Constraint Effects in Slow Crack Growth Test Methods for Service Life Prediction of High Density Polyethylene Piping

Abstract

High-density polyethylene (HDPE) pipe and piping components have been used successfully and safely for natural gas distribution around the world for several decades. The primary concerns for a 50-year life for buried HDPE piping involves designing against three primary failure modes—ductile fracture, rapid crack propagation (RCP), and slow crack growth (SCG) under sustained pressure loading. Although, design methodologies for preventing ductile fracture and RCP are well established, SCG remains to be a limiting failure mode in determining useful service life of HDPE piping as it may occur under sustained pressure and temperature. Although considerable amount of research has been conducted over the last two decades, SCG still remains less well understood than other failure modes. A critical evaluation of various test methodologies available to determine the SCG resistance of HDPE resins was conducted using finite element analysis (FEA) of various widely used laboratory test specimens. While there exist extensive information on the test methodologies and the applicability of each of the SCG testing methods, there is a growing concern as to whether any/all of these SCG tests give the same information akin to the industrial pipe application, particularly so when conflicting messages are obtained from time to failure predictions from two different SCG tests. While notched-pipe test (NPT) proves to be a direct approach to assess SCG resistance of the polyethylene (PE) pipe with the use of temperature as a test accelerating factor; in the case of newer grade PE resins, the failure time of NPT can still be considerably large (∼5000 to 10,000 h). For this reason, some of the other coupon SCG tests are focus of recent investigations and especially sought after for rapid ranking/assessment of resins and understanding the manufactured HDPE pipe performance. In this study, FEA was conducted to facilitate a direct comparison of leading SCG test methods, through determination of both the stress intensity factor, KI, and existing constraint factors in various widely used specimen geometries. These results are then compared to pipe specimen with an outer diameter (OD) or inner diameter (ID) surface notch. Since, constraint can have a significant role in SCG initiation, transverse/constraint stress (T-stress), and biaxiality ratios (β), these were compared along the crack fronts to arrive at definitive reasons for the smaller failure times observed when testing some of the SCG test specimens, and also reasons for SCG mode of failure observed even under large applied loads (large KI compared to that in a notched pipe) when testing some of the SCG test specimens. The use of stress intensity factor, KI, along with the T-stress and biaxiality ratio (β), was found to provide a complete picture on the broad spectrum of failure times observed from various SCG test specimens, and rationale for choosing a SCG test specimen when evaluating HDPE pipe or resins.