Experimental study of seismic behavior of two hilly sites in Tehran and comparison with 2D and 3D numerical modeling

• Published in: Soil dynamics and earthquake engineering (1984) Vol. 31; no. 5; pp. 737 - 756
• Main Authors: Khandan Bakavoli, Mohammad, Haghshenas, Ebrahim, Kamalian, Mohsen
• Format: Journal Article
• Language: English
• Published: Kidlington Elsevier Ltd 01.05.2011; Elsevier
• Subjects: Earth sciences; Earth, ocean, space; Earthquakes, seismology; Engineering and environment geology. Geothermics; Engineering geology; Exact sciences and technology; Internal geophysics; Natural hazards: prediction, damages, etc; Middle East; Iran; Asia; Iran, Tehran; experimental studies; topography; boundary element methods; digital simulation; resonance; vibration; frequency; instruments; numerical models; three-dimensional models; noise; hills; wavelet; two-dimensional models; transfer functions; soils
• DOI: 10.1016/j.soildyn.2010.12.003
• ISSN: 0267-7261

Two hilly sites were selected to study seismic site response due to topography effects. The sites were selected in a manner to be as much as possible homogenous and free of the soft soil layers effects. The hills were instrumented by nine velocimetric stations to record microtremors and the obtained data were analyzed using horizontal to vertical spectral ratios. Some standard spectral ratio tests were performed on noise as well. Then the instrumented hills were modeled (both 2D and 3D) assuming a linear elastic constitutive behavior subjected to vertically propagating SV and P Ricker wavelets. All calculations were performed in time domain using direct boundary element method. Different transfer function components, amplification patterns and spectral ratios were calculated in frequency domain. The frequency of vibration, obtained by experimental studies, is between 4 and 5 Hz for both of the hills. The spectral ratios derived by numerical simulations were compared with the observed spectral ratios. They show relatively good similarities between the results of these two methods. The frequencies of vibration derived from different methods seem to be nearly identical. The agreement in term of resonance frequency between microtremors and numerical modeling suggests that noise measurements could represent a simple, even if preliminary, tool in order to identify possible topographic amplification.