At present, the fatigue provision in most design specifications in this country is based upon test data presented in the form of modified Goodman diagrams. The recent AISC specification is based upon a constant-range criterion. In any event, neither the randomness of fatigue lives nor that of applied loads is being considered. In this paper, a simple and reliability-based procedure is presented for the fatigue design of structures. The probability distribution for repeated loads can be obtained from dynamic analysis or field measurements. In addition, the stress-life-probability relations are available from single-stress-level laboratory tests. With an initial configuration of the structure, the distribution of repeated stresses can be computed. Then, the life of the structure can be predicted using a cumulative-damage criterion and an acceptable probability of failure. The computed life is compared with the design life, and the design can be modified in an iterative manner. /Author/

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 00262496
  • Record Type: Publication
  • Report/Paper Numbers: Proc. Paper 10782
  • Files: TRIS
  • Created Date: Nov 27 1974 12:00AM