The interest in our paper shown by the discussers in the Short Communication [2] is appreciated. In the following, we respond to the issues raised by the discussers.
SHAKE TABLE RESULTSAs noted by the discussers, we used two well-established methods for determining the pounding force, with strain gauges installed on the attachment surfaces near the pounding interface, and accelerometers installed on the floor of each storey. In the former case, we calculated the pounding force F from Equation (1) in the reference [1].As shown in Figure A1, strain gauges are installed on the upper and under sides of the attachments. Because the specimen carry out insignificant irregular lateral vibration in spite of one-way shaking test because of the initial imperfections of the irremovable specimen, the strain at position 11 or 21 is not the same. For this reason, the strain was taken to be the mean value of the strains measured at positions 11, (11), 21 and (21). Therefore, because the treatment for the strain used here corresponds to full-scale contact, it underestimates the pounding effect, contrary to the assertion of the discussers.Thus, ε A and ε B in Equation (1) are, respectively, the average of the strain gauge readings on the upper and lower surfaces of the attachments, as clearly stipulated in the paper. It is known that the greatest damage produced by pounding occurs at the pounding interface, and it is also obvious from the fact that buildings that collide at an expansion joint generally undergo the greatest damage at that interface.The discussers claim that the results shown in Figure 4(a) and (b) in the reference [1] correspond to electrical noise. However, the experiments show that the presence of the shock-absorbing material greatly reduces the pounding force at the interface. In the case of the KOBE-NS waveform, which apparently inflicts only a very small amount of damage on either building, the pounding force is reduced by the shock-absorbing material to a level low enough to be regarded as noise. With regard to Figure 4(b), it should be noted that because the 2 mm gap is completely filled with rubber, the very small strain always occurs. The painting part in Figure 4(b) expresses the fact that there are many data in the prescribed scale of the transverse axis for time in the graph, as well-known observable event in graph used Excel. If the scale of the transverse axis in the graph is enlarged, the complete wave-form is confirmed visually.The discussers claim that Equation (2) is not valid. This assertion led us to re-examine Figure 10, and we discovered an error. The corrected Figure 10(a) and (b) is presented in the succeeding texts.