Monitoring Reinforced Concrete Structures with Coda Waves The Influence of Temperature on Ultrasound Velocity Changes calculated with Coda Wave Interferometry

Abstract

Monitoring of reinforced concrete structures to ensure their stability and increase their service-life is a crucial element of a modern infrastructural concept. With classical methods of non-destructive testing and inspection, repeated measurements under comparable conditions are difficult to conduct. Therefore, DFG research unit FOR 2825 CoDA researches the assessment of concrete damage using ultrasound coda wave interferometry and embedded sensors. Embedding the sensors into the monitoring target reduces human and non-human factors influencing repeatability. Using Coda Wave Interferometry (CWI), small velocity changes in the material can be detected by comparison of repeated measurements. The technique is sensitive to damaging changes like cracking as well as to reversible influences like material temperature. The understanding of these different influences on the signal is crucial for the analysis of long-term monitoring data to make an educated assessment of the structure and its integrity. With several laboratory experiments in a climate chamber and a long-term experiment recording an annual cycle in a large model on an outdoor test site in Horstwalde close to Berlin, we try to understand the influence of temperature on the CWI results. The results show that the velocity change calculated by CWI does closely follow the trend of concrete temperature. After one year of data recording with the large model being exposed to environmental variations only, the calculated velocity change resembles the annual temperature curve. The data shows a linear dependency between velocity and temperature change in a range of -0.03 percent per °K to -0.06 percent per °K - regardless of specimen size. An approach to remove temperature influence from the yearly cycle recorded in the large-scale experiment using this linear relation is unable to remove high-frequency variations - especially daily influences. Low-pass filtering the data can eliminate these variations while preserving permanent shifts caused by damages. Although we have shown that the influence of temperature on long term monitoring can be removed to a significant extent, there is still an influence of environmental changes remaining in the data. Possible nonlinear effects and influences not related to temperature need to be investigated in the future.