A Study on Hot Tearing in Direct Chill Casting of Al-Mn-Mg Alloys Using a Multi-scale Approach Thesis uri icon

  •  
  • Overview
  •  
  • Research
  •  
  • View All
  •  

abstract

  • Hot tearing is one of the most severe defects encountered in direct chill (DC) casting of aluminum alloys and is intimately linked with the constitutive behavior of the alloy in the semi-solid region. The experimental studies in this field are limited due to the high temperature at which hot tears form and the sensitivity of factors that cause hot tearing to the casting geometry; thus, modeling of the DC casting process at the macroscopic scale has become a popular practice in industry. However, such models do not consider the localization of deformation and liquid feeding between the grains that cause cracking. In this study, a new multi-scale approach to predicting the hot tearing defect in DC casting is proposed. In this approach, a thermomechanical model of the DC casting process has been coupled with a meso-scale coupled hydro-mechanical granular model. The thermomechanical model predicts the evolution of temperature distribution and displacement field for all locations within the DC cast billet for three different simulation cases with casting speeds of 46 mm/min, 56 mm/min, and 66 mm/min. Then, its output is used as input to the coupled meso-scale hydro-mechanical granular model to investigate the effect of different parameters on the tensile behavior of the mushy zone. The results show that the multi-scale approach can successfully simulate hot tearing formation at the regions in the DC cast billet which were found to be susceptible to hot tearing. Moreover, hot tearing susceptibility maps generated by this approach reveal that for the three DC casting simulations performed, the condition for hot tearing formation is favorable above the bottom block and near the center-line of the billet and as the casting speed increases, this region shifts to lower heights in the billet and a greater region becomes prone to hot tearing.

publication date

  • June 2021