Abstract
This study was undertaken to support the development of a Quantitative Risk Analysis (QRA) for tubulars. The QRA required a burst model that was relatively accurate compared to the normal burst models used in conventional design.
Conventional work-hardening plasticity theory was used to establish a computer program for the prediction of the burst pressure of a thin-walled tubular. The tubular material was assumed to be ductile and the burst phenomenon caused by plastic instability. Dimensions of the tubular, the constant, external, axial load, Poisson's ratio, and a stress–strain curve for the material are the required input data for the program.
In order to verify the predictions of the computer program, twenty-two full-scale bursts tests were performed. The test specimens had capped ends with five of the specimens having an external, axial load applied in addition to the pressure loading. A stress–strain curve was measured for each tubular used for the test specimens so that the computer program mentioned above could predict the burst pressure.
Agreement between the predicted and measured burst pressures was excellent. The standard deviation of the ratio of predicted to measured burst pressure for the twenty-two tests was 0.04 with a mean value of 1.04. The program was then run extensively with all of the stress–strain curves to provide data for the development of an elementary equation which predicts burst pressure in the presence of an axial load. The resulting equation is surprisingly accurate so the program was modified using the same theoretical basis to allow different load paths to be investigated. By this means the burst load dependence on load path was established and shown to be rather small.
Finally, the elementary equation predictions were compared to the API Internal Yield Pressure Equation which is sometimes used for burst pressure predictions. The comparison of the API equation predictions to the elementary equation predictions shows that the discrepancies can be significant depending on (1) the ratio of the ultimate strength to the yield point and (2) the diameter to wall thickness ratio. The first discrepancy occurs because the API equation is based on yield point and the elementary equation is based on ultimate strength. The second discrepancy occurs because the API equation is based on outer diameter, while the elementary equation is based on mean diameter.
