Seismic analysis of full-scale test models of reinforced concrete buildings using the finite element method

Since scaled experiments are conducted in the field or in a controlled laboratory environment, it is necessary to know and determine the limitations of design parameters, their compliance with governing physical laws, and the applicability of all relevant variables included in the physical problem from the beginning. While the reliability of experimental research results using scaled models close to the prototype decreases, the doubt on the data obtained from small-scale model tests increases. In this study, not only a geometric connection but also kinematic and dynamic similarity relationship was established between the prototype and the designed test models that can represent the vibration behavior of a full-scale structure at the desired level. Mathematical steps and design principles were systematically presented in a method for the development of steel frame test models that represent the dynamic behavior of the prototype in the dominant vibration mode during shaking table experiments. In this study, steel frame test models were established with different scale factors based on the dimensional properties and dynamic loading capacity of the small and large-scale shaking tables in the structural laboratory of the Civil Engineering Department at Sakarya University, which were selected as a reference. Dynamic analyses were performed using the Sap2000 finite element program with strong earthquake ground motion records. In addition, twelve earthquake records, including six near-field and six far-field earthquake records were used to observe the dynamic responses of structures depending on the distance of strong seismic ground motion. For the small-scale shaking table, it was determined that a scale factor of λ = 25 would not exceed the dimensional properties and dynamic loading capacity limits of the shaking table. Accordingly, three steel models were developed, including an 8-story scale test model, a single-story scale test model with dual-direction motion, and a single-story scale test model with single-direction motion. The single-direction scale test model was designed considering the weak direction of the prototype structure which is the y-axis. It was aimed to observe the extent of the ability of larger-sized model structures, which will be obtained by decreasing the scale factor using a large-scale shaking table, to represent the prototype structure. For this reason, a scale factor of λ = 14 has been selected for the shaking table in a way that it does not exceed the dimensional characteristics and dynamic loading capacity limits. Additionally, a scale factor of λ = 24 has also been considered in order to develop smaller-sized models for the large-scale shaking table. Three steel models similar to the models developed for the small-scale shaking table have been developed using these scale factors. The dynamic analysis results of the developed multi-storey shaking table test models were compared with the multi-storey prototype structure. However, instead of a multi-storey reinforced concrete prototype, an equivalent single-degree-of-freedom lumped mass system representing it has been compared due to the belief that numerical analysis results of single-storey shaking table test models provide a more accurate approach. When the dynamic analysis results of the equivalent single-degree-of-freedom lumped mass system were compared with the multi-storey prototype structure, it was observed that the error rate reached around 30%. The reason for this error rate reaching these levels is that we neglect the other mode shapes of the prototype structure and design it only based on the dominant mode shape. Additionally, since the design of the single-storey shaking table test models was based on the equivalent single-degree-of-freedom lumped mass system, it was observed that this error rate did not have an effect on the results. While dynamic analysis was performed using actual earthquake records in the Sap2000 finite element program for the reinforced concrete prototype structure, dynamic analysis was carried out for the steel models using the scaled versions of these earthquake records with the respective scale factors. As a result of the dynamic analyses maximum peak displacement, maximum base shear force and maximum overturning moment values occurring in the structures were examined and compared. It has been observed that as a result of the comparisons made, the values of ground acceleration of far-field earthquake records are smaller than those of near-field earthquake records, which leads to smaller displacements in prototypes and models. Although the peak ground acceleration values of Kocaeli, Chi-Chi, and Loma prieta near-field earthquake records were close to each other, it was observed that Kocaeli earthquake record caused larger maximum peak displacements compared to the Chi-Chi and Loma prieta earthquake records in all the prototype and model structures. When the Fourier amplitude-frequency graphs were examined, it was observed that the frequency content of the Kocaeli earthquake record was concentrated at levels close to the dominant frequencies of the prototype structure and the model structures. Resonance occurred due to the close frequencies, and it was observed that larger displacements occurred as a result of resonance. The same situation was observed for the N. palm springs and Morgan hill far-field earthquake records. When the Fourier amplitude-frequency graphs were examined, it was seen that the Morgan Hill earthquake record was close to the resonance frequency. Therefore, larger displacements occurred compared to the N. Palm Springs earthquake record. As a result of the comparisons, it was observed that the model structures generally represented the prototype structure with an error rate of less than 10% except for some outliers. The selected length scale factors were found to be effective in achieving complete similarity, including kinematic and dynamic interactions between the prototype and the models. Moreover, in this study, as a suggestion for researchers adjustable liquid damper energy isolation devices that do not require a separate power source and produce control forces by utilizing the motion of the structure have been designed among many existing passive vibration control systems. When designing tunable liquid column damping systems, the necessary coefficients are selected to keep the manufacturing cost and applicability at an acceptable level and the aim is to reduce the dynamic response of multi-storey prototype structures and single-storey single-directional scaled test models when subjected to strong earthquake ground motion. It has been concluded that the designed tunable liquid column damping systems can be an extremely effective method for dampening seismic energy transmitted to existing high-rise structures with high vibration capacity, especially under the influence of earthquake forces.