This series of articles is intended to highlight industry wide engineering experiences, practical opinions, guidance, and focused advice to practising technologists. The series is written by experts who have made significant contributions to the field of corrosion management. The article in this issue features a contribution from Alan Denney, who discusses weld corrosion problems.
Preferential corrosion of welds – an introduction
Welding of a metal results in a range of different metallurgical structures in the heat affected zone (HAZ); that part of the parent material adjacent to the weld which has not been melted and fused into the weld itself. Where a filler metal is used in welding (in the form of a welding rod or wire), such as in manual metal arc welding (MMAW), metal inert gas/metal active gas welding (MIG/MAG welding), flux cored arc welding (FCAW) or submerged arc welding (SAW) the weld metal will have an inherently different chemical analysis to the parent metal. This is because the parent metal has a wrought structure, with properties coming from mechanical working and heat treatment whereas the weld metal has a cast structure which achieves its strength mainly due to its additions of alloying elements.
When a weldment is exposed to an electrolyte (a conductive corrosive medium) one part of the weldment may corrode preferentially over another part, resulting in the weld being preferentially attacked. In real life situations the preferential attack may be in another discrete area such as the heat affected zone or even some distinct part of the heat affected zone which has a particular metallurgical structure. Where preferential attack occurs it can be rapid, difficult to detect by conventional monitoring methods, and potentially dangerous.
Galvanic corrosion
The corrosion may be caused or enhanced by galvanic action, galvanic corrosion being the electrochemical process where there is a different natural electrical potential between two metals or two different metallurgical structures, resulting in one part of the weldment being the cathode and another the anode of a corrosion cell. It is the anodic area which corrodes, but the cathodic reaction drives the corrosion process. The potential difference may be relatively small but if the parent metal is the cathode and the weld is the anode then there is a large cathodic area to drive the corrosion reactions and rapid weld metal dissolution can result.
Galvanic corrosion is a clear risk where a weld is made between significantly different metals, such as in welds between carbon-manganese steels and duplex stainless steel when exposed to a corrosive environment, but problems can also occur where there is less obvious difference in materials.
Preferential corrosion of welds in seawater lines
One well-known example of preferential corrosion is internal corrosion of pipelines transporting seawater. These are commonly used for water injection into oil wells. The corrosion occurs within the welds at the pipe joints made of carbon-manganese steel welded using welding consumables which would be considered as ‘matching’ in properties (albeit there are the wrought/cast structural differences mentioned above). One means of avoiding the corrosion is to reduce the oxygen level of the seawater to low levels of parts per million (which is extremely challenging) and/or to control the as-deposited composition of the weld metal to increase its cathodic potential thus minimising galvanic potential differences. This involves careful selection of the welding consumables, based on consideration of composition, but also the results of sophisticated corrosion testing of trial welds. A further solution is to install the line with a continuous internal plastic liner in the pipe, thus preventing seawater contact with the pipe or welds surfaces.
Preferential attack in welds in stainless steel
Stainless steels are cathodic when in combination with other metals such as carbon steels or aluminium alloys and hence unaffected by corrosion when in combination in a corrosive environment, but they are not immune to preferential corrosion effects. One example is when austenitic stainless steels are welded (to themselves), and exposed in a corrosive environment. They may then suffer from what is commonly termed ‘weld decay’. If this occurs in current times it is a consequence of an error in not selecting a modern resistant grade of stainless steel.
Stainless steels owe their corrosion resistance to a layer of chromium oxide a few microns thick which forms a passive protection layer. If breached and in a corrosive environment this layer does not repair itself and localised corrosion will occur. In their original form the austenitic stainless steels (the familiar 300 series of stainless steels) contained carbon up at levels of up to 0.08%. This carbon content and its distribution in the matrix becomes significant in welding the steel when ‘sensitisation’ can occur.
Sensitisation is due to the carbon preferentially combining with the all-important chromium to form chromium carbide which forms along the grain boundaries. This depletes the areas adjacent to the grain boundary of the chromium and this is no longer available to form the passive protection layer. In a corrosive environment, local corrosion cells will form, corrosion occurs around the grains and they can detach from the surface (Figure 3). This situation occurs within a range of temperatures experienced in the heat affected zone of the weldment.
This problem is readily avoided by selecting the low carbon stainless steels denoted by the L designation (e.g. 304L or 316L) which use modern steelmaking techniques to achieve carbon contents of less than 0.03%. There is a further material selection option which is to use the so-called stabilised grades of stainless steel, such as type 347, in which the carbon is ‘stabilised’ (i.e. chemically combined) by the addition of niobium.
In a component already fabricated from the higher carbon (unstabilised) grades it can be prevented by suitable post weld heat treatment.
In conclusion, there are numerous other examples of preferential corrosion of welds and this topic is well covered in web articles and technical papers which the reader can discover by performing online searches.
Welcome to the world of the welding engineer who has to select the welding and heat treatment parameters based on the materials and their application.
Alan Denney, Consultant.
References
1. Corrosion of welded components in marine environments (April 2003). Stuart Bond, TWI Ltd. Prevention and Management of Marine Corrosion, London, 2-3 April 2003. Lloyds List Events. (available on the TWI website https://www.twi-global.com/technical-knowledge/published-papers/)
2. Good welding practices for stainless steel. Presentation by Glenn Allen, TWI Ltd. Available to members on the TWI website.
Figure 1: An example fusion weld showing the heat affected zone.
Figure 2: Preferential weld metal corrosion in carbon steel (copyright TWI Ltd reproduced with permission). Source: Reference 1.
Figure 3: (copyright TWI Ltd reproduced with permission).
Source: Reference 2.
