Does black oxide process cause hydrogen embrittlement and require baking?

A. I think that the black oxide step per se would probably not cause hydrogen embrittlement, but any acid immersion in the preparatory steps well might. Baking should not remove the black oxide finish but might effect the wax or oil on it.

For a more authoritative insight, please see if your library can get you a copy of "HYDROGEN EMBRITTLEMENT OF STEEL IN METAL FINISHING PROCESSES OF BLACK OXIDE AND ZINC PHOSPHATIZE" .

Ted Mooney, P.E.
Striving to live Aloha
finishing.com - Pine Beach, New Jersey

A. If the black oxide is formed by the hot alkaline process there is no hydrogen entrapment. Hydrogen is formed in an electrolytic deposition system at the cathode in acid solutions, e.g., Nickel plating, etc., so you have no worries with the black oxide.

A. Hydrogen embrittlement perhaps not, but most steel must see some sort of pretreatment prior to black oxide. If the pretreat includes cathodic cleaning or acid pickling, hydrogen embrittlement might be a concern.

A couple of other interesting thoughts on this are --
(1) the new MIL-DTL-13924 [on DLA] for mil spec black oxide requires post coating baking presumably for hydrogen embrittlement relief. Black oxide has such poor corrosion resistance by itself, we often end up with rusted parts right out of the oven! Oiling the parts prior to baking is a stinky, smoky mess that I wouldn't suggest.
(2) Helicopter manufacturer Sikorsky no longer uses black oxide due to what they claim was CAUSTIC EMBRITTLEMENT of high strength steel. I'm not positive that the two had anything to do with the other but it's interesting that the new mil spec and Sikorsky's "no black oxide" edict coincided within months of one another! I think there's a conspiracy here and it begs the question: is CAUSTIC EMBRITTLEMENT removed by baking like HYDROGEN? Any insight into either the mil spec rationale for baking or the phenomenon of caustic embrittlement would be appreciated.

A. Well its kinda hard to say what you are seeing. I suspect they probably are polishing or grinding marks. I have been doing professional bluing for 25 years and I must admit I have never had anything crack. I have blued some pretty weird alloys over the year too. Everything from super hard gun parts to drills, springs, taps, dies chrome moly, chrome vanadium and HSS steels. Mind you I am using sodium hydroxide and Sodium Nitrate [affil links] and the hottest the solution ever gets is 310 °F. A lot of machine shops or production shops blacken with molten potassium and I'm not certain how hot that gets but I once watched an idiot spit into one of the tanks and the reaction was quite spectacular! I suppose the simple thing would be to do a dye penetrant test, the kits are cheep and they work quite well. If cracks do indeed show up then it's a simple matter of backtracking your procedures and making changes to the process or the materials. But as stated I have never experienced any cracking with caustic bluing.

A. Has the steel been carburised before you black oxided the parts? I had a quick look at the AMS as my initial thought was that maybe the steel would require a stress relief after grinding and found it to be a low alloy steel suitable for carburising.

If it has been carburised then you have a very hard casing around a soft core (in relative terms). Placing the parts into a hot caustic solution after grinding may well induce cracks in the surface.

I would suggest a low temperature stress relief after grinding, 275±10 °F for about six hours may help to mitigate some of the cracking you are seeing.

If the parts have not been carburised then I am not certain what is happening. The steel is of a low enough strength/hardness that you would not expect stress to build up in the surface and I can't imagine that you would be abusing the surface that badly during the grinding operation to cause actual cracking. You may be leaving striations, which may look like cracks, but these would be pretty well visible to the naked eye, so if you are inspecting these parts before you send them out you would probably catch them.

Here's an idea for you, if you have enough time to do it. Set off a couple of batches. MPI after each stage of the manufacturing process, e.g. after machining, grinding, black oxide. One batch try stress relieving as suggested above after grinding and then compare the results. Well, it's an idea to get you going.

Another suggestion is that you walk the process and see if there is something obvious amiss with your process route. Cheap and easy to do, but very effective at picking up some quite simple errors that you may well not have thought of.

A. Hi Sergio,

There are effects caused by immersion in hot black oxide processes. Sometimes we are lazy and label it as embrittlement, giving the impression that hydrogen is involved. In the case of black oxide it has to do with the highly stressful nature of the treatment to the base material. The baking operation is more like a low temperature stress relief rather than a de-embrittlement operation.

As Milt has previously observed, Sikorsky dropped black oxide without replacement due to what they called "caustic embrittlement", probably as they had no better way of describing the failure. In my previous company we built up plenty of evidence that black oxide will drastically reduce fatigue life if the parts are not baked (can be as much as 80%, dependent on the operating conditions!)

Hydrogen embrittlement is a completely different story, is time dependent and I have never heard of allowing parts to just sit around for 48 hours to allow natural dispersion of hydrogen. If you are using a zinc phosphate on steels with a strength greater than 1100 MPa (160 ksi) or a hardness greater than 36HRC then you should be actively de-embrittling to prevent parts suffering from damage.

A. Here is what is happening chemically: The black iron oxide, Fe3O4, is being oxidized (at least in part) to red iron oxide, Fe2O3. (This wouldn't happen in a reducing atmosphere). I wonder why the parts are being treated to reduce hydrogen embrittlement? First, black oxiding is not normally considered an embrittling process. Secondly, it is not typically used on high strength fasteners. Thirdly, are the parts over 32 Rockwell 'C' scale? If not, they are not subject to hydrogen embrittlement anyway.

A. Hi Basilio,

Are you sure it is the coating turning red or is it corrosion of the base material, which is then consequently showing through the thin black oxide coating?

One common problem with stoving of black oxide is that if the parts are stoved with even a hint of moisture the base material will corrode, causing the reddening that you may be seeing. One way one of my previous companies got over it was to use a water displacing fluid after hot water rinsing and blow drying followed by a vapour degrease.

We have a habit of calling the stoving operation a hydrogen embrittlement relief operation. In the case of black oxide it is due to something that has been labeled as caustic embrittlement. Essentially the process is severely fatiguing (see previous posts on this thread) and the operation acts more like a stress relief operation than a hydrogen embrittlement relief operation.

A. Hi Pravin. Your metal finisher has perhaps seen the parts and I haven't, so I'm not anxious to disagree with him :-)

Instead we appended your inquiry to a thread on the subject from which you may be able to learn the answer, or or least the issues to consider. Good luck.

Regards,