Mechanisms of interfacial heat transfer and nucleation of liquid steel on metallic substrates

A modified levitated droplet technique and an immersion technique were used to investigate the mechanisms of interfacial heat transfer and nucleation of liquid low carbon steel on metallic substrates. The aim of the experimental study was to examine the effects of process parameters of melt temperatures, gas atmosphere, melt sulfur content, substrate surface texture and melt/substrate interfacial oxide on the interfacial heat transfer during the first fifty milliseconds of contact and the nucleation behaviour of the melt. Levitation experiments were conducted to investigate the effects of melt temperature, melt sulfur content, gas oxygen content, substrate texturing and melt/substrate interfacial oxides on the initial interfacial heat transfer. The melt surface tension prior to contact was measured and the resulting interfacial heat flux determined. Immersion experiments were carried out using an experimental apparatus which approximated the initial roll/melt contacting conditions in the meniscus region of a twin roll caster. Thermal histories of the solidifying shell and the substrate were recorded and used to elucidate the mechanisms of interfacial heat transfer and nucleation. Experimental results demonstrated that the melt/substrate wettability strongly affected the interfacial heat transfer during the first 50 milliseconds of contact. Peak values of the interfacial heat flux occurring in the first 50 milliseconds of contact were found to increase with increases in melt temperature. This resulted from the increased driving force for heat transfer and the improved wetting due to the extended time in which a liquid/solid contact existed. The melt surface tension was strongly influenced by the gas atmosphere oxygen content and the melt sulfur content. Reductions in the melt surface tension were measured with increased gas oxygen and let sulfur contents. Reduced melt surface tensions resulted in increased gas oxygen and melt sulfur contents. Reduced melt surface tensions results in increased peak interfacial heat fluxes due to increases in the surface nucleation density through improved melt/substrate wetting. Surface nucleation patterns on smooth substrates were random, with textured substrates exhibiting patterns correlating to the ridge pitch. For both substrates, the interfacial heat transfer increased with nucleation density. Peak interfacial heat fluxes were greater for the smooth substrate due to the larger surface area available for direct contact. Large increases in the peak interfacial heat influx were found when oxide material deposited at the melt/substrate interface melted. This material was found to be manganese, silicon and iron based oxides with some sulfur present. The melt/substrate wettability was enhanced when these oxides melted, resulting in greater intimate melt/substrate contact area and higher heat fluxes. The influence of the melt/substrate wettability on the heat transfer was characterised by the resistance to heat transfer across the interface. Reductions in the minimum interfacial resistance were observed when the interfacial oxide/sulfide layer had melted. The interfacial resistance then increased sharply associated with the recrystallisation of the interfacial oxide/sulfide layer. Correlation of the peak heat flux with the minimum interfacial resistance yielded a peak heat flux of 581MW/m² for an interfacial resistance approaching zero. Estimated contact angles were between 50° and 110° as affected by oxide accumulation at the interface and the melt sulfur content. Each melt/substrate contact point also provided a site for nucleation of solid from the melt. The melt/substrate wetting was found to control the cooling rate of the melt surface during the initial contact. The degree of undercooling of the melt was found to be a linear function of the cooling rate approaching nucleation. Nucleation cooling rates in the range of 10⁴ to 10⁶ K/s were measured. Model predictions indicated that local cooling rates of up to 10⁷ K/s could be achieved and nucleation undercoolings of up to 1000 K were possible at the points of intimate melt/substrate contact. A nucleation undercooling of 975 K was measured using a 100μm optical fibre arrangement. According to the classical heterogeneous nucleation theory, improved wetting is expected to reduce the energy barrier for nucleation. Since the overall nucleation rate is controlled by both the rate of cluster formation and the rate of atom transfer to the nucleus, increasing the cooling rate above a critical level is expected to reduce the nucelation rate. The measured experimental data allowed the melt undercooling and the time for nucleation of the first solid phase to be determined and compared to the theoretical predictions. It was shown that under such rapid cooling conditions, a metastable glass phase would be likely to form initially without the evolution of latent heat. Subsequent transformation into a crystalline material would release latent heat and cause temperature recalescence. If the heat transfer rates remain high, the recalescence temperature would be substantially lower and an unique solidification path with the possibility of bypassing the δ to γ transformation in low carbon steels would be possible.