TY - JOUR
T1 - Mechanism-based representative volume elements (RVEs) for predicting property degradations in multiphase materials
AU - Xu, Wei
AU - Sun, Xin
AU - Li, Dongsheng
AU - Ryu, Seun
AU - Khaleel, Mohammad A.
PY - 2013/2
Y1 - 2013/2
N2 - Quantitative understanding of the evolving thermal-mechanical properties of a multi-phase material relies on the availability of statistically representative microstructure descriptions. Questions then arise as to whether a two-dimensional (2D) or a three-dimensional (3D) representative volume element (RVE) should be considered as the statistically representative microstructure. Although 3D models are more physically representative than 2D models in general, they are usually computationally expensive and difficult to be retrieved and/or reconstructed. In this paper, we evaluate the accuracy of a 2D RVE in predicting the property degradations induced by different degradation mechanisms with the multiphase solid oxide fuel cell (SOFC) anode material as an example. Both 2D and 3D microstructure RVEs of the anodes are adopted to quantify the effects of two different degradation mechanisms: humidity-induced electrochemical degradation and phosphorus poisoning induced structural degradation. The predictions of the 2D model are then compared with the available experimental measurements and the results from the 3D model. It is found that the 2D model, limited by its inability to reproduce the realistic electrical percolation, is unable to accurately predict the degradation of thermo-electrical properties. On the other hand, for the phosphorus poisoning induced structural degradation, both the 2D and 3D microstructures yield similar results, indicating that the 2D model is capable of providing computationally efficient yet accurate results for studying the structural degradation within the anodes.
AB - Quantitative understanding of the evolving thermal-mechanical properties of a multi-phase material relies on the availability of statistically representative microstructure descriptions. Questions then arise as to whether a two-dimensional (2D) or a three-dimensional (3D) representative volume element (RVE) should be considered as the statistically representative microstructure. Although 3D models are more physically representative than 2D models in general, they are usually computationally expensive and difficult to be retrieved and/or reconstructed. In this paper, we evaluate the accuracy of a 2D RVE in predicting the property degradations induced by different degradation mechanisms with the multiphase solid oxide fuel cell (SOFC) anode material as an example. Both 2D and 3D microstructure RVEs of the anodes are adopted to quantify the effects of two different degradation mechanisms: humidity-induced electrochemical degradation and phosphorus poisoning induced structural degradation. The predictions of the 2D model are then compared with the available experimental measurements and the results from the 3D model. It is found that the 2D model, limited by its inability to reproduce the realistic electrical percolation, is unable to accurately predict the degradation of thermo-electrical properties. On the other hand, for the phosphorus poisoning induced structural degradation, both the 2D and 3D microstructures yield similar results, indicating that the 2D model is capable of providing computationally efficient yet accurate results for studying the structural degradation within the anodes.
KW - Degradation mechanisms
KW - Finite element method
KW - Microstructure
KW - Statistically representative volume element
UR - http://www.scopus.com/inward/record.url?scp=84869878057&partnerID=8YFLogxK
U2 - 10.1016/j.commatsci.2012.10.026
DO - 10.1016/j.commatsci.2012.10.026
M3 - Article
AN - SCOPUS:84869878057
SN - 0927-0256
VL - 68
SP - 152
EP - 159
JO - Computational Materials Science
JF - Computational Materials Science
ER -