SCT85-10  excitation  (corresponding  to PGA  0.194g,  the  input  in  calculation).  The  test model had  higher  acceleration  responses  than  the  prototype building. The  difference  between  the  calculated  and  the measured  results  can  be  attributed  to  the  similarity between  the  reduced-scale  model  and  the  prototype building  in nonlinear  status,  and  to  the gap between the analysis  model  and the  real  building.  In  the  nonlinear stage, the model  and the prototype building had different stiffness degradation  and vibration properties. The floor slabs with increased thickness  in the test building model had  a  greater  effect on  its  stiffness  than  those  in  the prototype  building.  In  analysis,  the  effect  of  the  slab and  slab bars  to beam members were not  included,  and assuming  a  rigid  floor, did not reflect the  effect of floor slab deformation. Again,  the reinforced concrete column member bad evently distributed reinforcing bars  around the  four sides,  so  it did not have  a  clear  yielding point. The method used to determine  the yielding capacity and post-yielding  stiffness  of the members  influenced  the member  behavior,  especially  at  the  yielding  and  post- yielding  stages.  In  the  analysis  model,  the  nonlinear shear  deformation  of  beams, columns  and  beam- column  joints  could  not  be  taken  into  account,  while the  prototype  building  had  diagonal  cracks  on  these members. Additionally,  the analysis model was not able to  consider concrete  crush and  spall  off. These  factors may have reduced the calculated  results when compared with  the  actual  seismic  response  of  the  prototype building. 5  Conclusions Shaking  table  test  and  nonlinear  response  analysis are  two  important  approaches  used  to  understand  the seismic  response  and  damage  mechanism  of structures subjected  to  strong earthquakes.  In  universities  and research  institutes  in China,  shaking  table  facilities have been established to perform experimental  research. Through  evaluating  the  shaking  table  test  results presented  in this  paper  and  from  some  other  research reports,  the  following conclusions are obtained. The  shaking  table  test  results  are  affected  by  the specimen  size,  similitude  design,  input  and  excitation control,  and  precision  of  the  data  acquisition  system. The  test  of  a  carefully  designed  and  constructed  one- tenth-scale  RC  building  specimen  has  shown  that the  damage  pattern  of  RC  buildings  observed  in  past earthquakes  can  be  reproduced  to  a  certain  degree  of reliability  in the  laboratory. The  studies  also  indicate  the  possibility  of  using a  sophisticated 
3-D  nonlinear  numerical  analysis  to simulate  the  dynamic  response  of RC  frame  buildings for minor  and moderate  earthquakes,  but matching  the response to a  severe earthquake  is difficult. 振动台实验和动态响应地震后的钢筋混凝土建筑英文文献和中文翻译(2):http://www.751com.cn/fanyi/lunwen_52376.html