Component Importance Analysis Of Spatial Structures Based On Structural System Reliability

Strengthening the important members to improve the robustness of the structure is a widely advocated approach. In engineering practice, for a large number of members in the overall structure, in order to do monitoring, maintenance and reinforcement of individual members, a method to find important member is needed. Through the summary of study on the status quo of structural robustness, this dissertation proposed two methods to evaluating component importance based on system reliability which are vulnerability of probability and reliability importance factor.Vulnerability of probability utilizes responses to structures as performance criteria, the degree of degradation in system reliability after the removal of element is utilized to evaluating its component importance. For frame system, the displacement of top layer is utilized as performance requirements, while, for single-layer latticed shell, the maximum unit stress utilized as performance requirements. This dissertation gives the quantitative component importance coefficient based on vulnerability of probability under system performance requirements. During calculating process, Monte Carlo Method is utilized, besides, the Latin hypercube sampling applied to enhance the accuracy and efficiency of probability of failure. The results verified the applicability of probability vulnerability method in evaluating components importance. Moreover, the discrete coefficient of results is relatively large, which is advantageous to find a few key elements in the numerous bars.Component importance based on the reliability importance factor is defined as the sensitivity of system reliability to component reliability. This dissertation established a logical relationship of system reliability and component reliability by structural reliability block diagram method, then using the System-Subsystems-Components three levels to decompose the whole system, finally, solving the direct differentiation as reliability importance factor as component importance. For the spatial grid structure and cable dome structure, geometric stability is utilized as structural performance criteria to evaluating the component importance. The results showed that, for the spatial grid structure, the peripheral components of upper and bottom layer chords have higher component importance, as well as the web members in middle region. For the cable dome structure, the inner compressive bracing has highest importance, outer notochords, ring ropes, other bracings, etc. take second place, and inner notochords have lowest importance.