AN OVERVIEW OF IMPROVED PROBABILISTIC METHODS FOR STRUCTURAL REDUNDANCY ASSESSMENT OF INDETERMINATE TRUSSES

The redundancy of a structure refers to the extent of strength that is not considered in design. For an indeterminate structure a member failure does not necessarily induces the loss of integrity or functionality of the structure; rather it will affect its potential for safely carrying any future load. Numerous methods have been introduced in structural reliability literature to measure and implement the redundancy in design. However, in accordance with the semi-probabilistic approach of the codes which aim to develop design method providing consistent level of redundancy within the structure, the probability of failure of the structure has been proposed and is widely used as a redundancy measure. A classical method to compute the probability of failure of the structure based on failure paths is presented as a reference in this thesis. However, although extensively used, this method has major shortcomings which may lead to a misrepresentation of the structure redundancy. By using a geometric representation of members’ limit states associated with a loading regime, the research presented herein proposes an improved method for structural redundancy estimation that may be helpful to overcome problems associated with approximations and inconsistencies inherent in classics methods. Specific assumptions and/or procedures considered in the proposed method are described below. (1) An approximation is given to make the events of member failures mutually exclusive. (2) Geometric calculations are used to determine reliability indices and conditional reliability indices in order to establish closer bounds for the failure probability of individual structural members. (3) System’s failure probability is obtained using the assumption and procedures outlined in (1) and (2) above. (4) To further extend the method beyond geometrical redundancy, and to consider material redundancy, plasticity models commonly used in the structural analyses are considered in this studyM.S. in Civil Engineering, July 201