Detection of Anomalies In Drilled Shafts Using theImpedance Log Method

The impulse response method is a nondestructive technique used to check the length and continuity of deep foundations. Uncertainty may exist when interpreting the results to quantify the size and location of an anomaly, if present. The analysis can be enhanced by developing a two-dimensional graphical image of the shaft using the impedance log method. The method makes use of the simulated response of an ideal infinitely long shaft in soil and the experimental impulse response data of the in-situ shaft. Experimental data from impulse response tests performed on 24 drilled shafts constructed with and without anomalies at four National Geotechnical Experimentation Sites (NGES) were used in this study. Initial work focused on deriving the impedance log methodology from first principles and implementing the theoretical work of Paquet (1991) and Davis (1993) into a computer program in Labview. Digital filtering, windowing and back-shifting were added to the program to filter the experimental velocity data. A systematic procedure using control impedance logs was established to obtain the soil, shaft and amplification parameters used to obtain the impedance logs for the in situ shafts at the NGES sites. Each control impedance log was developed for a control shaft (in situ or simulated) which was straight and free of anomalies. The best fit soil and shaft parameters were found from a mobility curve matching procedure and were based on the soil and shaft parameters from design drawings and construction records. Criteria were established to determine the amplification parameters based on the decaying tendency of the reflection peaks in the experimental velocity, and criteria were established to identify the location of the toe in the impedance log based on the location of the toe peak trough in the relative reflectogram. After the control impedance log parameters were established, they were ported to obtain the impedance logs for the 24 in situ shafts at the four NGES. If anomalies were shown on the impedance log, the impedance ratio was calculated and compared to that of the best-fit profile. When the impedance ratio at the depth of interest was equal to the impedance ratio of the best fit profile, the impedance log was considered to be representative of the in situ shaft profile. If anomalies that were observed in the relative reflectogram were not properly shown on the impedance log, the impedance log was rescaled or regenerated using a new amplification factor. Overall, the procedure developed herein produced impedance logs that were representative of the in situ drilled shafts at the NGES sites. The impedance ratio of the anomalies in the impedance logs matched the impedance ratio of the best-fit profile within 0 to 19%. The locations of the mid-section depth of the necks in the impedance logs were generally lower (up to 29%) and their lengths were generally longer than that of the best-fit profile; whereas the mid-section depths of the bulges where either equal to or higher (up to 19%) than the best-fit profile. Using the criterion developed herein to identify the toe, 10 shafts had the same length as the best fit profile and 9 shafts had shaft lengths up to 9% shorter. A new amplifier called the inverse regression amplifier was investigated and produced similar results to the general amplifier. The implication of observing multiple reflections from shallow anomalies is also investigated