Modelling and simulation of solidification

• Grant number: 01/01342-5
• Support Opportunities: Research Projects - Thematic Grants
• Duration: October 01, 2001 - September 30, 2005
• Field of knowledge: Engineering - Materials and Metallurgical Engineering
• Principal Investigator: Amauri Garcia
• Grantee: Amauri Garcia
• Host Institution: Faculdade de Engenharia Mecânica (FEM). Universidade Estadual de Campinas (UNICAMP). Campinas , SP, Brazil
• Pesquisadores principais: Maria Clara Filippini Ierardi

Abstract

The main purpose of this project consists in promoting the integration of research groups working, directly or indirectly, with the Solidification of Metals and Alloys, in the following institutions: Department of Materials Engineering: DEMA - State University of Campinas: LINICAMP, Engineering College of São Carlos - EESC, Federal University of Rio Grande do Sul - UFRGS and Technical Institute of Lisbon - IST, and Technology Research Institute - IPT and Chemical Engineering. Faculty at Lorena -FAENQUIL (collaborator institutions). The Department of Materials Engineering at UNICAMP has been working in isolated research projects with each of these institutions during last years, and will be the nucleant agent of this new integrated research initiative. The selected subject of common research will be "modelling and simulation of solidification processing" spread into three different research modules, which will be developed in an integrated way, but according to the research interest and background of each institution. The research module 01 will treat the continuous casting of billets, blooms and slabs of steel, as well as the horizontal continuous casting of non-ferrous alloys and the strip casting of steel. Mathematical models will be developed permitting solidification to be analysed along the continuous caster, for any ingot geometry. In order to allow the developed numerical tools to be more versatile, an analogy between thermal and electrical circuits will be adopted to treat the generalised heat flow problem and a technique will be developed permitting the coupling of mesh elements of different sizes and geometries. The numerical models will be validated against experimental results and other numerical approaches existing in the literature, and the numerical simulations will be compared with available information from the steel industry. For the horizontal continuous casting, experiments will be carried out in a pilot scale equipment developed at the Foundry Laboratory - UFRGS. Artificial Intelligence techniques will be inserted in the continuous casting control algorithm, permitting an automatic search of optimised conditions of operation. In this research module, other two components of quality in the continuous casting of steel will be inserted: (i) the thermodynamic analysis of inclusions formation along the continuous casting operation and its correlation with thermal information provided by solidification mathematical models, and (ii) crack formation and its correlation with thermal parameters controlling solidification and the mechanical interaction ingot/ caster. The first component, which will be based on a thermodynamic analysis, will be conducted by using a computational tool named Thermo-Calc. It consists of a data base for inorganic chemistry and metallurgy, and at the same time is a powerful and flexible tool used for calculations concerning thermodynamic reactions and phase diagrams. The experimental validation will be based on samples collected from a Brazilian steel plant, along the continuous caster, and that will be characterised by using a procedure which dissolves the ferritic matrix and SEM analyses. The other component analysing crack formation, will be based on an interaction between solidification mathematical models, a knowledge base consisting of literature and industrial data and an artificial intelligence technique which uses a heuristic search to provide solutions which minimise crack formation. The industrial evolution of continuous casting of thin steel strips will be strongly dependent on the development of very efficient control systems, which must be able to perform quick modifications of process variables. The mathematical analysis of fluid flow, heat transfer and solidification will provide essential information to the control system, A physical model will be constructed in which the flow of liquid metal will be studied by using water as the fluid of simulation. Solidification will start when the liquid metal touches the surface of the cooled rolls, and it will be dominated by the metal/ roll thermal resistance. This thermal resistance must be characterised in terms of transient metal/ roll heat transfer coefficients. The mathematical model which will be developed to describe solidification in twin-roll strip casting, will depend on precise values of heat transfer coefficients in order to permit confident simulations to be generated. This numerical model will be the main component of the control system, and an artificial intelligence approach will be used in the functional structure of the algorithm. Predictions furnished by the numerical model and information from the simulations performed by the physical model will be used for settlements in a pilot scale twin-roll caster at the Technology Research Institute - IPT. The research module 02 will analyse the influence of heat and mass flow during solidification on macrostructure and microstructure formation. Numerical models will be developed involving coupled mass flow and heat transfer phenomena with a view to permit a clearer understanding of the influence of solidification rates on the solute redistribution. The macrostructure columnar to equiaxed transition will be studied in order to evaluate the quantitative influence of thermal parameters controlling solidification on the transition point. The more recent research articles on that subject, published last year, are still not convincing about the actual factors governing such structural transition. The microstructure will be analysed under the point of view of the cellular/ dendritic transition and dendritic growth, under conditions of unsteady state solidification. The literature reports a good number of studies concerning dendritic growth under controlled steady state solidification, but the contributions analysing dendritic microstructure formation under unsteady state heat flow conditions, i.e. under foundry or casting practical conditions, are rare. Quantitative expressions permitting the correlation between mechanical properties, dendritic structure parameters and thermal parameters controlling solidification will be developed. The research module 03 will focus on surface treatments using laser beams as the heat source. These treatments are already of industrial use, mainly in the automotive industry, and can be classified as solid state treatments, and treatments involving surface remelting, with or without the addition of metallic alloys or powders to promote modifications of chemical compositions on surface. The use of mathematical models for determining optimised laser operational variables can be considered a very effective tool in the improvement of process performance. Two different mathematical approaches are going to be considered in this research module: A first one describing the workplace laser heating, i.e. the interaction laser -material; and a second one describing the cooling process; including the rapid solidification which occurs when heating promotes surface remelting. An interaction between these mathematical models and an Artificial Intelligence heuristic search method will be used for providing the automatic search of the best set of operational conditions, as a function of geometry and dimensions of the section to be treated and desired microstructure... (AU)