The object of this investigation is to review the rational and safe use of plain, reinforced and prestressed concrete as structural media. Developments in technology seem to precede science, which in turn has to explain these achievements systematically and rigorously. This order of events, in the case of concrete technology, stretches back thousands of years, yet only centuries to the beginnings of conscious scientific postulations. Therefore, it should not be surprising that even today many inconsistencies, or indeed contradictions, persist to cloud fundamental issues with regard to safety. Following a systematic historical review, and an explanation of the present State of the Art, the Author presents simple design recommendations based on a full recognition of the fundamental role of the tensile strength of concrete, bearing in mind its inherent elasto-plastic properties. These recommendations acknowledge the rationality of the current ultimate limit state concepts with respect to flexure and postulate straightforward yet safe proposals for resolving the continuing confusion with regard to combined shear and flexure.

This thesis presents the dynamic behaviour of a rotor consisting of multidiscs on a solid or hollow shaft, of the type used in gas or steam turbines. The effect of shaft flexibility on the dynamic characteristics of discs and the coupling effects between the discs and shaft modes are investigated. The influence of the disc flexibility on the hehaviour of single span and multispan shaft system is also investigated. A new, thick, three-dimensinnal, Isoparametric, cylindrical element is developed to determine the axial, bending and torsional modes and their coupling effects on the vibrating shaft-discs systems. The element is chosen to have eight nodes at its corners with twelve degrees of freedom at each node giving ninety-six degrees of freedom for a complete element. The triple integration over the element of stiffness and mass matrices are evaluated algebraically and numerically by Gaussian Quadrature using a 4 x 4 x 4 mesh. The main objective of the research reported in this thesis is to fulfil the requirement for a theoretical solution capable of accurate analysis, and the development of finite element program using a double precision arithmetic to obtain the dynamic characteristics of a multirotor system. The effect of disc flexibility on the dynamic characteristics of a system where the discs are carried in the interior of thin-walled cylindrical hollow shafts or "drums" are also studied. This type of disc-drum assemblies are used in aeroengines. Advantage of the rotational periodicity and linear periodicity of the wave propagation technique is fully utilised in the analysis. A new technique of combination of the two methodsis developed to investigate the dynamic behaviour of multidiscs - multi span systems. The present investigation is divided in two parts. The first part deals with several applications of the study of vibration characteristics of circular and annular discs, sector plates, cylindrical shells, hollow cylinders and shafts for various boundary conditions. The results for such applications are obtained separately by using the thick , three-dimensional element, and are compared with those of existing results given by classical plate and shell theory,by exact and other numerical methods of analysis. They also show very good agreement with the experimental results. The objective of this part is to present the accuracy of the thick, three-dimensional element solution including the effects of rotary inertia and shear deformation on the vibration characteristics in the case of axisymmetric and non-symmetric modes of vibration. The effect of the radii ratio b/a and thickness variation of uniform and variable thickness disc on the frequencies of vibration is discussed. The effect of sectorial angle 0 of the sector plate, the ratio of L/R[m], R[1]./R[o] and R[m]/t of the shaft on the dynamic behaviour are also studied. A thin, two-dimensional annular sector element with twelve degrees of freedom is also developed to find the mass and stiffness matrices and to obtain the frequencies of a uniform and variable thickness disc. A comparison between the results of vibrating disc obtained by the thick three-dimensional element and by this element is made to show the superiority of the thick, three-dimensional element to the thin, two-dimensional element for the dynamic analysis and also to find the influence of shear deformation and rotary inertia on the dynamic behaviour which is not allowed for in the thin plate theory. The large size of the stiffness and mass matrices obtained in the case of a three-dimensional cylindrical element analysis are reduced by using the eigenvalue economizer technique. The second part of this investigation is to present the coupling influence and interaction between the discs and the shaft by using the thick, three-dimensional element. Results presented for various cases with differing flexibility and geometry show clearly the coupling effects in a multi disc-shaft system. An inference diagram is developed from which the dynamic behaviour of a system can be predicted for differing flexibility relationships between the solid or hollow shafts and the discs. The effect of various fixing conditions and boundary conditions on the coupling characteristics are also presented. The effectiveness, simplicitiy of use and significance of the element applied and its superiority over other elements in the dynamic analysis is studied. The theoretical results show very close agreement with the experimental results.

Computer-aided structural analysis processes are highly dependent on the use of computer graphics. The objective of this work is to evolve techniques that allow structural analysts, designers and architects to work with computer graphics in a convenient manner. The formex approach of data generation is explained through a number of examples. This approach enables data to be generated very conveniently for the purposes of structural analysis. Also, introduced are the main features of an interactive programming language which acts as a vehicle to implement the concepts of formex algebra. An attempt to investigate the possibility of using the concepts of formex graphics in postprocessing stages of structural analysis is presented. This enables output of structural analysis programs to be graphically displayed so that plots of structural configurations can be shown in both their deformed and undeformed shapes. It is also shown that it is possible to employ the concepts of formex graphics in order to produce axial force, shear force, bending moment and torque diagrams in a manner that they can be visualized conveniently.

A review is made of existing theories for predicting the tensile stress-strain behaviour and the mechanisms of fibre-matrix stress transfer for fibre reinforced cements. The applicability of the inherent assumptions in the theories to practical fibre cements and polyalkene cements in particular is discussed. A description is given of the structure, mechanical properties and production techniques of fibrillated polyalkene films in order that their properties may be related to those of the composites. Five types of fibrillated polyalkene film and a monofilament were incorporated in a high strength cement based matrix. The general composite tensile behaviour was examined with particular reference to the "first cracking" strain of the composite and the fibre-matrix stress transfer mechanisms. The films were mainly opened to four times their unopened width but two composites were made with unopened film to increase the axial alignment of the fibrils. The majority of testing was after 28 days water curing with two composites also being cured for 1 year. Composites were monitored for load,strain and acoustic emission measurements. A small number pull out tests were performed on narrow unfibrillated strips of film. It is shown that the general behaviour of the composite complies with existing theoretical predictions. However, a new model is proposed for predicting the "first cracking" strain of the matrix. The fundamental mechanism of fibre-matrix stress transfer is attributed to misfit resulting from the non-uniform cross section of the individual fibrils. The implications of this work are assessed and a tentative film specification is proposed for the production of improved composites.

New technologies for the manufacturing of glass reinforced plastics (G.R.P.) with possible outlets into the construction industry have broadened the capabilities of this material as well as the structural efficiency in design and construction. The pultrusion technique, although not a new production process, has not been used greatly in the structural field in spite of its enhanced strength and stiffness values, over the hand-lay production method, due to the unidirectional arrangement of the fibres. However, if a folded plate continuum G.R.P. structure of low modulus material requires to be stiffened by a compatible material, pultruded sections could be used to form a skeletal/continuum system with a much improved performance, the present work has investigated G.R.P. skeletal/continuum systems which are connected only at nodal points. In the present work the pultruded sections have a glass/polyester ratio of 65-35% by weight whilst the continuum component is manufactured by the hand lay-up method using chopped strand mat glass laminates with polyester resin with a glass/polyester ratio of 30-70% by weight. The theoretical analysis was undertaken by the finite element technique which provided a linear and a stability analyses of the skeletal/continuum G.R.P. space structures. The analytical procedure involved the combination of the stiffnesses of two types of elements in one overall stiffness matrix, these elements are the line element representing a two ended skeletal member in space of six degrees of freedom per node; three translational and three rotational (deltax, deltay, deltaz, thetax, thetay, thetaz) given by Livesley (9) and a rectangular plate element of four nodes and compatible degrees of freedom per node with the line element developed by Scordelis (15). The stability analysis was based on small displacements and the distinct bifurcation point related non-linearly only with the level of axial stresses in the skeletal members and the continuum. Small scale models were manufactured from perspex materials and tested; these were undertaken mainly to verify the theoretical analysis. The analytical solutions were then used to undertake parameter studies. Applications to G. R. P. composites included flat plates stiffened by pultruded members and a prototype vee sectioned composite roof system. All G. R. P. models were tested to failure to investigate experimentally the buckling characteristics of the composite systems; the results were compared with the buckling loads predicted theoretically. It was concluded that the theoretical method satisfactorily predicted the linear behaviour of skeletal/continuum systems as well as the stability behaviour of such systems in predicting the buckling loads of these composites.

The use of multi-component light-weight structures has increased considerably during recent years. Collapse of such structures may be propagated by progressive failures of their components and the load-deflection curves of these structures may be characterised by a number of limit points. The fundamental mechanics of sequential collapse behaviour of these structures is presented herein. Expressions relating the loads and deflections are derived for a general structural system. The family of curves obtained describes the essential characteristics of the sequential instability behaviour of large redundant structures. Simple structural systems displaying the complete set of characteristics are presented and investigated in detail. Large-deflection elastic buckling models and a plastic buckling model are considered. It is found that member behaviour and connectivity between the members have extremely important effects on the structure behaviour. The theoretical work is supported by experimental evidence. Tests are performed on a model structure subjected to large elastic deformations. High strength steel is used to construct the model to ensure elastic behaviour throughout. An extensive and thorough experimental investigation is carried out to obtain the complete set of equilibrium characteristics. The experimental results are in complete agreement with the general and detail theoretical work. The theoretical critical behaviour of the models investigated in this thesis is shown be governed by Thom's Swallowtail Catastrophe and this is verified by the experiments. These structures exhibiting a series of limit points in their load-deflection curves are prone to demonstrate dynamic snap under dead loading. As a consequence, the concept of dynamic snap is examined. The occurrence of premature global failure due to dynamic snapthrough is deduced from a study of the form of the potential surface. An investigation is also made of the important effect of initial imperfections on the sequential buckling process. It is shown that the presence of imperfections may markedly change the form of a collapse sequence for particular structural systems. The use of "Force Limiting Devices" to improve the collapse characteristics of structures liable to sequential instability is also investigated.

This work presents a practical approximate technique for the analysis of dense space structures, with particular reference to single layer space structures. The technique is exemplified in terms of barrel vaults and flat grids. Following a general introduction in Chapter 1, Chapter 2 provides a general survey of the methods employed in the analysis of space structures, highlighting improvements to the methods of analysis which have been made to reduce the cost of analysis and storage requirements. The concept of formex algebra is described in Chapter 3, together with its application to data generation for structural analysis. A technique for frontwidth minimisation for frontal solution technique is also presented, with a number of illustrative examples. In Chapter 4, the concept of structural factoring is proposed as a basis for approximate analysis of very dense space structures. The Chapter gives the procedure by which the concept of structural factoring is used to estimate the structural behaviour of a dense space structure from the result of analysis of a similar structure with fewer number of nodes and members. Chapter 5 gives the details of all the barrel vaults analysed and the parameters varied to assess the accuracy of the proposed method. In Chapter 6, the results of the analysis are discussed. The results obtained by the method are compared with those obtained by the 'exact' analysis of barrel vaults and flat grids. The conclusions are given in Chapter 7, together with the suggested extensions.

This thesis examines the behaviour of a 114.0 mm diameter segmental tubular steel pile jacked into loose sand, and loose sand overlying clay. The soil was placed under controlled conditions in a 3.0 m diameter by 3.0 m deep concrete tank. The variation in local unit shaft friction and radial effective stress was monitored along the pile shaft, together with the distribution of axial load within the pile. Density changes within the sand were recorded at the end of pile installation. Vertical displacements and vertical effective stresses within the sand were recorded. In the case of the layered soil profile the shear and vertical effective stresses generated on the sand/clay interface were monitored. Data from both the pile and soil instrumentation was recorded throughout pile installation and load testing, consisting of CRP, ML and CRU tests, by an Orion Data Logger which was interfaced with a Commodore PET micro computer. Each stage of the test was controlled by a 'Management' program, written by the author. This also recorded the incoming raw data on a floppy disc and reduced the raw data, outputting a hard copy as the test proceeded. The results showed: (i) The local unit shaft friction and radial effective stress is practically constant along a pile shaft in sand for a given pile embedment, and increases at a diminishing rate with pile embedment. (ii) The average coefficient of earth pressure, K , at ultimate load in loose sand exceeds Kp for shallow pile embedments. (iii) At full pile embedment and ultimate load the local coefficient of earth pressure, Kza, may greatly exceed K_ near the top of the pile and tend to a lower limiting value of 0.5 near the pile base. (iv) Axial stresses within the sand around the pile shaft are reduced by the development of arching. Adjacent to the pile shaft the radial effective stress is the major axial stress. (v) The development of shaft friction is directly related to displacements within the surrounding sand and on the sand/clay interface. (vi) The presence of an underlying clay layer effects the development of shaft friction to a limited height above the sand/clay interface. (vii) The drawdown of sand into the underlying clay had a direct effect the local unit shaft friction developed within the clay.

Non metallic bearing plates have been identified as the weakest component in GRP rockbolting systems used in UK coal mines. A literature review revealed a shortage of information on the design of rockbolting components, particularly the design of the non metallic bearing plates. Over twenty parameters concerning the design and testing of non metallic rockbolt bearing plates have been investigated using Finite Element Analysis and laboratory testing. The parameters included bearing plate depth, central hole diameter, external diameter, material, coned angle for both solid and webbed bearing plates. The designs have been evaluated by comparing the load/plate volume to consider the efficient use of the material. The experiments were designed using Taguchi and one factor at a time methodologies. Interactions between some parameters have been investigated. Reasons for the observed effect of parameters have been suggested. A modification of the BS 7861 test is proposed which more closely simulates the colliery failure mechanism and hence gives a better measure of a bearing plate's suitability for use in a coal mine. The modified test uses a 100 mm hole in the steel support plate as specified by American Standard F432 for steel rockbolt bearing plates, not the 55 mm hole as specified by BS 7861.Optimum values for important parameters have been determined for a 100 mm hole in the steel support plate as proposed in this research and for the 55 mm hole diameter specified by BS 7861. The results produced can be used to design a bearing plate for use in a coal mine with optimum depth and coned angle.

No description available.