The response of steel-framed structures to applied loading depends to a large degree on the behaviour of the joints between the columns and beams. Traditionally designers have assumed that these joints act either as 'pinned', with no ability to transmit moments from beam to column, or as 'rigid', providing perfect continuity between the connected members. Advances in analysis, and developments in modem codes of practice, permit designers to account for the real behaviour of steel joints where this is known or can be predicted. Even though experimental studies of joints conducted at many research centres around the world have provided a large bank of test data, the vast number of variables in joints (beam and column sizes, plate thicknesses, bolt sizes and spacing, etc.) often means that data for a specific joint arrangement does not exist. As a result, researchers have turned their attention to ways of predicting the behaviour of such joints. One approach which has gained acceptance is based on the "Component Method" in which overall joint behaviour is assumed to be produced by the responses of its various simpler components. To date, data on the response of joints at elevated temperatures has been gathered from full-scale furnace tests on cruciform arrangements, which have concentrated exclusively on moment-rotation behaviour in the absence of axial thrusts. However, when steel-framed structures are subjected to fire, the behaviour of the joints within the overall frame response is greatly affected by the high axial forces which are created by restraint to the thermal expansion of unprotected beams. If momentrotation- thrust surfaces were to be generated this process would require prohibitive numbers of complex and expensive furnace tests for each joint configuration. The alternative, and more practical, method is to extend the Component Method to the elevated-temperature situation. The basic theme of the Component Method is to consider any joint as an assembly of individual simple components. Each of these components is simply a non-linear spring, possessing its own level of strength and stiffness in tension, compression or shear, and these will degrade as its temperature rises. The main objective of this study was to investigate experimentally and analytically the behaviour of tension and compression zones of end-plate connections at elevated temperatures. A series of experiments has been carried out and a simplified analytical model has been developed, and this has been validated against the tests and against detailed finite element simulations. The simplified model is shown to be very reliable for this very common type of joint, although similar methods will need to be developed for other configurations. The principles of the Component Method can be used directly in either simplified or finite element modelling, without attempting to predict of the overall joint behaviour in fire, to enable semi-rigid behaviour to be taken into account in the analytical fire engineering design of steel-framed and composite buildings.

No description available.

Over the last decade it has become increasingly clear that the traditional methods of fire safety design can be unnecessarily conservative and therefore expensive. In 1995 a series of fire tests were carried out at Cardington, UK on a full-scale eight storey steel-concrete composite building. These tests produced an extensive body of data about the response of such structures to fire conditions and it is intended that this data be used to develop a clearer understanding of the structural behaviour involved. This thesis presents a method of analysing the behaviour of structures such as the Cardington frame using the commercial finite element package ABAQUS, with the addition of user defined subroutines; applies the method to two of the Cardington tests and analyses the results. FEAST, a suite of computer programmes that defines the behaviour of shell finite elements using a stress-resultant approach, was programmed for use with ABAQUS. The FEAST suite consists of two main programmes. The first, SRAS, is designed to model the behaviour of orthotropic plate sections at elevated temperatures. The second, FEAI, interfaces with the finite element package ABAQUS and allows realistic models of the behaviour of whole structures in fire conditions to be obtained. Phenomena modelled by FEAST include non-linear thermal gradients, non-linear material behaviour and coupling between membrane and bending forces. FEAST was used to analyse the behaviour of the Cardington Restrained Beam Test and the Cardington Corner Test. In both cases it was possible to produce a comprehensive set of results showing the variation of forces, moments and deflections in the structure under fire conditions. In addition, a number of parametric studies were performed to determine the effect of factors such as slab temperature and coefficient of thermal expansion on the behaviour of the structure. Special attention was given to the role of tensile mebrane action. The results showed that the behaviour of the heated structure was very different to that of an unheated structure. The response of the structure was shown to be very strongly governed by restrained thermal expansion and by thermal gradients. Degradation of material properties were found to have only a secondary effect on the structural behaviour.

In the analyses of steel framed structures it is customary to assume the beam to column connections as either fully rigid which means that all moments are transferred from beam to column with negligible rotation or ideally pinned that resists negligible moment. This assumption is reasonable for some types of connections. However when considering steel connections such as bolted-bolted double web angle connections it can be seen that the behavior of these connections is in between the two extreme cases. Thus a third connection type, namely semi rigid or partially restrained connection, is introduced. However this type of connection exhibits such a nonlinear behavior that modeling this behavior necessities a substantial effort. Moreover to perform a performance based analyses the true behavior of these connections should be incorporated as part of the modeling effort. Several researches dealing with these two topics have been undertaken in literature. Despite these efforts, modeling of the moment rotation behavior of these connections still requires improvement especially under cyclic loading conditions. In addition to this, performing an analysis with existing elements incorporating semi-rigid connections as a spring attached to beam ends is not practical because of the fact that displacement based formulation increases meshing significantly which requires substantial computational power. In this study a hysteretic (quadra-linear) moment rotation model considering pinching, damage and possibility of residual moment capacity is developed. The behavior is calibrated via experimental data available in the literature. Furthermore a force based macro element considering spread inelastic behavior along the element is presented. With this element several connections located anywhere along the beam could be incorporated in the analysis with less degree of freedom with respect to displacement based elements. Moreover the macro element model can be used in conjunction with corotational formulation for the capture of nonlinear geometric effects.