Numerical modelling and condition assessment of timber utility poles using stress wave techniques

University of Technology Sydney. Faculty of Engineering and Information Technology.Timber utility poles are traditionally used for electricity and telecommunication distribution and represent a significant part of the infrastructure for electricity distribution and communication networks in Australia and New Zealand. Nearly 7 million timber poles are in service and about $40-$50 million is spent annually on their maintenance and asset management. To prevent the ageing poles from collapse, about 300,000 electricity poles are replaced in the Eastern States of Australia every year. However, up to 80% of the replaced poles are still in a very good condition (Nguyen et al., 2004). Therefore, huge natural resources and money is wasted. Accordingly, a reliable non-destructive evaluation technique is essential for the condition assessment of timber poles/piles to ensure public safety, operational efficiency and to reduce the maintenance cost. Several non-destructive testing (NDT) methods based on stress wave propagation have been used in practice for the condition assessment of timber poles. However, stress wave propagation in timber poles especially with the effect of soil embedment coupled with unknown pole conditions below ground line (such as deterioration, moisture etc.) is complicated, and therefore it hindered the successful application of these NDT methods for damage identification of timber poles. Moreover, some stress wave based NDT methods are often based on over-simplified assumptions and thus fail to deliver reliable results. In the presented study, in order to gain an in-depth understanding of the propagation of stress waves in damaged poles and to develop an effective damage detection method, a solid numerical study of wave behaviour is undertaken and novel wavelet packet energy (WPE) method is investigated for damage identification. Numerical studies utilises finite element (FE) models to track the wave propagation behaviour characteristics considering different boundary conditions, material properties as well as impact and sensing locations. WPE is a sensitive indicator for structural damage and has been used for damage detection in various types of structures. This thesis presents a comprehensive investigation on the novel use of WPE for damage identification in timber utility poles using FE models. The research study comprises several aspects of investigations such as a comparative study between 2D and 3D models, a sensitivity study of mesh density for 2D models, and a study of the novel WPE-based technique for damage classification and detection in timber poles. Support vector machine (SVM) is imported for damage classification and particle swarm optimisation (PSO) is selected to achieve the classification. The results clearly show the effectiveness of the proposed novel WPE based damage identification technique. Damage prediction based on optimisation procedure is also carried out in this thesis. Several numerical models with different damage conditions are created and the damage size is predicted according to optimisation procedure based on information from sample damaged model. Genetic algorithm and artificial fish swarm algorithm are used as optimisation algorithms and the comparative study is conducted based on the prediction results. The influence of damage on the strength of timber utility poles is also studied in this thesis. The damage conditions are changes in diameter, length as well as location. Wind is considered as a main reason to cause the collapse of timber utility poles in this research. Wind load is defined based on Australian standards and the Ausgrid manual, and the corresponding stress is calculated through FE analysis. According to this analysis, it can be found that under specific damage conditions, some small damage may cause collapse; however, for certain conditions, the timber poles can still be safe even when large damage exists. In conclusion, a novel WPE based damage detection method has been successfully developed to address the limitations of existing methods for condition assessment of timber utility poles. The numerical verification has shown the method is effective for identification of the classification and severity of damage