Statistical flaw strength distributions for glass fibres: Correlation between bundle test and AFM-derived flaw size density functions
Résumé
The present paper investigates glass fibre flaw size distributions. Two commercial fibre grades (HP and HD) mainly used in cement-based composite reinforcement were studied. Glass fibre fractography is a difficult and time consuming exercise, and thus is seldom carried out. An approach based on tensile tests on multifilament bundles and examination of the fibre surface by atomic force microscopy (AFM) was used. Bundles of more than 500 single filaments each were tested. Thus a statistically significant database of failure data was built up for the HP and HD glass fibres. Gaussian flaw distributions were derived from the filament tensile strength data or extracted from the AFM images. The two distributions were compared. Defect sizes computed from raw AFM images agreed reasonably well with those derived from tensile strength data. Finally, the pertinence of a Gaussian distribution was discussed. The alternative Pareto distribution provided a fair approximation when dealing with AFM flaw size. © 2012 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
Mots clés
Fractography
Fracture mechanics
Glass fibers
Cement-based composites
AFM image
AFM
Tensile testing
Textures
Fibers
Atomic force microscopy
Tension tests
Statistical mechanics
Tensile strength
Pareto principle
Defect size
Failure data
Fibre surfaces
Flaw size distribution
Flaw sizes
Gaussians
Multi-filament
Pareto distributions
Single filaments
Strength distribution
Tensile tests