- Title
- Experimental model systems to investigate factors driving iron ore sintering coalescence
- Creator
- Andrews, Lauren
- Relation
- University of Newcastle Research Higher Degree Thesis
- Resource Type
- thesis
- Date
- 2018
- Description
- Research Doctorate - Doctor of Philosophy (PhD)
- Description
- Sintering is the main process used to agglomerate iron ore fines into an iron rich feed source for blast furnace ironmaking. The process is complex, involving many inter-related variables that cannot be studied in-situ. Studies into iron ore sintering typically involve pilot-scale investigations using sinter pot tests, and a range of laboratory-scale methods that allow specific aspects of the sintering process to be investigated. This study focuses on developing the fundamental knowledge of iron ore sintering coalescence, which encompasses the movement and reshaping of solid-melt-gas systems to form agglomerated regions known as sinter. A laboratory-scale approach was developed to investigate simplified sinter model systems that approximated a sinter granule; which typically comprise a nuclear iron ore particle surrounded by adhering fines material. The model systems comprised pressed sinter mix tablets that formed melts on heating, taking the role of the adhering fines, and tile substrates that represented nuclear particles. The study investigated the impact of altering the chemical composition of melt forming regions and iron ore particle porosity on coalescence behaviour. Chemical reagents were used to control chemical composition, with purchased alumina tiles and manufactured pressed hematite tiles used as substrates. A coal ash fusion furnace was used to sinter the model systems in an air atmosphere, which provided the ability to monitor movement and reshaping behaviour at sintering temperatures. Experimental techniques were developed to study the interaction between sinter mix tablets with both alumina and hematite tiles. Unique to the experimental design were: Creation of suspended model systems, where reshaping behaviour on melt formation provided insight into structural change and coalescence processes that cannot be observed in-situ. The model systems comprised pressed sinter mix tablets, with diameters of 6 and 13 mm, suspended adjacent to pressed hematite tiles, with diameters of 6 and 13 mm, in various configurations. Platinum wire was used to suspend the model systems from the furnace sample carrier. Heating rates of 3 and 7°C per minute were used to a maximum temperature of 1350°C. ; Development of a technique where sinter mix tablet reshaping behaviour on melt formation was investigated on both upward and downward facing tiles, which provided insight into the influence of melt formation sites on reshaping and microstructural changes during sintering. The model systems comprised 6 mm sinter mix tablets placed adjacent to 24 mm square alumina tiles or 28.5 mm diameter hematite tiles, with different porosities. A heating rate of 6°C per minute, to a maximum temperature of 1350°C was investigated. During sintering operations, the preheating stages before melt formation result in movement and shrinkage within the green sinter bed. The suspended model systems demonstrated how the coordination of the sinter bed just prior to melt formation influences the way the bed reshapes on melt formation. Contact between melt forming regions and ore will result in the formation of agglomerated regions as material pulls together and reshapes around ore type particles. While loss of coordination, due to dehydration, calcination and shrinkage, results in the formation of voids. The strength of surface forces was demonstrated by the way sinter mix tablets reshaped around hematite tiles on melt formation, when the tablet was placed below the tile. The extent of reshaping increased with increasing tablet basicity, which was varied between 1.5 and 3.0, and highlighted the need to understand the impact of tablet composition on reshaping. The results obtained from the suspended model systems, while exploratory and qualitative in nature, provided the basis for investigating composition factors driving tablet reshaping over tile substrates, and demonstrated a novel experimental approach for future work. A designed experiment approach was used to develop an understanding of sinter mix tablet composition factors on reshaping behaviour, for tablets placed adjacent to alumina tiles, for both upward and downward facing tiles. The midpoint sinter mix composition was chosen to reflect melt forming regions in sintering, and three composition factors were investigated including: a basicity range of 1.5 to 2.5; alumina in gangue levels between 8.5 and 12.5 mass percent; and gangue levels between 18.8 and 22.8 mass percent. Three tablets were tested for each composition and tablet/tile orientation. Basicity was statistically assessed to be the main factor driving tablet reshaping. Increasing tablet basicity increased reshaping, due to its influence on the amount of melt formed within the tablet, which correlated with FactSage melt mass predictions. The multi-component nature of the sinter mix tablets meant that partial melting occurred once the eutectic temperature was reached; forming complex three-phase systems with proportions and properties dependent on composition and temperature. The experimental technique that explored tablet reshaping on downward facing tiles demonstrated that melt surface forces were high; with all tablet compositions maintaining contact and reshaping over the alumina tiles even though tablet weight opposed reshaping. Mathematical models obtained from literature were used to predict the impact of composition and temperature on tablet properties, which indicated the main determinant of tablet reshaping was the apparent viscosity, of the melt and solid phases within the tablet; as surface forces while high, did not vary significantly between the experimental compositions. Since tablet reshaping at the macro level was accompanied by changes in the tablet microstructures, the study also investigated the resulting tablet microstructures at 1350°C to assess if the driver for macro reshaping reflected changes in tablet microstructure. To assess the impact of composition on the sintered tablet microstructures, a baseline porosity was established for tablet compositions heated to 1100°C, which was after dehydration and calcination, but before melt formation. The baseline porosity at 1100°C did not vary significantly between tablet compositions, so the average across all tablet compositions was used to compare with tablet porosities at 1350°C. Results showed that tablet porosity increased on melt formation, due to melt flow from the tablet over the alumina tile, with the extent dependent on tablet composition and orientation. The results obtained on downward facing alumina tiles were used to statistically assess the impact of composition factors on tablet porosity, since results on the upward facing tiles were confounded by bubbles being expelled from some tablet compositions, due to the free surface. Tablet porosity was found to be influenced by both basicity and the alumina in the gangue, where increasing basicity and decreasing alumina drove increased tablet porosity. The results highlighted the need to consider both macro and micro behaviour when assessing coalescence behaviour, with basicity found to be a significant factor driving both tablet reshaping and porosity; however, alumina in the gangue was also significant in terms of influencing tablet porosity. Investigations into the influence of alumina levels on melt properties and tablet porosity results were inconclusive, and it was postulated that alumina may impact the kinetics of melt formation. Tablet microstructures were also assessed in terms of pore size distributions, which were grouped into four size fractions: 5-50 micrometers; 50-100 micrometers; 100-200 micrometers; and greater than 200 micrometers. Both tablet composition and tablet/tile orientation were shown to influence the resulting pore size distribution. Pore coalescence increased when the weight of the tablet worked with tablet reshaping over the tile, and large pores were eliminated from tablets with low alumina in gangue levels, but only when there was a free surface in the direction of pore buoyancy. The effect of high alumina in the gangue on pore elimination reduced when tablet compositions allowed high amounts of melt to form. The experimental technique developed to assess tablet reshaping on downward facing tiles provided additional information, compared to data from upward facing tiles alone, which allowed factors driving pore elimination from the tablets to be determined. A mechanism for the development of porosity within the sinter mix tablets was proposed, which incorporated pore formation, coalescence and expulsion based on tablet composition factors and tablet/tile orientation. Simplified sinter model systems involving sinter mix tablets and hematite tile substrates, with varying porosities, were tested to ascertain the impact of tile porosity on the driving force for melt flow, tablet reshaping and microstructure. Three tablet compositions, which demonstrated low, mid and high amounts of reshaping over alumina tiles, were investigated. Pressed hematite tiles were manufactured using reagent grade hematite to which different levels of icing sugar were added. The tiles were preheated to remove the icing sugar, leaving pores. Mercury intrusion porosimetry was used to determine the hematite tile porosities, which were; 2%, 5% and 23%, with 23% porosity the tile had pore equivalent spherical diameters in the range 1-10 micrometers. Tablets were placed adjacent to both upward and downward facing hematite tiles and sintered to 1350°C. Changes to the driving force for melt flow and tablet reshaping were demonstrated. Reshaping behaviour of tablets adjacent to dense (2% porosity) hematite were similar to tablets on alumina tiles, being driven by wetting and spreading, with the extent of tablet rounding and reshaping influenced by the amount of melt formed in the tablet. While melt flow from tablets adjacent to porous (23% porosity) hematite was dominated by capillary forces that wicked melt from the bulk of the tablet into tile pores, at the expense of tablet reshaping. Melt flow from the tablets was also investigated using calcium mapping, which confirmed that melt formed within the tablets is mobile, with the extent of melt flow and segregation determined by the amount of melt formed within the tablet and the driving force for melt flow, due to substrate porosity. Melts formed in the sinter mix tablets demonstrated wetting behaviour when adjacent to dense tiles however, when tile porosity was high increases in the apparent contact angle occurred. In the case of the high flow sinter mix, the tablet balled-up due to increases in the contact angle on the porous tile, as the rate of melt formation exceeded the rate of capillary driven melt flow. An analysis of tablet microstructures showed that when capillary forces dominated, melt flow from the tablets increased, compared to wetting and spreading alone. The increased driving force for melt flow caused tablet porosities and pore coalescence to increase, with the larger pores being less rounded. By comparing tablet porosities and pore sizes between different tile orientations, pore elimination was also found to occur on downward facing porous tiles as bubbles were able to move into open tile pores, assisted by melt flow. The findings from this study were used to develop a mechanistic understanding of material coalescence behaviour for the simplified sinter model systems investigated, with potential implications to iron ore sintering discussed. The experimental techniques developed in this study provide avenues for further investigation into iron ore sintering coalescence fundamentals, and recommendations for future work are provided.
- Subject
- iron ore sintering; melt propeties; viscosity; surface tension; FactSage; surface forces
- Identifier
- http://hdl.handle.net/1959.13/1388139
- Identifier
- uon:32722
- Rights
- Copyright 2018 Lauren Andrews
- Language
- eng
- Full Text
- Hits: 1105
- Visitors: 1638
- Downloads: 488
| Thumbnail | File | Description | Size | Format | |||
|---|---|---|---|---|---|---|---|
| View Details Download | ATTACHMENT01 | Thesis | 9 MB | Adobe Acrobat PDF | View Details Download | ||
| View Details Download | ATTACHMENT02 | Abstract | 534 KB | Adobe Acrobat PDF | View Details Download |