STP1191

    Inelastic Stress-Strain Predictions for Multiaxial Fatigue Damage Evaluation

    Published: Jan 1993


      Format Pages Price  
    PDF Version (372K) 25 $25   ADD TO CART
    Complete Source PDF (8.8M) 25 $119   ADD TO CART


    Abstract

    The inelastic stress-strain response of a material subjected to complex loading must be known or estimated in order to predict the fatigue life of a component using many current multiaxial fatigue damage models. This paper presents stress-strain predictions obtained using two incremental plasticity algorithms for several nonproportional loading paths.

    The two algorithms both employ von Mises yield surfaces with kinematic hardening. However, one algorithm translates a single yield surface within a stationary limit surface using the hardening rules specified by Mróz and invokes a radial return for neutral loading. The other algorithm translates multiple yield surfaces (or a field of plastic moduli) according to a modified Mróz hardening rule proposed by Garud with no neutral loading provision. The algorithms are used to predict stress histories that are compared with measured responses from strain-controlled tests on normalized 1045 thin-walled tubes. A variety of nonproportional load paths were investigated. In general, relatively good agreement between predicted and measured stress-strain response for the two algorithms was observed for application to multiaxial fatigue evaluation. Strengths and weaknesses of each model are discussed.

    Keywords:

    nonproportional loading, multiaxial fatigue, kinematic hardening, two-surface, multisurface, von Mises yield surfaces, normality flow rule, cyclic plasticity, cyclic stress-strain stability, neutral loading


    Author Information:

    Tipton, SM
    Associate professor, The University of Tulsa, Tulsa, OK

    Bannantine, JA
    Staff engineer, IBM, San Jose, CA


    Paper ID: STP24807S

    Committee/Subcommittee: E08.05

    DOI: 10.1520/STP24807S


    CrossRef ASTM International is a member of CrossRef.