Columns are significant contributor to the lateral force resisting system in reinforced concrete moment frames. Failure of these structural members can cause stability issues in a framed structure, and in worse cases, can even lead to global collapse. Hence, the design of column carries an important place in modern seismic codes where it is emphasized that these structural members suffer close to no damage. Although the design of reinforced concrete moment frames in Pakistan is carried out using ACI-318, the design engineers do not have an effective control on the properties of steel bars used in the construction industry. The steel bars available in the local market, like in many other developing countries, lack homogeneity and routinely exhibit mechanical properties different than the specified mechanical properties. Due to this mismatch in the mechanical properties, the design of the structural members has the potential to be unreliable. While the problem is similar for all the structural members, the implications of an uncertain and unreliable design are more serious for columns, as overestimation of strength or deformability related parameters may initiate collapse. Hence, there is a need to study the behaviour of concrete columns reinforced with steel bars that have this mismatch between actual and specified mechanical properties. To cater to this need, an experimental investigation was carried out on spliced reinforced concrete columns having steel bars having higher than the specified yield strength, hence having a mismatch in the actual and specified mechanical properties. The variables explored were, axial load ratio, presence of cold joint at base, concrete strength, splice location, reinforcement ratio, type of steel (hot-rolled versus cold-twisted),  transverse reinforcement ratio, confined concrete core area, different drift interval and type of loading (cyclic versus monotonic). Higher degree of damage and strength degradation, lower ductility and rotation capacities were found for specimens having spliced steel bars with the un-warranted increase in the yield strength. Converse behaviour was observed for the specimen having spliced reinforcement without the un-warranted increase in the yield strength. These observations were attributed towards the splice deficiency which originated due to the un-warranted increase in the yield strength. Formulations pertaining to estimation of rotation capacity and lateral load capacity were found to be unsafe for spliced columns with steel having higher than specified yield strength, conversely, these were found to be safe for the spliced columns having steel bars without the un-warranted increase in the yield strength. Numerical modeling was done using a modified steel stress-strain curve to accommodate strength degradation in the load-displacement response and good correlation was noticed between the model and experimental results. Formulations for estimation of maximum developable stress in the splice zone and ultimate rotation capacity have been proposed so that safe estimation of these parameters can be made for spliced columns having steel bars with un-certain yield strength.

The rapid global urbanisation has led to increased construction activity, resulting in a high demand for raw materials and subsequent environmental degradation. To address this issue, recycled aggregate concrete (RAC) has emerged as a viable solution by utilising concrete waste from demolished structures. However, the incorporation of RAC into structural applications requires a better understanding of the behaviour of RAC under multiaxial states of stress. This can be achieved by developing a constitutive model that can simulate its behaviour under various stress conditions. This study, therefore, focuses on developing a constitutive model for RAC within the framework of damage mechanics, using artificial intelligence (AI) techniques to estimate the model parameters. This study modifies existing elasto-damage natural aggregate concrete (NAC) model proposed in literature by Khan and Zahra to incorporate the effect of replacing natural coarse aggregates with recycled coarse aggregates. The proposed constitutive model incorporates essential features of concrete, such as different behaviour in tension and compression, stiffness degradation, strain softening, strength gain under confinement, and volumetric dilatation. Also the proposed model considers incorporation of plastic strains along with damage. Four parameters of the model i.e. compressive strength (f_c^'), Elastic Modulus (E), and calibrated parameters that control the phenomenoologically known peak strengths to model the different behaviour of concrete in tension and compression (𝛼 and 𝛽), were modified and predicted using artificial neural network (ANN). Compressive strength and elastic modulus were defined as a function of different parameters (water to cement ratio, amount of cement, water, natural and recycled coarse aggregate, fine aggregate, fly ash and superplasticizers contents, and the replacement ratio of recycled aggregate concrete). 𝛼 and 𝛽 are defined as functions of concrete compressive strength, elastic modulus, and stress paths. Experimental investigations were conducted at different percentages of recycled coarse aggregate replacement i.e. 0%, 30%, 50%, 70% and 100% at three different water to cement ratio of 0.4, 0.5 and 0.6 to determine mechanical properties and develop experimental stress-strain curves. Experimental results along with existing published data were used to predict the aforementioned parameters. The constitutive model was validated with the existing experimental data. The resulting constitutive model was designed to be easily integrated into finite element codes. Furthermore, experiments were conducted reflecting multiaxial state of stress conditions. These were simulated in nonlinear FEM software (ATENA-GiD) after incorporating proposed constitutive model in the software. The results showed that the proposed model can predict peak stress, initial stiffness, post peak behaviour in a good manner and overall results were found to be in a close agreement with the experimental and published results. 

The rapid global urbanisation has led to increased construction activity, resulting in a high demand for raw materials and subsequent environmental degradation. To address this issue, recycled aggregate concrete (RAC) has emerged as a viable solution by utilising concrete waste from demolished structures. However, the incorporation of RAC into structural applications requires a better understanding of the behaviour of RAC under multiaxial states of stress. This can be achieved by developing a constitutive model that can simulate its behaviour under various stress conditions. This study, therefore, focuses on developing a constitutive model for RAC within the framework of damage mechanics, using artificial intelligence (AI) techniques to estimate the model parameters. This study modifies existing elasto-damage natural aggregate concrete (NAC) model proposed in literature by Khan and Zahra to incorporate the effect of replacing natural coarse aggregates with recycled coarse aggregates. The proposed constitutive model incorporates essential features of concrete, such as different behaviour in tension and compression, stiffness degradation, strain softening, strength gain under confinement, and volumetric dilatation. Also the proposed model considers incorporation of plastic strains along with damage. Four parameters of the model i.e. compressive strength (f_c^'), Elastic Modulus (E), and calibrated parameters that control the phenomenoologically known peak strengths to model the different behaviour of concrete in tension and compression (𝛼 and 𝛽), were modified and predicted using artificial neural network (ANN). Compressive strength and elastic modulus were defined as a function of different parameters (water to cement ratio, amount of cement, water, natural and recycled coarse aggregate, fine aggregate, fly ash and superplasticizers contents, and the replacement ratio of recycled aggregate concrete). 𝛼 and 𝛽 are defined as functions of concrete compressive strength, elastic modulus, and stress paths. Experimental investigations were conducted at different percentages of recycled coarse aggregate replacement i.e. 0%, 30%, 50%, 70% and 100% at three different water to cement ratio of 0.4, 0.5 and 0.6 to determine mechanical properties and develop experimental stress-strain curves. Experimental results along with existing published data were used to predict the aforementioned parameters. The constitutive model was validated with the existing experimental data. The resulting constitutive model was designed to be easily integrated into finite element codes. Furthermore, experiments were conducted reflecting multiaxial state of stress conditions. These were simulated in nonlinear FEM software (ATENA-GiD) after incorporating proposed constitutive model in the software. The results showed that the proposed model can predict peak stress, initial stiffness, post peak behaviour in a good manner and overall results were found to be in a close agreement with the experimental and published results. 

The rapid global urbanisation has led to increased construction activity, resulting in a high demand for raw materials and subsequent environmental degradation. To address this issue, recycled aggregate concrete (RAC) has emerged as a viable solution by utilising concrete waste from demolished structures. However, the incorporation of RAC into structural applications requires a better understanding of the behaviour of RAC under multiaxial states of stress. This can be achieved by developing a constitutive model that can simulate its behaviour under various stress conditions. This study, therefore, focuses on developing a constitutive model for RAC within the framework of damage mechanics, using artificial intelligence (AI) techniques to estimate the model parameters. This study modifies existing elasto-damage natural aggregate concrete (NAC) model proposed in literature by Khan and Zahra to incorporate the effect of replacing natural coarse aggregates with recycled coarse aggregates. The proposed constitutive model incorporates essential features of concrete, such as different behaviour in tension and compression, stiffness degradation, strain softening, strength gain under confinement, and volumetric dilatation. Also the proposed model considers incorporation of plastic strains along with damage. Four parameters of the model i.e. compressive strength (f_c^'), Elastic Modulus (E), and calibrated parameters that control the phenomenoologically known peak strengths to model the different behaviour of concrete in tension and compression (𝛼 and 𝛽), were modified and predicted using artificial neural network (ANN). Compressive strength and elastic modulus were defined as a function of different parameters (water to cement ratio, amount of cement, water, natural and recycled coarse aggregate, fine aggregate, fly ash and superplasticizers contents, and the replacement ratio of recycled aggregate concrete). 𝛼 and 𝛽 are defined as functions of concrete compressive strength, elastic modulus, and stress paths. Experimental investigations were conducted at different percentages of recycled coarse aggregate replacement i.e. 0%, 30%, 50%, 70% and 100% at three different water to cement ratio of 0.4, 0.5 and 0.6 to determine mechanical properties and develop experimental stress-strain curves. Experimental results along with existing published data were used to predict the aforementioned parameters. The constitutive model was validated with the existing experimental data. The resulting constitutive model was designed to be easily integrated into finite element codes. Furthermore, experiments were conducted reflecting multiaxial state of stress conditions. These were simulated in nonlinear FEM software (ATENA-GiD) after incorporating proposed constitutive model in the software. The results showed that the proposed model can predict peak stress, initial stiffness, post peak behaviour in a good manner and overall results were found to be in a close agreement with the experimental and published results. 

Reinforced concrete (RC) structures may lose their strength over their design lifetime period due to various reasons. The change of mode of usage and introduction of more rigorous code requirements and loss of strength due to any reason would require the structure or elements of structure to be strengthened. The use of fibre composites such as textile reinforced mortar (TRM) has been an effective recent solution which has increased the interest of researchers and scientists to further investigate its potential in strengthening of reinforced concrete structures. Although carbon, glass, and PBO fibres have been explored significantly and have shown upshots that are encouraging, the use of basalt fibres has mostly been limited to strengthening of masonry structures. Infrequent small-scale data can be found for TRM that has used basalt as its fibres, and the data for full scale beam is inadequate. This thesis presents an experimental and numerical study of two schemes of full-scale strengthened RC beams along with corresponding control beams. The control beams and the two schemes are a set of 6 beams each with shear span to depth ratios varying from 1 to 6. The beams were strengthened using TRM with basalt fibres. Modelling of the testing beams was also done on ATENA software along with the parametric study. The results show that basalt fibres are effective in improving the performance of RC beams in terms of serviceability, crack and deflection control, load carrying capacity, initial and post cracking stiffness and ductility. The parametric study showed that the number of layers of TRM also improves the performance in terms of load carrying capacities, while varying the depth of U-shaped TRM wraps up to the neutral axis from the top of the beam does not affect the performance of TRM material for flexural performance enhancement.