Role of Crack Closure in Crack-Size-Dependent Corrosion-Fatigue Crack Growth of X65 Steel Exposed to Sour Brine Environment

Abstract

Corrosion-fatigue in sour brine (SB) environments is a significant design consideration in deepwater floating production systems. Extensive testing over the past 20 years has shown that sour brine environments can reduce the fatigue life of line pipe steels by factors of 10× to 50× compared to fatigue lives measured in laboratory air; moreover, the extent of material degradation depends on a multitude of loading, environmental, and materials variables. Thus, in 2010 Southwest Research Institute (SwRI) embarked on an industry-supported Joint Industry Project (JIP) to develop a quantitative model to predict the effects of these variables on corrosion-fatigue crack growth rate (CFCGR) in offshore structure steels exposed to sour brine environments. Phase 1 of this JIP had successfully developed and validated such a model in the intermediate fatigue crack growth rate regime — i.e., with CFCGRs between 10−4 ∼ 10−2 mm/cycle. However, the Phase 1 model gave overly conservative CFCGRs at rates in the low growth rate regime below 1 × 10−4 mm/cycle, corresponding to S-N corrosion-fatigue lives in the high-cycle fatigue regime. It was hypothesized that these conservative predictions might result from the fact that the model did not consider effects of crack closure that could significantly reduce the effective crack-driving force in this low growth rate regime, a process that might also give rise to crack-size effects. Thus, the primary objective of the current study was to assess whether or not crack closure is responsible for the conservativism in the Phase 1 CFCGR model, as well as to explore related crack-size effects that in theory would not be predictable with conventional linear elastic fracture mechanics. Both of these possible effects are explored here using critical CFCGR experiments on X65 steel in sour brine under loading conditions for which the nominally applied mechanical driving force (ΔK), as well as the stress ratio (Rσ) and loading frequency were held constant, while crack closure measurements were made as the crack grew from 2 mm to about 10 mm. The crack closure measurements were made using elastic compliance measurements made with a specially designed, high-sensitivity clip gage. Results indicate that a crack-size dependence of CFCGR did occur and could be correlated using a crack-closure-corrected effective stress intensity factor (ΔKeff). These results have provided a foundation for extending the JIP’s Phase 1 CFCGR model into the low growth rate regime in the ongoing Phase 2 of the JIP.