Abstract
Due to production imperfection, substantial spatial variability is present in carbon-based composite materials. Moreover, there exists a high contrast of thermal properties between fibers and the void phase. This paper answers the question of how the effective thermal properties of the macroscale composite material change with respect to the geometric variability and to the conductivity discontinuity at the interfaces between fibers and air. The Porous Microstructure Analysis (PuMA) software is used in this work to artificially generate 3D random carbon fibers. Geometric variability is characterized by uncertian parameters ranging from fiber dimensions to orientations. This uncertainty model is then solved using Global Sensitivity Analysis method. We have found that the porosity of the material has the most significant influence on the effective property.
| Original language | English (US) |
|---|---|
| Journal | UNCECOMP Proceedings |
| State | Published - 2023 |
| Event | 5th ECCOMAS Thematic Conference on Uncertainty Quantification in Computational Sciences and Engineering, UNCECOMP 2023 - Athens, Greece Duration: Jun 12 2023 → Jun 14 2023 |
Keywords
- ANOVA
- Anisotropic Materials
- Carbon fiber
- Global Sensitivity Analysis
- Porous Media
- Thermal Conductivity
ASJC Scopus subject areas
- Computational Theory and Mathematics
- Computer Science Applications
- Modeling and Simulation
- Statistics and Probability
- Control and Optimization
- Discrete Mathematics and Combinatorics