| About this Abstract |
| Meeting |
2024 ASC Technical Conference, US-Japan Joint Symposium, D30 Meeting
|
| Symposium
|
2024 ASC Technical Conference, US-Japan Joint Symposium, D30 Meeting
|
| Presentation Title |
Design of a Reduced Order Finite Element Model to Study the Effects of Microstructure Morphology on Composite Properties |
| Author(s) |
Eric Carey, Jamal Husseini, Scott Stapleton |
| On-Site Speaker (Planned) |
Eric Carey |
| Abstract Scope |
Fiber reinforced composites are hierarchical materials where microscale features influence the mechanical response at higher scales. Capturing the stochastic effects of fiber morphology can provide significant insight towards explaining the variability often seen from experimental tests. However, generating large amounts of data can be prohibitively expensive when using traditional finite element methods or equivalent techniques. The aim of the current work is to provide a computationally efficient method of characterizing composite stiffness, strength, and fracture toughness through a new reduced order microscale finite element model referred to herein as the Fixed Triangulation-Mesh Model (FxT). Statistically equivalent composite microstructures were generated with varying volume fractions for an IM7-8552 composite material system using a microstructure generator. A triangulation-based meshing algorithm with fixed element sizes was developed to discretize the model where a combination of higher order 8-noded quadrilateral and 6-noded triangular elements were used. An efficient homogenization technique to characterize all components of the composite’s stiffness matrix at each load step was also developed, and a smeared crack approach for higher order elements was used to model progressive failure in the matrix. First, 100 randomly generated RVEs with varying volume fractions were generated and the results for homogenized stiffness components were computed. Then, the transverse strength for an RVE under different loading scenarios was predicted using the progressive damage model. Results were compared against NASA’s High Fidelity Generalized Method of Cells (HFGMC). Both models show a near 1:1 agreement between stiffness components, and the strength predictions between both models were also very close. RVEs that were characterized using the FxT model were on average 3-5 times faster than HFGMC. |
| Proceedings Inclusion? |
Definite: Post-meeting proceedings |