No description available.

Steel beam-to-column connections are vulnerable structural elements when a fire strikes a building, as observed in fire incidents and full-scale fire tests. Existing techniques allow researchers to model the behaviour of different types of connections in fire but are difficult to use when conducting simulations on full-scale frames with multiple connections due to time and computation requirements. Therefore a need for a simplified connection modelling approach that can significantly reduce the computational time required without compromising on the accuracy of the simulation results so that large-scale simulations of structures with multiple connections in fire can be performed. A simplified spring connection modelling approach for steel flush endplate beam-to-column connections in fire has been developed in this research project so that the realistic behaviour of connections can be incorporated into full-scale frame analyses at elevated temperature. The proposed modelling approach divides the connection into two or three (depending on the connection size) T-stubs and employs ABAQUS as a pre-processor to generate the force-displacement characteristics for each T-stub by detailed finite element modelling. These characteristics are then input into specialised software (VULCAN) to simulate the behaviour of structure in fire including realistic representation of the steel beam-to-column connections. As a result of its simplicity and reliability, the proposed approach permits full-scale frame analysis in fire to be conducted efficiently. The proposed simplified spring connection modelling approach has been used to investigate the influence of connection ductility (both axial and rotational) on frame behaviour in fire. 2D steel and 3D composite frames across a range of spans were modelled to aid the understanding of the differences in frame response in fire when the beam-to-column connections have different axial and rotational ductility assumptions. The research study highlights that adopting the conventional rigid or pinned connection assumptions does not permit the axial forces acting on the connections to be predicted accurately, since the axial ductility of the connection is completely neglected when the rotational ductility is either fully restrained or free. By including realistic axial and rotational ductility of the beam-to-column connections, the frame response in fire can be predicted more accurately, which is advantageous in performance-based structural fire engineering design.

The present study is concerned with the assessnent of structural steel fracture toughness, as close to real service loading conditions as practically possible in the laboratory, using small scale specimens. The effects of stored strain energy content is evaluated for slow-static and dynamic COD tests for maximum load and cleavage instability. The literature reviews elastic-plastic fracture mechanics and goes on to study the effect of stored strain energy, the COD technique and dynamic testing procedures presently available. Static and dynamic fracture toughness testing using the COD technique is carried out on BS4360 - 50D structural steel in its normalised state. The testing procedures used closely relate to either the BS5762 COD standard or BS5447 plane strain standard. The specimen size tested is 2B =U= 24 mn, with a fatigue notch size of approximately a/W between 0.48 and 0.57. Photographic-macros and SEM fractography were carried out after the specimens were tested to assess the micromechanism processes operative during a fracture test. It is believed the present work is of special significance to determinate structural design using structural steel, for example with liquefied gas pressure vessels. The resulting test data available from this thesis is envisaged to be the closest approach to real service 'true limit severity', and consequently is beneficial to fracture prevention technology.

This thesis describes specific new applications of Z-Fibre ® pinning and focuses the attention onto the failure modes of locally reinforced (z-pinned) structures. Design implications of the use of localised reinforcement on structures, rather than laboratory coupons, are considered. Z-pinning reinforcement is applied to cylindrical crash tubes and I-section patch joints manufactured from woven carbon / epoxy pre-pregs. Z-pinning is shown to improve the Specific Energy Absorption (SEA) of the crash tubes by up to 76 %, the exact amount depending on the tube geometry. For the I-section patch joints, an initial increase in ultimate load carrying capability due to z-pin use is observed and, as the quantity of z- pins increases, a change in failure mode is induced. Z-pinning is also shown to enhance the damage tolerance of these. The ability to predict major changes to the structural response due to use of z-pins, and design for them accordingly, is the next step in the understanding of the technology. The design element of this study is contained in the development of a new Finite Element model using cohesive interface elements. The provision of mode II input data for this model comes from End Loaded Split (ELS) testing of the woven laminates and continued development of the Z-shear test. A new analysis for quantifying the crack sliding displacement, based on the ELS test, is developed. Z-shear testing has shown that the z-pin ‘mode II’ fracture energy is strongly affected by the amount of mode I opening of the shear surfaces. Here, new data are obtained for a fully constrained, pure mode II case. Using this modelling tool, changes in failure mode due to z-pin use can be predicted. Verification is provided by a new simulation of the I-section patch joint geometry.

The feasibility of enhancing the post-buckling load bearing capability of carbon/epoxy composite plate/panel structures utilizing embedded activated near equiatomic nickel-titanium, Ni-Ti, shape memory alloy, SMA, wire actuators has been investigated. Enhanced post-buckling is achieved through utilization of the unique shape memory phenomenon inherent within the Ni-Ti material. The unique phenomenon requiring a thermal stimulus. Within this investigation, such a thermal stimulus is provided for by an electrical current. Several host laminates, varying in lay-up architecture, have been considered. Two control strategies have been employed that utilize the unique SMA response at an elevated temperature. Control strategy 1 features embedded SMA actuators located within tubes that run along the specimens neutral plane. Here, the SMA's are constrained to external boundaries. Control strategy 2 also features embedded SMA actuators. For this control strategy, however, the actuators are partially constrained to the host laminate. For each strategy, upon SMA energization, shape memory constraint results with the formation of a recovery force within the SMA material. It is this recovery force that is employed to control the post-buckling response of the selected laminated specimens. A requirement for control strategy 2 is that the SMA/host interface must be of sufficient quality to sustain an elevated temperature as well as the imposed recovery force. Pertaining to control strategy 1, for the associated specimens, activation of constrained pre-strained SMA wire actuators can result with a significant specimen post-buckled deflection alleviation while under the influence of an external compressive load that is approximately three times the critical buckling value. While not as effective as control strategy 1, the concept behind control strategy 2 has been shown to work. Its efficient, or optimal, utilization, however, has yet to be demonstrated. For all the specimen configurations, the constrained SMA response acts to pull the specimens back to their flat configuration. This is true even when employing a low SMA volume fraction. Depending on the magnitude of the in-plane compressive load, however, this can result with post-buckled instability. SMA restoration recovery forces not only reduce the peak displacement amplitude, they also alleviate high stress levels, local to the boundary supports, that are typical to postbuckled plate/panel configurations. The tendency of adaptation is to redistribute the loading back towards the plates central region, such that, a more uniform stressed state exists. The stability of the adapted shape is dependent upon the laminate stacking sequence. Due to the elevated temperature required for SMA energization, the stacking sequence chosen should be such that temperature effects have minimal influence on the structural performance. SMA utilization would certainly be of benefit when such components are subjected a thermal environment by a means other than electrical energization. As an example, heating, associated with skin friction, may be sufficient to drive the actuators through their phase transition such that they exert stabilising recovery forces on the skin sections of high speed aircraft. The performance benefits of the SMA/carbon/epoxy composites materials, however, must carefully be assesseda gainst issueso f technical risk, producability, maintainability, reliability, and, of course, cost. The improved performance must be at an affordable price.

Translational slides in clays are often characterized by long-lasting intermittent movements associated with the fluctuations of pore-water pressures. These are in turn associated with either the rise or fall of the groundwater table for the case where the failure surface develops in the saturated zone below the phreatic surface or the infiltration of rainwater for the case where the failure surface is located above the phreatic surface in the unsaturated zone. Physically-based models designed to support hazard analysis of landslide movements and early warning systems require the integration of time-dependent (viscous) constitutive models for the shear displacements because landslide movements are typically controlled by the viscous behaviour of the clay geo-material. However, little information is available on the creep response of clay geo-materials under unsaturated conditions concerning different water content and matric suction values. This thesis presents an investigation of the viscous response of a clay geo-material under saturated and unsaturated conditions. Creep and relaxation tests have been first carried out on saturated clay samples by means of direct shear box. For the creep tests, the shear force was increased to a target value and maintained constant while monitoring the shear displacements. For the relaxation tests, shear displacement was applied to a target value and maintained constant while monitoring the shear stress decay. To gain a conceptual understanding of the viscous response of the clay in shear, analogue models were developed based on combinations of springs and dashpots. The aim of this modelling simulation was to identify a single mechanical model to simulate both creep and relaxation response using single set of parameters. Tests on unsaturated samples at different water content were finally carried out to gain insight into viscous response of the clay under unsaturated conditions.

Rock discontinuities are fundamentally important to most rock engineering projects but predicting or measuring their properties such as roughness, aperture, shape and extent (persistence) are fraught with difficulty. So far the solution of how to measure or predict persistence is poorly researched partly because the concept of how to investigate the extent of rock discontinuities within a rock mass seems intractable, by any economical methods. In the majority of engineering applications it is a fairly widespread practice to follow a conventional approach, assuming a 100% persistence value. However that is certainly incorrect even if usually a conservative assumption. This project is a small step towards resolving this issue. A series of laboratory and field research activities were carried out to investigate incipient nature of rock discontinuities and the extent of rock bridges. Uniaxial tensile strength of incipient discontinuities was quantified in the laboratory using cylindrical rock samples. The tested samples included incipient joints, mineral veins and bedding. It has been confirmed that such visible yet incipient features can have high tensile strength, approaching that of the parent rock. Factors contributing to the tensile strength of incipient rock discontinuities have been investigated. It is concluded that the degree of incipiency of rock discontinuities is an important factor that should be differentiated as part of the process of rock mass classification to inform more realistic engineering design and that this might best be done with reference to the tensile strength relative to that of the parent rock. An original methodology has been developed in the laboratory using expansive chemical splitters in drillholes, to quantify the tensile strength of large-scale incipient rock joints. In these tests, smaller tensile strengths were obtained, which probably was the result of localised stress concentration, low pressurization rate and unavoidable variations of expansive tensile force arising from the chemical splitter. A technique ‘Forensic Excavation of Incipient Rock Discontinuities (FEIRD)’ was established and employed to investigate areal extent and incipient nature of discontinuities in the field. Large rock blocks, containing incipient features, were split using similar expansive grout techniques as developed in the laboratory. Test results were interpreted and discussed with respect to fracture mechanics, fractographic features (such as hackle and rib marks), as well as geological conditions affecting the incipiency of the tested discontinuities including degree and extent of weathering and mineralisation. One common observation from the tests conducted is that breakage of non-persistent sections of incipient rock joints (rock bridges) leads to the development of rough surfaces over those freshly broken areas, and this may have implications for rock fracture development more generally. Despite rock bridge failure (say as part of rock slope mass movement), the freshly formed surfaces might be expected to have relatively high strength compared to the pre-existing persistent sections. An important conclusion from this research is that areal extent of open rock discontinuities (persistence) can be investigated realistically using the FEIRD technique. It has been found that estimates of persistence from trace mapping on rock exposures can be wildly inaccurate and it is concluded that field studies using FEIRD techniques (perhaps at a larger scale than used for this research to date) can be used to understand and quantify better the true nature of rock mass fracture network connectivity and extent that are important parameters for many rock engineering endeavours.

Vibro stone column techniques create an improved composite foundation in fine grained soils because of: (1) the installed load bearing columns of well-compacted, coarse-grained material and (2) the improvements to the surrounding soil due to the construction of the stone columns consolidating the surrounding soil. Extensive research work has been carried out over the last 20 years to understand the improvement in the composite foundation performance due to the consolidated soil. Few of these studies have quantified the changes in the stiffness and stress state of the treated soil, or have considered the impact that these changes have upon the performance of the composite foundation. Consequently, empirical and conservative design methods are still being used by ground improvement companies leading to a significant range of results in engineering practice. Based on cylindrical cavity expansion theory, two-dimensional finite element study to develop an axisymmetric model of a single stone column reinforced foundation was performed using PLAXIS 2D to quantify the effect of the vibro installation of this column in soft saturated clay by producing the load settlement response of the foundations. An updated mesh was used to cope with the large deformation of the soft clay around the installed column caused by the lateral expansion due to the Vibro technique. Different amounts of lateral expansion were simulated to determine the change in the stress state, stiffness and load settlement response. It was found that the radial expansion increases the pore pressure in the clay that starts to dissipate immediately after finishing the column installation leading to a permanent improvement of the stiffness of the soil which decreases with distance from the column. The radial stress acting on the column also changes creating a new coefficient of lateral earth pressure K, a key design parameter. The effect of these altered soil characteristics were assessed by applying a load to the composite foundation and calculating the resulting settlement. The previous model results have been validated and applied for a well-documented field case of stone column groups using Plaxis 3D after adopting a conceptual model for accumulating the installation effect of two adjacent stone columns. A very good agreement between the recorded and simulated load-settlement curves was achieved after performing few calculation cycles of different degrees of expansion cavity. A simplified design framework base on numerical analysis in how to account for the stone column installation and the recommended degree of applied radial cavity during stone column installation was the main output of this research to achieve more efficient composite foundations.

No description available.

No description available.