Thermal‐Mechanical Model Calibration with Breakout Shell Measurements in Continuous Steel Slab Casting

Abstract

Heat-flow and thermal-stress models of continuous steel slab casting are calibrated with detailed measurements of a breakout and applied to predict longitudinal off-corner crack formation. First, a fluid mass balance is applied together with the measured slide-gate position, mold level, casting speed histories to reconstruct the transient events that occurred during the breakout, including the flow-rate and solidification time histories. An efficient one-dimensional (1-D) heat transfer model of the mold, CON1D, is calibrated to match the measured mold heat flux and thermocouple temperatures, with the help of a full 3-D finite-element model. Using these results, a finite-element thermal-stress model of the solidifying shell was able to match the measured shell thickness profiles, and was applied to reveal insights into interfacial gap conditions and other effects on the formation of off-corner longitudinal cracks and breakouts. Introduction Computational models of heat transfer and thermo-mechanical behavior are useful tools to understand quality problems such as longitudinal cracks in continuous casting of steel. Making accurate, quantitative predictions is difficult, however, because many of the phenomena, such as interfacial heat transfer and crack formation, depend on empirical parameters. One way to meet this challenge is to calibrate the models with measurements on the commercial process. A breakout is the ultimate casting defect. This work performs comprehensive measurements of a breakout at Nippon Steel Yawata Works No.2 strand caster, and presents the casting conditions and data as a benchmark for understanding breakout events, and for model calibration. Finally, the breakout is treated as a longitudinal crack, where the conditions that caused it can be estimated with sufficient accuracy to use as a basis for evaluating hot-tear crack criteria. Breakout Analysis The breakout occurred while casting a 252 x 1360 mm slab of plain carbon steel (0.162%C, 0.71%Mn, 0.016%P, 0.006%S, 0.02%Si, 0.039%Al), under generally steady conditions, given in Table 1, 104 minutes after changing heats. Liquidus and solidus temperatures are 1515.4 and 1479.8 oC. The mold was 258mm thick (top) x 900mm long with 9.5mm taper/side. The mold powder had CaO/SiO2 ratio of 1.2, with a flux solidification temperature of ~1160 oC, and viscosity of 0.6P at 1300oC. The oscillation marks averaged 0.37mm in depth and 3.55mm in width. Supplemental Proceedings: Volume 2: Materials Properties, Characterization, and Modeling TMS (The Minerals, Metals & Materials Society), 2012