A computerized database was compiled with results from over 3000 CT fatigue and mechanical tests conducted over a 10 year period. The database contains detailed information about tensile properties (yield and ultimate strengths) for individual samples that were tested to failure under constant internal pressure with specialized CT fatigue testing fixtures. This paper presents a survey of those data to assess the variability in the mechanical properties within a specific CT material grade and to investigate the effect of this variation on CT fatigue behavior.
The growth of the coiled tubing industry can be partially attributed to the successful development of analytical techniques for modeling the ultra-low cycle fatigue behavior exhibited by CT under routine use [1–5\. Analytical fatigue life prediction models for CT are semi-empirically formulated for specific material grades. Parameters are computed based on constant pressure fatigue data generated on specialized testing fixtures using samples made from a specific CT material grade.
Models developed using this approach are focused on the fatigue properties of the material without regard to the conventional tensile properties (e.g., yield and ultimate strength) obtained from monotonic tensile testing. More advanced models [1,3\, also utilize uniaxial strain-controlled low-cycle fatigue test data [6\ but such data are also generated from a specific sample of the same alloy, and do not explicitly make use of the tensile properties for that alloy. The philosophy underlying this approach is that a particular material grade has consistent fatigue properties and variation within the grade is a second order effect.
However, mechanical properties for any alloy can vary within that particular alloy, from one heat of material to the next. Coiled tubing material is no different [7\. Nominal yield strengths for CT are minimum values and actual values typically exceed nominal values, sometimes considerably. The variability in tensile properties is examined in this paper from a database of over 3000 samples compiled over a 10 year period. This database includes standard, commercial grades, and unusual experimental grades, including special tests on samples with welds, defects, prior field service and other anomalies. In this paper, only data from commercial grades are presented.
Fatigue testing of coiled tubing (and all other engineering) materials is subject to scatter. Two nominally identical samples, taken from the same string, tested on the same machine by the same operator, on the same day, under identical conditions can fail in a significantly DIFFERENT number of cycles. When utilizing data taken from different heats over a long time period, even more variation can occur. Another goal of this paper is to present an assessment of how variation in fatigue behavior could be caused by variation in mechanical strength, as reflected in the tensile properties. Since relations have been observed between monotonic tensile properties with fatigue properties in engineering alloys, it is possible that the tensile properties could be used to refine coiled tubing fatigue life prediction routines to account for variations and reduce scatter. The potential for this is also explored.
It should be noted that the fatigue testing program that generated the data presented in this paper was not conducted for purposes of developing a fatigue life prediction model. The testing was conducted as a product development tool, in order to gain an understanding of the differences in fatigue behavior between different material grades. The testing is also conducted routinely for purposes of quality control.