This report is concerned with the problem of designing civil engineering structures for adequate reliability, and examines a number of different approaches proposed for this purpose. In the context of limit-stage design, the problem is that of making an appropriate choice of safety factor for each combination of loading, structural material, type of member and limit-state considered in practice, so that the risks of occurrence of the different ultimate and serviceability limits are acceptably small. Solutions can be found only by adopting a probabilistic view of structural behaviour and by statistical analysis of loads, strength and other design variables. Three semi-probabilistic safety formats are compared with regard to their suitability for adoption in codes of practice. The basis for a fully-probabilistic treatment of structural safety is also outlined, but a number of theoretical and practical difficulties preclude the use of the latter at the present time. The adoption of a semi-probabilistic safety format necessitates 'calibration' of the new code to an existing code, until such time as acceptable risks of failure and unserviceability can be objectively defined. 'code optimisation' is discussed as a possible future method of establishing risk levels, but a prerequisitie is a significant development of techniques for assessing the reliability of multi-member structural systems. A number of recent studies are discussed. Finally, the need for further research is considered. (A) /TRRL/

  • Corporate Authors:

    Construction Industry Research & Information Assoc

    6 Storey's Gate
    London SW1P 3AU,   England 
  • Authors:
    • Baker, M J
  • Publication Date: 1973-7


  • English

Media Info

  • Features: Figures; References;
  • Pagination: 83 p.

Subject/Index Terms

Filing Info

  • Accession Number: 00131495
  • Record Type: Publication
  • Source Agency: Transport and Road Research Laboratory (TRRL)
  • Report/Paper Numbers: No. 50 Tech. Note
  • Files: ITRD, TRIS
  • Created Date: May 14 1976 12:00AM