Keywords: reliability, reliability index, steel frames, semi-rigid connections, probability of failure.

The influence of random variables of semi-rigid beam-column and column-base connection parameters on the reliability, with respect to the ultimate load, of a steel frame previously experimentally tested, is studied. The effect of the random variables on the failure mode is also presented. The random variables considered are the yield stress of the steel, and the initial stiffness and moment resistance of the semi-rigid connections. The histograms of the ultimate load multiplier demonstrate the sensitivity of the reliability index to the adopted random variables.

This study augments an experimental investigation on full-scale frames aimed at determining the effect of flexibility in the beam-column and column-base connections on the behavior of the frames. The frame test presented in this paper was carried out in the Building Research Laboratory, Department of Steel Structures, Technical University of Budapest (currently Budapest University of Technology and Economics).

The reliability of a structure means the structure's ability of satisfying its requirements within a given risk. Using the Monte Carlo simulation method this study deals with the sensitivity of the reliability index of the load carrying capacity of a semi-rigid steel frame to the parameters of the connections.

In this paper the Cornell reliability index [1] is used. In accordance with Eurocode 3 [2], a second-order elastic-plastic in-plane analysis was applied to analytically determine the load caring capacity of the semi-rigid steel frame. The design value was taken for probability (this value is often used as reference in specifications), and is called characteristic value [3].

For the purpose of this sensitivity analysis the mean values for the connection parameters were taken from the tests and the deviation was taken as 10% of the mean value [4]. The reliability index and the failure probability were determined for two loading levels, kN and kN. The sensitivity of the reliability index of the semi-rigid frame considered was studied adopting random variables corresponding to the yield stress and moment resistance of the semi-rigid beam-column and base-column connections.

The extensive experimental study into the behaviour of semi-rigid connections, showed how the complexities arising from the three-dimensional nature of full-scale tests can be addressed [5]. The sensitivity study of the reliability index of the load carrying capacity of the frame showed that the cumulative effect of multiple random variables erode the value of the reliability index. The effect of base-column connections on the reliability index was less significant than the effect of the beam-column connections.

1

J.R. Benjamin, C. Cornell,. "Probability, Statistics, and Decision for Civil Engineers", McGraw-Hill Book Company, 1970.

2

EUROCODE 3 Editorial Group, "Design of Steel Structures. Part 1: General Rules and rules for Buildings", ENV 1993-1-1, Brussels, 1993.

3

J.P. Muzeau,. "Modele de l'influence d'imperfections sur la securite des structures metalliques en comportement non lineaire. Comparaison de reglements internationaux". These de doctorat d'etat es Sciences Physiques, Universite Blaise PASCAL, Clermont-Ferrand, Mai 1987.

4

A. Mebarki, A. Colson, "Sensitivity of Steel Frames Reliability to Semi-Rigid Connections Parameters", First International Speciality Conference on Coupled Instabilities in Metal Structures, Timisoara, Romania, October 1992.

5

M. Iványi, "Full-Scale Tests of Steel Frames with Semi-Rigid Connections", Engineering Structures, 22, 168-179, 2000. doi:10.1016/S0141-0296(98)00106-0

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £135 +P&P)