No abstract

An investigation consisting of plant trials and water modelling studies has been carried out with the aim of gaining a better understanding of flow characteristics of a blast furnace taphole stream and its effect on the trough refractory wear. It has been found that the entrainment of the blast gas into the taphole is the most likely root cause for splashy taphole stream, which is believed to greatly contributes to refractory wear on the trough in the region above the liquid level and on the trough cover. Based on the findings from the present studies, changes were made to furnace operation and casting practices at BHP Steel's Flat Products Division No. 6 Blast Furnace (BF6) to minimise splashy taphole streams. As a result, the refractory wear on the trough and its cover has been significantly reduced. The trough cover life of BF6 has been improved, from 60 000 to 300 000 tonnes of hot metal.

The aim of this study was to better understand flow characteristics in the taphole stream impingement region of a blast furnace trough, and its effect on localized trough refractory wear. A 1/5th scale perspex model was used, and oil and water were adopted to simulate the molten iron and slag, respectively. Velocities and turbulence intensities in the region adjacent to the trough wall were measured by means of laser doppler velocimetry (LDV). This study highlighted the entrainment of bubbles by the impinging taphole stream, resulting in a buoyancy-driven flow pattern within the trough. The identified buoyancy-driven flow resulted in high velocities and turbulence intensities in the region where maximum refractory wear occurred. Methods for minimizing the influence of the buoyancy-driven flow, and resultant high velocities and turbulence intensities are proposed in this paper.KEY WORDS: ironmaking; blast furnace; trough refractory wear; oil/water modeling; LDV.from the denser phase while traveling along the trough due to the density difference, and flowed into the oil tank via the 'slag' runner. Water flowed through the gap underneath the separating baffle, and then into the water tank. The taphole diameter was 0.0104 m ID, and taphole angle was 10 degree for all the tests, except those examining the effect of taphole angle on the fluid flow in the trough. The width of the worn trough model was 0.265 m at its widest point, with a gradual reduction to the dimension of the standard trough over the impingement region. Experimental conditions are given in Table 1. The choice of the water flowrates in the table was based on the similitude analysis described in the Section 3. MeasurementExperimental program involved two separate systems. The first system involved water/oil to simulate metal/slag within the blast furnace trough. In this setup, characteristics of the three-phase (metal, slag and entrained gas) flow in the impingement region of the taphole stream was qualitatively investigated. In particular, identification of the buoyancy-driven flow and mixing of the three phases was determined. Initially, it was planned to undertake LDV measurements with the water-oil system to quantify the flow pattern. However, due to difficulties in data acquisition, caused by the oil phase adhering to the trough wall, the LDV measurements were not undertaken in the three-phase system. The inability to take these measurements was not a major drawback, given that it was found that the water-only system, which was used for LDV measurements, as described below, behaved essentially the same as for the water-oil system (see Sec. 4.1).The second system, involving water only, was used to undertake LDV measurements, in which, the slag phase was not considered due to the reason described above. The LDV measurements were carried out in a grid (6ϫ10) pattern, beginning at 0.8 m from the taphole. The LDV measurement area is shown in Fig. 1, i.e. 0.45 m along the trough and 0.05 m down the wall. The top of the measurement window was 0.01 m belo...

Unexpected discharge of blast gas in the early stages of castingfrom an ironmaking blast furnace has a signrficant impact on the furnace operations. A physical modelling study was carried out, using both a Hele-Shaw cell and porous bed, in order to understand the mechanism of the gadliquid interface instability during the hearth drainage. The proposed hypothesis that the early discharge of blast gas is due to the occurrence of "viscous fingers" at the gas/liquid interface was tested. The experimental evidence from both the Hele-Shaw and packed bed models proved conclusively that viscous fingering did take place and resulted in the premature release of gas. However, viscous fingering occurred when the drainage velocity was an order of magnitude higher than those used for normal blast furnace operation. These conditions can occur due to local variations in the bed permeability, and further experiments were performed to demonstrate this situation.

Blast gas discharge from the taphole in the course of the blast furnace hearth drainage was experimentally studied using a packed bed cold model. It was found that gas break-through time was strongly influenced by the furnace operating conditions and coke bed structure. Gas break-through time decreases with (a) increasing draining rate; (b) decreasing slag and iron levels in the hearth; and (c) increasing slag viscosity. It increases with an increase in the coke-free layer depth and coke-free space width. Under certain conditions, the gas-liquid interface in the region directly above the taphole becomes unstable, leading to viscous finger formation and subsequently early blast gas discharge from the taphole. The amount of blast gas entrained into the taphole due to viscous fingering, when it occurs, is sufficient to cause a splashy taphole stream.KEY WORDS: blast gas discharge; splashy taphole; viscous fingering; hearth drainage; blast furnace; ironmaking.