Modeling and Optimization of Direct Chill Casting to Reduce Ingot Cracking

Abstract

A successful four-year project on the modeling and optimization of direct chill (DC) casting to reduce ingot cracking has been completed. The project involved close collaboration among private industries, national laboratories, and universities. During the four-year project, 16 quarterly meetings brought the industrial partners and the research team together for discussion of research results and research direction. The industrial partners provided guidance, facilities, and experience to the research team. The research team went to two industrial plants to measure temperature distributions in commercial 60,000-lb DC casting ingot. The collaborative research resulted in several major accomplishments or findings: (1) Surface cracks were shown to be a result of hot tearing rather than cold cracks, as was thought before this project. These cracks form on the surface of a DC cast ingot just above the impingement point of the secondary cooling water jets. The cracks form along dendrite and grain boundaries, where solute and impurity elements are highly segregated. This understanding led to the development of a new technique for determining the mechanical properties in the nonequilibrium mushy zone of alloys and to thermodynamic predictions of the hot tearing propensity of DC cast ingots. (2) The apparent heat transfer coefficient (HTC) at the ingot surface in the water cooling region during DC casting was determined on the basis of temperature measurements in commercial DC casting ingots and an inverse heat transfer analysis. HTCs were calculated as a function of temperature and time, and covered the different regimes of heat transfer expected during DC casting. The calculated values were extrapolated to include the effect of water flow rate. The calculated HTCs had a peak at around 200 C, corresponding to the high heat transfer rates during nucleate boiling, and the profile was consistent with similar data published in the literature. (3) A new method, termed the reheating-cooling method (RCM), was developed and validated for measuring mechanical properties in the nonequilibrium mushy zones of alloys. The new method captures the brittle nature of aluminum alloys at temperatures close to the nonequilibrium solidus temperature, while specimens tested using the reheating method exhibit significant ductility. The RCM has been used for determining the mechanical properties of alloys at nonequilibrium mushy zone temperatures. Accurate data obtained during this project show that the metal becomes more brittle at high temperatures and high strain rates. (4) The elevated-temperature mechanical properties of the alloy were determined. Constitutive models relating the stress and strain relationship at elevated temperatures were also developed. The experimental data fit the model well. (5) An integrated 3D DC casting model has been used to simulate heat transfer, fluid flow, solidification, and thermally induced stress-strain during casting. A temperature-dependent HTC between the cooling water and the ingot surface, cooling water flow rate, and air gap were coupled in this model. An elasto-viscoplastic model based on high-temperature mechanical testing was used to calculate the stress during casting. The 3D integrated model can be used for the prediction of temperature, fluid flow, stress, and strain distribution in DC cast ingots. (6) The cracking propensity of DC cast ingots can be predicted using the 3D integrated model as well as thermodynamic models. Thus, an ingot cracking index based on the ratio of local stress to local alloy strength was established. Simulation results indicate that cracking propensity increases with increasing casting speed. The composition of the ingots also has a major effect on cracking formation. It was found that copper and zinc increase the cracking propensity of DC cast ingots. The goal of this Aluminum Industry of the Future (IOF) project was to assist the aluminum industry in reducing the incidence of stress cracks in DC castings from a current level of 5% down to 2%. This could lead to energy savings in excess of 6 trillion Btu by the year 2020 given full-scale industrial implementation of the results. The project indicates that ingot cracking can be minimized by reducing the casting speed or by controlling the composition of the alloy. These results can be incorporated into industrial applications to achieve significant energy savings.