COMPARISON OF NUMERICAL ANALYSIS OF A SINGLE-SPAN STEEL PROTOTYPE STRUCTURE AND A SCALE MODEL STRUCTURE UNDER THE EFFECT OF SEISMIC LOADS

Abstract

One of the biggest problems encountered in many experimental studies is examining a real-size structure in the field or in a laboratory environment. With today's technological opportunities, it is possible to experimentally examine a real-sized structure in the field or in a laboratory environment. However, to do this, the manufacture of a large, real-size structure, experimental setup and measuring devices are required, which are costly. For this reason, it is not always possible to reach such a laboratory environment. It is very difficult to experimentally examine large-scale structures both economically and in terms of time saving. In this context, in this study, a scaling factor (λ) widely accepted in the literature was used to design a scaled model to represent a real-size structure. λ=10 was used in this scaling approach. A real-size three-story single-span steel prototype building was scaled to a laboratory-scale model structure and analyzed digitally with the Sap2000 program. The natural period/frequency values of the real-size prototype structure and the scaled model modeled in the Sap2000 program were examined. Later, Time history analyzes were performed using real earthquake records from El Centro (1940), Kobe (1995) and Northridge (1994). While real earthquake records were used as they were in the analysis of the prototype structure, these real earthquake records were used by scaling them depending on the scaling factor λ in the analysis of the scaled model. Subsequently, the digital analyzes of the prototype and scaled structure were compared by looking at the acceleration and displacement values of each floor. It was observed that the results were close to each other when scaled according to the scaling factor (λ). This situation demonstrated the accuracy of the scaling rates applied within the scope of the study. Thus, it has been shown that a real-size structure can be scaled to a model in a laboratory environment with correct scaling methods and that this prototype structure can be analyzed with more economical and simple methods.