TY - CONF

T1 - Modelling of a steel ladle and prediction of its thermomechanical behavior by finite element simulation together with artificial neural network approaches

AU - Hou, Aidong

AU - Jin, Shengli

AU - Gruber, Dietmar

AU - Harmuth, Harald

PY - 2019/7/4

Y1 - 2019/7/4

N2 - Refractory linings of industrial vessels are of decisive importance for hightemperature industries. To facilitate the lining design for various material properties and lining configurations, quantitative prediction of thermomechanical responses is of importance prior to industrial application. 192 lining configurations including 10 geometrical and material property variations of a steel ladle lining were defined by six orthogonal arrays for finite element (FE) simulations. The maximum compressive stress at the hot face of the working lining and the maximum tensile stress at the cold end of the steel shell were the selected responses of interest. The impact of geometrical and material property variations on thermomechanical performance of the selected ladle was assessedby analysis of variance (ANOVA) and signal-to-noise (S/N) ratio using 32 lining concept results from one out of six orthogonal arrays. Two optimized lining concepts were proposed accordingly. Their responses were well predicted by a three-layer backpropagationartificial neural network (BP-ANN) model.

AB - Refractory linings of industrial vessels are of decisive importance for hightemperature industries. To facilitate the lining design for various material properties and lining configurations, quantitative prediction of thermomechanical responses is of importance prior to industrial application. 192 lining configurations including 10 geometrical and material property variations of a steel ladle lining were defined by six orthogonal arrays for finite element (FE) simulations. The maximum compressive stress at the hot face of the working lining and the maximum tensile stress at the cold end of the steel shell were the selected responses of interest. The impact of geometrical and material property variations on thermomechanical performance of the selected ladle was assessedby analysis of variance (ANOVA) and signal-to-noise (S/N) ratio using 32 lining concept results from one out of six orthogonal arrays. Two optimized lining concepts were proposed accordingly. Their responses were well predicted by a three-layer backpropagationartificial neural network (BP-ANN) model.

KW - Steel ladle

KW - Finite element simulation

KW - Thermomechanical behavior

KW - Artificial Neural Network

M3 - Paper

SP - 1109

EP - 1116

T2 - Congress on Numerical Methods in Engineering 2019

Y2 - 1 July 2019 through 3 July 2019

ER -