United States Department of Transportation
i
Sinkhole Detection With 3-D Full Elastic Seismic Waveform Tomography
-
2020-05-01
-
[PDF-10.92 MB]
Details:
-
Creators:
-
Corporate Creators:
-
Corporate Contributors:
-
Subject/TRT Terms:
-
Resource Type:
-
Geographical Coverage:
-
Edition:Final Report 3/1/18 – 5/15/2020
-
Corporate Publisher:
-
Abstract:A new 3-D full waveform inversion (FWI) method was developed, optimized, and verified using synthetic and field test data. The goal was to overcome some of the limitations of a previously developed 2-D approach. Particularly, it was expected that the 3-D FWI method could detect out-of-line voids and 3-D features (pinnacles, zones, etc.), be easier to implement for a large volume, and increase overall detection accuracy. To ensure robustness of the developed method, well-known and established techniques for forward modeling and optimization were implemented. Elastic wave equations were solved using a finite difference method implemented in time domain to simulate wave propagation. Boundary truncation techniques of perfectly matched layer (PML) and image technique were used to truncate the testing medium and reduce computational demand. The Gauss-Newton optimization technique was used to minimize the error and update the VS and VP model parameters independently. Synthetic tests were carried out to find the best testing configuration and establish maximum void detection depth. These tests revealed optimal receiver-source spacing as equal to the maximum void diameter. Maximum void detection depth was established as 3 times the void diameter from the ground surface. Field verification was carried out at various locations to ensure the 3-D FWI method result validity on noisy field data. Initial tests on a stormwater pipe located at the UF campus proved that the method can successfully be used to detect the depth, direction, and overall shape of the pipe. Further testing at a site in Newberry, FL, showed that the method can detect variable soil-rock layering and identify unknown anomalies. Three shallow voids were detected and verified using SPTs. Overall, there was good agreement between the SPT N-values and inverted results, and the presence of the voids was verified. An attempt was made to extend the void detection depth of the developed method. This was achieved through the application of a modified source to generate more energy at lower frequencies (higher wavelength). The modified source was used at a bridge construction site in Miami, FL, where a large deep void was detected. Final inversion results showed that using the modified source with the developed method could help increase its void detection capability. Void features including depth and position were successfully detected and characterized. Finally, a novel SPT seismic source method was proposed for the first time. It was based on using SPT device as the source and gathering data with receivers on the ground surface. The SPT seismic method is useful for increasing void detection depth when there is limited access on the surface and the source does not produce enough energy at low frequencies (0-10 Hz). Tests with the SPT seismic approach at the Newberry and Miami, FL, sites showed that it can detect shallow voids and very deep voids where there is limited access on the surface (e.g., right of ways). All the results were compared and verified using invasive tests performed at the test site.
-
Format:
-
Funding:
-
Collection(s):
-
Main Document Checksum:
-
Download URL:
-
File Type:
