Substructuring technique is a method that an original structural model is divided into two parts; experimental and numerical substructures, and then its dynamic characteristic is replicated with only experimental substructure which is manufactured as a physical model and is tested. This paper proposes a shaking table testing method based on the substructuring technique and discuss its experimental verification. By applying the substructuring technique to an original structural model, it is decomposed into two parts; an upper experimental substructure with multi degrees-of-freedom and an lower numerical substructure. At the moment, interface force becomes to act between their interfaces due to artificial dividing into two parts. In this paper, numerical substructure corresponding to the lower part with single or multi degrees-of-freedom of the original multi degrees-of-freedom is incorporated in control computer of shaking table, to produce the interface acceleration by inputting the interface force which is experimentally measured from the upper experimental substructure. Finally, shaking table is used for exciting the upper experimental substructure with the motion of the interface acceleration. Experimental results show the validity of the proposed method that an experiment can be implemented by feedback of the interface acceleration.

This paper proposes substructuring method using experimental model of only upper MDOF structure and verifies its applicability to shaking table test. The structure considered in the test is a five-stoty shear type building model which is divided into two parts, of which one is lower three floors for a numerical substructure and the other is upper two floors for a experinlental substructure. The experimental substructure is excited by the shaking table and the acceleration data for the input of shaking table is obtained numerically from lower numerical part. For implementation of the proposed substructuring method, a controller of the shaking table is designed reflecting experimental model. In comparison with acceleration responses, which are measured from experinlental part of 3rd, 4th, and 5th floor and are numerically calculated from the original five-stoty structure, it is observed that the experimental results match well with the numerical results in both time and frequency domain.

지진력을 받는 건물의 응답을 능동적으로 제어하기 위해서는 건물의 응답을 측정하고, 이것을 바탕으로 제어력을 산정하여야 한다. 제어력의 산정 방법에는 여러 제어 알고리듬이 적용이 될 수 있는데, 2차 성능지수를 이용하는 LQ제어는 해석의 용이함과 제어의 효율성으로 인하여 널리 쓰이고 있다. 그러나, LQ제어에는 실시간으로 계측이 된 건물의 지반과의 상대 변위 및 속도를 필요로 하나 이러한 상태 변수를 계측하기가 매우 어려워 건물의 제어를 위한 적용에 한계가 있다. 따라서, 계측이 용이한 건물의 절대 가속도를 바탕으로 관측이 용이한 건물의 절대 가속도를 바탕으로 관측기를 이용하여 상태 변수를 추정하여 제어력을 산정하는 LQC 제어 알고리듬이 지진력을 받는 건물에 대한 실용적인 알고리듬이 될 수 있다. 본 연구에서는 이러한 LQC 알고리듬의 성능을 검증하기 위하여 능동제어추진기가 설치된 축소 1층 모형에 대한 해석 및 실험을 수행하였으며, 그 결과 LQC 알고리듬의 제어 효율을 확인할 수 있었다.