Effect of spatially correlated initial geometric imperfection on reliability of spherical latticed shell considering global instability. (January 2020)

• Record Type: Journal Article
• Authors: Cui, X.Z.; Li, Y.G.; Hong, H.P.
• Is Part Of: Structural safety. Volume 82(2020)
• Publication Date: 2020-01
• Subjects: Spatial correlation -- Simultaneous autoregressive model -- Reliability analysis -- Lattice -- Spherical lattice shell -- Initial geometric imperfection; Structural stability -- Periodicals; Safety factor in engineering -- Periodicals; Reliability (Engineering) -- Periodicals; Constructions -- Stabilité -- Périodiques; Coefficient de sécurité en ingénierie -- Périodiques; Fiabilité -- Périodiques; 620.86
• DOI: 10.1016/j.strusafe.2019.101895
• Languages: English
• ISSNs: 0167-4730

Abstract

Highlights: Used simultaneous autoregressive (SAR) model for the initial geometric imperfection. Assessed statistics of initial imperfection by using the SAR model for a dome. Estimated the probability distribution of critical load for global instability. Estimated reliability by considering global instability of spherical latticed shell. Abstract: Latticed shells are extensively used for large-span structures because of their good structural behaviour, lightweight and ease of construction. Their codified design is often governed by the global instability requirement for gravity load and specified shape and amplitude of maximum initial geometric imperfection (IGI). The reliability basis for the assigned shape and maximum IGI is unknown. In the present study, it is proposed that IGI is modelled using the simultaneous autoregressive model. It is shown that this model is flexible in modelling spatially varying correlated imperfection and can cope with irregular grid system. By using this model, the effect of the spatially correlated IGI on the probability distribution of the load capacity for global instability and the reliability of spherical latticed shell designed according to the Chinese design code is assessed. It is shown that the use of the prefixed shape of the IGI to assess the probability distribution of the load capacity is unconservative as compared to that by using spatially varying random imperfection. The shape of the probability distribution for the former differs from that for the latter. The reliability analysis results indicate that the use of a critical load factor of 2 implemented in the code leads to a failure probability less than about 10 −6 for global instability if the specified maximum IGI defined in the code equals about two to three times the standard deviation of IGI. This failure probability becomes less than about 10 −5 if a critical load factor of 1.5 is used for design. It is suggested that the code recommended critical load factor may be reduced to 1.5 while achieving target reliability index often adopted for design code calibration.