Optimization — the basis of code-making and reliability verification

As early as 1971 Rosenblueth and Mendoza (Rosenblueth E, Mendoza E. Reliability optimization in isostatic structures. J Eng Mech Div, ASCE 1971;97(EM6):1625–42) published a paper on structural optimization the concepts of which have been refined later by Hasofer in 1974 (Hasofer AM. Design for infrequent overloads. Earthquake Eng and Struct Dynamics 1974;2(4):387–8) and Rosenblueth in 1976 (Rosenblueth E. Optimum design for infrequent disturbances. J Struct Div, ASCE 1976;102(ST9):1807–25) particularly in the context of earthquake resistant design. In essence, these authors proposed to distinguish between structures that can fail upon completion or never and structures which can fail under rare ‘disturbances’. Furthermore they distinguished between ‘single mission structures’ and structures which are systematically rebuilt after failure. The consequences of their findings for code making, especially for setting safety targets apparently have been overlooked since then. In fact, it is rather a yearly failure rate that has to be specified and verified and not a failure probability for an arbitrary reference time. The paper thoroughly reviews Rosenblueth's and Hasofer's developments and extends the concepts to failures including ultimate limit state failure under normal and extreme conditions, serviceability failure, fatigue and other deterioration and, finally, obsolescence. Some newly needed computational tools are addressed. Partial safety factors are derived for stationary failure processes and a new verification format for fatigue and other deterioration is proposed. Tools for optimization of structural components are presented.