Abstract
The flow properties of line-pipe steels control the failure resistance of the pipe, and as such are key in successful pipeline design, and in understanding the factors controlling failures when they occur. As first-principals predictive models are challenged to quantify the flow response in typical line-pipe steels, engineers must rely on empirically developed properties to support numerical analysis for purposes of design and/or integrity management. Stress-based design logically relies on a limiting stress, whereas strain-based design used to address issues like ground movement relies on a limit strain. Post-yield these limits are coupled through the steel’s stress-strain curve and strain-hardening response. Because the burst-pressure of pipes has been shown to depend on the steel’s collapse stress as well as its strain-hardening exponent, n, engineers will need more that the yield stress, Y, or the tensile stress, T, to adequately characterize a pipeline’s resistance to failure.
This paper presents results for the mechanical properties of line-pipe steels developed up to the ultimate tensile stress, or beyond. These stress-strain curves reflect 1) Grades ranging from vintage A25 through recent X100 production. These results have been analyzed to quantify n, Y, and T. These results have further been trended to relate commonly available metrics like Y/T and n, and provide a rational basis for the choice of properties input to numerical analysis.
It is apparent from this work that current correlations between n and Y/T diverge from the trend for the lower-strength Grades. Further, these results show that within a Grade the value of n is a strong function of the ratio of the actual yield stress (AYS) normalized by SMYS, with this dependence indicative of differences in the chemistry and processing used to achieve the Grade. The effects of n and its dependence on the ratio AYS/SMYS are illustrated regarding the predicted response of line pipes subject to increasing pressure. These predictions have been validated by comparison with results for about 20 full-scale tests to illustrate the viability of this technology.
