STRUCTURAL SAFETY, RELIABILITY AND RISK ASSESSMENT

A credible risk assessment by which alone the safety and reliability analysis can be integrated into the framework of a valid benefit-risk analysis that necessarily underlies any rational decision process including optimization and insurance, presupposes the use of realistic distribution functions. However, the distributions of the statistical variables involved in such analysis (loads, resistance, times to failure, defects, etc.) are neither known nor obtainable by direct statistical inference within the probability range that is significant for safety and reliability analysis. An effective way to circumvent this difficulty, which is the principle obstacle to the development of a safety and reliability analysis associated with a credible, realistic risk assessment, is the formulation of the arising problems which involve only statistical variables the extremes (largest or smallest) of which (rather than all values) are significant. The formulation of the safety factor as a ratio between an "extremal" (high) load and an "extremal" (low) resistance, of the (lowest) brittle fracture strength as the result of an "extremal" (large) crack, of the shortest time to fatigue or creep fracture as the result of an "extremal" (high) crack propagation or secondary creep rate or of the design wave for a ship-hull, an oil-drilling platform or a sea-wall as one of a (small) population of "extreme" (high) waves rather than as the highest of the unknown population of (all) waves, are examples of this approach. Its result is the introduction of the small number of extreme value distributions as the basis for a realistic assessment of safety, reliability and risk involving "extremal" conditions, which constitute the majority of conditions encountered in the design of high-risk structures, as well as significant conditions in the design of low- and medium-risk structures for the purpose of disaster-control and insurance.

  • Corporate Authors:

    George Washington Transportation Research Institute. Center for Intelligent Systems Research

    ,    
  • Authors:
    • Freudenthal, A M
  • Publication Date: 1974-5

Media Info

  • Features: References;
  • Pagination: 23 p.

Subject/Index Terms

Filing Info

  • Accession Number: 00072524
  • Record Type: Publication
  • Source Agency: Ship Structure Committee
  • Report/Paper Numbers: No. 20 Tech. Rpt.
  • Contract Numbers: N00014-67A-0214-0011
  • Files: TRIS
  • Created Date: Dec 19 1974 12:00AM