Abstract. We present laboratory experiments employing differential scanning calorimetry as well as flow cell microscopy to study the microphysics of aqueous NaC1 and sea-salt solutions and droplets at temperatures below 273 K. The freezing and melting points of ice and other precipitates were determined in NaC1 and sea-salt bulk samples as well as in emulsion samples. Using flow cell microscopy, we have determined the deliquescence and efflorescence relative humidities of NaC1 and sea-salt droplets at temperatures between 249 and 273 K, extending the existing room temperature data to polar conditions. Our measurements suggest that sea-salt aerosols will most likely be liquid most of the time under polar marine boundary conditions. In addition, we show that sea-salt aerosols or seawater spray deposited on the polar ice pack will remain partly liquid down to 230 K, with concentrations of C1-and Br-increasing by more than an order of magnitude upon cooling when compared to normal seawater concentrations. This is likely to enhance the rate at which heterogeneous bromine activation reactions occur in the sea-salt deposits. Such reaction rate enhancements with decreasing temperatures are currently not implemented in chemical models, and might help explain the fast bromine activation and subsequent ozone destruction observed during ozone depletion events in the polar marine boundary layer in spring. , 1994]. Information on the microphysics of sea-salt aerosols and patches of sea salt deposited on the ice pack under these extrcme polar conditions is still lacking. In this paper we present results on the microphysics of NaC1 and sea-salt solution droplets at low temperatures. We have preformed calorimetric freezing experiments with emulsified droplets as well as flow cell microscopy of deposited droplets to investigate the behavior of sea-salt aerosols under polar conditions. In addition, we have conducted bulk freezing experiments to mimic the behavior of patches of sea-salt aerosols and seawater deposited on the polar ice pack. Experimental Procedure CalorimetryWe have employed differential scanning calorimetry (DSC) to study phase transitions occurring in bulk and emulsion samples of aqueous NaC1 and sea-salt solutions. The experimental setup used in this paper has been described in detail previously [Chang et al., 1999;Koop et al., 1999]. Solutions of NaC1 were prepared by adding deionized water to NaC1 crystals (Mallinckrodt, >99%). The NaC1 crystals were dried in an oven at a temperature of about 380 K for at least 2 hours prior to preparing the solutions. Sea-salt solutions were prepared by diluting an aqueous stock solution of a synthetic sea-salt mixture (Instant Ocean ©, Aquarium Systems). Since sea salt is a very hygroscopic salt mixture, the composition of the stock 26,393

A unique in blow sampling system has been applied to a blowing converter to retrieve simultaneously representative bulk metal bath and slag/metal emulsion samples from seven specified positions and every 2 min from start of blow. Full sample datasets from 20 heats have been grouped according to differences in the bulk bath phosphorus removal profiles and analysed with respect to relative refining ability of the slag/metal emulsion and the bulk metal bath. The complexity of the thermokinetic relationships behind the removal of carbon and the transfer of silicon, phosphorus, manganese and sulphur between the metal and slag is highlighted and the metal circulation rate in the emulsion is derived.

No abstract

No abstract

A thermodynamic (equilibrium) model is developed for the BOF process. The predictions of this model show the trend of reactions when the process is considered to be at thermodynamic equilibrium. In the case of a real process, however, some tuning and adaptation becomes necessary to make more accurate predictions. A dynamic model is developed in which the kinetics of scrap dissolution is also incorporated. A comparison of the results of the equilibrium and dynamic models (made with some tuning parameters) reveals that mixing is the prime factor which can alter the course of reaction at any particular instant. Mixing is greatly affected by oxygen flow rate, lance height and the nature of scrap. The understanding of the secrets of process dynamics becomes clearer with this approach, providing a good insight into the process.

Excess slag foam growth is a frequent problem in the BOS process. In the worst case, foam is forced out of the vessel and this phenomenon, commonly called slopping, not only results in loss of valuable metal yield but also in equipment damage and lost production time. In order to minimize slopping, accurate estimation of the foam level inside the vessel is an important part of BOS process control. In the top blown BOS vessel, slopping control is achieved using both static and dynamic measures. The most common implemented technique for dynamic foam height estimation and slopping control is the audiometer system. An alternative method, vessel vibration monitoring, has been investigated as part of the work in a RFCS funded research project called IMPHOS. In order to judge the usefulness of this method, parallel vibration and audio measurements have been carried out on 130 tonne as well as on 300 tonne BOS vessels. The results show that during stable process conditions there is good agreement between the two methods with regard to foam height estimation and, as vessel vibration and audiometry are largely independent of each other, a combination of the two is likely to increase significantly the accuracy of slopping prediction.

No abstract

Selected IMPHOS heats, 1,2 have been used to make observations on decarburising and dephosphorising performance, scrap melting and slag foaming characteristics during BOS refining. If it is assumed that decarburisation takes place solely in the slag/metal emulsion then maximum metal residence time in the emulsion is just under 9 seconds and at peak decarburisation time, the maximum amount of metal in the emulsion is y50% of the total metal content in the converter. To evaluate the effects of changes in slag component chemistry on phosphorus refining it is necessary to account for changes in slag weight, which can change substantially throughout a heat and be significantly different heat-to-heat. Dephosphorising performance depends on the thermodynamic stability of slag phases that are able to take up phosphorus and the distribution of phosphorus between these thermodynamically stable phases. The application of proprietary thermodynamic models such as MTDATA and FACTSage has helped to clarify such events. Skull build-up on the scrap pile is at a maximum when the bulk bath temperature is y1460uC. At this time, the solid scrap and skull component of the bulk bath makes-up just over 60% of all the metal charged to the converter. All scrap and skull is melted out at a bulk bath temperature of y1610uC. The stability of the foamy slag/metal emulsion changes over the period of the blow. Slag height increases with an increase in FeO(tot)wt-% and decreases with a decrease in decarburisation rate and the collapse of the foamy slag.