Mechanics of Structure Genome

Our views on how one can leverage machine learning techniques to help constitutive modeling of composite materials are summarized in this paper

Xin Liu, Su Tian, Fei Tao and Wenbin Yu, A review of artificial neural networks in the constitutive modeling of composite materials, Composites Part B, vol. 224, 2021,109152. 

https://www.sciencedirect.com/science/article/pii/S1359836821005321.

After almost seven years development, we have achieved more and more clarity with the concept of structure gene and it governing principles (mechanics of structure genome). My recent talk on MSG summarizes its relations with materials genome, common structural theories, micromechanics, multiscale modeling, as well as its current application to deployable structures, homogeneous 3D elements, machine learning assisted multiscale modeling. You can watch this talk at Youtube https://www.youtube.com/watch?v=7d5LuQrLpCM&t=5s.

RVE analysis becomes a routine exercise in material modeling. Usually it is carried out using finite element codes such as ABAQUS or ANSYS. The main thing one should pay attention is to applying the right boundary conditions. It is settled that periodic boundary conditions are the preferred boundary conditions to be applied. The BCs are that u_i-\epsilon_{ij}x_j should be equal on the corresponding edges. This type of boundary conditions can be applied using coupled equations constraints. This requires that one creates a mesh with corresponding nodes on periodic edges.