Evaluation and Prediction of Hydrogen Assisted Cracking of Dissimilar Metal Welds PDF Download

Author: James R. Rule
Publisher:
ISBN:
Category : Dissimilar welding
Languages : en
Pages : 217

Get Book

Book Description
This work builds upon the previous research regarding hydrogen assisted cracking (HAC) of low alloy steel to nickel-base filler dissimilar metal welds (DMWs). In particular, this work is focused on DMWs commonly experienced in offshore oil and gas production systems in subsea use. The HAC tendency of these welds has been attributed to formation of susceptible microstructures at the fusion boundary during welding. As such, a post-weld heat treatment (PWHT) is utilized to temper these microstructures as well as relieve residual stresses. However, these microstructures can persist even after PWHT due to the steep compositional gradient driving migration of carbon from the base metal toward the fusion boundary and into the partially mixed zone (PMxZ) of the weld. The degree to which this migration occurs is a function of materials selection (base metal and filler metal) as well as weld and PWHT procedure. Due to this phenomenon, there is a balance that must be found to provide tempering of the susceptible microstructures that form during welding and limiting the formation of new susceptible microstructures during PWHT. Previous research has established a test method in the form of the delayed hydrogen cracking test (DHCT) which can delineate the effects of materials combination, weld procedure, and PWHT on HAC of DMWs. This test's qualitative ranking of susceptibility agreed well with industry experience. The current study worked towards refining the test methodology investigating the effects of test parameter influence on realized results. Of the investigated variables, it was found that how the test samples are coated is of primary importance where a consistently exposed fusion boundary scheme providing the most repeatable result in test. Additionally, a comparison was made between the test hydrogen charging condition which uses a dilute acid and constant current density of 10mA/cm2 and the service environment which is seawater with a constant potential (-850 to -1100mVAgIAgCl). Through this comparison it was understood that the dilute acid is indeed an accelerated charging environment where the level of acceleration scales with nascent hydrogen concentration differences as indicated by pH and charging current density differences with the dilute acid providing roughly a 1000x acceleration factor. Further work focused on establishing a pass/fail criterion which would transform the DHCT method from qualitative to quantitative. This was done by measuring the diffusible hydrogen content of each DMW at various charging times to find the saturation time. This diffusible hydrogen saturation time was then compared with DHCT results from previous and current work to show that samples which sustain load beyond saturation do not fail due to HAC. This methodology proved successful for sound welds without prior defects and correlated well with service experience. The final focus of the work related to modeling and prediction of DMW microstructures towards predicting HAC susceptibility. The modeling involved thermodynamic and kinetic simulations to model the diffusion of alloying elements both during weld and PWHT thermal cycles. The model was validated using quantitative measurements of the composition through electron probe microanalysis as well as through hardness and microstructural evaluation. The results were correlated with HAC experience to provide a microstructural/character map to facilitate identifying trends which led to susceptibility. The findings confirmed previous research showing fresh martensite to be the main driver for behavior followed by precipitation of M7C3 carbides. This model was applied to a previously untested DMW to predict the microstructure and gauge the relative HAC susceptibility. The predictions proved to be accurate and aligned with both microstructure and HAC susceptibility. The framework of the model can be used as an engineering tool early in the design stage for materials selection and weld procedure development. The final focus of the work related to modeling and prediction of DMW microstructures towards predicting HAC susceptibility. The modeling involved thermodynamic and kinetic simulations to model the diffusion of alloying elements both during weld and PWHT thermal cycles. The model was validated using quantitative measurements of the composition through electron probe microanalysis as well as through hardness and microstructural evaluation. The results were correlated with HAC experience to provide a microstructural/character map to facilitate identifying trends which led to susceptibility. The findings confirmed previous research showing fresh martensite to be the main driver for behavior followed by precipitation of M7C3 carbides. This model was applied to a previously untested DMW to predict the microstructure and gauge the relative HAC susceptibility. The predictions proved to be accurate and aligned with both microstructure and HAC susceptibility. The framework of the model can be used as an engineering tool early in the design stage for materials selection and weld procedure development.