Viscosity is one of the most important physical properties of slags, which needs to be controlled to improve the efficiency of operation of a wide range of industrial processes. The present investigations deal with viscosity measurements in the molten metal oxides as applicable to the pyrometallurgical and power generation industries. The effect of temperature and composition on the viscosities of the silicate melts was the topic of primary interest during present investigations. Viscosity measurements were carried out in the systems ‘FeO’ – SiO2, ‘FeO’ – MgO – SiO2, and Al2O3 - ‘FeO’ – SiO2 at metallic iron saturation conditions; and the CaO – K2O – SiO2 system during the course of the present program. Upon critically reviewing the existing literature in the systems of interest, it was found that there was a serious lack of reliable and systematic experimental data. The compositions selected for measurement in the present investigation will enable the systematic study of the effects of the different components in the melt on the viscosity of the melt. The temperature range was chosen keeping in mind the limiting factors, such as, the liquidus temperatures of the various slag compositions and the maximum operating temperature of the furnace. The high-temperature viscosity measurement apparatus that was recently constructed by the Pyrometallurgy Research Centre at The University of Queensland, was commissioned to operate under controlled atmospheres and used for experimental measurements during the present program. An experimental methodology, inclusive of sample preparation methods, viscosity measurement procedure and post-experimental analysis of samples, was established for conducting successful experiments using the viscosity apparatus. Rotational cylinder viscometry was employed as the technique to measure viscosity, as it was found to be suitable for the range of viscosities under consideration for present investigations. The viscometers were calibrated against standard liquids to determine the equipment constant to be used in converting the raw viscometer data into absolute values of viscosities. The viscosity measurements were typically conducted at 50 K intervals, after equilibrating the sample at each temperature. The sample was quenched from above the liquidus temperatures to analyse the phase and compositional homogeneity of the melt, using the unique quenching mechanism available on the viscosity apparatus. The microstructures of the slag samples after viscosity measurements were studied using Scanning Electron Microscopy (SEM); composition analysis was done using Electron Probe X-ray Micro Analysis (EPMA); and the bulk compositions of the slag samples were determined using X-ray Florescence (XRF) technique, to determine the phase homogeneity and bulk compositions of the quenched samples. This approach provided a way of determining any possible deviations from target compositions and presence of any crystal phases during the course of the experiment, and enabled the reduction of uncertainty associated with the measurements. Slag compositions in the systems ‘FeO’ – SiO2, Al2O3 – ‘FeO’ – SiO2, ‘FeO’ – MgO – SiO2 and CaO – K2O – SiO2 have been selected for investigations to enable systematic studies of the effects of compositional variation to be undertaken. In the ‘FeO’ – SiO2 system at metallic iron saturation, measurements were conducted for four compositions with SiO2 concentration varying between ~31 mol % to ~41 mol %, over the temperature range 1523 K – 1773 K. In the Al2O3 – ‘FeO’ – SiO2 system at metallic iron saturation, measurements were conducted for seven compositions over the temperature range 1473 K – 1773 K. In the ‘FeO’ – MgO – SiO2 system at metallic iron saturation, measurements were conducted for seven compositions over the temperature range 1523 K – 1773 K. Measurements in the CaO – K2O – SiO2 system were conducted for three synthetic slag samples and two natural slag samples for compositions pertaining to biomass ash slags in the temperature range 1270 K – 1873 K. The new experimental viscosity data obtained from the experiments were critically analysed and were found to match well with expected trends in terms of dependence of viscosity on composition and temperature variations. The data obtained ivfrom the present investigations were compared with results reported by other researchers. The present data were shown to be self-consistent and reproducible, the uncertainties in absolute values of viscosity were demonstrated to be significantly lower than most previous studies. The new data enable the individual effects of changing Al2O3, CaO, FeO, K2O MgO and SiO2 concentrations on the viscosities of fully liquid silicate slags to be identified in a systematic way. This systematic approach is particularly useful in resolving the effects of complex interactions in the melts, such as, charge compensation reactions (Me+, Me2+)O – Al2O3. The modified-Urbain viscosity model, one of the predictive models developed at the Pyrometallurgy Research Centre, was used to compare with results from experiments and validate the model parameters. In most cases, the absolute viscosities from present measurements were found to be slightly higher than model predictions, although the slopes of plots of logarithm of experimental viscosities as a function of inverse temperature were found to match well with the predictions by the model. The new data from the present investigations will be used to optimise the modified Urbain model parameters to improve the accuracy of viscosity predictions. The present investigations have resulted in: • Improved procedures for slag viscosity measurements • Systematic measurement of uncertainties in viscosity as a function of geometry of the crucible/bob assembly • New experimental data in a number of important slag systems, data of importance to both industrial applications and the improvement of predictive slag viscosity models.
11
