Intelligent design method of high volume modulus composite structure
By combining finite element simulation, machine learning surrogate models, and conditional diffusion generative networks, the problem of designing composite microstructures close to the theoretical stiffness limit was solved, achieving efficient and automated composite material design, and generating structures with excellent and reliable performance.
CN122224360APending Publication Date: 2026-06-16CHINA NORTH VEHICLE RES INST
View PDF 0 Cites 0 Cited by
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHINA NORTH VEHICLE RES INST
- Filing Date
- 2026-02-09
- Publication Date
- 2026-06-16
Smart Images

Figure CN122224360A_ABST
Abstract
The application belongs to the technical field of composite material microstructure design, and provides an intelligent design method of high bulk modulus composite structure, which is based on the voxel representation of two-dimensional two-phase composite material microstructure, combines finite element calculation, a proxy model of physical constraints and conditional diffusion generation network, and intelligently designs the composite material microstructure. Under the condition of a given volume fraction, the application obtains the equivalent bulk modulus of the microstructure through finite element simulation, quickly predicts the performance by using the proxy model, introduces the Hashin-Shtrikman theoretical upper bound constraint, guides the diffusion generation network to generate a candidate structure, and gradually approaches the Hashin-Shtrikman bulk modulus upper bound through closed-loop screening iteration optimization, so that the composite structure close to the theoretical limit bulk modulus is finally obtained. Compared with the traditional topology optimization and generative design method, the application can efficiently design the composite structure with the stiffness close to the theoretical upper limit under the given material consumption, the design process is intelligent and self-consistent, and the result has physical rationality.
Need to check novelty before this filing date? Find Prior Art