The progressive collapse of a structure occurs when local failure of a primary structural component leads to the collapse of adjoining members which in turn leads to further collapse through a chain reaction. Often the total damage is disproportionate to the original cause and is associated with a low probability event. The destruction of the World Trade Centre in 200 I due to aircraft impact is an example. Several studies on progressive collapse, mainly of steel structures, have resulted in changes to analysis and design guidelines. In precast concrete structures, connections play an important part in ensuring the safety of the whole structure. Current design practice is to assume these as pinned or rigid but this cannot be relied upon for safety against progressive collapse. The aim of this thesis is to examine the vulnerability of the behaviour of concrete frames with semi-rigid connections. The numerical responses of precast concrete frames with billet connections were studied. The connection was modelled by zero-length spring elements with rotational stiffness values at the ends of beams. These values were obtained from moment-rotation (M-θ ) relationships. A 3-dimensional finite element model was built and used to develop M-θ ) relationship of precast billet connection. Results demonstrate that rotational stiffness of billet connection (23138kNm/rad) leads to a fixity factor of 0.4 which is significantly different from the current practice of pinned connection. A 5-storey concrete building with different types of connection was analysed for the progression of damage after an accidental action. Linear static and nonlinear dynamic analyses associated with alternative load path method were performed. The collapse potential was assessed through demand capacity ratio, maximum deflection, ductility and rotation demand. Analyses show that precast frame with semi-rigid connections has higher potential than rigid connection, when one internal load-bearing element is damaged. Several parametric studies were conducted to investigate the sensitivity of progressive collapse. Results show that number of floors, location of column failure, connection flexibility and duration of column removal do affect the collapse potential results. Of these, the location of column failure is the most significant and the duration of column failure the least. In addition, the dynamic impact factor for the frame with semi-rigid connections were found to be 35% less than the usual guideline of 2. Finally, a new global damage index is proposed to estimate damage as a function of column location, fixity factor and local damage measure. A good relationship was found between the index, connection rigidity and damage.

A structure is robust if it can res'ist any action without disproportionate consequences. If it is vulnerable to any action under any circumstance it cannot be robust. Progressive failure of Ronan Point building in London in 1968, Alfred Murrah building in Oklahama in 1995 and World Trade Centre in 200 I have highlighted the need for reducing vulnerability. There are many measures of structural vulnerability and robustness, each with its own strengths and weaknesses. Some only depend upon the structural form and the others more concentrate upon the loading. The objectives of this research are to investigate the vulnerability and robustness of structures through an analysis of the form of the structure and the distribution of strain energy and to propose measures to improve it. A theory of structural vulnerability, previously developed at Bristol, focuses on the form of structure to identify vulnerable failure scenarios. A measure of structural well-formedness which uses member stiffness matrices and connectivity is central to it. An improved measure of structural well-formedness is developed. This better accounts for the supports and leads to better results for many cases such as the loss of a ground storey column which is common in progressive collapse studies. This is a hazard independent analysis and helps to identify low probability, high consequence scenarios. The consequences of an initial damage to a structure can be strongly sensitive to the applied loads. A structure will remain stable if the structural elements are able to redistribute released energy and dissipate any excess energy. The distribution of energy depends on both the form of the structure and the loading. Hence, a new approach to vulnerability based on strain energy distribution is developed. Numerical experiments are carried out to examine energy transition characteristics and the ability of the structural members to absorb energy. These show that the sequence of damage progression can be obtained from energy distribution. Higher gains in strain energy densities of members after the loss of a member indicate higher damage potential. A robust structure would have a smaller value of structural strain energy density and an even distribution of energy amongst members. It is also found that strengthening certain combination of members can significantly lower the changes in strain energy after damage and thus prevent collapse propagation. Several other indices to evaluate structural vulnerability are studied and a comparison is made with the newly developed energy approach. Numerical examples show that the new energy approach works well. It captures the practical structural responses better than stiffness-based indices. The vulnerability of a structure can be reduced by controlling the flow of energy and dissipating energy as damage propagates. For this purpose, the use of shape memory alloys (SMA) is investigated experimentally and numerically. Small models of concrete beams and columns incorporated withmartensite and austenite SMA cables are tested. The austenite SMA concrete members show high recovery and resistance capacities. The martensite SMA concrete members are found to be good at dissipating energy. The tests suggest that is possible to improve the robustness of structures by reinforcing with both martensite and austenite SMAs and larger scale tests may be carried out in future work.

No description available.

No description available.

No description available.

The design and construction of large $1.2$~$m$ diameter lysimeters has been implemented to monitor the soil water retention behaviour and permeability characteristics of contaminated soils under remediation. The work was carried out as part of a larger project focussing on the sustainable remediation of low value brownfield land. Three lysimeters have been filled with lead contaminated soil: one control; one with a \ac{WTR} amendment; and one with a \ac{WTR} and compost amendment. A new software system was built to control the \ac{TDR} point water content measurement and irrigation system, which could log data to an online unified data repository; provided an interface for connectivity to any serial port device; deal with templating for simplified setup; and realtime feedback for the end user. High capacity tensiometers were used in conjunction with the \ac{TDR} point water content measurement system to read volumetric water contents and suctions in the large control lysimeter over a series of wetting and drying cycles, each lasting several months. The results demonstrate that there was a difference between small scale laboratory tests and the data obtained from the lysimeters, particularly in the near surface soil due to cracking. Where cracking was not present, the agreement was stronger, but differences suggested that the drying curves in the lysimeter was predominantly scanning behaviour whereas the element tests were likely more representative of primary drying behaviour.

Engineered Cement Composite (ECC) materials have the potential to be used in civil engineering applications where a level of pseudo-ductility is required. Of particular interest is the possibility of eliminating the steel from reinforced cementitious structures ensuring that no long-term corrosion exists, which is especially relevant for hydraulic tunnels. Uncertainties remain, however, with regard to the mechanical performance, physical properties, durability and shrinkage of these materials, especially when they are used in thick sections for large scale engineering structures. The current work has studied ECCs in this light. The physical properties such as density, porosity and fibre dispersion and orientation are also of interest: this forms the classic materials engineering triangle of the links between material composition and manufacturing process, the microstructure and mechanical properties. A cementitious matrix, reinforced with polymeric fibres, has been manufactured using two different processes and fibre types. Specimens have been tested in tension and flexure, and multiple matrix cracking has been observed, which leads to a pseudo-ductile behaviour. Enhanced mechanical performance in tension is in line with a greater fibre alignment and higher levels of porosity do not necessarily lead to a loss of pseudo-ductility. Flexure testing shows a pseudo-ductile behaviour maintained for over three years. The fibre surface coating and the interfacial properties are relatively stable, which is in line with the maintaining of the pseudo-ductility. Theoretical models suggest that the results are in line with the ACK model, particularly for a fibre volume fraction which considers fibre orientation and fibre pull-out is likely to be responsible for the pseudo-ductile behaviour of the material. ECC materials tend to exhibit a high shrinkage on cure; this could result in cracking, which could compromise the longevity of structures. Methods for controlling shrinkage include the controlling of the environment and use of additives such as powder micro-silica.

No description available.

No description available.

In the design of reinforced concrete structures, a designer must satisfy not only the strength requirements but also the serviceability requirements, and therefore the control of the deformation becomes more significant. To ensure serviceability criterion, it is necessary to accurately predict the cracking and deflection of reinforced concrete structures under load. For accurate determination of the member deflections, cracked members in the reinforced concrete structures need to be identified and their effective flexural rigidities determined. The effect of concrete cracking on the stiffness of a flexural member is largely dependent on both the magnitude and shape of the moment diagram, which is related to the type of applied loading. The aim of the present investigation is to study the cracking behaviour (crack spacing and crack width) of reinforced concrete beam elements constructed from high performance concrete and subjected to bending. An initial optimization was performed to identify a suitable mix design, which has the different required performance attributes of high performance concrete, such as strength, workability, dimensional stability and durability. It is likely that this stage will confirm two mixes, which can be used to compare and contrast the cracking behaviour and to isolate the effect of creep on the long-term performance of such elements constructed from high-strength concrete. Typically, a concrete compressive strength of approximately 100MPa will be sought. High performance concrete is made by partially replacement of cement by silica fume. Once the mixes identified, a series of beams (typically 4.2 m span with a constant moment zone of one and half metres) were tested under 2 point loading (also called 4-point loading including the reactions). These elements were subjected to a constantly applied load for at least three months. The intention was to isolate the effect of creep on the long-term performance of such elements constructed from high-strength concrete. This has not been achieved before. Complimentary tests were also performed on the concrete itself (strength, stiffness, shrinkage, creep (compressive and tensile) etc.). Finally, the deflection, crack width and spacing calculation procedures in several major international concrete codes, including EC2-92 and MC 90, were assessed and some deficiencies in the existing approaches exposed. A comparison between the experimental results, and international codes was presented for all beam specimens time-dependent behaviour. Overall, the programme provided much needed data, which verify the influence of creep and shrinkage on the long-term performance (deflection) of high strength reinforced concrete beam elements.