Filtern
Erscheinungsjahr
- 2019 (2) (entfernen)
Dokumenttyp
Sprache
- Englisch (2)
Referierte Publikation
- ja (2)
Schlagworte
- Deconvolution (1)
- L-curve (1)
- NDT (1)
- Pile integrity testing (1)
- Pile testing (1)
- Shaker (1)
- Tikhonov regularization (1)
Organisationseinheit der BAM
In this study, an Impulse Response Function analysis of pile response to sine-sweep excitation by a low cost, portable Shaker was used to identify defects in piles. In straightforward impact-echo methods, echoes from the pile toe and defects are visible in the time domain measurements. However, these echoes are not present in the time domain records of piles subjected to sine-sweep excitations, due to interactions between the input and output signals. For this reason, the impulse response function in the time domain has been calculated and is able to identify the echoes from pile impedance changes. The proposed methodology has been evaluated both numerically and experimentally. A one-dimensional pile-soil interaction system was developed, and a finite difference method used to calculate the pile response to sine-sweep excitation. The numerical simulations indicate that impulse response measurements with a synthesized logarithmic, sine-sweep excitation could be an effective tool for detecting defects in piles. The methodology was further tested with field trials on 6 cast in situ concrete test piles including 1 intact pile and 5 defective piles subjected to sine-sweep excitations by a shaker. In 5 of the 6 cases the echoes from the pile toe could be identified in the deconvoluted waveforms—the impulse Response functions. Damage detection is more difficult and dependent on the selection of the optimal regularization parameter. Further research and optimization of the deconvolution process is needed to evaluate the effectiveness compared to standard pile integrity testing methods.
Concrete piles are used as a foundation when the load capacity of the soil is insufficient or when the sustainable soil is found at a higher depth. Among other impact factors, the load capacity depends on the pile’s integrity and length.
Therefore, verifying these parameters using adequate methods is recommended. The most common procedure is the low-strain integrity test using the hammer impact method. Developed and established in the 1970s, this method uses stress waves induced by a hammer impact at the pile head and its reflections at impedance changes (length, defects, geometry changes) to estimate the length and defect locations. Although this method is widely used due to its low cost and fast conduction in situ, one disadvantage is its inability to classify the exact type of defect, i.e., crack, change in diameter, or concrete quality. Furthermore, very long and slender piles are difficult to test and small defects cannot be detected. In addition, it is necessary for the test engineer to hold a high level of experience and expertise in this field. The European Union–funded PileInspect project (2013–2016) aimed to compensate for these disadvantages by using a low-cost shaker as the excitation source and sophisticated artificial intelligence algorithms for damage detection (higher-order spectra method). Because this technology lacks the capacity to localize damages and verify the pile length, an additional impulse response (IR) measurement technique was developed using vibrational excitation and regularized deconvolution to extract the depth information from the data in a similar manner as the hammer method.
Simulations and subsequent experiments conducted at a test facility on 90-cm-diameter bored piles 11 m in length and containing cracks at approximately 4 m below the pile head confirmed the capacity to determine the pile length.
Damage diagnosis and localization, however, are more difficult than for the hammer method. Although the damaged piles could be distinguished from the intact piles, in a blind test, this method might lead to misinterpretations caused by perturbations arising from the deconvolution process. The results also indicated that the low-cost shaker used for these measurements might be inappropriate for
the transferal of sufficient energy. Although the IR method cannot compensate for the disadvantages of the hammer method by itself, it may enable the possibility of using long and fully controllable and repeatable signals (chirp, synthetic impacts, even noise, etc.) for pile integrity testing in the future.