| Περίληψη: | Combining good material properties and low weight, composites have become increasingly popular over the past decades among conventional, well-studied engineering materials. With their anisotropic nature and laminated structure allowing for enhanced design potential compared to metals, the widening use of composites in operating structures has turned the need for reliable inspection and condition assessment into an issue of great importance.
Unlike metals, failing due to a propagating critical crack, the inhomogeneous and anisotropic nature of composites renders a more complicated behavior: composite structures bear the applied design loads during the entire service life, while damage accumulates. High damage tolerance is thus another important advantage over metals.
The main source of damage in a composite is mechanical and/or environmental loading. Several failure mechanisms are encountered during service: matrix cracking, debonding of the fibre-matrix interface, delaminations and fibre breakage are common damage modes. Although this is their actual temporal sequence in general, propagation and coalescence of failure mechanisms are often simultaneous and therefore, damage in the composite can be regarded as the superposition of various failure modes. Damage accumulation, either localized or distributed throughout the volume of the composite, leads to degradation of the composite material mechanical properties.
Matrix cracking is one of the major damage mechanisms encountered in FRP composites during service. Although seemingly less critical than delaminations and fibre breakage, propagation and coalescence of matrix cracks precede and promote more severe damage modes. Characteristic consequences of matrix-dominated failure are debondings at the trailing edge or between stiffening components and the skin, in wind turbine rotor blades, and also material degradation due to ingress of fluids, in composite pipes. However, since formation of matrix cracks begins at sub-critical loading stages, appropriate non-destructive tools should contribute to reliable damage assessment throughout service.
Aiming to damage assessment in composite materials, non-destructive inspection (NDI) relished rapid and broad development. Acoustic emission and acousto-ultrasonics are listed among well-established NDI techniques. These acoustic methods are able to reflect the integrated damage state of a structure. The scope of this dissertation is NDI assessment of distributed damage in glass/epoxy (Gl/Ep) fibre-reinforced (FRP) composites, using acoustic emission and acousto-ultrasonics.
Most research on the use of acoustic methods for non-destructive inspection is concentrated on the detection of localized defects, generated either during fabrication or in-service. A considerable amount of publications is also focused on the more complicated, distributed damage, e.g. due to fatigue. In most cases, however, acoustic emission and acousto-ultrasonics are not suggested as stand-alone tools, but are rather used to indicate qualitative trends or to complement other methods in the investigation of damage progression. Although a common outcome from this approach is that AE and AU signal parameters are, in general, correlated with damage accumulation, no robust models for remaining life or strength prediction have been proposed.
Such NDI tools for the assessment of strength degradation, due to fatigue, in fibre-reinforced composites, exclusively via acoustic non-destructive measurements, are proposed in the present work. Reliable engineering models, based on acoustic emission and acousto-ultrasonic measurements, are established and validated in dedicated chapters. Residual strength prediction in composite specimens, featuring matrix cracking due to fatigue, is thus accomplished.
This thesis is based on experimental work performed on an improved Gl/Ep composite, used in the manufacturing of new generation wind turbine rotor blades. The work included thorough material characterization as well as a dedicated experimental series aiming to understand, model and assess the axial, transverse and shear strength degradation of the unidirectional composite.
Besides preliminary and benchmark testing, the exhaustive experimental schedule included 713 valid mechanical tests. From these 713 specimens, 222 were tested in tension/compression, 236 were subjected to constant-amplitude fatigue loading and 29 to spectrum loading. Another 217 specimens were used to investigate strength degradation due to constant-amplitude loading and 9 due to variable-amplitude loading. To execute this grand experimental plan, our 4-member team occupied 3 testing machines for 52 months.
Although scrupulous indeed, the material characterization stage was just a prerequisite for the residual strength experimental task. In common practice, residual strength tests are a combination of a damaging process, e.g. fatigue loading, and a static test to failure. However, the aim of this dissertation was strength degradation assessment using non-destructive techniques. Residual strength tests were thus accompanied with acoustic emission monitoring, stiffness degradation measurements and acousto-ultrasonic scanning. This increased the duration of the experiments at least 4-fold, while rendering the procedure much more complicated. However, a unique database was formed, including data from all discrete steps. This extensive and combined information is a novel contribution in the field of non-destructive inspection.
Acoustic emission monitoring and acousto-ultrasonic measurements were herein used to assess material strength degradation due to fatigue-induced matrix cracking. The goal was accomplished with remarkable success and reliable engineering AE and AU-based models were introduced. These validated schemes were based on the largest experimental database so far produced. Moreover, the proposed models were generalized, i.e. applicable in all damage states examined. As obvious this could seem for acousto-ultrasonics, this is not the case regarding acoustic emission measurements. Thus, from the acoustic emission side, this generalization renders an original contribution. AE-based models proved able to assess tensile and also compressive strength degradation. This is another novel achievement.
In this thesis, the proposed AE models were superior to the respective descriptor-based AU schemes. However, although performance of the second, using novel descriptors, was more than adequate, wave propagation in the specimen under consideration was also studied. This area failed to produce new descriptors or schemes, however indicated damage-associated qualitative trends in the recorded signals. Several issues related to the acousto-ultrasonic experimental technique were underlined and the complexness of the problem depicted.
The experiments presented herein were performed in the frame of EC research project "OPTIMAT BLADES: Reliable Optimal Use of Materials for Wind Turbine Rotor Blades", ENK6-CT-2001-00552. Partial funding was provided by the Greek Secretariat for Research and Technology, F.K. 6660. It is emphasized that no other partner of the OPTIMAT BLADES project, engaged in non-destructive condition assessment, managed to propose successful engineering NDT models.
|
|---|
