Assume a chrome-plated alloy-steel component that can be susceptible to hydrogen embrittlement (>40 HRC). Questions:
1.Would hydrogen always be present in identifiable quantities if hydrogen embrittlement occurred either because of insufficient bake-out or baking starting too late?
2. If so, what amount of hydrogen (in ppm) would be a minimum threshold as an identifier that hydrogen embrittlement occurred?
3. Any references for this threshold?
This is a complex subject because the level or threshold of dissolved hydrogen to cause embrittlement will vary depending on strength level and alloy type. In failures I have seen over the years, the remaining hydrogen contents were as low as 2 ppm, yes, some hydrogen escapes upon failure.
Take a look at the publication below. I believe there is good information on hydrogen contents in ppm that can cause hydrogen embrittlement in high strength steels and other metals.
Very interesting paper and a colleague of mine carried out some work years ago using X-70 Pipeline Steels.
He was reporting 4ppm as having an influence on ductility.
The measurement of hydrogen were made using a Balzers Thermo-Balance and the value reported was around or just below the limits that this machine so I never liked the results.
I understand that using neutron scattering techniques it is possible to measure 2ppm by weight but wonder if anyone has good data with regard to this.
There are a couple of cases that can cause difficulty.
One is if the residual stress in the parts is high enough that combined with H you initiate cracking. In this case the bake out can’t be started soon enough to prevent cracks.
The other is an in service condition where H is generated, usually by a corrosion mechanism. They often correctly identify these as HIC failures, but then look into the fasteners production and not the application.
The threshold in H and stress have a lot to do with the alloy and the stress level.
Considering that there are not many Fe based alloys that at full H charging will have more than 10ppm H the low end of the range is very low.
In answer to your original questions,
The amount of hydrogen required to cause embrittlement depends on the material in question and it’s condition. For example, high strength alloy steels and tool steels with tempered martensitic microstructures and yield strengths between 175 and 320 ksi can be severely embrittled by less than 5 ppm of hydrogen. But austenitic stainless steels, unhardened steels, and steels hardened below about 35 Rockwell C are generally only mildly affected, or not affected at all regardless of the hydrogen concentration.
Here is a very good paper on the subject of hydrogen assisted cracking:
- It depends on the technique. Hydrogen embrittlement is due to mobile hydrogen, so total hydrogen concentration as measured by Leco combustometric method is not a reliable method to detect susceptibility. A better method would be thermal desorption spectrometry (TDS), which has been used to study fasteners, see here:
- Here are hydrogen concentration data, but not threshold values, for high strength steel fasteners such as you mentioned:
as-received: < 1 ppm
after acid pickling: < 1 ppm
after acid zinc electroplating (higher probability to embrittle): 7 ppm
after 4-hour bake: 5.5 ppm
after 8-hour bake: 5 ppm
as-received: < 1 ppm
after acid pickling: < 1 ppm
after alkaline zinc electroplating (lower probability to embrittle): 5.5 ppm
after 4-hour bake: 3.5 ppm
after 8-hour bake: 2 ppm
- SAE 960312 Hydrogen Embrittlement in Automotive Fastener Applications
Hydrogen is trapped in steel. Some trap sites are weak/reversible, and others are strong/irreversible. Baking treatments not only increase the speed (rate with respect to time) that hydrogen leaves the steel, it also increases the total amount of hydrogen that leaves the steel.
Hydrogen embrittlement can be divided into internal hydrogen embrittlement (IHE) and environmental hydrogen embrittlement. EHE certainly can occur if the fastener is not stressed, and can take longer than hours/days to occur. The recent problem with fasteners on the San Francisco-Oakland Bay Bridge took years to occur.
It’s a short 9 min video made by DNV and Spectra Energy about hydrogen diffusion. I had never seen this demo before and it was an eye opener. It gets real interesting from about the 6 min mark.
The equilibrium amount of hydrogen in martensite depends primarily on the dislocation density, which is in turn reflected by the hardness. Hydrogen binds exit heroically to mixed dislocation (65 kcal per mole). This is greater than the dissociation energy of water vapor, so at room temperature in air for an as-quenched martensite with 0.40%C you will have several hundred PPM of trapped hydrogen. If you heat it to 400F you will drive off the hydrogen but when it returns to ambient, hydrogen will spontaneously return unless blocked by a impervious coating.
What is generally considered the hydrogen content is the free, dissolved interstitial hydrogen, which is the trivially small - one or two PPM is usually cited.
Because the bound hydrogen seeks dislocation, it loves to go to the high dislocation density crack tips and cause its famous trouble.