ABSTRACT
To design new structures and to assess the safety of existing ones structural engineers rely on models, based on practical experience, scientific underpinnings and the hard lessons of failure, as well as accumulated knowledge. They cannot rely, as others often can, on trial and error to obtain, eventually, a satisfactory product. Similar situations exist for other major infrastructure, such as pipelines, particularly those conveying hazardous materials, nuclear facilities and nuclear waste containers, arbour facilities, major dams, canals, etc. A distinguishing feature of engineering in these domains is the development of theoretical, conceptual models with which to predict the outcome of the engineering project with an acceptably high (usually very high) degree of accuracy. Usually predictability has to be high because the possibility of failure, with whatever consequences might result, must be kept under some degree of control, that is, at an acceptable level.
