High-rise truss towers consist of steel tubular section members have high risks to be damaged by seismic input, because their hysteretic characteristics deteriorate after buckling and easy to fracture after local buckling ; However present design does not include the effect of such member fracture. Authors have established algorism to predict member fracture after buckling for tubular members using macro-model, derived from experimental and analytical studies. In this paper, this algorism is applied for time-history analyses and seismic responses of detailed tower structures are analyzed. The validity of proposed method is firstly compared with actual earthquake damages, followed by researches on the effect of member fracture in various truss towers.
