PROBABILISTIC APPROACHES TO BRIDGE DESIGN LOADS

Load uncertainties form a major part of the probabilistic basis for structural design. This paper is part of a state-of-the-art report on probabilistic design for bridges. The emphasis is on vehicle-induced loads for short- and medium-span bridges. Unlike environmental loads, vehicle loads can evolve according to imposed regulatory controls. Loads for railway bridges are relatively well controlled, and methods for deriving probabilistic load factors for repetitive spectra (fatigue) and maximum lifetime loading (strength) have been presented. Statistical data for highway bridges, however, are relatively sparse, especially for extreme vehicle weights and spacing configurations. Computer simulation and convolution models for deriving probabilistic load factors for strength are described, and safety indices for fatigue--including uncertainties in vehicle weights, truck dimensions, impact, headway, volume, girder distribution, and material fatigue life--are also presented. These probabilistic approaches are limited by the available data base and the continuing growth in truck loads. The paper also presents a system-reliability approach to model behavior beyond the simple element capacity check. Bridge damage versus load curves should be investigated for different geometries and configurations. Superimposing load and resistance probability distribution will provide damage costs for deriving optimum load factors. Because load growth would be partially absorbed in the bridge performance range between initial element failure and extreme damage, design ductility and redundancy need lower factors. Methods for establishing these goals are discussed. (Author)

Media Info

  • Media Type: Print
  • Features: Figures; References;
  • Pagination: pp 14-23
  • Monograph Title: Probabilistic design, redundancy, and other bridge papers
  • Serial:

Subject/Index Terms

Filing Info

  • Accession Number: 00302392
  • Record Type: Publication
  • ISBN: 0309029589
  • Files: TRIS, TRB
  • Created Date: Jan 30 1980 12:00AM