TWI Knowledge Summary

Design and structural integrity - Corrosion

Corrosion is the degradation of a material arising from exposure to its environment. There are many different forms that corrosion may take but many may be broadly categorised as either progressive material loss or progressive cracking so that, over a period of time, the material may become unfit for the intended service. Common examples include:

rusting of steel
dissolution of metals in acid media
localised pitting attack in chloride media
oxidation at elevated temperatures
degradation of polymers exposed to UV radiation
environmentally assisted (stress corrosion) cracking
corrosion fatigue

Design against corrosion

When designing against corrosion, one of two general philosophies may be adopted:

If attack proceeds fairly slowly and at a reproducible and predictable rate, a 'corrosion allowance' may be built into the design, so that a safe material thickness remains at all points after the anticipated lifetime of the structure. Typical phenomena that may be addressed in this way include dissolution of steel in acid media and elevated temperature oxidation. A limitation of this approach is the accuracy of the predictive tools or 'corrosion models' available and it may be appropriate to perform non-destructive evaluation of the actual rate of corrosion during service or at shutdowns, so that the corrosion model can be refined periodically. Thus the actual useful life of the item may be determined with greater accuracy than via use of a model alone. To assist this approach, the environment may be modified to reduce corrosion rates, where practicable.
If attack may become localised and propagate rapidly once initiated, e.g. pitting and crevice corrosion of corrosion resistant alloys, it is normal to design so that attack can never initiate. This requires appropriate material selection. The simplest approach to materials selection is to rely on previous experience, although it is essential to understand the operating environment, so that changes between past and future service regimes can be accounted for. Where previous successful experience does not exist, materials testing may be required to demonstrate fitness for purpose. Testing typically seeks to recreate the principal components of the service environment, whilst increasing the aggressiveness, to obtain a suitably short test exposure period. This may be achieved by increasing temperature, increasing the concentration of certain environmental species or by forcing electrochemical changes.

A similar approach may be taken when designing against stress corrosion cracking, e.g. chloride stress corrosion cracking of austenitic stainless steel and sulphide stress cracking of carbon steels, where any crack initiation is considered unacceptable due to the rapid rate of subsequent propagation. In this case, three factors contribute to cracking

material type and microstructure,
applied stress and
the environment.

Therefore, avoidance of the problem may be achieved through appropriate material selection or by keeping the stress below the critical level for cracking. Where practicable, the environment may be modified but this is only occasionally possible.

Corrosion fatigue, although similar outwardly to stress corrosion, tends to propagate more steadily and reproducibly than stress corrosion and consequently may be designed against by assigning a lifetime, based on predictions made from experimental data, i.e. in a very similar way to fatigue in the absence of an aggressive environment.

Further information

Corrosion Services from TWI

Other relevant items include -
FAQ: What is the effect of corrosion on the fatigue strength of welded steel structures?

The effect of low H ₂S concentrations on welded steels

TWI: Corrosion, welds and pipelines - An update (July 2000)