CFD Calculation of Nonmetallic Inclusion Removal by Bubble Attachment in Continuous Casting Mold

摘要

A water model has been carried out in the present study to observe bubble behavior and particle removal rate depending on gas flow rate, and a mathematical model based on various turbulent-particle model has been carried to predict bubble behavior in casting mold. It has been found out particle density does not have significant effect on particle removal rate in water circulating system. When particles have low contact angle with water, the particle removal rate in bubbling system is similar with in no-bubbling system. However when particles have high contact angle with water, the particle removal rate is depending on gas bubbling rate. These quantitatively measured data can be used in theoretical research of inclusion-bubble collision. In CFD calculation of bubbles, the drag coefficient is very important variable. In present paper, 4 different models are used. One is standard drag coefficient model, which is widely used in non-turbulent solid particle system. Other two model are based on empirical equations, which is found by Brucato et al. and Shimasaki et al. The other model is the efficient Reynolds moded, which is based on eddy viscosity concept. The comparison between CFD calculation and water model result has shown the followings. (a) In the standard drag coefficient model, bubbles float up very quickly. It shows standard drag coefficient is much lower than actual value. (b) In the Brucato model, the bubble behavior is too much sensitive with gas flow rate. In 1.7L/min gas flow rate, the prediction agrees with the water model. However, in 2.5L/min gas flow rate, the bubble floats up too quickly. (c) In the Shimasaki model, basically, the prediction agrees with the water model. However, in 2.5L/min gas flow rate, the bubbles are scattered in the mold. (d) In the efficient Re model, the prediction agrees with the water model. They show small discrepancy between prediction and water model. It is considered that the simplifying process of bubble size distribution makes the discrepancy. The efficient Re model agrees with the water model well. And it is derived by theoretical research not by empirical equation. I consider the efficient Re model is the most suitable model in bubble calculation in continuous casting mold. In near future, I hope to simulate the inclusion removal process by gas bubbling based on this research.