Sensitivity of results inferred from microtremor observations
Date of Issue2010
School of Civil and Environmental Engineering
Microtremor survey method has some advantages over conventional drilling method because the former method is non-invasive and can be carried on easily without disturbing the environment. The general array configuration used in microtremor field measurement is equilateral triangle. However, in Singapore context, to find a large space to do the equilateral triangle measurement seems to be challenging. Hence, further studies on different array size and configuration were done. Firstly, the author will focus on suitability of Spatial Autocorrelation (SPAC) and Frequency-Wavenumber (f-k) method, by doing the data processing of small array using f-k method and comparing it with Cross-hole PS Logging data. The result shows that f-k method overestimates the phase velocity. Hence, SPAC is the most reliable method for small array processing. The data processing of large array using SPAC method is not possible because SPAC method needs a precise array for length combination input. To ensure the consistency of microtremor method from time to time, three series of measurement were done in Katong Park: 1st October 2008, 2nd December 2009, 6th February 2010 and two series of measurement were done in Beatty Secondary School: 3rd October 2008 and 19th March 2010. After comparing the result to Cross-hole PS Logging data, it can be concluded that microtremor method yields consistent result in estimating soil phase velocity for these two different locations. To ensure the applicability of microtremor method on estimating phase velocity of soft soil, another location which lies on Kallang Formation was chosen besides Katong Park, namely Beatty Secondary School. The result from microtremor observation in Beatty Secondary School was then compared to Cross-hole PS Logging data and the result from microtremor observation shows a comparable result to Cross-hole PS Logging data. During field measurement, there is possibility of presence of defective sensor(s). In this study, three scenarios were analyzed: assuming the ideal array is equilateral triangle with 7 sensors; one defective sensor was located on external triangle corner (scenario 1), one defective sensor was located on internal triangle corner (scenario 2), and the combination of first and second scenarios (scenario 3). SPAC and f-k analysis were done for these three scenarios, and the result shows that SPAC still gives reliable result in estimating soil phase velocity under these three scenarios, while f-k only gives reliable result in estimating soil phase velocity under scenario one. The first step in estimating subsurface soil phase velocity from microtremor observation from time history plot is to find several segments which contain optimum number of data to be analyzed without compromising the accuracy. Data length of 4096, 8192, and 16394 are analyzed. From the analysis, it can be seen that the use of data length of 4096 is still reliable for SPAC method, but not for f-k method. For longer data length of 16384, the result from both methods is reliable and the resolution is high; however, it is not advisable to use data length of 16384 in the analysis due to the presence of spikes (disturbances) on the time history plot. In this study, the effect of various small array configurations on consistency of SPAC result was investigated. The objective of the investigation was to find the simplest configuration that can be constructed in the field without compensating SPAC result consistency. The array configurations used were line in East West (EW) direction, line in North South (NS) direction, square, circular, and rectangular with diameter of 12 m and 30 m for each configuration. Even though there are some discrepancies from Cross-hole PS Logging data, in the context of estimating phase velocity, all array configurations yield comparable and reliable result. To investigate the effect of near source effect in microtremor observation, an observation was done in Katong Park using seven sensors placed following the order of linear array with the same direction as traffic. This study shows that under near source condition, the use of microtremor survey method may lead to significant distortion of surface waves detection. This phenomenon is due to the one directional noise. Hence, it is advisable to locate the sensors faraway from traffic noise. The applicability of microtremor survey method in hard soil stratum phase velocity estimation is also investigated. Microtremor observation was done in NTU, which is underlain by Jurong formation (stiff to hard clay). It can be seen that microtremor method gives consistent result for stiff clay compared to bore-log data.
DRNTU::Engineering::Civil engineering::Structures and design
Final Year Project (FYP)
Nanyang Technological University