A D V A N C E D M A T E R I A L S & P R O C E S S E S | S E P T E M B E R 2 0 2 2 2 5 initiated from an etch pit by some kind of stress corrosion cracking mechanism. It seems clear that the etch pit is far too small to constitute a stress raiser in components that have been generously designed with a safety factor as high as five. Furthermore, the steels almost exclusively used for such purposes are those produced by vacuum arc remelting (VAR), a process that the author has shown to result in huge bifilm cracks occupying, at times, a significant percentage of the area of the ingot[2,5]. At the present time, we must accept that cracks pre-exist in most metals (with interesting and important exceptions mentioned later) as a result of current unsatisfactory melting and casting processes[1,2,5]. INVASIVE CORROSION: STRESS CORROSION CRACKING (SCC) The network of bifilm cracks through some metals can, as discussed for pitting, lead to deep penetration of a corrodent into the interior of the metal. Only low tensile stress is generally required to initiate extensive internal corrosion which will lead to tensile failure[6]. This is probably the result of corrodents being drawn into the narrow “air gaps” of bifilms; partly by the tensile stress, and partly by capillary attraction. Extra rapid corrosion would be expected at localized highly stressed regions, possibly suffering plastic flow, in the ligaments separating bifilms. As ligaments fail, one by one, the tensile stress would progressively open the crack, sucking in fresh supplies of corrodent. As is to be expected, the process is arrested by eliminating the tensile load, and no cracking is observed when a compressive load is applied[6]. Corrodents are typically aqueous solutions of chlorides or sulfides. However, paradoxically, water is a surprisingly good promoter of SCC for a number of metals. SCC by such a seemingly harmless fluid as water illustrates the counter-intuitive nature of SCC. As such, it represents a challenge to explain its surprising behavior. Two examples of SCC by water are taken. Japanese workers[7] studying steel for the nuclear industry found different crack rates in the Ni-base alloy 690, in simulated primary circuit water of pressurized water reactors (PWRs) that they were unable to explain. The cracks were associated with TiN and M23C6 inclusions on grain boundaries. This wrought alloy had relatively slow crack growth rates because of its bifilms both partly closed by the plastic working, and their general alignment with the working direction. Boeri and Martinez[8] observe that the strength of austempered spheroidal graphite (ductile) cast iron is impaired by contact with water, but failure becomes practically instantaneous when applied to the metal stressed to its yield point. The failure is especially fast for the cast iron because the bifilms will tend to be at right angles to the surface, and the numerous second phases and inclusions from impurities which decorate the bifilms will aid chemical dissolution, and release hydrogen gas. The hydrogen would be expected to pressure the water-filled network of cracks. In addition to pressurization by the hydrogen, the naturally entrapped gas (probably mainly air), in the central bifilms surrounded by the advancing liquid, will be compressed, adding to any hydrogen pressure and the applied tensile stress. Boeri and Martinez found the SCC failure was completely eliminated by first applying a mineral oil to the surface. We can assume the entrances to the bifilms become sealed with the oil, preventing the ingress of the water. It seems possible that capillary attraction would draw at least some oil deep into the matrix, despite its somewhat higher viscosity. However, the lack of any failure points to its inertness, in contrast to the action of water to produce hydrogen. In summary, once again, in this proposed mechanism, the SCC process does not create the cracks that are typical of the process and that lead to failure. The population of cracks is formed in the liquid state during casting. The cracks therefore pre-exist as bifilms throughout the solid metal and can suck in corrodent when the bifilms are opened by one or more of the following factors: applied strain; inclusion precipitation; gas pressurization by capillary pressure of the corroding fluid; or gas pressurization from hydrogen evolution. Regrettably, space does not permit a discussion of the central role of bifilms in hydrogen embrittlement. The reader is referred to Reference 2. ELIMINATING BIFILMS There are a number of quite different technologies for eliminating bifilms[1,2]. Several suggestions are summarized here. Prior to casting, liquid metals are first required to be treated to reduce their huge bifilm populations (accumulated from years of recycling and/or poor melt handling). Techniques include filtering, flotation, and sedimentation[1,2]. Gravity-poured castings have designs of filling systems, only recently developed[1] that reduce air entrainment in the liquid metal during pouring, and provide conditions for filling of the mold cavity at the correct velocity, free of air bubbles and surface turbulence. The ultimate filling system technique is counter-gravity casting, in which the whole melting and casting process is designed so that all fall of the metal under gravity is eliminated, and the rate of vertical rise in the mold cavity is controlled to avoid any danger of reintroduction of bifilms[1,2]. The wholesale conversion of foundries to counter-gravity casting is strongly recommended. Some alloys, when molten, have surface oxide films that are liquid. Others can be tailored to be liquid by careful alloying. The folding over of the liquid surface, or impingement of drops and splashes, now results in liquid-to-liquid impingement of oxides. Bifilms cannot be formed. Only oxide droplets are formed as spherical inclusions which can float out rapidly. The admirable toughness and failure resistance of the 13%Mn Hadfield steel is an example. Its properties are usually attributed to its austenitic structure, but seems more probably the result of the low melting
RkJQdWJsaXNoZXIy MTYyMzk3NQ==