calibrated to the travel speed and joint design. It is also frequently seen in plasma arc welding (PAW) of turbine components and in orbital TIG welding of pipes where access is limited.

Underfill is when the entire weld face is below the base metal surface, indicating insufficient filler material in the joint. Undercut is a sharp groove melted into the base metal at the toe of the weld, which is a stress concentration point, even if the weld face itself is properly filled. Both are dimensional defects, but they occur in different locations and have different root causes.

Most welding codes and standards set limits on the amount of underfill that is acceptable. For example, AWS D1.1 (Structural Welding Code – Steel) specifies maximum allowable underfill depths depending on the thickness of the material and the type of loading (static vs. cyclic). ISO 5817 also classifies underfill by quality levels (B, C, D) with specific dimensional tolerances.

Incomplete penetration leaves unfused areas within the joint that act as stress concentrators. These hidden discontinuities compromise the joint’s ability to handle tensile and bending loads and can lead to crack initiation under cyclic stress. In critical applications, this defect demands full rework to ensure structural reliability.

This defect is often a result of insufficient heat input or improper joint setup. Key causes include:

• Low current settings that fail to reach the required joint depth

• High travel speed that reduces the arc’s ability to achieve full penetration

• Improper joint fit-up, such as a misaligned joint or a root gap that is too small.

Ensure proper joint preparation with an adequate root opening. Use a higher welding current and a slower travel speed to increase heat input and allow the arc to penetrate the root. Proper cleaning of the joint is also essential. For automated processes, precise joint tracking is critical to maintain consistent torch-to-seam alignment.

In automated welding, ensuring the torch is perfectly aligned with the weld seam is crucial for achieving consistent penetration. The Xiris SeamMonitor tool within WeldStudio™ 3 Pro can be used to track the seam in real-time, ensuring the arc is always directed at the root of the joint, which is fundamental to preventing lack of penetration.

Since lack of penetration is an internal defect, it is not visible from the surface. It must be detected using non-destructive testing (NDT) methods such as radiographic testing (X-ray), which shows the unfused root as a dark line, or ultrasonic testing, which detects the discontinuity as a reflected signal.

Lack of penetration is when the weld fails to reach the root of the joint — it’s a depth issue. Lack of fusion is when the weld metal fails to fuse with the sidewalls of the joint or with a previous weld pass — it’s a bonding issue. Both are serious internal defects, but they occur in different locations within the joint.

Absolutely. A joint with a very narrow groove angle, a large root face (the flat part at the bottom of the groove), or an insufficient root gap can physically prevent the arc from reaching and melting the root of the joint, making complete penetration impossible regardless of the parameters used. Proper joint design is the first line of defense.

Yes. Processes with lower heat input, such as short-circuit GMAW, are more prone to lack of penetration on thicker materials. Conversely, high-energy processes like SAW and plasma arc welding are less susceptible but can still produce this defect if the joint fit-up is incorrect.

Repair typically involves back-gouging the root of the weld from the opposite side (if accessible) to remove the unfused area, then re-welding. If back-gouging is not possible, the entire weld may need to be removed and re-done. This is one of the most costly defects to repair.

Beyond aesthetic concerns, spatter can affect component fit-up in assemblies and requires post-weld cleanup. While spatter does not directly weaken the weld, it can compromise the precision and surface finish required in high-quality manufacturing. In some cases, spatter can also damage sensitive nearby components or coatings.

Spatter is often caused by incorrect machine settings, such as:

• Excessive current that expels metal particles from the weld pool.

• Improper polarity that affects arc stability.

• Insufficient shielding gas flow that results in scattered particles on the surface. A dirty or rusty workpiece can also increase spatter significantly.

To prevent spatter, optimize your welding parameters. Lower the voltage and current, shorten the arc length, and ensure you are using the correct shielding gas composition and flow rate for your material. Using a consistent wire feed and maintaining a proper torch angle are also key. Anti-spatter spray can be applied to the workpiece as an additional measure.

Spatter can be easily monitored in real-time. A Xiris Weld Camera provides a clear view of the arc and weld pool, allowing an operator to see the amount of spatter being generated and adjust parameters to minimize it. Additionally, the WeldMic™ can detect the distinct sound of spatter, providing another layer of process monitoring to help operators tune the process for maximum stability.

Yes, significantly. Short-circuit transfer mode is prone to spatter as the wire repeatedly touches the workpiece, causing small explosions. Spray transfer is much smoother with less spatter, but requires higher energy. Pulsed GMAW offers a good compromise, providing a stable arc with low spatter across a wide range of parameters.

Not necessarily a bad weld in terms of strength, but it is a sign of a poorly controlled or non-optimized welding process. Excessive spatter indicates instability that could also be causing other, more serious defects like porosity or lack of fusion.

The cost of spatter goes beyond the obvious cleanup time. It includes labor for grinding and removal, potential damage to the workpiece surface, consumption of anti-spatter products, and the risk of quality issues in painted or coated assemblies. In high-volume production, even small amounts of spatter per part can add up to significant costs.

Overlaps compromise the weld surface, reducing its corrosion resistance and creating potential points of weakness. This flaw is particularly detrimental in applications that require strict surface integrity, as it undermines the weld’s longevity and resistance to environmental degradation.

Overlap is typically caused by:

• High welding speed that allows molten metal to flow onto the surface without bonding.

• Excessive filler metal deposited without enough fusion control

• Inadequate heat distribution that prevents bonding with the base metal. An incorrect torch angle can also push molten metal onto the cold base plate.

Maintain proper heat input to ensure the weld metal fuses with the base metal. Avoid excessive filler metal deposition and use consistent wire feed or rod addition. Keep a correct electrode or torch angle to guide the molten metal into the joint instead of onto the surface. Adjusting the travel speed to match the deposition rate is key.

Overlap is a surface defect that is readily visible. Using a Xiris Weld Camera, an operator can monitor the weld bead profile in real-time. The camera’s clear view allows the operator to see if the weld pool is flowing correctly and fusing with the base metal, or if it is rolling over the edge, enabling immediate correction of travel speed or angle.

Overlap and cold lap are very similar defects. Both involve weld metal flowing over the base metal without fusing. The term “overlap” is more commonly used in codes and standards, while “cold lap” emphasizes the thermal cause (insufficient heat for fusion). In practice, they are often used interchangeably.

Yes. The overlapping metal must be removed by grinding, and if necessary, the area is re-welded with correct parameters to ensure proper fusion at the toes. Simply grinding the overlap smooth is not sufficient if the underlying metal is not fused.

Yes. Most welding codes consider overlap a rejectable defect because the unfused metal at the toe acts as a crack initiator. It is classified as a lack of fusion defect at the weld toe and must be repaired to meet quality requirements.

Cracks represent a serious risk to structural integrity, serving as focal points for failure under cyclic or static loading. Unlike volumetric defects like porosity, cracks are planar defects that can propagate rapidly under stress, leading to catastrophic failure. In industries such as aerospace and nuclear, the presence of any crack is grounds for immediate rejection.

Cracking is caused by high levels of residual stress combined with other factors. Specific causes include: 

• Rapid cooling that induces internal stresses in the metal.

• Material incompatibility where materials with different thermal expansion rates cause tension

• Residual stresses from improper welding techniques. Hydrogen contamination from moisture is a major cause of cold cracking.

Prevention strategies depend on the type of crack. For cold cracks: use low-hydrogen consumables, preheat the material to slow the cooling rate, and ensure the joint is clean and dry. For hot cracks: choose a filler metal that is less susceptible to cracking and control the weld pool shape. Proper joint design to minimize stress and post-weld heat treatment (PWHT) for high-strength materials are also crucial.

Controlling the cooling rate is critical for preventing cracks in many high-strength steels. The Xiris XIR-1800 Thermal Camera can measure the weld’s thermal profile in real-time, including the critical t8/5 cooling time (the time to cool from 800°C to 500°C). By monitoring this parameter, operators can adjust heat input or travel speed to control the cooling rate and prevent the formation of brittle microstructures that lead to cracking.

Hydrogen-induced cracking (HIC), also known as cold cracking or delayed cracking, occurs when atomic hydrogen diffuses into the weld metal or HAZ during welding and becomes trapped as the metal cools. The hydrogen creates internal pressure that, combined with residual stress and a susceptible microstructure (martensite), causes cracks to form. This can happen hours or even days after welding.

Crater cracking occurs at the end of a weld bead when the arc is extinguished too abruptly. The crater (the depression at the end of the bead) solidifies rapidly from the outside in, creating tensile stresses that cause star-shaped cracks. It is prevented by using a crater-fill technique, where the current is gradually reduced at the end of the weld.

Cracks must be completely removed before re-welding. The crack is typically ground or gouged out, and the area is inspected (often with dye penetrant or magnetic particle testing) to confirm complete removal. The area is then re-welded using a qualified procedure, often with preheat to prevent the crack from recurring.

Preheating the base metal before welding slows the cooling rate after welding, which reduces the formation of hard, brittle microstructures (martensite) in the HAZ. It also helps hydrogen diffuse out of the weld before it can cause cracking. Preheat temperatures are specified based on the material type, thickness, and hydrogen level of the consumables.

Incomplete fusion weakens the weld by creating internal planes where stress can concentrate. Under cyclic or static loading, these planes can initiate cracks, leading to premature failure of the joint. Because of its impact on structural integrity, incomplete fusion is strictly limited by fabrication standards and typically requires full rework when detected.

Causes include:

• Improper electrode or torch angle that limits heat transfer and prevents full bonding.

• Low heat input or excessive travel speed that reduces penetration into the joint surfaces.

• Contaminated base metal where oxides, oils, or dirt block proper fusion.

Maintain correct electrode or torch angle to ensure heat is directed to the bonding surfaces. Adjust current and travel speed to provide adequate heat input for complete fusion. Clean the base metal thoroughly to remove all oxides, oil, and other contaminants. In multi-pass welding, ensure each pass adequately remelts the surface of the previous pass.

Visual monitoring is key. A Xiris Weld Camera provides a detailed view of the weld pool’s interaction with the joint. Operators can watch to ensure the molten metal is wetting the sidewalls and achieving a good tie-in, which are direct indicators of good fusion. This real-time feedback is invaluable for preventing this defect.

Incomplete fusion is best detected using ultrasonic testing (UT), which can identify the planar discontinuity as a strong reflected signal. Radiographic testing (RT) can also detect it, but it may be harder to see on film if the defect is oriented parallel to the X-ray beam. Phased array UT is particularly effective for this defect.

In practice, the terms are often used interchangeably. However, some codes distinguish between them: "lack of fusion" may refer specifically to the failure to fuse with the sidewall or a previous pass, while "incomplete fusion" may be a broader term that also includes the root. The key point is that both describe a failure to achieve a complete metallurgical bond.

Maintaining the correct inter-pass temperature is important for preventing incomplete fusion in multi-pass welds. If the previous pass cools too much before the next one is applied, the new weld metal may not have enough energy to remelt the surface and achieve fusion. Conversely, if the inter-pass temperature is too high, other problems like hot cracking can occur.

Arc strikes can create small heat-affected zones that may harden and become brittle, especially in high-strength steels. These localized zones can act as stress concentrators, increasing the risk of crack initiation and propagation. In structural or pressure applications, arc strikes are considered unacceptable and often require grinding and inspection.

Arc strikes are almost always a result of poor technique or carelessness:

• Accidental electrode contact that creates localized melting and surface marks.

• Poor tool or cable control that allows unintended touching of the base metal.

• Sticking and breaking electrodes that leave small scars or depressions when detached.

Each arc strike creates a small area that has been rapidly heated and then quenched. This can create localized hard, brittle spots (martensite) on the surface of the metal, which can act as initiation points for cracks, especially under fatigue loading. They are considered a serious defect in critical applications such as pressure vessels and structural steel.

Arc strikes are repaired by grinding the affected area smooth and then inspecting it (typically with magnetic particle or dye penetrant testing) to ensure no micro-cracks have formed. If cracks are found, the area must be ground deeper until all cracks are removed, and then re-inspected.

In critical applications (pressure vessels, structural steel under cyclic loading, aerospace), arc strikes are always rejectable and must be repaired. In less critical applications, some codes may allow minor arc strikes if they are ground smooth and inspected. The applicable code and the service conditions of the structure determine the acceptance criteria.

Yes. The rapid heating and cooling cycle of an arc strike can create a hard, hydrogen-susceptible microstructure in the base metal. If hydrogen is present (from moisture in the air or on the surface), the arc strike can serve as an initiation point for hydrogen-induced cracking, particularly in high-strength or high-carbon steels.

Cold lap creates weak points along the weld toes where cracks can initiate under stress. Because the metal has not fused properly, the joint’s load-bearing capacity is reduced, and the defect can lead to premature failure if subjected to cyclic or dynamic loads.

The primary causes are:

• Low heat input that prevents the molten metal from bonding to the base metal.

• Improper travel speed, where moving too slowly piles up metal without full fusion.

• Incorrect electrode angle or technique that directs the molten pool away from the joint surfaces. It is particularly common in short-circuit MIG welding.

Increase heat input to achieve proper fusion between filler and base metal. Maintain an even travel speed to allow the weld pool to wet and bond correctly. Hold the electrode or torch at the correct angle to ensure fusion along the weld toes. A slight weaving motion can also help to preheat the area just ahead of the weld pool and promote better fusion.

Like overlap, cold lap is a surface defect that can be detected visually. A Xiris Weld Camera gives the operator a clear, close-up view of the weld pool and the leading edge of the bead. The operator can see if the molten metal is flowing smoothly and fusing with the base metal, or if it is sluggish and simply rolling over the surface, allowing for immediate correction.

The terms are closely related and often used interchangeably. “Cold lap” emphasizes the thermal cause of the defect (insufficient heat for fusion), while “overlap” is the more formal term used in welding codes and standards. Both describe weld metal that has flowed over the base metal without achieving a metallurgical bond.

Yes. Cold lap is particularly common in short-circuit GMAW (MIG) welding, where the heat input is inherently lower than in spray or pulsed transfer modes. It can also occur in SMAW (stick) welding with low-hydrogen electrodes if the amperage is set too low for the electrode diameter.

The unfused metal must be removed by grinding. If the remaining weld is insufficient, the area is re-welded with higher heat input and correct technique to ensure proper fusion at the toes. Simply grinding the surface smooth is not a valid repair if the underlying metal is not fused.