Subscribe to our newsletter to receive the latest news and events from TWI:

Subscribe >
Skip to content

Hydrogen Embrittlement of High Strength Nickel Alloys

Hydrogen embrittlement of high strength precipitation hardenable nickel alloys

TWI Core Research Project 1128/2020

Precipitation hardenable (PH) nickel alloys are often used in subsea applications where environmentally-assisted cracking is a potential risk. In this project a review of hydrogen embrittlement of PH nickel alloys was undertaken, followed by environmental tensile testing and characterisation, to better understand this mechanism of failure.


The high strength and corrosion resistance of nickel-chromium-iron alloys, such as Alloys 718 (UNS N07718), 945 (UNS N09945) and 945X (UNS N09946), make them particularly good candidates for use in demanding environments in the upstream oil and gas industry. These materials generally perform well where resistance to sulphide stress cracking, chloride stress corrosion cracking, and hydrogen embrittlement is required. However, environmentally-assisted failures can still occur.

It is generally accepted that for hydrogen cracks to initiate, threshold conditions of stress, susceptible microstructure and hydrogen concentration must be exceeded. Logically, adequate control of any one of these variables would prevent failure altogether. Of course this is not always practicable in the field, and a simpler approach is often to understand how these variables interact, such that the risk of failure can be managed.

The effects of stress, hydrogen concentration and microstructure have been explored in isolation by a number of authors, however, there does not appear to be a unified source of information on the interaction between each variable.  In this project, the effect of microstructure was explored by heat treating Alloys 718, 945 and 945X to standard and non-standard conditions.  Tensile specimens were slow-strain-rate-tested in air and under cathodic protection (CP) to explore sensitivity to hydrogen embrittlement.  Finally, the effect of a severe stress concentration, in the form of a sharp notch, was used to determine whether there was an enhanced susceptibility to hydrogen embrittlement, due to the presence of local stress raisers. The results were compared with tests undertaken by other authors under various hydrogen-charging conditions.


To evaluate the influence of stress concentration and microstructure on susceptibility of precipitation hardenable nickel alloys to hydrogen embrittlement.


Unnotched and acutely notched circular cross-section tensile specimens were tested in air and under hydrogen-charging conditions in a solution of 3.5% sodium chloride with CP, to simulate subsea service. After testing, the fracture surfaces were examined at high magnification to determine the fracture morphology at various radial positions.

The materials were also characterised using a combination of metallography, hardness testing, and light and scanning electron microscopy.


Testing of the unnotched specimens under CP did not result in a significant reduction in proof strength. However, there was a clear relationship between increasing material strength, as measured in air, and reduced ultimate tensile strength (UTS) and elongation, when tested under CP.

The materials revealed an increased notch sensitivity, when tested in the presence of hydrogen, particularly for the higher strength materials, such as Alloy 945X (UNS N09946). Notch sensitivity in hydrogen was manifested mainly by reduced UTS. The increased notch UTS sensitivity in hydrogen is attributed to the interplay between strain localisation within the notch, and the propensity for hydrogen to diffuse towards highly strained and plastically deformed regions. It would appear as though the onset of plasticity was the point of divergence in material properties, and that the materials behaved similarly in equivalent tests in air within the elastic regime.

Figure 1 shows the fracture morphology of the hydrogen-charged specimens consisted of a ring of  brittle faceted fracture which corresponded to the area into which hydrogen had diffused during pre-charging and testing. Towards the centre of the specimen, the fracture morphology became increasingly ductile.

High magnification inspection of the embrittled portions of the fracture surface revealed the ‘brittle’ intragranular facets to be populated by slip band traces, the intersections of which were shown to be nucleation sites for micro- and nano-voids. At high strains, it is anticipated these voids will coalesce, resulting in hydrogen crack propagation. Most importantly, these results show that hydrogen embrittlement of these alloys, whilst macroscopically brittle, is fundamentally a high strain and dislocation activated plastic process.

Figure 1. Fracture surface from a notched tensile specimen after slow strain rate testing (SSRT), tested under CP.  A ring of brittle intergranular and transgranular fracture is shown around the circumference near the root of the notch.
Figure 1. Fracture surface from a notched tensile specimen after slow strain rate testing (SSRT), tested under CP. A ring of brittle intergranular and transgranular fracture is shown around the circumference near the root of the notch.


This project has shown that the susceptibility of precipitation hardenable nickel alloys to hydrogen embrittlement is governed by microstructure, strength, and the presence and geometry of stress raisers. However, it is understood that materials used in this project are sensitive to the test variables, such as strain rate, pre-charging duration, etc. It is also apparent that the tensile SSRT method cannot quantify the point of crack initiation, or easily distinguish between crack initiation and propagation. These are two fundamental aspects of environmental cracking that need to be better understood. In other words, environmental tensile SSRT, as conventionally carried out and without quantified acceptance criteria, is useful only for basic screening or ranking.

Appropriate guidance should be sought before applying these methods during design, fabrication and service to ensure that components are fit-for-service.  Where possible, quantitative methods should be explored, allowing crack initiation to be captured and its significance to the environmental performance of these materials to be determined. These aspects are currently being studied in follow-on programmes at TWI.



We would like to thank Special Metals and Titanium Engineers Ltd. for their in-kind contributions to the project. Gisle Rorvik from Equinor is thanked for his assistance in mentoring the project.


This project was funded by TWI’s Core Research Programme.

Avatar Mike Dodge Principal Project Leader, Technology

Mike first joined TWI as a PhD student working to understand hydrogen-assisted cracking mechanisms in dissimilar metal welds. He now works within TWI’s Materials and Structural Integrity Technology Group where he leads failure investigations and research projects for TWI Industrial Members. He has a particular interest in environmentally-assisted cracking, and has authored a number of papers and technical reports on this topic.