.png){kind=icon}
.png){kind=icon}
Return
Unified functional based data-model-coupling computing for composite materials and structures
.png){kind=icon}
PRE
AI
DOI：10.1016/j.compstruct.2023.116840.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The coupling of model-free data-driven computing and model-driven computing has been proposed recently, in which the model-driven computing refers to the constitutive model-based simulations. Its basic idea is to combine data-driven with traditional finite elements in the same mesh. Specifically, the data-driven computing is applied for the local region of the structure to avoid empirical material modeling, while the constitutive model-driven computing based on traditional FEM is employed for the rest regions of the structure to improve the numerical efficiency. This article aims to further prove that data-driven and model-driven computing can share the same distance-based functional. Thus, both algorithms can be freely converted to each other, which is the key basis for realizing data-model-coupling computing. Moreover, data-model-coupling computing based on extended FEM is applied for the analyses of crack propagation problem. A sandwich beam structure with a metallic foam core layer, a three-phase concrete material with an interface layer and a concrete-rock composite beam with initial interface crack are considered respectively to demonstrate the effectiveness of the proposed coupling strategy for the analyses of composite materials and structures.
Keywords:
Data-driven
Model-driven
Finite elements
Extended finite elements
Coupling
Journal
IF:
7.1
Papers:
1.8W
Citations:
8.0W