ABSTRACT
A classification system for foundations and earth structures has been proposed based on the various levels of uncertainty associated with the design process. In order to obtain a valuable probabilistic representation of analyses from fundamental data, it is necessary to include less tangible components of uncertainty that require subjective judgment. For some cases, a satisfactory design may be achieved from this alone. However, frequently, some reduction in this uncertainty is essential for economic design and one way to achieve this is through prototype testing. This approach may be formalized through Bayes’ theorem. For some cases, empirical correlations may be used to reduce the uncertainty to a lower level. For highly sophisticated, one-off projects, that warrant additional design expenditure, a higher level of expertise in design and control of construction can reduce these less tangible uncertainty components.
