The Pipeline and Hazardous Materials Safety Administration (PHMSA) Notice of Proposed Rulemaking (NPRM), with Docket No. PHMSA-2011-0023, substantially revises 49 CFR Part 191 and 192. Notable among these changes was the addition of §192.607, verification of pipeline material. This section calls for the verification of material properties of pipe and fittings located in either high consequence areas, class 3, or class 4 locations where traceable, verifiable, and complete records do not exist. Material properties include grade (yield strength, YS, and ultimate tensile strength, UTS) and chemical composition. The proposed regulations include an independent third-party validation for non-destructive testing (NDT) methods to determine material strength and require an accuracy of within ±10% of an actual strength value.
Among the NDT technologies currently available to pipeline operators to estimate material strength is instrumented indentation testing (IIT). IIT is based on the principal that there exists a relationship between the indentation response of a material and its stress-strain curve. The indentation response is measured during the IIT process whereby an indenter is sequentially forced into the material during testing. The link between the indentation response and the material stress-strain curve is established often through the use of iterative Finite Element Analysis (FEA). The IIT vendor’s proprietary software performs this calculation, converting force-displacement measurements into an estimate of YS and UTS.
In this study we extracted force-displacement data from IIT performed using FEA on an idealized steel. This data was then coupled with literature algorithms developed at Seoul National University (Kwon et al.). Parametric sensitivity analysis was then performed on estimated YS with respect to the algorithm parameters. Preliminary results indicate that while variations in the indenter constant, ω, used to estimate surface deformation do not significantly alter the predicted UTS or YS, the sensitivity to deviations in the empirical constant, Ψ, relating normal load to representative stress was more pronounced due to an effect on the calculated power-law constant, K.
PHMSA’s NPRM accuracy requirements for NDT to establish yield and tensile strength should be driven by a rigorous understanding of material inhomogeneities, uncertainties in actual tensile strength determination, experimental uncertainty, and modeling uncertainties. The analysis performed in this paper provides part of this rigorous framework to establish realistic accuracy requirements for NDT that must drive federal rulemaking. In addition, this research highlights the need for pipeline operators to establish controls on the algorithms adopted by commercial NDT vendors.
