Met. Mater. Int., Vol. 16, No. 3 (2010), pp. 501~506 doi: 10.1007/s12540-010-0601-y Published 26 June 2010 Effect of Refractory Properties on Initial Bubble Formation in Continuous-Casting Nozzles Go-Gi Lee 1 , Brian G. Thomas 2 , and Seon-Hyo Kim 3,* 1 Research Institute of Industrial Science and Technology (RIST), San32, Hyojadong, Pohang-si, Gyeongbuk 790-784, Korea 2 Department of Mechanical Science and Engineering, University of Illinois at Urbana-Champaign, Urbana, IL 61801, USA 3 Department of Material Science and Engineering, Pohang University of Science and Technology, San31, Hyoja-dong, Pohang-si, Gyeongbuk 790-784, Korea (received date: 14 August 2009 / accepted date: 5 November 2009) A water model has been applied to investigate initial bubble behavior using specially-coated samples of porous MgO refractory to simulate the high-contact angle of steel-argon refractory systems with different permeabilities. Air is injected through the porous refractory and travels through many inter-connected pores to exit the surface through “active sites”. An active site is a pore where bubbles exit from the surface of the porous refractory. The effect of refractory properties has been investigated in both stagnant and downward-flowing water. The number of active sites increases with increasing gas injection flow rate, permeability, and velocity of the downward-flowing water, and lower contact angle. Keywords: porous refractory, permeability, bubble, nozzle, continuous casting 1. INTRODUCTION Argon gas is injected into downward-flowing liquid steel through the upper tundish nozzle (UTN), which connects the tundish bottom and slide gate system. The fluid flow in the submerged entry nozzle (SEN) is highly turbulent, and depends greatly on the amount and size of injected gas. Fur- thermore, the flow pattern in the mold is strongly affected by the fluid flow at the nozzle ports. Gas bubbles injected through the UTN could penetrate deep into the mold and become entrapped in the solidifying steel shell [1], where they cause blisters and other costly defects [2,3]. Knowledge and interpretation of the size of bubbles forming in the noz- zle, therefore, are essential for the prediction and understand- ing of multiphase fluid flow behavior and related defects in the continuous casting process. Due to high operating temperature, it is difficult and expensive to directly investigate bubble formation in contin- uous steel casters [4]. Physical water model experiments with transparent plastic walls, therefore, have been employed to gain insight into single-phase fluid flow in steel casting processes [5-10]. For these models, Froude dimensionless number similarity is usually applied due to the nearly equal kinematic viscosities of molten steel and water. Extensive studies [11-16] of bubble formation have been performed on aqueous systems both experimentally and the- oretically. Recently, Wang et al. [17] used water models to study air-bubble formation from gas injected through a porous refractory into an acrylic nozzle with flowing water. The wettability was reduced by waxing the walls, which caused the gas to form large pockets that travel along the wall and break up into many uneven-sized bubbles. With an unwaxed surface, relatively uniform-sized bubbles were formed and detached from the wall to join the liquid flow. Although most previous studies focused on bubble forma- tion from an upward-facing orifice or nozzle, some authors [14,18] have observed that bubbles formed from a horizontal orifice behaved in a highly similar manner to that in stagnant flow. Bai and Thomas [15] developed a correlation to predict average bubble size in both water model experiments and steel–argon systems, as a function of downward water veloc- ity and gas flow rate injected horizontally through drilled holes. There has been no reported study, however, on the effects of refractory properties and major processing param- eters on bubble formation in the nozzle. The present study was conducted to quantify bubble size and distribution during gas injection into downward flowing *Corresponding author: seonhyo@postech.ac.kr KIM and Springer