UBC Theses and Dissertations
Measurements of the effective thermal conductivities of granular solids and oxide melts under shear strain by a new technique Cao, Feng
Heat transfer is fundamental to pyrometallurgy processing and the viability of new or existing processes often hinges on achieving high rates of heat transfer into the material being processed. Slags are a feature of extractive and refining processes and heat transfer through liquid slag is an important part of many operations, for example utilizing energy from port-combination in the EAF for scrap melting. However, relatively little work has been directed at characterizing heat transfer through non-stagnant slag layers. Heat transfer through the slag includes conduction, radiation and, with stirring, advection due to flow within the slag. The combined effect of all three mechanisms is convection but it is often more convenient to combine all three as an effective thermal conductivity. A review of methods currently employed to measure the conductivity of stagnant oxide melts, which involve numerical solution of the transient conduction problem for interpretation of the data, did not indicate that any method was readily adaptable to stirred systems nor were they broadly applicable to a wider range of materials such as liquid metals or granular solids. Based on these factors the objectives of the work were to develop a broadly applicable steady-state method for directly measuring the effective thermal conductivity of a variety of materials ranging from stationary granular solids to liquid oxides and metals (with or without stirring) for temperatures covering the spectrum encountered in extractive metallurgy, and to use the technique to determine the extent to which the effective thermal conductivity of liquid oxides might be increased by flow-induced stirring. The methodology covers a range of issues related to the design of the apparatus, interpretation of the raw data and general validation of the technique. For stagnant systems the technique was validated at low temperature against existing conductivity data for both water and silicon oil and good agreement was obtained. For intermediate temperatures trials were also carried out using several granular materials at temperatures up to about 800°C and results were shown to be within the range indicated by some common models for predicting conductivity of packed beds. As expected, the measured conductivity of packed beds increases with temperature due to the radiative contribution to heat transfer. For the limited range of sites tested no clear link between particle site and conductivity was observed. High temperature validation was obtained against existing data for oxide melts (40%CaO - 40%SiO₂ - 20%Al₂O₃ ) and again good agreement was shown. For non-stagnant systems, of the liquids tested, water, silicon oil and oxide melts, only the former showed the rotation-induced flow to have any significant effect on effective thermal conductivity. This was explained by the calculated peripheral Reynolds and Taylor numbers for the systems that indicted laminar flow for oil and oxide melts and turbulent flow for water. For water with turbulent flow and at shear rates up to 0.55 sec⁻¹, the effective conductivity increased by a factor up to ~ 2.7 which is well short of the order of magnitude increase deemed desirable for heat transfer through the slag layer in the rotary scrap-melting furnace. For silicon oil with paddle mixers and at shear rates up to 0.55 sec⁻¹, the effective conductivity showed only a small increase.
Item Citations and Data