2D Modeling Problem of a Infinite Slope

[BhuiyanMH]

Hi All!
Hope every-ones doing great this summer. I have brought a problem for you which I couldn't solve for months now!

I have been doing seismic site response analysis of a 2-D sloping ground (infinite) problem. I have a very deep profile of 137 m to the bedrock. For model generation, I have been using the OpenSees example "Dynamic 2D Effective Stress Analysis of Slope" by Dr. McGann and Dr. Arduino and the GiD template files they have provided there. Problem is: for very soft profile (low Vs) and steeper slopes (>5 degrees), the recorded time-history at the surface show unusually high vibration at the very beginning of the time-history although the peak vibrations are still few seconds away! The same model with stiffer layers (large Vs) do not show this problem at all. I have tried with zero loading at first few seconds of the input motion and still got high amplitude in the recorded acceleration time history at the surface in the earlier periods. Could not think of any solution to this problem. Please help!!!

Thanks.
MD. ARIFUL HAQUE BHUIYAN
GRADUATE RESEARCH ASSISTANT
CLEMSON UNIVERSITY
EMAIL: mhbhuiy@g.clemson.edu