Computer simulations of realistic microstructures: implications for simulation-based materials design

Simulation;Microstructures;Composites;Discontinuously reinforced aluminum

Singh, Harpreet ; Materials Science and Engineering

University:Georgia Institute of Technology

Department:Materials Science and Engineering

关键词: Simulation;    Microstructures;    Composites;    Discontinuously reinforced aluminum;   

Others  :  https://smartech.gatech.edu/bitstream/1853/22564/1/Singh_Harpreet_200805_phd.pdf

美国|英语

来源: SMARTech Repository

【 摘 要 】

The conventional route of materials development typically involves fabrication of numerous batches of specimens having a range of different microstructures generated via variations of process parameters and measurements of relevant properties of these microstructures to identify the combination of processing conditions that yield the material having desired properties. Clearly, such a trial and error based materials development methodology is expensive, time consuming, and inefficient.Consequently, it is of interest to explore alternate strategies that can lead to a decrease in the cost and time required for development of advanced materials such as composites. Availability of powerful and inexpensive computational power and progress in computational materials science permits advancement of modeling and simulations assisted materials design methodology that may require fewer experiments, and therefore, lower cost and time for materials development. The key facets of such a technology would be computational tools for (i) creating models to generate computer simulated realistic microstructures; (ii) capturing the process-microstructure relationship using these models; and (iii) implementation of simulated microstructures in the computational models for materials behavior. Therefore, development of a general and flexible methodology for simulations of realistic microstructures is crucial for the development of simulations based materials design and development technology. Accordingly, this research concerns development of such a methodology for simulations of realistic microstructures based on experimental quantitative stereological data on few microstructures that can capture relevant details of microstructural geometry (including spatial clustering and second phase particle orientations) and its variations with process parameters in terms of a set of simulation parameters. The interpolation and extrapolation of the simulation parameters can then permit generation of atlas ofvirtualmicrostructures that covers the complete range of variations of processing conditions of interest. These simulated andvirtualmicrostructures can then be used in the micromechanical models such as FEM to analyze their constitutive properties