Prediction of the strength distribution for unidirectional fiber reinforced composites
- Technical Univ. of Darmstadt (Germany). Dept. of Materials Science
A statistical model is proposed which predicts the strength distribution of composites. It accounts for the following parameters: fiber strength distribution, fiber spatial distribution, fiber-matrix interface properties, matrix plasticity, load transfer at broken fibers and load profile of the specimen (bending or tension). The model is solved via Monte Carlo simulation for small specimens and via an analytical approach for large parts. Both techniques predict well the strength distribution of the composite. Aluminium matrix composites reinforced with unidirectional ceramic fibers show a large scatter in three point bending strength. This scatter is caused by the probability that a cluster of broken fibers of a critical size occurs at a given stress level. When the limiting stress is reached the critical cluster grows in an unstable way leading to catastrophic failure of the specimen. The results obtained by this model agree favorably with the experimental findings and indicate that strength prediction using other concepts, like fracture mechanics, global load sharing, a simple rule of mixtures or Weibull statistics, is unreliable for predicting the strength of large composites.
- OSTI ID:
- 415431
- Journal Information:
- Acta Materialia, Journal Name: Acta Materialia Journal Issue: 11 Vol. 44; ISSN 1359-6454; ISSN ACMAFD
- Country of Publication:
- United States
- Language:
- English
Similar Records
Bounds on the Strength Distribution of Unidirectional Fiber Composites
A micromechanics model for predicting the tensile strength of unidirectional metal matrix composites