Buckling analysis of tubular wind turbine towers by means of the applied energy method

Ma, Yang ORCID: 0000-0002-9015-2731 (2021). Buckling analysis of tubular wind turbine towers by means of the applied energy method. University of Birmingham. Ph.D.

Ma2021PhD.pdf
Text - Accepted Version
Restricted to Repository staff only until 1 April 2026.
Available under License All rights reserved.
Download (10MB) | Request a copy

Abstract

Tubular steel towers are currently the most practical type of supporting structure for wind turbines used in wind energy harvesting. The surging demand for energy has necessitated an increase in the scale and capacity of wind power plants. However, higher towers result in greater threats from external loadings, as historical records of failure incidences reveal the severity of tower collapses due to various buckling modes of the shell structures. It seems, therefore, necessary to thoroughly investigate the collapse mechanisms of the respective structures in response to environmental actors.

The first part of this dissertation discusses a historical case study which indicates that buckling is a common type of tower failure. The buckling mode shown in this case study displays a range of thin-walled structure buckling patterns, in which the wavenumber on the compression side is between 1 and 9. A new approach to the buckling behaviour of cylindrical steel shells under combined axially asymmetric compressive loads and bending moment is presented, employing the energy method.

Within the proposed framework, in the second section, the classical buckling analysis was conducted to determine the governing equilibriums which express the displacement function of a shell’s mid-surface in terms of stress–strain elementary expressions. The equations describing curvatures were obtained at the same time. The outcomes of classical buckling theory analysis provided a basis for the application of energy methods.

The governing partial differential equations have been rescaled using energy methods, in order to obtain dimensionless expressions which show stored energy taking place along the shell surface by surface displacement functions. For the perfect shell under axial compression (pure bending moment and combined load), equilibrium buckling paths close to bifurcation points have been investigated by energy functions to establish the critical points. The analysis shows that bifurcation buckling is at the infimum of energy levels, whereas multiple longitudinal and circumferential wavenumbers may occur with the energy magnitude jumping over adjacent buckling eigenmodes.

The initial geometric imperfections have been introduced to the shell by trigonometric deformation functions, updating the previous governing functions for imperfection sensitivity analysis. The diverse final buckle point numbers verified according to the buckling types of historically observed incidents are determined by shell geometry and external load magnitude. The respective imperfection sensitivity analysis indicated lower energy bound levels and fewer diversities in circumferential wavenumbers.

The last section comprises numerical analysis, performed by modelling twenty shell sections, with five loading scenarios for each. Shell models were implemented with initial imperfections embedded using the finite element method via the Riks method for tracking buckling paths. The results show that buckling evolution paths are significantly affected by bending parts, which leads to section ovalisation that in turn changes the internal strain energy dissipation routine and section curvature. The cylinder’s geometrical parameters, the length–radius ratio (L/R) and radius–thickness ratio (R/t), also show a strong influence on buckling behaviours, seemingly linked to a noticeable reduction in the shell buckling bearing capacity during the combined loading scenarios. The results of the numerical simulation are also interpreted and validated the derived theoretical equations under various conditions.

Type of Work:

Thesis (Doctorates > Ph.D.)

Award Type:

Doctorates > Ph.D.

Supervisor(s):