The letter from Stephen A. Knox on stainless steel fuel tanks is fundamentally correct ("Problems with stainless steel tanks," Issue No. 61); however, his understanding of the metallurgy of corrosion needs some updating.
The metallurgically-upsetting sentence stated, "Welded areas tend to be more susceptible to work hardening and subsequent cracking because of the heat effects of welding." The sentence would have been correct if it had omitted the reference to work hardening, since work hardening is actually strengthening of the material caused by permanent deformation and no permanent deformation is involved in the welding process.
However, something occurs in the steel during the welding process to increase the susceptibility to "weld decay" (as it is called). Weld decay is a process that we now understand and that can usually be avoided either by the choice of a correct heat treatment after welding or, more easily, by the choice of a stainless steel specifically formulated to avoid susceptibility to cracking after welding.
First, allow me to explain the mechanism by which austenitic stainless steels (typically type 304 or 18-8, which contain up to 0.08% carbon) become susceptible to cracking after welding, and also to explain how three special alloy formulations allow the problem to be avoided in service. The susceptibility is caused by heating of the metal near the welded joint to a temperature in the range 950° to 1,450° F where a chromium carbide, Cr, can form in the grain boundaries of the steel. At these temperatures, the diffusive velocity of the carbon atoms is very high while that of the chromium atoms is at least 1,000 times slower. The carbide that forms appears preferentially in the grain boundaries because these are high energy, open regions where the carbide phase may easily be nucleated. The rapid growth of the carbide to form a nearly connected series of tiny particles draws carbon atoms from a wide area, but the necessary chromium atoms, due to their slower movement, are drawn exclusively from the area near the boundary. Thus, the formation of the chromium carbide in the boundaries leaves a chromium-depleted zone of the material on either side of the boundary and parallel to it. Steels are alloyed with more than 10% chromium to make them passive to corrosive environments (passivity being due to the formation of a protective oxide layer on the steel). Because of the high chromium content of the carbide (about 94%), the nearby grain boundary regions become depleted in chromium to a degree that they are no longer passive, while the bulk of the grain remains passive. This causes galvanic action, called weld decay or intergranular corrosion, to occur between the chromium-depleted grain boundary region and the higher chromium-containing regions within the grain. The result in this case is that the susceptible grain boundary regions corrode preferentially (intergranular corrosion) and very rapidly, causing entire grains of the steel to fall out of the material. In the case of a fuel tank, the result is that the tank begins to leak slowly at first, but at an accelerating rate.
There are three well-known solutions to the problem of weld decay in austenitic stainless steels. This type of corrosion can be avoided by heating the entire welded structure to about 2,000° F to redissolve the carbides and rehomogenize the chromium. This treatment must be followed by rapid quenching to avoid the reformation of the carbides. For large structures or for weld repair in the field, it is not a good solution.
This intergranular corrosion may usually be avoided by using 18-8 type steels to which a small amount of a very strong carbide has been added, or by specifying a very low carbon level. The strong carbide-forming materials niobium or titanium are added to the 18-8 types of steel, yielding type 347 or type 321, respectively. These elements are added in sufficient quantity to combine preferentially with all the carbon in the steel. Such "stabilized" steels do not require a heat treatment after welding fabrication or weld repair. Type 347 is subject to another type of attack (knife-line attack), so type 321 may be the better material for marine service. Another steel, type 304L, is formulated to contain less than 0.03% carbon, so that even though some carbides do form on welding, the few isolated carbides that appear in the grain boundaries are not destructive for many of the applications where a nearly continuous carbide network would be disastrous.
Harry A. Lipsitt is a professor of materials science and engineering at Wright State University in Dayton, Ohio.