STP927: Evaluation of a Stochastic Initial Fatigue Quality Model for Fastener Holes

    Yang, JN
    Professor, School of Engineering and Applied Science, George Washington University, Washington, DC

    Manning, SD
    Engineering specialist senior, Structures Technology Staff, General Dynamics, Fort Worth, TX

    Rudd, JL
    Aerospace engineer, Air Force Wright Aeronautical Laboratories, Flight Dynamics Laboratory, Wright-Patterson Air Force Base, OH

    Pages: 32    Published: Jan 1986


    Abstract

    An initial fatigue quality (IFQ) model, based on stochastic crack growth and the equivalent initial flaw size (EIFS) concept, is described and evaluated for the durability analysis of relatively small cracks in fastener holes [for example, <2.54 mm (0.10 in.)]. The IFQ model uses a stochastic crack growth rate model which accounts for crack growth rate dispersion. Procedures and concepts are also described and evaluated for optimizing initial flaw size distribution parameters based on pooled EIFS results. Fatigue crack growth test results for 7475-T7351 aluminum specimens subjected to fighter and bomber load spectra are used to evaluate the proposed IFQ model and model calibration procedures. The cumulative distribution of crack size at any given time and the cumulative distribution of the time-to-crack initiation (TTCI) at any given crack size are predicted using the derived EIFS distribution and a stochastic crack growth approach. The predictions compare well with the actual test results in the small-crack-size region. The methods described are very promising for durability analysis applications.

    Keywords:

    durability, fatigue, extent of damage, small crack size, equivalent initial flaw size (EIFS), crack size distribution, crack exceedance probability, stochastic crack growth, probabilistic fracture mechanics, initial fatigue quality (IFQ), time-to-crack initiation (TTCI)


    Paper ID: STP29058S

    Committee/Subcommittee: E08.04

    DOI: 10.1520/STP29058S


    CrossRef ASTM International is a member of CrossRef.