Seismology-based approaches for the quantitative acoustic emission monitoring of concrete structures
Date
2015
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
Over the last few decades there has been sustained interest in learning about fracture processes to achieve a better understanding of how concrete structures behave and what leads to failure. Such understanding would greatly benet the concrete industry as it would provide ecient engineering solutions to improve the quality, durability, strength, and behavior of concrete structures and answer more challenging questions such as how, where, and when cracks initiate and propagate. The type of cracking also matters. For example, shear-type failure happens suddenly without warning and means for early detection would be useful. Due to the complex nature of concrete, however, monitoring approaches are challenging to apply and are often subject to signicant measurement noise.
This PhD research investigated the feasibility of employing moment tensor inversion (MTI), a powerful quantitative seismology-based technique, combined with the capabilities of the acoustic emission (AE) technique for the quantitative monitoring of crack initiation and propagation in concrete materials. The goal was to monitor fracture mechanisms in concrete structures in real time. For this reason a hypothesis was put forth that MTI techniques applied to AE data can produce pertinent information about cracks such as location, type, orientation, and intensity, and thus can distinguish flexural from shear and mixed-mode cracks. The motivation behind using MTI techniques is the strong analogy between AE and seismic events, albeit at dierent scales. To test the hypothesis identied above, a methodology that capitalizes on the strengths of both the AE technique and the MTI technique was proposed and applied according to which AE waveforms recorded from concrete fracture were inverted using a MTI code, called MTI Toolbox, in order to study the sources of fracture and infer their properties. The MTI Toolbox was applied to estimate source mechanisms from mining-induced seismicity and was fairly extensively applied to mining research problems. It was used in this research because of the great features it oers that make it applicable to performing inversions on AE data recorded by networks consisting of uniaxial sensors. A number of experiments were performed in the laboratory in a civil engineering environment in order to assess and apply the methodology proposed. In all the experiments, high-delity Glaser/NIST point-contact sensors, which represent an important aspect of this research, were used to collect AE data. An algorithm based on the Geiger's location method was implemented in MATLAB and used for all the source location problems.
First, a preliminary study was conducted that aimed at evaluating the usefulness of the MTI-based methodology in characterizing AE sources. Articial AE sources with known location, type, and orientation were produced on two dierent specimens: a UHMW Polyethylene specimen and a normal-weight concrete specimen. The goal was to establish whether these parameters can be estimated from the recorded AE signals using MTI techniques. The results of this preliminary study were satisfactory as it was possible to reproduce the parameters of the predened externally applied sources. This proved that the methodology proposed had the potential to be applied in the monitoring of real fractures in critical concrete structures. Second, a case study was conducted on a 6 in 6 in 21 in concrete beam. A notch was cut at mid-span to serve as a crack initiator and the beam was loaded to failure. The aim of the experiment was to produce tensile fracture AE sources. An analysis of the results of this experiment showed that 20% of the sources located with high accuracy (= 71) out of a total of 262 sources located were dominantly tensile, 7% were dominantly implosive, and the remaining 73% were dominated by shearing. Even though it was anticipated that a pure flexural crack would produce AE events with high positive isotropic (i.e. opening) components, this was not conrmed and it was speculated that this is due to the heterogeneity of concrete. The methodology however showed potential in distinguishing between the dierent types of cracks. Third, two large-scale experiments were conducted on two dierent reinforced normal-weight concrete beams of the same dimensions (12 in 24 in 16 ft). One beam was designed to fail in flexure and the second one to fail in shear. The aim of these experiments was to evaluate the ability of the methodology to characterize and monitor fracture in large-scale concrete structures. The MTI results obtained from this experiment showed that the cracks produced from the fracture of the flexure beam are dominantly tensile (%81) while the cracks obtained from the fracture of the shear beam were dominated by shearing (%73). These results satised our expectations from the methodology and we were able to prove the hypothesis as we were, in fact, able to (1) distinguish between flexure and shear cracks during the fracture of large concrete structures and (2) answer some of the challenging questions regarding the mechanics of concrete materials.