Hydrogen embrittlement

From Wikipedia, the free encyclopedia

Hydrogen-Induced Cracks (HIC)

 

Steels were embrittled with hydrogen through cathodic charging. Heat treatment (baking) was used to reduce hydrogen content. Lower bake times resulted in quicker fracture times due to higher hydrogen content.

 

Hydrogen embrittlement is the process by which hydride-forming metals such as titanium, vanadium, zirconium, tantalum, and niobium become brittle and fracture due to the introduction and subsequent diffusion of hydrogen into the metal. 

Susceptibility to hydrogen-induced cracking ('embrittlement') is often a result of the introduction of hydrogen during forming, coating, plating, cleaning, and finishing operations, often referred to as 'internal embrittlement'. Hydrogen also may be introduced over time (so-called 'external embrittlement' through environmental exposure (soils and chemicals, including water), corrosion processes (especially galvanic corrosion), cathodic protection, and/or from hydrogen generated by corrosion of a coating. To be susceptible, a combination of three factors is required: presence of (and diffusion of) hydrogen, susceptible material, and stress. For susceptible materials (such as coated high-strength bolts, where hydrogen is often present after manufacturing and may also be added over time by cathodic protection, galvanic connections, etc.), cracking will initiate when a sufficient stress has been reached; this is known as the threshold stress or Ki_SCC. 

The 'hydrogen embrittlement' phenomenon was first described in 1875.

Process

During hydrogen assisted-cracking (also known as 'hydrogen embrittlement'), hydrogen is introduced to the surface of a metal and individual hydrogen atoms diffuse through the metal structure. Because the solubility of hydrogen increases at higher temperatures, raising the temperature can increase the diffusion of hydrogen. When assisted by a concentration gradient where there is significantly more hydrogen outside the metal than inside, hydrogen diffusion can occur even at lower temperatures. These individual hydrogen atoms within the metal gradually recombine to form hydrogen molecules, creating pressure from within the metal. This pressure can increase to levels where the metal has reduced ductility, toughness, and tensile strength, up to the point where it cracks open (hydrogen-induced cracking, or HIC).

Although hydrogen atoms embrittle a variety of substances, including steel, aluminium (at high temperatures only), and titanium, however these metals are still affected in high concentrations, hydrogen embrittlement of high-strength steel is of the most importance. Austempered iron is also susceptible, though austempered steel (and possibly other austempered metals) display increased resistance to hydrogen embrittlement. Steel with an ultimate tensile strength of less than 1000 MPa (~145,000 psi) or hardness of less than 23 HRC is not generally considered susceptible to hydrogen embrittlement. In tensile tests carried out on several structural metals under high-pressure molecular hydrogen environment, it has been shown that austenitic stainless steels, aluminium (including alloys), copper (including alloys, e.g. beryllium copper) are not susceptible to hydrogen embrittlement along with a few other metals. As an example of severe hydrogen embrittlement, the elongation at failure of 17-4PH precipitation hardened stainless steel was measured to drop from 17% to only 1.7% when smooth specimens were exposed to high-pressure hydrogen.

However, recent computational research (using Parrinello-Rahman molecular dynamics) has shown that instead of leading to a decrease in ductility, there is local enhancement of ductility in areas that are hydrogen saturated. This increase in ductility leads to areas where there is a reduction in the critical tensile stress. This, in turn, enables failure to occur at lower-than-expected stresses.

Hydrogen embrittlement/hydrogen-assisted cracking can occur during various manufacturing operations or operational use - anywhere that the metal comes into contact with atomic or molecular hydrogen. Processes that can lead to this include cathodic protection, phosphating, pickling, and electroplating. A special case is arc welding, in which the hydrogen is released from moisture, such as in the coating of welding electrodes. To minimize this, special low-hydrogen electrodes are used for welding high-strength steels. Other mechanisms of introduction of hydrogen into metal are galvanic corrosion, as well as chemical reactions with acids or other chemicals. One of these chemical reactions involves hydrogen sulfide in sulfide stress cracking (SSC), a significant problem for the oil and gas industries.

As the strength of steels increases, the susceptibility to hydrogen embrittlement increases. In high-strength steels, anything above a hardness of HRC 32 may be susceptible to early hydrogen cracking after plating processes that introduce hydrogen. They may also experience long-term failures anytime from weeks to decades after being placed in service due to accumulation of hydrogen over time from cathodic protection and other sources. Numerous failures have been reported in the hardness range from HRC 32-36 and more above; therefore, parts in this range should be checked during quality control to ensure they are not susceptible.

Counteractions

Hydrogen embrittlement can be prevented through several methods, all of which are centered on minimizing contact between the metal and hydrogen, particularly during fabrication and the electrolysis of water. Embrittling procedures such as acid pickling should be avoided, as should increased contact with elements such as sulfur and phosphate. The use of proper electroplating solution and procedures can also help to prevent hydrogen embrittlement.

If the metal has not yet started to crack, 'hydrogen embrittlement' can be reversed by removing the hydrogen source and causing the hydrogen within the metal to diffuse out through heat treatment. This de-embrittlement process, known as "baking", is used to overcome the weaknesses of methods such as electroplating which introduce hydrogen to the metal, but is not always entirely effective because a sufficient time and temperature must be reached. Tests such as ASTM F1624 can be used to rapidly identify the minimum baking time (by testing using design of experiments, a relatively low number of samples can be used to pinpoint this value). Then the same test can be used as a quality control check to evaluate if baking was sufficient on a per-batch basis. 

In the case of welding, often pre-heating and post-heating the metal is applied to allow the hydrogen to diffuse out before it can cause any damage. This is specifically done with high-strength steels and low alloy steels such as the chrome/molybdenum/vanadium alloys. Due to the time needed to re-combine hydrogen atoms into the hydrogen molecules, hydrogen cracking due to welding can occur over 24 hours after the welding operation is completed. 

Another way of preventing this problem is through materials selection. This will build an inherent resistance to this process and reduce the need of post processing or constant monitoring for failure. Certain metals or alloys are highly susceptible to this issue so choosing a material that is minimally affected while retaining the desired properties would also provide an optimal solution. Much research has been done to catalog the compatibility of certain metals with hydrogen. Tests such as ASTM F1624 can also be used to rank alloys and coatings during materials selection to ensure (for instance) that the threshold of cracking is below the threshold for hydrogen-assisted stress corrosion cracking. Similar tests can also be used during QC to more effectively qualify materials being produced in a rapid and comparable manner.

Examples

• In 2013, six months prior to opening, the East Span of the Oakland Bay Bridge failed during testing. Catastrophic failures occurred in shear bolts in the span, after only two weeks of service, with the failure attributed to embrittlement, possibly from the environment.

• In the City of London, 122 Leadenhall Street, generally known as 'the Cheesegrater', suffered from hydrogen embrittlement in numerous steel bolts, with three bolts failing in 2014 and 2015. Extensive remediation works were initiated. 

Related phenomena

If steel is exposed to hydrogen at high temperatures, hydrogen will diffuse into the alloy and combine with carbon to form tiny pockets of methane at internal surfaces like grain boundaries and voids. This methane does not diffuse out of the metal, and collects in the voids at high pressure and initiates cracks in the steel. This selective leaching process is known as hydrogen attack, or high temperature hydrogen attack, and leads to decarburization of the steel and loss of strength and ductility. 

Copper alloys which contain oxygen can be embrittled if exposed to hot hydrogen. The hydrogen diffuses through the copper and reacts with inclusions of Cu₂O, forming H₂O (water), which then forms pressurized bubbles at the grain boundaries. This process can cause the grains to literally be forced away from each other, and is known as steam embrittlement (because steam is produced, not because exposure to steam causes the problem). 

A large number of alloys of vanadium, nickel, and titanium absorb significant amounts of hydrogen. This can lead to large volume expansion and damage to the crystal structure leading to the alloys becoming very brittle. This is a particular issue when looking for non-palladium based alloys for use in hydrogen separation membranes.

Testing

Most analytical methods for 'hydrogen embrittlement' or more accurately hydrogen-assisted cracking involve evaluating the effects of (1) internal hydrogen from production and/or (2) external sources of hydrogen such as cathodic protection. For steels, it is important to test specimens in the lab that are at least as hard (or harder) than the final parts will be. Ideally, specimens should be made of the final material or the nearest possible representative, as fabrication can have a profound impact on resistance to hydrogen-assisted cracking. 

There are numerous ASTM standards for testing 'embrittlement' (actually hydrogen-assisted cracking) due to hydrogen: B577, B839, F519, F1459, F1624, and F1940, G142.

• ASTM B577 focuses on hydrogen embrittlement of copper alloys, including a metallographic evaluation (method A), testing in a hydrogen charged chamber followed by metallography (method B), and method C is the same as B but includes a bend test.
• ASTM B839 is the 'Wedge test'
• ASTM F519 is the method for evaluating 'hydrogen embrittlement', although this term is a misnomer in steels. There are 7 different samples designs and the two most commons tests are (1) the rapid test, the Rising Step Load (RSL) test per ASTM F1624 and (2) the sustained load test, which takes 200 hours. The sustained load test is still included in many legacy standards, but the RSL method is increasingly being adopted due to speed, repeatability, and the quantitative nature of the test. The RSL method provides an accurate ranking of the effect of hydrogen from both internal and external sources.
• ASTM F1449 is the Standard Test Method for Determination of the Susceptibility of Metallic Materials to Hydrogen Gas Embrittlement (HGE) Test, uses a diaphragm loaded with a differential pressure.
• ASTM G142 is the Standard Test Method for Determination of Susceptibility of Metals to Embrittlement in Hydrogen Containing Environments at High Pressure, High Temperature, or Both uses a cylindrical tensile specimen tested into an enclosure pressurized with hydrogen or helium.
• ASTM F1624 is the incremental step loading (ISL) or rising step load (RSL) method for quantitatively testing for the Hydrogen Embrittlement threshold stress for the onset of Hydrogen-Induced Cracking due to platings and coatings from Internal Hydrogen Embrittlement (IHE) and Environmental Hydrogen Embrittlement (EHE) - F1624-06 Standard Test Method for Measurement of Hydrogen Embrittlement Threshold in Steel by the Incremental Step Loading Technique. F1624 provides a rapid, quantitative measure of the effects of hydrogen both from internal sources and external sources (which is accomplished by applying a selected voltage in an electrochemical cell). The F1624 test is performed by comparing a standard fast-fracture tensile strength to the fracture strength from a rising step load test where the load is held for hour(s) at each step. In many cases it can be performed in 30 hours or less.
• ASTM F1940 is the Standard Test Method for Process Control Verification to Prevent Hydrogen Embrittlement in Plated or Coated Fasteners. While the title now explicitly includes the word fasteners, F1940 was not originally intended for these purposes. F1940 is based on the F1624 method and is similar to F519 but with different root radius and stress concentration factors. When specimens exhibit a threshold cracking of 75% of the net fracture strength, the plating bath is considered to be 'nonembrittling'.

and ASTM STP 962, "Hydrogen Embrittlement: Prevention and Control." The step loading test offers a quantitative measure of hydrogen effects in a much reduced time in comparison with much longer methods and has been proven repeatedly to provide matching results. 

There are many other related standards for hydrogen-assisted or hydrogen-induced cracking (aka hydrogen embrittlement):

• NACE TM0284-2003 (NACE International) Resistance to Hydrogen-Induced Cracking
• ISO 11114-4:2005 (ISO)Test methods for selecting metallic materials resistant to hydrogen embrittlement.
• Standard Test Method for Mechanical Hydrogen Embrittlement Evaluation of Plating/Coating Processes and Service Environments

Numerical modeling

Reliable modeling of hydrogen embrittlement is hindered by its complexity and the uncertainties that surround the understanding of the underlying physical mechanisms. Numerous hydrogen-interaction mechanisms have been postulated, spanning a wide range of scales. As argued by several authors, the most plausible scenario is that several mechanisms are acting in concert, with one dominating others within specific regimes. The predictive capabilities of hydrogen embrittlement models have improved significantly through the years, with current models being able to reproduce laboratory test data with few fitting parameters. Most of these models rely on the hydrogen-enhanced decohesion mechanism, the hydrogen-enhanced localized plasticity mechanism, or a combination of both.