3D Reconstruction of Carbon Nanotube Composite Using Statistical Correlation Functions

Microstructure reconstruction from statistical microstructure descriptors is a strong field of interest for researchers worldwide, due to its importance in material design. A new methodology is presented in this paper to reconstruct microstructure with a large number of representative volume elements. The proposed methodology provides a stable input for a deterministic method which can be used to simulate performance and effective properties. The Monte Carlo technique is used as the basis of the reconstruction methodology in this work. In this paper statistical correlation functions are extracted from the images of the experimental samples. The proposed algorithm reconstructs new samples which has similar statistical correlation functions in comparison to the experimental samples. The information of the geometric distribution of the nanotubes of the composites is stored in a database of the node locations of the unit cylinder segments and the corresponding waviness, Instead of using a discrete image matrix. These node locations are attractive results which can be utilized in the simulation software in order to obtain the properties of the studied composite. In this way, robust microstructures with a large number of representative volume elements were reconstructed for the future evaluation.