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 3 to fracture which is theoretically predicted. In this author’s opinion, metals could be relied upon not to fail by fracture. TENSILE FRACTURE When a tensile fracture sample is pulled, it might be expected that in the absence of defects of any kind (not even microscopic “Griffith cracks” present) the metal would be incapable of failing by fracture because a crack could not nucleate[2]. This follows from the surprising fact that metals have no intrinsic (built-in) atomic initiation mechanism for cracks, simply because the inter-atomic bonds in metals are too strong. For instance, the widely accepted dislocation pile-up mechanism for the initiation of a crack is now discredited[2]. For metals without any crack-initiating defects, cracking could not occur; so the only failure mode available would be ductile failure, in which the tensile sample would plastically neck down to a perfectly regular conical point corresponding to 100% reduction in area. The fact that perfectly regular smooth cone failures are rarely seen is a clear indication that defects are contributing to the failure process. In experiments with the common aluminum alloy Al-7Si-0.3Mg elongation to failure has been proven to be hyper-sensitive to the presence of the bifilm population of the metal (Fig. 2). For instance, it is common for foundries to achieve elongations to failure of 3 or 4%. These achievements contrast with current efforts to reduce the bifilm content by improved melting and casting techniques, which immediately increase the elongation to 20%. Even so, it is certain that elongations much higher are achievable[2]. Turning briefly to brittle fracture as a failure mode, it seems to this author that brittle failure requires the pre-existence of cracks to avoid the initiation difficulty. Such pre-existing cracks will normally be bifilms from the casting process (bifilms are everywhere!). Once initiated, propagation is now well described and quantified by the field of fracture mechanics, and needs no further discussion in this short article. FATIGUE The fact that fatigue failures continue to be suffered in components that are designed to withstand repeated stresses is, regrettably, a clear illustration of the presence of some as-yet unknown factor. The well-known observation that fatigue failure of cast products needs no initiation stage follows naturally if the component is pre-cracked with a dense population of bifilms. In the case of castings, the bifilms are organized by the growth of the solidifying grains, so that bifilms tend to be trapped between columnar dendrites growing inward at right angles to the surface of the casting. Thus, the bifilms become transgranular, and oriented precisely to maximize their damage potential[1]. For wrought products, rolled or extruded, the grains, together with the bifilms, become oriented along the working direction, so that wrought products take time for the fatigue process to start. Note that the bifilms may also become partially bonded during plastic working of metals, but in general seem to survive well as a result of their inert, highly chemically stable oxide internal interface. The large scatter that characterizes fatigue data is explicable by the variation of bifilm population from melt to melt and casting to casting. No two casts of ingots or lengths of continuous cast steel are expected to be similar. All other aspects of casting such as Fig. 2 — Foundry data sourced from the literature[3] for elongation to failure versus yield strength for Al-7Si-Mg alloys, showing general commercial and aerospace castings and projections for future properties as bifilms become increasingly better controlled.
RkJQdWJsaXNoZXIy MTYyMzk3NQ==