ESTIMATION OF S-WAVE VELOCITY USING LEVENBERG-MARQUARDT INVERSION ON RAYLEIGH WAVE DISPERSION CURVE, CASE STUDY: ALLUVIAL AREA, TANI KOTA
In geotechnics, elastic parameters of the rocks are important to characterize near surface. Surface wave method can be utilized to characterize near surface layer since it is a function of the S-wave velocity, P-wave velocity, density and layer thickness. In heterogenous medium, Rayleigh wave is dis...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Project |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/31132 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | In geotechnics, elastic parameters of the rocks are important to characterize near surface. Surface wave method can be utilized to characterize near surface layer since it is a function of the S-wave velocity, P-wave velocity, density and layer thickness. In heterogenous medium, Rayleigh wave is dispersive because Rayleigh wave phase velocity depends on frequency. We can construct dispersion curve from phase velocity and frequency. Inversion is one of the most important stages in surface wave method. The inversion method used in this experiment is Levenberg-Marquardt using Singular Value Decomposition (SVD) technique. Damping factor is determined iteratively based on objective function. This experiment begin with acquisition using refraction seismic methods with different receiver spacing, one meter, three meter and five meter. Then, experiment data used to extract observation dispersion curve using phase shift method. Observation dispersion curve to be inverted to get 1D shear wave velocity profile. Parameters which include in inversion process are shear wave velocity and layer thickness because those parameters give big effect to dispersion curve. The 2D cross-section is obtained by interpolating on 1D S-wave velocity profile. The result of Levenberg-Marquardt inversion algorithm test on synthetic data shows good result on four layer model and model with noise up to 10% and able to detect low velocity layer. The 2D cross-sectional results from the experiment data show three zones. First zone at a depth of 0-7 meters with S-wave velocity of 80-220 m/s, interpreted as a layer of weathered clay. The second zone at a depth of 7-13 meters with S-wave velocity of 260-380 m/s, interpreted as a claystone. The third zone at a depth of more than 13 meters with a velocity of 420-620 m/s, interpreted as claystone with higher stiffness than second zone. |
|---|