학술논문

Multi-length scale modeling and analysis of microstructure evolution and mechanical properties in polyurea.
Document Type
Article
Source
Journal of Materials Science. Mar2011, Vol. 46 Issue 6, p1767-1779. 13p. 4 Color Photographs, 4 Diagrams, 2 Graphs.
Subject
*MATERIALS analysis
*CONSTITUTION of matter
*MICROSTRUCTURE
*ELASTOMERS
*MECHANICAL behavior of materials
Language
ISSN
0022-2461
Abstract
number of experimental investigations reported in the open literature have indicated that polyurea possesses a very complex nanometer-scale material microstructure consisting of hydrogen-bonded discrete hard (high glass-transition temperature, T) domains and a soft (low T) matrix and that the mechanical properties of this type of elastomer are highly dependent on the details of this microstructure. To help elucidate the internal processes/mechanisms associated with the microstructure evolution and to improve the understanding of microstructure/properties in this material, a multi-length scale approach is developed and utilized in the present work. This approach combines well-established and validated atomic, meso, and continuum length-scale techniques and spans around six orders of magnitude of length (from nanometers to millimeters). While within the atomic-scale approach, the material is modeled as a collection of constituent atom-size particles, within the meso-scale approach this atomistic description of the material is replaced with a collection of coarser particles/beads which account for the collective degrees of freedom of the constituent atoms. One of the main efforts in the present work was the derivation of accurate input parameters for meso-scale simulations from the associated atomic-scale material model and results. The meso-scale analysis provided critical information regarding the material microstructure and its evolution (from an initially fully blended homogeneous state). To obtain quantitative relationships between material microstructure and its mechanical properties, the computed meso-scale material microstructures are combined with a finite element approach. The predictions of the present multi-scale computational approach are found to be in good overall agreement with their experimental counterparts. [ABSTRACT FROM AUTHOR]