Much research has been performed over the past twenty-five years to refine our basic understanding of tubular stability, which includes bifurcation, imperfect systems, factors influencing tubular stability and post-buckling behaviour. Tubular instability resulting from load combinations is not a trivial topic, particularly when inelastic material behaviour occurs. Many influencing factors must be considered when attempting to understand (and predict) the onset of instability.
Many existing collapse predictive methods are either simplistic or involve advanced plasticity or finite element methods. Simplistic methods are typically semi-empirical, and contain a degree of uncertainty resulting in conservative collapse predictions. Nonetheless, they are generally considered satisfactory for design purposes. Advanced methods normally involve high-end calculations using specialized software programs that might not be available for general use. Therefore, a relatively easy-to-use method that accurately predicts the actual collapse resistance is, in many cases, the most desirable option.
This paper presents a collapse predictive methodology, developed from a variety of research projects performed over the last fifteen years. The prediction method, which can easily be entered into a spreadsheet program, is applicable to most forms of tubular members, including pipelines. Applicable load combinations include external pressure, axial tension and bending. An overview of the parameters influencing collapse resistance is also provided, including manufacturing history, material modelling, and tubular geometry and imperfections. Also presented is a summary of accuracy of the method to predict some test results. The test database largely contains results of collapse tests on tubular members subject to only external pressure, and axial tension with external pressure. The adaptation of the method to include external pressure with bending is summarized, and the accuracy of the prediction method is demonstrated by predicting the results of the Oman-India and Blue Stream pipeline collapse test programs, and comparing these predictions with those of other well known methodologies.