| Summary: |
Public authorities and general public agree that structural safety and performance assessment of the built environment is a priority. The need for rational and cost effective interventions on existing constructions led to the worldwide development of norms and guidelines addressing this topic. Given the impact of earthquakes on society, seismic safety assessment (SSA) and retrofit of constructions is fundamental and the development of methods as those in Eurocode 8 Part 3 (EC8-3), one of the first codes to address such issues, is a logical step.
EC8-3 includes modelling, analysis and verification methods that are more detailed and lengthy than those considered in the design of new buildings. It is widely agreed that extensive application studies of such methods must be developed to observe their adequacy and to determine if some of them need to be modified in a second generation of the Eurocodes. Alas, few studies concerning EC8-3 methods in reinforced concrete (RC) buildings, but none in steel ones, were performed to date. The same applies to studies on the reliability of the code methods based on a probabilistic framework.
In our research on this topic, we recently identified several issues on the applicability of the code methods. For example, on the applicability of linear analysis, we found that the code stringent applicability condition is not expected to be satisfied in many cases. As it is unclear if such condition is not excessively conservative, further studies are needed. We also found that difficulties can be faced in the demand quantification, e.g. for the member chord rotation. Research is being carried out by our group to develop alternative/simplified approaches addressing such issues. In the context of the reliability evaluation of the code methods, we conducted a study on the probabilistic evaluation of the EC8-3 proposed values for the Confidence Factor (CF). The CF addresses the knowledge level available on 3 main construction aspects (geome  |
Summary
Public authorities and general public agree that structural safety and performance assessment of the built environment is a priority. The need for rational and cost effective interventions on existing constructions led to the worldwide development of norms and guidelines addressing this topic. Given the impact of earthquakes on society, seismic safety assessment (SSA) and retrofit of constructions is fundamental and the development of methods as those in Eurocode 8 Part 3 (EC8-3), one of the first codes to address such issues, is a logical step.
EC8-3 includes modelling, analysis and verification methods that are more detailed and lengthy than those considered in the design of new buildings. It is widely agreed that extensive application studies of such methods must be developed to observe their adequacy and to determine if some of them need to be modified in a second generation of the Eurocodes. Alas, few studies concerning EC8-3 methods in reinforced concrete (RC) buildings, but none in steel ones, were performed to date. The same applies to studies on the reliability of the code methods based on a probabilistic framework.
In our research on this topic, we recently identified several issues on the applicability of the code methods. For example, on the applicability of linear analysis, we found that the code stringent applicability condition is not expected to be satisfied in many cases. As it is unclear if such condition is not excessively conservative, further studies are needed. We also found that difficulties can be faced in the demand quantification, e.g. for the member chord rotation. Research is being carried out by our group to develop alternative/simplified approaches addressing such issues. In the context of the reliability evaluation of the code methods, we conducted a study on the probabilistic evaluation of the EC8-3 proposed values for the Confidence Factor (CF). The CF addresses the knowledge level available on 3 main construction aspects (geometry, construction details and materials) and represents a partial factor in the safety verification stage. EC8-3 defines 3 knowledge levels with corresponding CF values. Our study focused on the adequacy of the CFs for the evaluation of material properties and defined a probabilistic method to assess the EC8-3 values. Of the 3 values, 2 were considered adequate while the other was not.
Given the previous research, an extensive investigation and calibration study addressing EC8-3 methods for SSA of existing buildings is clearly needed. As more applicability issues will be identified in this project, our goal is to establish results contributing to a more consistent application of EC8-3, defining modifications when needed. To attain this goal, the project will comprise 3 different parts. The first aims to calibrate EC8-3 structural survey procedures to determine their practicability and the validity of the proposed minimum knowledge conditions. This study will be based on the survey of real RC and steel buildings and the real data obtained from the surveys will lead to the definition of uncertainty levels for posterior probabilistic modelling.
The second part corresponds to a deterministic calibration of EC8-3 numerical analysis methods. Using the surveyed buildings, a numerical comparison of the available possibilities for the structural analysis of the buildings will be addressed. The further probing on the applicability of linear analysis, the influence of the parameters of the ground motions matching EC8-3 criteria, the influence of different numerical modelling approaches are some of the aspects addressed. The third part of the project is the probabilistic calibration of the SSA results obtained with EC8-3 and is divided in 3 tasks. The first is the development of a probabilistic framework for calibration of EC8-3 results, allowing for local (structural member) level (LL) and global (building) level (GL) applications. The LL approach will be based on a previous work by our group while the GL approach will be developed in this project and calibrated by a methodology we are currently developing. For the GL assessment, an extension of the EC8-3 method, that only allows LL verifications, will be defined based on an existing method. The second and third tasks are the probabilistic evaluation of the safety levels obtained by EC8-3 at the LL and GL using the referred methods, considering all relevant uncertainties. Results will lead to the definition of probabilistic confidence levels associated to the EC8-3 deterministic SSA results.
This project gives us the opportunity to conduct research with a paramount influence in design practice and in the future evolution of European norms. Moreover, such norms are expected to considerably influence both people's safety and the economic sustainability of existing constructions. Although valuable contributions to the evolution of the Eurocodes are foreseen, this project is also likely to identify influential factors requiring further investigation, towards which new research should be directed in the future. |