A novel high-accuracy strain measurement system for robust vibration-based damage identification

Student thesis: Doctoral Thesis

Abstract

Vibration-based damage identification is a non-destructive method that enables health monitoring of civil engineering structures. It aims to detect the presence and growth of damage by measuring changes of natural frequencies, damping factors or modal displacements of these structures. This approach typically suffers from the low sensitivity of these natural frequencies and modal displacements to certain types of damage. Modal strains and curvatures can be more sensitive to local damage. However, when considering ambient excitation, the strain amplitudes in concrete civil structures can be in the sub-microstrain range (<1με). Unfortunately, the use of commercially available strain sensors combined with an entry-level interrogator does not allow for quasi-distributed strain measurements below 1 με. Therefore, the first challenge we will focus on in this PhD is to develop a highly sensitive strain sensor for monitoring bridges. Owing to many advantages of optical fiber sensors we have decided to use them in our research. Nonetheless, one of the drawbacks of optical fiber sensors is their cross-sensitivity to various physical parameters which can make an independent measurement of one quantity using only one sensor impossible. For the moment, additional temperature sensors are always required for this purpose. Therefore, the second challenge we will focus on in this PhD is to develop a novel optical fiber enabling the simultaneous and independent measurement of strain and temperature using commercially available interrogators. On the way to achieving these objectives first we design a novel strain amplifying transducer using fiber Bragg grating based sensors. The basic idea is to utilize a symmetric cantilever structure to enlarge the strain measured by the FBG sensor compared to the strain applied to the transducer itself. We build a 3D FE model for determining the strain level on the FBG sensor installed in the center of the transducer. We verify the operation of the transducer by means of tensile test experiments and by static and dynamic tests on a concrete beam equipped with the novel strain amplifying transducers. We obtain an average experimentally derived strain amplification of 32 for this first version, which agrees very well with the finite element modelling. Second, we perform two experiments on two concrete beams using strain amplifying transducers to gain experience in measurements on concrete structures. The purpose of these experiments is also to use modal analysis to extract the modal strains of the beams using FBG sensors on our strain amplifying transducer and compare them with regular FBG strain sensors at low excitation levels close to those experienced by bridges under ambient excitation. We conclude from these experiments that FBGs mounted on our strain amplifying transducers in combination with entry-level interrogation equipment allow strain mode identification in ambient conditions yielding dynamic strain level amplitudes below 1 �휀. Then, for the first time we carry out the field test on the Arbre high-speed train rail viaduct using our strain amplifying transducers. To perform this experiment, we build on the knowledge gained from the concrete beams. The purpose of this experiment is to verify if it is possible to extract the modal strains of the viaduct using our strain amplifying transducer (and compare them with regular FBG strain sensors) under ambient excitation. We equip a 53 m long span with 20 transducers with a strain amplification of 80 and we succeed to identify 6 strain modes under extremely low ambient excitation level using a high-resolution FBG interrogator whereas this was not possible with FBG sensors only. Finally, we design a novel polarization maintaining optical fiber (PM-fiber) for simultaneous measurement of strain and temperature. We build a 2D FE model to determine the birefringence of the PM-fiber and to analyze and minimize stresses that occur in different designs of PM-fiber. We propose 3 different designs which improve the PM-fiber's birefringence and have decreased stress levels inside the fiber. The shared feature of these designs are 4 additional air holes around the core of the fiber. We decide to fabricate and characterize one of the designs. Our PM-fiber features higher birefringence and differential temperature sensitivity compared with commercially available PM-fibers, which allows for the simultaneous and independent measurement of strain and temperature with a higher resolution than before. While additional work is still needed to bring our technology to its full maturity, we can state that our novel strain amplifying transducer and our novel PM-fiber have great potential in the field of structure health monitoring and beyond.
Date of AwardMar 2019
Original languageEnglish
SupervisorLincy Pyl (Jury)

Cite this

'