Papers by Author: David J. Young

Abstract: When an alloy component is selectively oxidised but cannot reach the surface quickly enough to form a scale, then internal oxidation results. In this process, a gas phase oxidant dissolves in an alloy and diffuses inwards, reacting with a dilute solute metal to precipitate metal oxide or carbide, etc. Penetration kinetics are parabolic, the rate being controlled by oxidant diffusion and the concentration of reacting metal. Rates are predicted from classical oxidation theory on the basis that the reaction product is exceedingly stable, no solute metal remains in the reacted alloy, and oxidant diffusion is via a solvent metal matrix. This paper is concerned with situations where these approximations fail: the development of low stability precipitates and the growth of elongated precipitates which allow interfacial diffusion of the oxidant. Effects on the rates of internal oxidation are discussed.

Abstract: Wagner’s 1959 diffusion model of the internal oxidation process provided a method of predicting the rate at which a binary alloy was penetrated by dissolved oxygen as it precipitated the more reactive (but dilute) alloy component. Parabolic kinetics were predicted to depend on oxygen permeability in the unreacted alloy solvent and also, in cases where the reactive component was sufficiently mobile, the diffusion coefficient of the latter. The model has proven very successful, but is restricted to single oxidant-binary alloy systems, in which the precipitated oxide has extremely low solubility. This paper reviews recent results on a number of internal precipitation processes which cannot be described with the Wagner theory. These include formation of low stability carbides and nitrades; internal precipitation driven by multiple oxidants; the templating effects of prior precipitates on subsequently formed corrosion products; cellular precipitation morphologies; internal interface diffusion effects; volume changes in the reaction zone and the effects upon them of simultaneous external scaling.

Abstract: Iron and model alloys containing 2.25, 9, and 20 wt% Cr, 2, 4 and 6 wt% Al, 1, 2 and 3 wt% Si, and dilute Fe-Si-Al ternaries were reacted in dry and wet Ar-CO2 gases at 800°C. External oxide scales grew on Fe according to fast, linear kinetics in dry CO2. Additions of H2O accelerated the reaction until steady-state parabolic kinetics were achieved. High Cr content alloys developed slow-growing chromium-rich oxide scales. Dry CO2 mixtures produced faster rates than wet gas mixtures. Lower Cr alloys developed thicker iron oxide scales, featuring cavities, cracks and poor adherence, and sustained internal oxidation. The presence of H2O led to even higher oxidation rates. Aluminium additions to iron of up to 4 wt% provided no protection, but instead caused internal oxidation. A level of 6 wt% significantly slowed oxidation by forming a continuous Al2O3 layer. Silicon additions had little effect, apart from promoting internal oxidation. However, simultaneous alloying with aluminium and silicon strongly depressed corrosion rates. The effectiveness of different alloy additions is discussed, along with the effects of water vapour and carbon activities, in the context of oxyfuel combustion technology.

Abstract: Fe-Cr binary model alloys (Cr: 25 wt %) with additions of 0.09 wt% lanthanum were subjected to cyclic oxidation experiments at 700oC. All model alloys were exposed in five different gases; Ar-20O2, Ar-20O2-5H2O, Ar-5O2-20H2O, Ar-10H2-5H2O (pO2 = 3.64 x 10-22 atm) and Ar-10H2-20H2O (pO2 = 7.37 x 10-21 atm) all in volume %. Very low weight gains were observed in all gases of Fe-25Cr-0.09La. However, breakaway oxidation occurred on La-free alloy experienced increased weight gain in Ar-5O2-20H2O due to formation of iron-rich oxide. The addition of La (<0.1 wt%) to the Fe25Cr retarded the growth of iron-rich oxide in Ar-5O2-20H2O.

Abstract: Dusting of iron results from partial disintegration of a cementite scale which grows on the metal surface during reaction with carbon-supersaturated gas. Scaling kinetics are shown to be consistent with diffusion of carbon through the cementite. Further diffusion of carbon into the iron supersaturates it to a very high degree. Dusting of nickel and austenitic alloys leads to no carbide formation. Instead graphite grows into the metal, supported by diffusion of dissolved carbon to growth sites. Variations in rate with alloy iron content reflect the known effects of iron on carbon solubility and diffusivity. Alloying with copper also changes coking and dusting rates, although copper does not affect carbon permeability. The effect is shown to be due to interaction of copper with graphite nucleation sites.

Abstract: Water vapour interacts with growing chromia scales in several different ways. Formation and volatilisation of Cr2O2(OH)2 is shown to account quantitatively for chromium loss from thin alloy foils reacted with air-steam mixtures over periods of 103 h. In the shorter term, water vapour is shown to refine the grain structure of Cr2O3 scales grown on Ni-25Cr. Scaling kinetics are at the same time accelerated by an additional, larger contribution to diffusion by a grain boundary species, either OH- or H2O. A slight increase in scaling rate observed at low water vapour partial pressures in H2/H2O gases is thought to be due to hydrogen doping.

Abstract: Iron and nickel, model alloys of Ni-Cu and Fe-Cr, and commercial heat resisting alloys were exposed at 650-680oC to flowing CO-H2-H2O gases which were supersaturated with respect to carbon. All ferritic materials, including chromia and alumina formers, developed a coke deposit of carbon nanotubes, the growth of which was catalysed by nanoparticles of Fe3C. Austenitic materials formed graphite filaments and clusters in association with nanoparticles of austenite. Graphite cluster formation was suppressed by alloying copper with nickel. The sensitivity of coking kinetics to alloy copper content was consistent with a mechanism involving graphite nucleation within the subsurface metal. Chromia forming alloys resisted dusting until damage to the scale could no longer be repaired by Cr2O3 regrowth, and carbon gained access to chromium – depleted metal.

Abstract: Although not standard, Forming Limit Diagrams, FLD’s, are used throughout the automotive industry as a preliminary tool to determine if a sheet metal forming process is capable of forming a good part. FLD’s show a limited range of strains on the diagram; typically the range is 0 to 1 on the major strain axis. A new rapid prototyping process called Single Pont Incremental Forming, SPIF, experiences strains over 3. As FLD’s do not typically cover that level of strain, a new method for developing FLD’s is needed. Such a method is proposed in this paper. Research has been conducted with five different shapes, formed using Single Point Incremental Forming. The part shapes utilized contain the most common combinations of angles and curves observed in formed sheet metal products. The strains encountered in forming each of these parts are measured and the strain data is then plotted on the same FLD. These new FLD’s can then be utilized as a predictive tool for engineers to determine if their design can be produced using the SPIF process.