Large diameter bored piles are commonly used nowadays in many construction projects in Hong Kong, they usually carry large structural loads and are relatively few in number compared to a driven pile foundation to support the same total load. The integrity of each shaft is critical to the overall performance of the structure. Crosshole Sonic Logging (CSL) is a standard test method (ASTM-D6760) for evaluating the integrity of large diameter bored pile foundation structures. However, results of CSL are typically displayed with one-dimensional (1D) representation. When a defect or anomaly is detected, the only information about the defect location is the elevation within the shaft. One opportunity for innovation in the construction industry is the Cross-hole Sonic Logging Tomography (CSLT) test. This test is an enhancement technique to the currently adopted CSL method. CSLT provides a visualization of the shape, size, and location of irregularities in the shaft. The CSLT also provides much more qualitative and quantitative information about the bored piles’ integrity.

The aim of the thesis is to evaluate and implement the CSLT system in diagnosis of defects in bored piles in Hong Kong. Since the CSLT is not yet a standard test method, we have established a testing procedure for the CSLT in accordance with the general requirements of the ISO/IEC 17025 international standards and designed as an extension to the existing ASTM-D6760 standards. (See Chapter 5).

In addition, we have performed verification studies, performance experiments, variability studies in Chapter 6 and case studies in Chapter 7. The test is found to be fit for purpose whereas individual qualified operators were able to follow the procedure and produce accurate enough results for engineering purposes. Therefore, we recommend using the procedure for all future CSLT tests. The procedure can also be used for HOKLAS accreditation in the near future.

With the testing procedure established, a verification of the CSLT results was conducted by quantifying the resulted two or three-dimensional visual images produced by the CSLT technique. A test pile with known defects was constructed and experiments were performed on the test pile to quantify the measurement performance. A performance criteria analysis was then carried out. From our experiment, we found that the CSLT method serves as both a practical and a robust tool for defects detection and diagnosis in bored piles, it provides accurate enough information regarding the location and size of the defects for defect evaluation. We can recommend using the CSLT to diagnose defects identified with the CSL, especially when the measurements are performed at least 28 days after concreting.

We have included a case study in the very end to showcase the implementation of the CSLT in a construction site in Hong Kong. We found that the CSLT is an invaluable and robust tool for defect diagnosis, as it provides visualization of the shape, size, and location of irregularities in the bored pile which aids engineers carrying out remedial actions on bored pile with defects.

Overall, this dissertation has contributed to the innovation of the construction industry in Hong Kong by introducing and establishing the CSLT method for diagnoses of defects in bored piles. The accuracy and precision of the CSLT has been fully evaluated through our performance criteria analysis and Gage R&R variability study. The implementation of the CSLT has been demonstrated through the establishment of the CSLT test procedure and, the application and execution of CSLT in field as detailed in our case study.