A new counterweight design method for shaking table test models of single-layer spherical latticed shells

Shaking table test is one of the effective methods to study the dynamic behaviors of single-layer latticed shells, and the dynamic behavior of a shaking table test model is significantly affected by the size and installation way of the counterweight. Considering the characteristics of the single-layer latticed shells and the requirement of the test model similarity design, the size of the counterweight needed is normally large. To accommodate the counterweight in the test model, one way is to install the counterweight on the nodes directly and the other way is to enlarge the node size directly to account for the counterweight. Both the methods change the size and configuration of the nodes significantly, which brings additional bending moment and torque forces to the bars or causes a non-negligible reduction of the effective length of the bars. Those problems caused by the counterweight will have considerable influence on the evaluation of the seismic behavior of the single-layer latticed shells. A new counterweight design method is proposed by using multiple layers of upper and lower paired spherical steel shells assembled around a node, which improves the shortcomings of the traditional counterweight in the shaking table test model of single-layer spherical latticed shell (SLSLS). Thin rubber cushions are adopted between two layers of steel counterweight layers to help accommodate the local defects of the counterweight layers when they are not manufactured perfectly so that the counterweight layers can be tightly fixed to the node. With the new method, the improved counterweigh is installed around the node to avoid producing additional forces and keep the satisfaction with the design of the effective length of the bars. The effectiveness of the proposed counterweight design method is verified through theoretical analysis and numerical simulation of the test models. To carry out the finite element (FE) analysis of the test models, the seismic waves input way and the FE model updating method are improved. The static stability capacity of the bars and SLSLSs, and dynamic characteristics and dynamic stability capacity of SLSLSs are discussed to compare the improved counterweight with the traditional counterweight. The applicability of the new counterweight design method to the similarity ratios of the dynamic behaviors of the test models and the prototypes of SLSLSs is also studied under different rise-span ratios and grid-making.