Modeling Study On Flow Field Optimization And Transient Behavior Of Ladle Change For Five-strand Bloom Tundish

The failure rate of flaw detection for inclusion reason was a bit high in casting process offive-strand bloom caster in a certain steelmaking plant, especially in the fraction of the No.3strand significantly higher than other four strands. As well as the internal quality of billet defectsduring unsteady casting was obviously higher than steady casting. For the five-strand bloomtundish, the mathematical and1:3physical models were established to investigate the effect offlow control devices on flow field, temperature field and inclusions removal during the period ofsteady-state. The result of mathematical simulation was verified by water modeling experimentto optimize the structure of the flow control devices. Besides, multiphase VOF （volume of fluid）model was applied to investigate the steel-slag-air three phases fluid flow in the tundish duringthe change of ladle, in which the steel-slag-air interfacial behaviors in filling process wasfocused on. The results show that:（1） The flow control devices of prototype tundish is unreasonable, the similarity among thefive strands is poor and the difference of response time and average residence time of strands isgreat. Also, the short circuit flow in No.3strand is obvious and the dead zone around the No.5strand is larger.（2） Through using optimized baffle, the flow field is greatly improved, the response timeand average residence time of No.3strand have been significantly prolonged, and the shortcircuit flow has been eliminated. At the same time, the distribution of molten steel flow isreasonable, and the similarity has been enhanced remarkably among the strands. Furthermore,the turbulent energy and the velocity of molten level can be distinctly decreased in impact zoneby using type B （round shape with eaves） turbulent controller.（3） Case B4is ascertained as the optimum scheme of flow field optimization. Afteroptimization, the average residence time of tundish has been increased to595.0s, the plug zonevolume fraction is raised by4.8%, the dead zone volume fraction is decreased by0.2%, and theresponse time and average residence time of No.3strand are prolonged42.5s and264.6srespectively. Besides, the overall average standard deviationS Nof tracer concentration foreach strand is only0.0057and has been decreased by75%. The similarity among the strands hasbeen greatly improved. Meanwhile, the temperature field is greatly improved in the tundish，thetemperature dropping of molten steel reduces by20.3%, the maximum temperature differenceamong the five strands decreases from3.7K in A0to1.4K. Moreover, the inclusion removal rateof diameter size in10¹00μm increases by0.9%⁹.3%correspondingly, and reach up to94.4%for the size of100μm.（4） At the end of emptying process, the liquid steel level is dropped to the lowest position,the molten steel flow is not obvious except the region of around the stopper rods, and thesteel-slag-air interface is smooth and clear, the mutual entrapment phenomenon of multiphase is not appeared. During filling process, the stirring of molten steel flow is strong, and thefluctuation of liquid level is fluctuated with two wave crests in front and back of the nozzle.Simultaneously, there are the entrainment phenomenon of steel/air and steel/slag and exposedmetal. The maximum exposed area of steel is252cm²at4.0s of the filling time, and thengradually reduces.（5） While all else the same in the filling process, the exposed area increases with theincreases of injection flow rate of nozzle. However, it can be completely avoided when the flowrate is1.5Q_steady. With the increasing of nozzle depth, the height difference of steel level slowlyincreases. And the exposed area is relative minimum in310mm. After optimization by case B2,the two wave crests are eliminated in the impact zone, and the fluctuation of the steel level isgreatly weaken，also the exposed surface of molten steel is not occurred. The exposed areagradually increases with the increasing of ladle change time, and increasing amplitude is notobvious with the time of2min and3min.While the ladle change time adds to4min, the exposedarea is maintained at500cm²above during stability period in filling process, which is far morethan150¹70cm²those time of2min and3min. In addition, the level of molten steel drops toolow, which led to aggravate steel/slag entrapment through the diversion hole in the filling process.The nozzle is protected by using slag breaker at the start of recasting, which can avoid largeamounts of slag dropping into bath of tundish.