Design by testing: A procedure for the statistical determination of capacity models

Worldwide research has now reached a level of integration where an effort towards the harmonization of procedures is absolutely needed. Such harmonization may regard, for example, the various steps that lead to the definition of capacity models to be included in design codes, specifically: definition of the test setup, quantities to be measured, identification of the basic variables influencing the phenomenon, distinction between average values and other fractiles, disaggregation of the model in different parts accounting for mechanics, fine-tuning and randomnesses, and, finally, assessment of the model against the experimental results. Test results and ensuing model developed according to this procedure would naturally lend themselves to be easily shared among the scientific community and would facilitate the task of calibrating the partial coefficients, with the ambitious aim of attaining a uniform reliability level among all capacity equations. © 2008 Elsevier Ltd. All rights reserved.