delay in baking after black oxidation of hardened HSLA 1

[fhadzafar]

High strength low alloy steels Jobs were quenched and tempered to a hardness of 40-41 HRC. Subsequently they were cleaned by degreasing and pickling, then oxidized for corrosion protection. After the surface treatment they were supposed to be baked as soon as possible as per MIL-DTL-13924D. However, the baking process was carried out after 48 hrs (190 Celsius for 3 hrs and then cooed to RT).Parts are physically hardness tested individually and hardness was found b/w 39.5- 41 HRC. All parts are currently free of any surface cracking(after lapse of 27 days). Are they still susceptible to hydrogen embrittlement and cracking? how we can qualify by destructing some of the parts to assure that they are fit for use and are not prone to hydrogen embrittlment cracking? or they can only be rejected?

 

[tbuelna]

MIL-DTL-13924 sec 3.10 requires embrittlement relief of ferrous parts with surface or thru hardness of Rc40 or higher. Section 4.4.4 specifies embrittlement relief in accordance with AMS 2759/9. AMS 2759/9 sec 3.3.3.1 requires the parts to be baked within 4 hours after completion of the plating process unless temper etch inspection is required. If temper etch inspection is performed, the max delay allowed before baking is 24 hours.

The QA provisions of AMS 2759/9 require documenting the time a batch of parts completes the plating process and when that batch of parts begins the embrittlement relief process. If your engineering/manufacturing documentation specifies black oxide coating in accordance with MIL-DTL-13924, then the QA record would show that this batch of parts was not processed in accordance with the relevant specs. So QA should first reject the batch as being non-conforming. Your manufacturing folks can then submit a DR to engineering asking if there is a way to salvage the batch of parts. Using destructive testing to validate the condition of the discrepant batch would likely require testing a large number of specimens, so it may not be practical or economical.

 

[fhadzafar]

Thanks tbuelna for the clarification. These are total 18 jobs only. Please let me know the suitable verification testing, And supporting standard/source literature for this verification, if any.

 

[tbuelna]

fhadfazar-

I believe there are some NDI techniques that can check for this condition but I don't know much about them- how accurate they are or the cost involved. I would suggest doing some research on non-destructive inspection processes for hydrogen embrittlement.

Some people claim that in certain limited cases parts like yours can be salvaged with some re-processing. However, I work in the aerospace industry and based on my own experience it would not be acceptable to rework parts with this condition. The consensus is that the metallurgical damage produced by a delay in post-process embrittlement relief cannot be undone. So the only acceptable disposition for these parts would be to scrap them.

Even if your engineering/MRB thinks it might be acceptable to rework the parts for your particular application, then what you really will need to decide is whether the added cost and effort involved in re-processing and verifying that your discrepant batch of 18 parts is good ends up being less than what it will cost to scrap them. I definitely feel for you, since I've been in similar situations. I know how much a company's management hates to scrap parts, and how much pressure they can put on you to agree with a salvage/rework approach.

Good luck to you.
Terry

 

[fhadzafar]

Thanks indeed Terry. i'll check out the NDE methods.