ABSTRACT
The necessity of recalculating existing engineering structures has inspired a bunch of methods to ensure reliability. Among them is probabilistic modelling of material properties for the probabilistic structural analysis or methods of Bayesian System Identification that allow for the verification of the model parameters in a finite element calculation. Both methods rely on probability distribution functions to describe scatter in the behavior of an engineering structure. Of special interest in this field are concrete structures as they show a great variance in the material characteristics. Using a Monte-Carlo-Simulation, an approximate lower bound for the truncation of the probability function for compressive strength is given, based on the regulations of conformity testing that were in effect at the time of construction. The introduced methodology is elaborated using examples from different periods of quality management. The lower bound provided by the methodology avoids excessively low values of compressive strength in the probability distribution as they exist in boundless functions. The results of this simulation are afterwards transferred to the concrete’s Young’s modulus, so that the truncated probability function can be used as a prior in a Bayesian system identification framework.
