2019 Pittsburgh AISTech

Sessions Speakers Exhibitors Products Press Releases Show Specials Floor Plan Buzz

Development of a Predictive Tundish Open Eye Model Using Artificial Neural Network (ANN) (Room 411)

The presence of "open eyes" in steelmaking ladles and/or tundishes leads to reoxidation, and thus, steel quality deterioration. All the previous models, such as those based on computational fluid dynamics, dimensional analysis, mechanistic modeling and simple regression, allow a fair understanding of the problem; however, they can only provide static recommendations. In the present work, artificial neural network (ANN) models have been used to predict non-dimensional open eye sizes in tundish. These models are accurate, fast and do not deviate with the introduction of new data, which makes them a strong contender to be used in real time in industrial processes.