Design by Testing and Statistical Determination of Capacity Models

In this chapter, the procedure proposed in EN1990 is adopted and extended to the case of EBR FRP systems, with the aim of attaining a uniform reliability level among all equations developed in this technical report. This approach will allow comparing experimental results and theoretical predictions in a consistent manner, and also identifying possible sources of error in the formulations. Any capacity model should be developed on the basis of theoretical considerations and subsequently fine-tuned through a regression analysis based on tests results. The validity of the model should then be checked by means of a statistical interpretation of all available test data. The formulation should include in the theoretical model a new variable that represents the model error. This variable is assumed to be normally distributed whit unit mean and standard deviation to be evaluated from comparison with experimental results. Once the statistical parameters of the model error are known, it is possible to define the statistical parameters of the capacity model and to evaluate its characteristic value, which is the aim for application in design. Some applications are shown to prove the feasibility of the proposed procedure.