Özet:

Experimental techniques are of great importance in determination of earthquake behavior of structures and structural members. Structural behavior could be determined by means of quasi-static tests in which prescribed displacement protocols are applied, by real time shake table tests realized with real acceleration records or by pseudo-dynamic tests in which experimental and numerical simulation efforts work together. In pseudo-dynamic test, the reactional forces measured from test specimen (stiffness matrix) are utilized for the numerical solution of dynamic equilibrium equation and for the determination of target displacement vector to be applied to test specimen in the next step. In pseudo-dynamic test method, the through interaction of the analytical and experimental parts is very effective on the success. For the success of the test; in addition to the sensitivity of the measuring devices used, the stability of the numerical integration technique to be used in the solution of the dynamic equilibrium equation is also of great importance. In order to ensure the stability of the numerical integration, the selected time interval should be small enough. In this study, the central difference method (CDM) and linear acceleration-based force separation method (Li-OSM) used as a numerical integration tool in pseudo-dynamic test method are compared with numerical and experimental studies. The performed numerical studies; regardless of the constitutive model used, showed that Li-OSM method produces more stable results than CDM method in the case of the integration step interval selected up to half of the earthquake record interval. From the experimental studies; instead of using “quarter” integration step interval in CDM method, it was concluded that in case of using “half” integration step interval in Li-OSM method, the total testing time will decrease by half and the stability of the test will increase.

Anahtar Kelimeler: