10 Common Weld Defects, Their Causes and How NDT Finds Them

The 10 most common weld defects, from cracks and lack of fusion to porosity and undercut: causes, risk, prevention and which NDT method finds each one.

Welder in protective gear working with welding equipment in an industrial workshop
Photo: Hoang NC / Pexels

Key takeaways

  • Most weld defects come from a small number of causes: poor joint preparation, wrong parameters, contamination and unsuitable technique.
  • Surface defects are found with VT, PT and MT; internal defects need UT, PAUT, TOFD or RT.
  • Planar defects such as cracks and lack of fusion are the most dangerous and are best found by ultrasonic methods.
  • Prevention through qualified WPSs, trained welders and in-process inspection costs far less than repair.

Every inspector sees the same weld defects again and again. Knowing what causes them, how dangerous they are and which NDT method finds them is the fastest way to reduce repairs and to agree realistic inspection plans. Defects are classified in ISO 6520-1, and their acceptance limits for steel welds in ISO 5817. Here are the ten you will meet most often.

1. Cracks

Cracks are the most serious weld defect because they are sharp and can grow under load. Hot cracks form during solidification, often from high sulphur or phosphorus, a deep narrow bead or high restraint. Cold (hydrogen) cracks appear hours later in hardened zones when hydrogen, a susceptible microstructure and stress combine. Found by: MT and PT at the surface, UT, PAUT and TOFD inside. Prevented by: correct preheat, low-hydrogen consumables, controlled heat input and suitable joint design.

2. Lack of fusion

The weld metal does not fuse with the base metal or the previous pass, leaving a flat, crack-like gap. Typical causes are low current, fast travel speed, wrong torch angle or oxide on the joint faces. Found by: ultrasonic methods, especially PAUT with angles aligned to the bevel; radiography can miss it. Prevented by: correct parameters and welder technique, and clean bevels.

3. Incomplete penetration

The root is not fully fused through the joint thickness. Causes include a small root gap, a thick root face, low current or poor fit-up. Found by: RT clearly, UT and PAUT, and VT where the root is accessible. Prevented by: accurate joint preparation and root pass parameters.

4. Porosity

Gas pores trapped in the weld metal, isolated, clustered or elongated. Common causes are moisture, contamination, poor gas shielding and wind on site. Found by: RT very clearly, VT for surface pores. Prevented by: dry consumables, clean surfaces and protection from draughts.

5. Slag inclusions

Non-metallic slag trapped between passes, usually from poor cleaning or incorrect bead placement in processes that use flux. Found by: RT and UT. Prevented by: thorough interpass cleaning and good bead sequencing.

Welder in a fabrication shop working to prevent common weld defects
Most weld defects are prevented at the workstation, not found at inspection.

6. Undercut

A groove melted into the base metal at the weld toe and not filled, reducing the section and creating a stress concentration. Caused by excessive current, travel speed or a wrong electrode angle. Found by: VT with gauges. Prevented by: correct parameters and technique.

7. Overlap

Weld metal flows onto the surface of the base metal without fusing, forming a notch at the toe. Usually from low travel speed or too much filler. Found by: VT, and MT or PT to check for an associated notch.

8. Excess reinforcement and poor profile

Too much weld metal or a sharp angle between weld and base metal. It wastes filler and, more importantly, reduces fatigue life. Found by: VT with weld gauges against ISO 5817 limits.

9. Burn-through and excess penetration

Too much heat at the root melts through the joint, leaving a hole or an excessive root bead. Common on thin material and pipe roots. Found by: VT inside the pipe where possible, RT.

10. Misalignment and distortion

Linear misalignment (hi-lo) between parts and angular distortion after welding. Both add secondary bending stresses and can make the joint unacceptable even when the weld metal is sound. Found by: dimensional checks and VT. Prevented by: good fit-up, tack welding and a planned welding sequence.

Which NDT method finds which defect

DefectVTPT / MTUT / PAUT / TOFDRT
Surface cracksPartlyYesYesPartly
Internal cracksNoNoYesDepends on orientation
Lack of fusionNoOnly if open to surfaceYesOften missed
Incomplete penetrationIf root is visibleIf root is visibleYesYes
PorositySurface onlySurface onlyPartlyYes
Slag inclusionsNoNoYesYes
Undercut, overlap, profileYesHelpsNoPartly

For a detailed comparison of the volumetric methods, read PAUT vs TOFD vs RT. Acceptance levels B, C and D for these defects are explained in our guide to ISO 5817 quality levels.

Practical tip: track defects by type, welder and WPS. A rising repair rate for one defect type usually points to one cause that can be fixed in a day.

Repair rate too high?

Our welding engineers and inspectors find the root cause, adjust procedures and set up in-process inspection that stops defects before NDT.

Talk to a welding engineer

Frequently asked questions

What is the most dangerous weld defect?

Cracks and lack of fusion, because they are planar and sharp. They concentrate stress and can grow in service, especially under fatigue loading.

Is porosity always a reason to reject a weld?

No. Standards such as ISO 5817 allow limited porosity depending on the quality level. Only porosity above the limits must be repaired.

Which standard classifies weld defects?

ISO 6520-1 classifies geometric imperfections in fusion welds, while ISO 5817 sets acceptance limits for steel welds by quality level.

Written by

Nicolae Ciubotaru

Let’s build together.

Your partner in welding, inspection and engineering. Tell us about the scope, materials and schedule, and a welding engineer from our team will get back to you.