ering their extraordinary chemical effects and prospective benefits for engineering applications. Even more enticing is that RSP magnesium alloys also display enhanced resistance to corrosion18 but again the fundamentals are not yet clear.
No less graphic is the effect of RSP on the oxidation resistance of high-temperature alloys, including superalloys; the oxide scales that normally develop on such alloys then become more adherent and protective. With RSP stainless steels,19 the fine grain size provides extra paths for chromium atoms to reach the surface quickly and so maintain a chromium oxide layer. If the chromium supply from the base metal is deficient, iron oxide begins to form in the outer scale and renders it less protective. Because of the grain-boundary diffusion mechanism, oxide phases form to some extent at the outer grain-boundary junctures in continuity with the surface oxide layer, perhaps helping to lock the scale to the base alloy. There are also indications that finely dispersed stable phases, such as hafnium and yttrium oxides or sulfides, may participate in the oxidation resistance. Although adequate understanding of the relevant phenomena is still lacking, the effects can be dramatic. Figure 1120 shows the results of a severe, cyclic oxidation test on two RSP superalloy specimens, with and without the addition of yttrium. It is obvious here that a fertile field of basic research on oxide-scale formation lies ahead, with farreaching technological implications.
The second-phase precipitates in RSP alloys can be refined even further if the rapid solidification occurs under conditions of solute trapping—i.e., when the interfacial velocity is fast enough (whether by supercooling or by sufficiently rapid heat extraction) to prevent equilibrium partitioning between the liquid and the solid at the growth front. If precipitation is then allowed to take place from the resulting highly supersaturated solid solution rather than during the prior solidification process, the particles can be a factor of 10 smaller than in the absence of solute trapping. As indicated by Figure 12,21 solute trapping (k>k0) sets in when the interfacial velocity reaches within an order of magnitude of the diffusive velocity (DL/λ, where DL is the solute diffusivity in the liquid phase and λ is the diffusion jump distance). The solute trapping becomes substantially complete (k≅1) when the interfacial velocity reaches about an order of magnitude greater than the diffusive velocity. It is evident, then, that one of the future thrusts of RSP research lies in the direction of hitherto unexplored alloy systems and compositional regimes that can be accessed by faster solidification rates, thereby opening up new ranges of structure-property relationships. In this approach, the formation of exotic metastable phases (such as the aforementioned metallic glasses and the “forbidden” fivefold structure in the aluminum-manganese system) should become more common in view of the suppressed equilibrium reactions and the greater opportunities for less stable phases to compete in the kinetics of nucleation and growth. Although the emergence and nature