1. Monitoring Applied Stress in Concrete Using Ultrasonic Full-Waveform Comparison Techniques
Portland State
University
Ali Hafiz and Thomas Schumacher
Civil and Environmental Engineering Department, Portland State University
iSTAR-NDT
Laboratory
INTRODUCTION
DATA ANALSIS
Concrete is the most widely construction material worldwide. In addition, an increasing demand exists for ways to monitor the performance concrete structures.
Concrete it is a complex heterogeneous material because it is made of aggregates of different sizes, cement, and water.
When an ultrasonic wave propagates through concrete, it experiences multiple scattering, which makes the propagation paths longer and more complex.
The recorded ultrasonic waveform can be divided into two portions: the early part of the signal, which is referred to as the coherent portion, and the later part, which is referred to as the diffuse portion or Coda wave.
Objective: the objective of this research is find correlation between applied stress and the changes in recorded ultrasonic full- waveform.
Correlation Coefficient
The correlation coefficient (Pearson correlation coefficient, ρ) was used to describe the similarity (or difference) between two waveforms.
The Pearson correlation coefficient varies from -1 to 1
Similarity index, R vs. applied stress.
where Y0 is the reference waveform recorded at zero stress and Yi is a waveform recorded at certain level of applied stress i.
The results of the four specimens show the same general trend, and this trend can be seen that two portions exist for SI: a linear and a non-linear portion.
EXPERIMENTAL SETUP
Energy Filter
Four plain concrete cylinders with identical dimensions of 6 in diameter and 12 in height were cast and used.
The specimens were loaded up to failure.
Throughout the loading process, an ultrasonic pulse was transmitted and recorded by two identical ultrasonic normal-wave transducers.
During loading of the concrete cylinder specimens, Acoustic emissions may occur and appear in the recorded waveforms. P
Similarity index, R vs. normalized stress.
Two sample waveforms: one with and one without an AE.
Since waveforms with AE have considerable higher signal energy, a filter was designed to exclude them P
Preamplifier
Three samples of recorded waveforms for different level of applied stress (Specimen A2).
Illustration of experimental setup with specimen.
The recorded waveforms were not only much longer but also consisted both of a coherent and a diffuse portion.
(a) Signal energy and (b) corresponding correlation coefficient for a selected number of recorded waveforms.
Results and Discussion
The correlation coefficient was squared in this study and referred to as the similarity index, SI:
∆TOF vs. normalized stress for Specimen A2.
Sample waveforms transmitted pulse
CONCLUSIONS
The coherent (early) portion of a recorded waveform, which is typically used for ultrasonic testing, does not show any significant changes before reaching 50% of ultimate stress.
The diffuse (or Coda) portion is very sensitive to small changes in the applied stress.
SI was sensitive to very small changes in the applied stress at very low stress levels.
Spontaneous acoustic emissions (AE) were filtered based using a signal energy filter.
The relationship between the similarity index and applied stress for all specimens observes that SI correlates well with applied stress.
Sample waveforms: recorded waveforms at two different levels of applied stress.