All the chapters of this thesis are linked with the concern of aeolian vibration. At locations where the motion of overhead power line conductors is restrained (e.g. at suspension clamps), the presence of aeolian vibration may result in another phenomenon called ``fatigue of overhead conductors'. The latter being responsible for serious damage to overhead power lines.
The first part of the document is devoted to a review of the basic concepts of aeolian vibration and fatigue, of the tools available to model, measure or predict vibrational damage, and what are the remedial measures.
In a second part, a series of experimental studies are related:
- A review of the usual fatigue indicators, based on measurements performed on a 63.5m laboratory test span (chapter 2). The aim is to better understand how to perform a correct vibration risk analysis, knowing e.g. what to measure and at which locations.
- An evaluation of conductor self-damping properties based on real outdoor measurements, using a new type of monitoring device, able to perform continuous measurement on power lines (chapter 3). Unlike laboratory tests, such on-site measurements permit to take into account e.g. the effect of span ends, the spatial and time fluctuations in wind, leading to a more realistic vibration risk assessment.
- A study of the vibratory pattern associated to the failure of a conductor wirefootnote{Overhead power line conductors are made out of ``twisted' (``stranded') wires} (chapter 4). If not detected early, the presence of conductor fatigue may eventually lead to the failure of some conductor wires. This chapter investigates the possibility to detect such an event and eventually to use it as a fatigue indicator.
The third part of the thesis gathers modelling studies, starting with a few basic model validations (chapter 5), which permit to understand some interesting phenomena observed experimentally and to highlight the presence of amplitude fluctuations in the computed time response. The latter are believed to be linked to tension fluctuations. The hypothesis is further developed in chapter 6, comparing the results of constant versus variable tension models and discussing the potential impact of tension fluctuations, when the vibrational behaviour of a real line is being extrapolated from tests performed on laboratory test spans. The modelling part ends up with the mining of experimental curves measured at University of Liège by A. Godinas (chapter7). Once this input has been adequately processed, the followings have been achieved:
- A formula to compute the self damping per unit length within the conductor has been deduced. The predicted self damping values are consistent with those deduced from the literature, using more complicated measurement techniques,
- The parameters of an equivalent viscoelastic material have been deduced. Once implemented within a non-linear beam element, it permits to model both the conductor variable bending stiffness and self damping properties,
- A new method is proposed to measure the conductor self damping properties. The required test set-up is the one used by Godinas. The proposed method is much simpler than others already published in the literature in the sense that: the load can be applied quasi-statically, only a limited number of measurements is required, the test set-up is simple, simple to operate, less expensive than others...The perspective shown by this work is the following: knowing the conductor properties and the shape of some moment versus curvature cycles, the proposed formula would permit to estimate the conductor self damping in any kind of aeolian vibration without any dynamic testing.
In the last and fourth part of this chapter, the modelling results have been compared to experimental ones. To be more precise, the compatibility between the conductor variable bending stiffness deduced in chapter 7 and other results either published in the literature or measured by the author is checked in chapter 8. Then, chapter 9 illustrates the difficulties faced to reproduce resonance conditions on an experimental test bench. Some of them could be due to tension fluctuations.
Last, in chapter 10, an energy balance is used to figure the order of magnitude of the self-damping due to non linearities.
![[Public/Internet]](/ETD-db/images/unrestricted.gif "[Public/Internet]")