Compression behavior of ultra-high modulus carbon/epoxy composites and fracture energy characterization of intermediate modulus carbon/epoxy composites
Open Access
- Author:
- Glath, Michael Charles
- Graduate Program:
- Engineering Mechanics
- Degree:
- Master of Science
- Document Type:
- Master Thesis
- Date of Defense:
- May 09, 2014
- Committee Members:
- Kevin L Koudela, Thesis Advisor/Co-Advisor
Charles E Bakis, Thesis Advisor/Co-Advisor - Keywords:
- ASTM D6641
SACMA SRM 1R-94
ASTM D5528
end-notched flexure
ANOVA
finite element analysis
damage
fracture energy - Abstract:
- Among compression test methods, ASTM D3410, SACMA SRM 1R-94, and ASTM D6641 are currently used to determine the compression modulus and compression strengths for ultra-high modulus (UHM, >350 GPa) composites. In this study, the compression modulus and compression strength of a proprietary ultra-high modulus carbon/epoxy composite were experimentally determined using SACMA SRM 1R-94 and ASTM D6641 for laminates with stacking sequences of [(0/±60)S]2 and [(+60/0/-60)S]2. The moduli of both laminates were experimentally shown to be statistically equivalent; the strength of the [(0/±60)S]2 laminate was shown to be significantly less than the [(+60/0/-60)S]2 laminate. A non-linear finite element model, using the commercial software package Abaqus, was developed for each compression test to predict the compression modulus and compression strength using the built-in damage progression algorithm in Abaqus. The models were used to identify failure modes for each compression test. Premature end-crushing was predicted at 2,200 µε in the modeled modulus specimen of SACMA SRM 1R-94 for the [(+60/0/-60)S]2 laminate. End-crushing in the modeled SACMA SRM 1R-94 strength specimen was predicted to occur slightly after ultimate laminate failure for the [(+60/0/-60)S]2 laminate. End-crushing was predicted to be highly unlikely for both laminates in the ASTM D6641 model. The stress-strain response of both models matched relatively well with experimental data. The predicted strengths for both test methods were underpredicted for the [(+60/0/-60)S]2 laminate and overpredicted for the [(0/±60)S]2 laminate. The fracture toughness of a proprietary intermediate modulus carbon/epoxy [0]10 laminate was experimentally determined using ASTM D5528 and a U.S. Government provided end-notched flexure compliance calibration test method for modes I and II, respectively. Significant fiber bridging was noted for the ASTM D5528 experimental test specimens, especially at the crack tips. A finite element model was developed for each test method using the virtual crack closure technique. A boundary was developed for the modeled load-displacement plot to encompass the experimental behavior of test specimens. This predicted boundary agreed well with mode I of ASTM D5528 experimental data. The predicted load-displacement response of mode II had a stiffer response than the experimental data, but the model did not consider potential stiffness degradation due to 5 preload calibrations used on each test specimen. A statistical sensitivity analysis (ANOVA), based on the previously developed finite element model of ASTM D6641, was used to vary parameters assumed to influence the compression modulus, compression strength, and first ply failure of ASTM D6641. Of the chosen parameters, fiber misalignment was predicted to significantly affect the FEA-based results of the compression modulus, compression strength, and first ply failure for the chosen levels of variation. The laminate thickness, adhesive thickness, and enforced bolt displacement were predicted to not significantly affect the FEA-based results of the compression modulus, compression strength, and first ply failure for the chosen levels of variation.