The effect of a gas flow field on the size of raceway has been studied experimentally using a twodimensional (2-D) cold model. It is observed that as the blast velocity from the tuyere increases, raceway size increases, and when the blast velocity is decreased from its highest value, raceway size does not change much until the velocity reaches a critical velocity. Below the critical velocity, raceway size decreases with decreasing velocity but is always larger than that for the same velocity when the velocity increased. This phenomenon is called "raceway hysteresis." Raceway hysteresis has been studied in the presence of different gas flow rates and different particle densities. Raceway hysteresis has been observed in all the experiments. The effect of liquid flow, with various superficial velocities, on raceway hysteresis has also been studied. A study of raceway size hysteresis shows that interparticle and particle-wall friction have a very large effect on raceway size. A hypothesis has been proposed to describe the hysteresis phenomenon in the packed beds. The relevance of hysteresis to blast furnace raceways has been discussed. Existing literature correlations for raceway size ignore the frictional effects. Therefore, their applicability to the ironmaking blast furnace is questionable.S. SARKAR, Graduate Student, and G.S. GUPTA, Assistant Professor, are with the

It has been reported in the literature that raceway measurement made during the decreasing gas velocity is relevant to operating blast furnaces. However, no raceway correlation is available for decreasing gas velocity and none of the available correlations either in increasing or decreasing gas velocity take care of frictional properties of the material. Therefore, a systematic experimental study has been carried out on raceway hysteresis. Based on experimental data and using dimensional analysis, two raceway correlations, one each for increasing and decreasing gas velocity, have been developed. Results of these correlations have been compared with the data obtained from literature on the cold models and plant data along with some experimental data. A good agreement exists between the correlations and other data.

When a gas is introduced at high velocity through a nozzle into a packed bed, it creates a raceway in the packed bed. It has been found that the raceway size is larger when it is formed by decreasing the gas velocity from its highest value than when it is formed by increasing the gas velocity. This phenomenon is known as raceway hysteresis. A hypothesis has been proposed to explain the hysteresis phenomenon based on a force-balance approach which includes frictional, bed-weight, and pressure forces. According to this hypothesis, the frictional force acts in different directions when the raceway is expanding and contracting. In this article, the entire packed bed has been divided into radial and Cartesian co-ordinate systems, and the forces acting on the raceway have been solved analytically for a simplified one-dimensional case. Based on the force-balance approach, a general equation has been obtained to predict the diameter of the raceway for increasing and decreasing velocities. A reasonable agreement has been found between the theoretical predictions and experimental observations. The model has also been compared with published experimental and plant data. The hysteresis mechanism in the packed beds can be described reasonably by taking into consideration the direction of frictional forces for the increasing-and decreasing-velocity cases. The effects of the particleshape factor and void fraction on the raceway hysteresis are examined.

No abstract

Prediction of raceway shape and size has always been a challenging task. Based on the previous finding that raceway boundary is an iso-stress boundary, its shape and size has been predicted using a continuum approach. A relation has also been developed to get the stress at the raceway boundary at different blast velocity. Predicted raceway boundary is compared with the experimental raceway boundary/shape and size along with the comparison of theoretical and experimental pressure distribution. A reasonable agreement between the theoretical and experimental raceway shape and size is found. Extension of the work in real blast furnace case is also discussed.