In regions prone to earthquake-induced ground shaking, buildings with insufficient seismic capacity may suffer severe damage and in the worst case collapse. Collapse of buildings lead to loss of many human lives. Thus, collapse prevention of the buildings is the first priority in realizing earthquake-resilient community in such regions. This study proposes the concept of a base shear capped building for collapse prevention as one of promising seismic retrofits. The structural system detaches the superstructure from the foundation in order to prevent collapse by letting the superstructure slide under huge earthquakes. When small earthquakes occur, the proposed structure does not slide against the foundation and behave like a fixed-base structure. When huge earthquakes occur, the structure starts to slide because the horizontal force which develops between the sliding base and foundation exceeds the maximum static friction force.
The friction coefficient is one of key parameters in the sliding system. In this paper, graphite lubricant is used between steel and mortar, materials commonly used in building construction in order to obtain a target friction coefficient, 0.2.
Previous numerical studies with a SDOF system revealed that the mass ratio, the weight of the superstructure over the total weight, was one of the ruling parameters. The objectives of this study are as follows: (1) Shaking table tests were conducted to evaluate the basic sliding behavior; and (2) The maximum base shear coefficient was estimated by numerical simulation when the proposed structure is subjected by pulse-type ground motions.
In the experimental part, two types of specimens were designed: a basic specimen that simulates a SDOF system having a sliding system and a frame specimen that consisted of beams and columns. With those specimens, the stability of friction coefficient and the effects of the mass ratio and the variation of axial forces on the sliding behavior were investigated.
Based on the test results, a simplified SDOF model featured with sliding behavior is developed, and the maximum base shear is estimated under pulse-type ground motions. Two types of input motions are used: an impulse motion and a set of SAC20 ground motions. Even if an existing building is brittle, it has some post-yielding ductility before collapse. To reflect this, additional numerical simulation is conducted using inelastic models and estimate the maximum required base shear coefficient when slight ductility, say two in ductility ratio, is permitted.
The major findings are as follows: (1) Experimental studies verified the stability of friction coefficient between steel and mortar lubricated with graphite powder throughout all loading cases, and the dynamic friction coefficient is 0.16. (2) Even when the variation of axial force reached 50%, four column bases displaced equally, indicating that the effect of axial force variation on sliding behavior was minimal. (3) Numerical studies verified that there is an upper limit of the maximum base shear coefficient when the model is subjected to impulse motion, and the value of the upper limit is twice the friction coefficient. (4) For SAC20 ground motions, no obvious upper limit was present. However, even for earthquakes in level 2 (10 percent exceedance for 50 years), the maximum base shear coefficient remains at most 2.5 times the friction coefficient. (5) If slight ductility is permitted, the required maximum base shear coefficient can be reduced about twice the friction coefficient.
