- Material Performance
- Process Innovation, Development, and Additive Manufacturing
- Weldability Testing and Evaluation
You are here
Hydrogen Assisted Crack in Dissimilar Metal Welds for Subsea Service under Cathodic Protection
Dissimilar metal welds (DMWs) are routinely used in the oil and gas industries for structural joining of high strength steels in order to eliminate the need for post weld heat treatment (PWHT) after field welding. There have been reported catastrophic failures in these DMWs, particularly the AISI 8630 steel - Alloy 625 DMW combination, during subsea service while under cathodic protection (CP). This is due to local embrittlement that occurs in susceptible microstructures that are present at the weld fusion boundary region. This type of cracking is known as hydrogen assisted cracking (HAC) and it is influenced by base/filler metal combination, and welding and PWHT procedures.
DMWs of two material combinations (8630 steel – Alloy 625 and F22 steel – Alloy 625), produced with two welding procedures (BS1 and BS3) in as welded and PWHT conditions were investigated in this study. The main objectives included: 1) evaluation of the effect of materials composition, welding and PWHT procedures on the gradients of composition, microstructure, and properties in the dissimilar transition region and on the susceptibility to HAC; 2) investigation of the influence of microstructure on the HAC failure mechanism and identification of microstructural constituents acting as crack nucleation and propagation sites; 3) assessment of the applicability of two-step PWHT to improve the resistance to HAC in DMWs; 4) establishment of non-failure criterion for the delayed hydrogen cracking test (DHCT) that is applicable for qualification of DMWs for subsea service under cathodic protection (CP).
This study revealed that welding procedure BS1generated higher heat affected zone (HAZ) hardness in as-welded condition, formed a larger number of partially mixed islands of base metal (swirls) in the dissimilar transition zone, and produced lack of fusion defects. PWHT resulted in hardening of the planar growth region of 8630 steel - Alloy 625 DMWs due to carbon migration from coarse-grained heat affected zone (CGHAZ) and in formation of fresh martensite in the swirls of BS1 welds. No significant hardening was found in the planar growth region of F22 steel – Alloy 625 DMWs. Base metal and filler metal compositions affected the solidification morphology in the fusion boundary region and the gradient of chemical potential across the fusion boundary that causes carbon migration during PWHT. Increasing the carbon content in the Alloy 625 filler wire and controlling the size of the dissimilar transition zone were predicted as potential solutions for mitigating the carbon accumulation in the planar growth zone.
The susceptibility to HAC in the investigated DMWs was evaluated using the delayed hydrogen cracking test (DHCT). Worst case scenario non-failure criterion was established for the testing conditions of the DHCT: sustaining a constant tensile load of 90% yield strength (YS) with no failure under continuous electrolytic charging with hydrogen for at least 1,120 hours. The only DMW that qualified to this criterion was BS3 PWHT F22 steel – Alloy 625, samples of which sustained 90%YS and 115%YS for with no failure for 1,248 and 1,317 hours correspondingly. Non-failure condition / threshold stress for failure was determined for the BS1 PWHT F22 steel – Alloy 625 DMW that sustained a load of 60% YS with no failure for 1130 hours. Non-failure condition / threshold stress for failure has not been determined for the 8630 steel – Alloy 625 post weld heat treated DMWs, even after reducing the load to 40%YS for 8630/625 BS1 and to 60%YS for 8630 BS3. In terms of the effect of materials composition and welding procedure on HAC susceptibility, F22/625 performed better than 8630/625 and BS3 performed better than BS1.
Fractography was performed using a scanning electron microscope (SEM) along with energy dispersive spectroscopy (EDS) and revealed that brittle fracture morphologies of flat transgranular with grain boundary penetrations and cleavage occurred in the CGHAZ, planar growth region, weld metal swirls and cellular dendritic regions of the DMW. The relative location of the fracture morphologies observed was determined by the percent dilution calculated from the EDS measurements. Ductile fracture (micro void coalescence (MVC)) occurred in the cellular dendritic region of the DWMs. High speed camera recorded the failure of as welded 8630 BS1 sample 2 and revealed that cracking nucleated and propagated through lack of fusion defects. Interrupted testing coupled with low angle micro sectioning (LAMS) techniques showed that fracture occurred along the planar growth region and may have nucleated at carbides within the planar growth region.
Two-step PWHT was developed where the first step tempers the general HAZ. The second step tempers the diluted regions of the weld, particularly weld metal swirls. Based on the thermodynamic and diffusion simulations, PWHT of 10 hours at 650 °C + 20 hours at 300 °C and 10 hours at 650 °C + 20 hours at 500 °C were developed. Metallurgical characterization and hardness mapping were utilized to determine if the second step of the PWHT tempered the microstructure of the DMWs. Based on the metallurgical characterization that was performed, there was no obvious difference in the microstructure when the second step of the PWHT was applied. The hardness values however indicated that the two-step PWHT had a minimal tempering effect in all DMW conditions. Dictra simulations were conducted on each DMW condition and supported the minimal tempering effect of the second step of the PWHT by predicting that relatively small amounts of carbon (up to 0.16 weight percent difference) diffusing towards the fusion boundary. The DHCT was used to test the susceptibility to HAC on samples with the two-step PWHT. The results of the DHCT experiments conducted on samples with the PWHT of 10 hours at 650 °C + 20 hours at 300 °C exhibited a reduction in HAC susceptibility in F22/625 BS1 and 8630/625 DMWs. The PWHT of 10 hours at 650 °C + 20 hours at 500 °C shows an increase in HAC susceptibility in 8630/625 BS3 DMW and a slight reduction in HAC susceptibility in F22/625.
Industry Sponsor: Exxon Mobil
Faculty: Boian Alexandrov (OSU)
Graduate Student: Desmond Bourgeois
Industry Contact: Jamey Fenske