top of page

Different types of Cracks in Welded Structures

If a weld fails by cracking, it is important for a designer to know the reason why.

If the crack can be catagoriced as a Fatigue Crack or a Brittle Fractur Crack, the designer normally is responsible.

On the other hand, the production unit often is reponsible for other types of cracks.

Fatigue Cracks

Cracks in welds

Cracks caused by fatigue normally appear (are detected) after a long period of operation.

These cracks start from a stress concentration point, such as a weld toe or weld root, and grow perpendicular to the direction of the greatest principal stress as the examples shown in the figure.

Fatigue cracks

Fatigue cracks are easy to recognize if the section is sawn open thanks to their typical "beach mark" pattern, also called ”mussel lines”, which begins from the crack initiation point and grows into the material, see figure.


Corrosion attacks are often seen in the crack as the crack grows so slowly that the metal has time to corrode before the crack has grown through the entire thickness of the sheet.


There is often also a sharp boundary between the beach mark pattern and the final (brittle or ductile) fracture, when the remaining material is incapable of transferring the static load.

Fatigue crack

A fatigue crack at a poor start/stop position for a fillet weld between a web and a flange in an I-beam.

Fatigue crack in flange

A longitudinal fatigue crack at the toe of a fillet weld between a web and a flange in a bent I-beam.

Cracks caused by Brittle Fracture

Like fatigue cracks, brittle fracture cracks start from a point of stress concentration and also grow perpendicular to the direction of the greatest principal stress.  How they occur and how to avoid them are discussed in Chapter 10 of the Design Handbook. To summarize, cracks caused by brittle fracture may occur in the case of high strain rate in metals with low impact strength.


Occasionally, brittle fracture can occur even without any external loads if the conditions are unfavorable, such as sudden drop in temperature combined with poor welding quality. 

Brittle fracture cracks can be distinguished from fatigue cracks in that cracks caused by brittle fracture do not have the characteristic beach marks and the surface of the fracture is clearly brittle in nature, normally with a "glittery" appearance. There is no corrosion attack in cracks caused by brittle fracture, unless the inspection takes place a long time after the fracture has occurred. Rust in cracks caused by fatigue stresses, on the other hand, is common.

Cracks caused by brittle fracture often run over large areas in the structure. The structure often has failed totally when the crack is detected. Fatigue cracks, on the other hand, may be of widely differing lengths when detected as a result of their relatively slow growth process.

Brittle fracture in ship

Many spectacular brittle fractures have occurred in bridges and ships. Perhaps most famous are the "Liberty ships" that were mass produced between 1941 and 1945 for shipping general cargo and dry bulk between the United States and Europe during the Second World War. These ships were built in around 40 days by untrained personal using new production methods such as welding instead of riveting.

On January 16, 1943, the Liberty tanker "Schenectady" fractured in half with a loud crack. The ship was completely new and was moored at the fitting dock. The weather was calm.

This fracture was caused by a rapid drop in temperature, brittle steel and thick welded sheets. Upon later inspection of all the ships, it was discovered that there was also a large number of cracks in and adjacent to most of the welds. The notch effects caused by those cracks, together with the low temperatures, brittle steel and high residual welding stresses, were sufficient to cause brittle fracture even though there were no major external stresses. This historical case triggered a large number of research projects and affected the standardization and production of structural steels better suited for welding.

Solidification Cracks

Solidification cracks, also called hot cracks, normally occur in the middle of the weld. They often break through to the surface , but may also appear within the weld and don’t reach the surface.

Solification crack

Solidification cracks are caused by a high impurity content (primarily sulfur) in the metal. The risk of solidification cracks can be reduced by selecting steels with a low UCS - (Units of Crack Susceptibility) value.

Welding methods which involve a high degree of fusing of the parent material increases the risk of solidification cracking.

The risk of solidification cracking is also affected by the solidification process, and therefore, indirectly by the welding parameters as well.

The risk increases in the case of deep, narrow welds.

The risk of solidification cracking can be reduced if:                                                                                      


- Modern steels with low impurity levels are used.
- Welding is carried out using basic electrodes.

- The joints are made wider and less Deep.

- Joint gaps are kept to a minimum.

- The welding sequence is selected to keep tensile stresses to a minimum during cooling.

Hydrogen cracks

Hydrogen cracks (cold cracking, hardening cracking) normally forms in the heat-affected zone (HAZ), but it can also occur in the weld metal itself, both transversely and longitudinally. Crack formation is often delayed and can appear up to a couple of days after welding. Hydrogen cracks do not always reach up to the surface. The cracks are always sharp. If the crack is branched, it is definitely a hydrogen crack. Hydrogen-induced cracks in the HAZ indicate that the fracture toughness in the HAZ is low. This is a serious welding defect as it also acts as a starting point for a fatigue crack or a brittle fracture. 

Hydrogen crack

The following elements increase the risk of hydrogen cracking:

- Presence of hydrogen. Hydrogen comes from moisture and impurities which
  form atomic hydrogen in the arc. Electrodes and granular fluxes can readily
  absorb moisture from the atmosphere, particularly if the humidity is high.

- High residual welding stresses, which occur in constrained structures and
  structures having large joint gaps.

- Brittle microstructure in the heat affected zone HAZ.

How should hydrogen cracking be avoided?

- When welding, all sources of hydrogen must be kept to a minimum. This means that welding should take place in a warm, dry environment with dry
  electrodes. Coated electrodes should be stored in drying cabinets.

- Basic electrodes generally have less of a hydrogen content than rutile electrodes.

- Welding at elevated temperature according to EN 1011-2. The preheating temperature is dependent on sheet thicknesses, diffusible hydrogen
  content, carbon equivalent and heat supply.

- The workpiece can be kept hot after welding so that the hydrogen can diffuse out of the steel easier.

- MIG/MAG, TIG and plasma welding result in lower diffusible hydrogen contents than coated electrode welding (MMA).

- Residual welding stresses can be reduced by having a good fit in the welded joints, without large joint gaps.
  The welding sequence can be adapted so that residual stresses are kept to a minimum.

-  Because of the risk of brittle microstructure, the Carbon Equivalent Value (CEV) should be no greater than approximately 0.4.

Lack of fusion

Lack of fusion, also called cold lapping or cold shuts, is normally oriented along the original groove surface, failing to fuse during welding. Lack of fusion should be regarded as crack-like defects (sharp notch), and so they are particularly hazardous in structures subject to fatigue. Such defects can also occur between weld beads. They frequently do not reach the surface.

MIG/MAG welding involves a particularly high risk for lack of fusion. Using the right welding technique can reduce the risks, but you should be aware of the problem. The Swedish Welding Commission has published guidelines on how to avoid lack of fusion during MIG/MAG welding.

Lack of fusion

The following measures reduce the risk for lack of fusion:

- Good accessibility.

- Avoid large weld beads. It is better with several smaller ones.

- Grind out undercuts.

- Make sure that the joint angle is sufficiently large.

- Clean sheets and groove surfaces before welding.

Read about this and much more in the

Design Handbook for welded steel products 

bottom of page