The influence of the incident angle of seismic waves on the amplification effect of rock slopes is obvious in near-source earthquakes. Based on the theory of viscous-spring artificial boundary and the principle of equivalent nodal force

a seismic input method for oblique incidence of seismic SV and P waves is established and implemented in the finite element method. Various influencing factors such as slope inclination

incident angle of seismic wave

and type of incident wave are considered

and then a numerical study is conducted to investigate the amplification effect of rock slopes. The results of the peak ground acceleration (PGA) amplification factor of the slope show that the amplification effect of the rock slope may be significantly underestimated if only the vertical incidence of seismic waves is considered. In addition

the slope inclination and the type of incident wave also have a significant influence on the amplification effect. By comparing the calculated results of the PGA amplification factor with those suggested by the Code for Seismic Design of Buildings (GB50011―2010)

it is found that the values suggested by the seismic code may underestimate the amplification effect in some cases. The results of the spectral characteristics of the acceleration at the slope crest reveal that the incident angle of the seismic waves has little effect on the main frequency of the Fourier spectrum but has a significant effect on the Fourier amplitude. As the incident angle of the SV-wave increases

the Fourier amplitude of vertical acceleration at the slope crest increases

while the variation of Fourier amplitude of horizontal acceleration is not obvious. As the incident angle of P-wave increases

the Fourier amplitude of vertical acceleration at the slope crest decreases while that of horizontal acceleration increases. Finally

the spectral ratio curve is introduced to quantitatively evaluate the effects of different influencing factors on the spectral amplification of slope crest

and the variation pattern of the peak spectral ratio under different incident conditions is obtained.