Thesis (M. Phil.)--University of Hong Kong, 1974.

Thesis--M. Phil., University of Hong Kong, 1978. / Errata slip inserted.

Thesis (M. Sc.)--University of Hong Kong, 1960. / Type-written copy. Includes bibliographical references.

Thesis--M. Phil., University of Hong Kong.

published_or_final_version / Civil Engineering / Doctoral / Doctor of Philosophy

The stress-strain characteristics of concrete developed in flexure is very important for flexural strength design of reinforced concrete (RC) members. In current RC design codes, the stress-strain curve of concrete developed in flexure is obtained by scaling down the uni-axial stress-strain curve to account for the strain gradient effect. Therefore, the maximum concrete stress that can be developed under flexure is smaller than its uni-axial strength, and the use of which always underestimates the flexural strength of RC beams and columns even though the safety factors for materials are taken as unity. Furthermore, the value of strength underestimation was different for RC beams and columns, which indicates that the extent of strain gradient will affect the maximum concrete stress and stress-strain curve developed under flexure. To investigate the maximum concrete stress, 29 column specimens were fabricated and tested in this study. They were divided into 9 groups, each of which was poured from the same batch of concrete and contained specimens with identical cross-section properties. In each group, one specimen was tested under concentric load while the rest was/were subjected to eccentric or horizontal load. To study the strain gradient effects, the ratio of the maximum concrete compressive stress developed in the eccentrically/horizontally loaded specimens to the maximum uni-axial compressive stress developed in the counterpart concentrically loaded specimens, denoted by k3, is determined based on axial force and moment equilibriums. Subsequently, the concrete stress block parameters and the equivalent rectangular concrete stress block parameters are determined. It is found that the ratios of the maximum and equivalent concrete stress to uni-axial cylinder strength, denoted respectively by k3 and , depend significantly on strain gradient, while that of the depth of stress block to neutral axis depth, denoted by , remains relatively constant with strain gradient. Design equations are proposed to relate and  with strain gradient for strength calculation, whose applicability is verified by comparing the strengths of RC beams and columns tested by various researchers with their theoretical strengths predicted by the proposed parameters and those evaluated based on provisions of RC codes. Based on the test results, the stress-strain curve of normal-strength concrete (NSC) developed under strain gradient is derived using least-square method by minimising the errors between the theoretical axial load and moment and the respective measured values. Two formulas are developed to derive the flexural stress-strain curve, whose applicability is verified by comparing the predicted strength with those measured by other researchers. Lastly, the application of the proposed stress-block parameters and stress-strain curve of NSC will be illustrated by developing some charts for flexural strength design of NSC beams and columns. The application will further be extended to develop strength-ductility charts for NSC beams and columns, which enable simultaneous design of strength and ductility. By adopting the proposed design charts, the flexural strength design, as well as that of the plastic hinge forming mechanism during extreme events, will be more accurate. The resulting design will be safer, more environmentally friendly and cost effective. / published_or_final_version / Civil Engineering / Doctoral / Doctor of Philosophy

The creep and shrinkage of concrete and the relaxation of cables cause long-term redistribution of internal forces and time-dependent deformations in concrete structures, steel-concrete composite structures and concrete cable-stayed bridges. They should be properly modelled for accurate prediction of their long-term behaviour and its effect on instantaneous structural responses at service. An equivalent stress relaxation model is proposed for prestressing tendons based on the intrinsic stress relaxation, from which the equivalent creep coefficients can be obtained through a recursive algorithm. Based on the equivalent stress relaxation model, an accurate finite element analysis of time-dependent behaviour by time integration has been devised considering concrete creep, concrete shrinkage and cable relaxation. Concrete members are modelled by beam-column elements while tendons are modelled by truss elements with nodes connected to the beam axis by perpendicular rigid arms. It is found that the proposed relaxation model with time integration can provide a reliable method as well as benchmark solutions for time-dependent analysis. The numerical results obtained indicate that the interactions among these factors should be properly considered in analysing the long-term performance of concrete bridges. Although time integration provides a reliable method for time-dependent analysis, both the computing time and memory requirement increase drastically with the number of time steps as the time-dependent strains of concrete and tendons within a time interval depend on the loading history up to that time. It is therefore necessary to develop a more efficient method to conduct time-dependent analysis. The relaxation-adjusted elasticity modulus is introduced on the basis of equivalent creep coefficients of tendons. Then, an efficient tendon sub-element is put forward to cope with cables with arbitrary profiles. Finally, a more general single-step method is devised using the classical age-adjusted elasticity modulus to account for external loading and creep effect, the shrinkage-adjusted elasticity modulus to consider shrinkage effect and its interaction with concrete creep, and the relaxation-adjusted elasticity modulus to consider the effect of cable relaxation based on the finite element method. The numerical results obtained indicate not only the accuracy of the single-step method but also the significance of interaction among various time-varying factors. Based on the time integration or single-step method, a systematic method is developed to monitor the long-term variations of dynamic properties of cable-stayed bridges taking into account various time-varying factors and geometric nonlinearities. Numerical studies show that, although geometric nonlinearities tend to reduce the natural frequencies, the time-dependent behaviour of concrete more than offsets it and tends to increase the natural frequencies in the long run. A generic method is further presented to investigate the long-term dynamic response of vehicle-bridge interaction systems taking account of time-dependent behaviour. The vehicles are represented by a combination of mass-spring-damper systems while the bridge is modelled by finite elements. The surface roughness of bridge deck is simulated by spectral representation method and introduced to the coupled system properly. Based on the method, the individual and combined effects of various time-varying factors are studied in detail using various numerical examples. / published_or_final_version / Civil Engineering / Doctoral / Doctor of Philosophy

Existing reinforced concrete (RC) structure can be strengthened upon the addition of externally bonded high-strength light-weight fibre-reinforced polymer (FRP) composites. An abundance of research over the last two decades has established the effectiveness of the externally bonded FRP via extensive experimental testing. Perhaps the most commonly occurring failure mode though is premature debonding of the FRP and debonding generally occurs at strains well below the strain capacity of the FRP. Debonding failures are undesirable as they are typically brittle and represent an under-utilisation of the FRP material. A straightforward means to prevent or at least delay debonding is by the addition of mechanical anchors, however, research to date on anchors is extremely limited. Of the various anchor concepts examined to date by researchers, this dissertation will focus on anchors made from FRP which are herein referred to as FRP anchors. The details and results of a program of research on the performance of FRP anchors in FRP-strengthened structures are presented in this dissertation. An extensive review of exiting literature helps establish knowledge gaps which serve to justify the need and the scope of the research reported herein. A novel bow-tie FRP anchor concept is then proposed and tested in smaller-scale single-shear FRP-to-concrete joint assemblages as well as larger-scale simply-supported FRP-strengthened RC slabs. The anchors are shown to increase the strength and slip capacity of the joints by up to 41 % and almost 600 %, respectively, in comparison with unanchored control joints. The anchors are then shown to increase the load and deflection capacity of slabs by 30 % and 110 %, respectively, above an unanchored control slab. In addition to strength, it is the ability of FRP anchors to introduce deformability into FRP-strengthened RC slabs which is particularly beneficial in order to produce safer structures. An analytical model is then developed which is based on a novel quad-linear moment-curvature response which can capture the complete load-deflection response of the FRP-strengthened slabs anchored with FRP anchors. The analytical modeling approach enables closed-form equations to be derived which can then be used by design engineers to relatively easily construct load-deflections responses and accurately predict member responses. Following the concluding comments for the project as a whole, future research topics of relevance are identified. / published_or_final_version / Civil Engineering / Master / Master of Philosophy

A nonlinear finite element program to simulate the behavior of reinforced concrete (RC) members under the action of monotonic increasing loading has been developed. The nonlinear response of the RC members is mainly due to the nonlinear material characteristics including nonlinear biaxial stress-strain relations and cracking of concrete and yielding of steel reinforcement. A constitutive model of concrete under biaxial stress state is adopted in this thesis. In this model, concrete fails and critical cracks occur when the tensile strain of concrete exceeds the limiting tensile strain. The complete stress-strain relationship of concrete under compression and tension are employed in the study to investigate the post-peak behavior of reinforced concrete members. An elaborate cracking model has been implemented which allows concrete to crack in one or two directions. The tension stiffening effect of cracked concrete is also incorporated into this model by including a descending branch in the stress-strain curve of concrete under tension. Other nonlinear effects such as crushing of concrete in compression and yielding or strain hardening of steel reinforcement are also taken into account. A nonlinear finite element program was developed, in which the abovementioned nonlinear effects have all been included in modeling the reinforced concrete structures. The nonlinear equations of equilibrium are solved using an incremental-iterative technique performed under displacement control. The validity of the model including the confinement effect of secondary reinforcements has been examined by analyzing three reinforced concrete beams. The performance of the numerical model was assessed by comparing results with those from available experimental data. / published_or_final_version / Civil Engineering / Master / Master of Philosophy

Concerns over the future availability of traditional SCM sources, such as fly ash, have left the concrete industry in need of alternative sources of SCMs. The research presented here has evaluated natural pozzolans such as pumice, perlite, vitric ash, zeolites, shale and calcined clay as alternative sources of SCMs. Unlike previous research that has only concentrated on empirically evaluating the performance of natural pozzolans in concrete, the research presented in this dissertation has measured both the performance of the pozzolans in cementitious mixtures as well as their physical and chemical characteristics, to draw meaningful relationships between pozzolan properties and performance. The physical and chemical characteristics of these natural SCMs were measured using techniques like particle size analysis, Brunauer–Emmett–Teller (BET) surface area, scanning electron microscope (SEM) imaging, x-ray fluorescence (XRF), x-ray diffraction (XRD), and thermal gravimetric analysis (TGA). The performance of the pozzolans as alternative SCMs was examined by looking at their effect on mortar strength and mixture workability, as well as by their ability to mitigate expansions from durability problems like alkali silica reaction (ASR) and sulfate attack. The performance of the pozzolans was related back to their physical and chemical characteristics to gain an understanding of the underlying mechanisms of cement and pozzolan interaction, and to draw insights as to why some pozzolans perform better than others in cementitious mixtures. Using this knowledge, some of the under-performing pozzolans were modified to see if changes in their properties could improve performance. Results of the research showed that other than the two coarse zeolites, the rest of the pozzolans tested could be used as Class F fly ash replacements in concrete, with the pumice, perlite, metakaolin and fine zeolite being the best performers in terms of mortar strength and durability. Although the pumice mortar had lower strengths than the control at early ages, results from the performance improvement studies showed that the reactivity of pumice could be enhanced by grinding the pozzolans to a finer particle size distribution. Zeolites were found to negatively affect mixture workability, but calcination of the zeolites helped to improve the workability of zeolite mixtures. / text