TY - JOUR
T1 - Determination of size of representative volume element for coke using homogenization method based on digital image of actual microstructure
AU - Zhang, Xiaoqing
AU - Yamazaki, Yoshiaki
AU - Hiraki, Kenichi
AU - Kanai, Tetsuya
AU - Uchida, Ataru
AU - Saito, Yasuhiro
AU - Shoji, Masakazu
AU - Aoki, Hideyuki
AU - Miura, Takatoshi
PY - 2012
Y1 - 2012
N2 - When the homogenization method based on the digital image of actual microstructure is used to analyze the strength of coke, the representative volume element (RVE) is a crucial input to predict reliable results. The shape of RVE was identified as square, due to the two-dimensional numerical analysis with the research object being a section image of a coke sample was studied. The variation of multi-scale mechanical properties of coke according to a large number of unit cells with various scales was investigated to determine the proper RVE size. The resolution of calculation equaling original resolution was considered as Case 1. The other low resolution of calculation was considered as Case 2. The results showed that, in the macro-scale (homogenization) analysis, the proper size of RVE on coke's sample should be chosen larger than 874.8 μm, due to the mean calculated homogenized Young's modulus was almost constant in the case of unit cell larger than 874.8 μ m. In the micro-scale (localization) analysis, in Case 1, unit cell of a large scale was prone to stress concentration which happened at the relative thin parts of structure around large pores. In Case 2, it was considered that high resolution was necessary, because low resolution changed the structural characteristic. Finally, the optimum size of RVE on coke's sample was determined as 2 624.4 μ m, which could represent the structural relativity of the research object.
AB - When the homogenization method based on the digital image of actual microstructure is used to analyze the strength of coke, the representative volume element (RVE) is a crucial input to predict reliable results. The shape of RVE was identified as square, due to the two-dimensional numerical analysis with the research object being a section image of a coke sample was studied. The variation of multi-scale mechanical properties of coke according to a large number of unit cells with various scales was investigated to determine the proper RVE size. The resolution of calculation equaling original resolution was considered as Case 1. The other low resolution of calculation was considered as Case 2. The results showed that, in the macro-scale (homogenization) analysis, the proper size of RVE on coke's sample should be chosen larger than 874.8 μm, due to the mean calculated homogenized Young's modulus was almost constant in the case of unit cell larger than 874.8 μ m. In the micro-scale (localization) analysis, in Case 1, unit cell of a large scale was prone to stress concentration which happened at the relative thin parts of structure around large pores. In Case 2, it was considered that high resolution was necessary, because low resolution changed the structural characteristic. Finally, the optimum size of RVE on coke's sample was determined as 2 624.4 μ m, which could represent the structural relativity of the research object.
KW - Coke strength
KW - Homogenization method
KW - Microstructure
KW - RVE
KW - Stress concentration
UR - http://www.scopus.com/inward/record.url?scp=84864964489&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84864964489&partnerID=8YFLogxK
U2 - 10.2355/isijinternational.52.1250
DO - 10.2355/isijinternational.52.1250
M3 - Article
AN - SCOPUS:84864964489
SN - 0915-1559
VL - 52
SP - 1250
EP - 1256
JO - ISIJ International
JF - ISIJ International
IS - 7
ER -