PROBABILISTIC METHODS FOR THE ANALYSIS OF METAL-MATRIX COMPOSITES Academic Article uri icon

abstract

  • A probabilistic micromechanics-based non-linear analysis procedure is developed to predict and quantify the variability in the properties of high-temperature metal-matrix composites. Monte Carlo methods and assumed probabilistic distributions of the constituent level properties including fiber, matrix, and interphase properties, volume and void ratios, strengths, fiber misalignment, and non-linear empirical parameters are used to predict the resultant ply properties and quantify their statistical scatter. Graphite/copper and silicon-carbon/titanium-aluminide (SCS-6 T115) unidirectional plies are considered to demonstrate the predictive capabilities. The non-linear material response of continuous-fiber-reinforced metal-matrix composite structures has also been studied by using the macroscopic elastoplasticity theory and a continuum shell element. The procedures discussed have a high potential for use in material characterization and selection (to precede and assist in experimental studies) and determination of the structural response of new high-temperature metal-matrix composites. 1994.

published proceedings

  • COMPOSITES SCIENCE AND TECHNOLOGY

author list (cited authors)

  • ENGELSTAD, S. P., & REDDY, J. N.

citation count

  • 21

complete list of authors

  • ENGELSTAD, SP||REDDY, JN

publication date

  • January 1994