STP580: Probabilistic Design of Composite Structures

    Maxwell, RE
    Group supervisor of reliability engineering and lead engineer, Hercules Incorporated, Magna, Utah

    Toland, RH
    Assistant professor of mechanics, Drexel UniversityPa., Philadelphia,

    Johnson, CW
    Group supervisor of reliability engineering and lead engineer, Hercules Incorporated, Magna, Utah

    Pages: 19    Published: Jan 1975


    Abstract

    The method of statistical trials (Monte Carlo simulation) in combination with a sensitivity analysis by variance separation is presented as an efficient design and analysis system for each specific mode of failure. The method has several advantages: (1) it provides a realistic reliability prediction for each failure mode of a composite structure; (2) it determines the material, geometrical, and manufacturing variables which have a significant role upon the structure's reliability with respect to each failure mode; and (3) it determines which material, geometrical, and manufacturing variables should be controlled in order to manufacture the structure at minimum product assurance costs. Application is made to the graphite-epoxy truss tubes of NASA's applications technology satellite.

    Keywords:

    composite materials, reliability, composite structures, simulation, variance separation, significance, probability distribution functions, optimum design


    Paper ID: STP32298S

    Committee/Subcommittee: D30.05

    DOI: 10.1520/STP32298S


    CrossRef ASTM International is a member of CrossRef.