Statistically-based survey plans to estimate the concrete strength in existing RC buildings

• Main Author: Nuno Pereira
• Publication Date: 2014
• Other Authors: Xavier Romão, Raimundo Delgado
• Format: Book
• Language: eng
• Source: Repositórios Científicos de Acesso Aberto de Portugal (RCAAP)
• Download full: https://hdl.handle.net/10216/84587
• topic: Engenharia estrutural, Engenharia civil; Structural engineering, Civil engineering

In the proposed study, a framework for the definition of a safety factor leading to a reliable estimate of the mean concrete strength in existing reinforced concrete (RC) buildings is presented. The main objective is to provide a methodology capable of establishing a reliable correlation between a given level of knowledge that is required (which implies that a number of destructive tests have to be carried out) and the confidence in the corresponding mean estimate of the strength that is obtained. The development of this approach requires dividing the building into areas with a potentially homogeneous concrete strength (e.g. each building storey). Furthermore, the methodology also uses the notion of discrete structure where the concrete strength of each structural element of a given homogeneous building section is assumed to be defined by a single strength value, which can be different than those of the remaining structural elements. The reliable value of the mean concrete strength is established by a boundary value defined according to an admissible variation with respect to the true mean value and includes the effect of uncertainty due to sampling. This admissible variation is defined by a procedure that uses both destructive and non-destructive test results. The referred boundary value allows the definition of a safety factor which represents the maximum admissible ratio between the true mean value and the corresponding reliable estimate. The presented study also proposes an alternative approach regarding the definition of the Eurocode 8-Part 3 knowledge levels based on the relation between the number of structural members that were tested and the total number of structural members. An adaptive confidence factor for the mean value of the material strength is then provided for each knowledge level as well as an approximation for the admissible global variability of the concrete strength in each homogeneous area of the discrete building. The proposed framework can be seen to be a more consistent statistically-based alternative to the confidence factor values proposed by Eurocode 8-Part 3.