Abstract
Federal rule changes governing natural gas pipeline operation and safety have driven operators in the United States to establish materials verification programs (MVP) for pipelines that lack reliable records. The ongoing MVP at the Pacific Gas and Electric Company (PG&E) developed an extensive database of laboratory (destructive) tensile testing on over 144 coupons from pipe samples to quantify, in addition to yield strength and ultimate tensile strength, strain hardening exponent (n). A material’s strain hardening exponent quantifies the stress-strain behavior in work hardening and is also used in fracture mechanics-based tools, such as CorLAS™, for assessing flaws in cylindrical bodies, such as pipelines.
There is limited strain hardening data for pipeline steels in available literature. This led to the development of a correlation utilizing the ratio of yield stress to ultimate tensile strength (R-ratio) as part of NG-18 research in the 1970s. The original publication for the CorLAS™ model, by Jaske and Beavers in 2002, used this correlation to propose a quadratic relationship between n and R-ratio; however, this relationship was based on a total of four data points each from a different grade of line pipe. As a result, the uncertainty of this empirical relationship is large.
A more robust relationship that would be applicable across multiple grades and vintages was sought. Using the PG&E’s materials verification database, this paper presents an improved relationship for strain hardening exponent based on 58 pipe features ranging from the 1930’s to the 2010’s and grades ranging from Grade A to X70. Several case studies highlighting the impact of this change on downstream failure pressure evaluations using CorLAS™ are provided. In some instances, the authors have found that using this new model for the strain hardening exponent in conjunction with CorLAS™ can result in differences of approximately 10% in the calculated failure pressure.
