Best test for 885 embrittlement?

By Stephan (***)

Date 09-12-2009 10:16

OBEWAN,

as 885°F embrittlement has a strong time-temperature dependence, what does this particular 'ASTM procedure', you have mentioned, recommend?

My humble assumption is, it will always require an appropriate amount of time and temperature, in order to find how long it will take the delta ferrite to be dissolved.

Isn't there any information available on these details coming from the steel suppliers?

On the other hand, I am almost sure that Henry will be able to provide you with tons of material on this topic.

Stephan

By OBEWAN

Date 09-12-2009 11:11

I saw it in an ASTM test for sensitization. The main procedure was for weight loss after cooking in acid. It made some reference to 885 embrittlement too. I think it is because embrittlement is related to sensitization. I have the procedure at work and can not get to it until Monday morning. It is the standard practice test for ferritic stainless sensitization, practice X. It is the first time I have seen it in that type of test. As a welding engineer, I have always relied on mechanical tests like impacts to judge a weld process. The challenge is, I am looking for a lab test that can be done on samples without doing impacts. Lets say I want to compare a specimen from a cracked field weld to a good field weld specimen without doing impacts.

By Stephan (***)

Date 09-12-2009 16:07

I see...

It will be interesting to pursue this discussion.

With respect to the detail you have posted "...885 embrittlement in a 430 SS weld...". Are you welding this grade similarly, i.e. using a ferritic filler, or dissimilar, i.e. using an austenitic filler?

Or is it an autogenous weld?

I'm just interested in the weld metal analysis. I mean it could play a part with the embrittlement susceptibility.

Anyway, like I said, I guess there will follow clarification short-term as soon as the stainless-experts step into this.

Thanks!
Stephan

By OBEWAN

Date 09-12-2009 16:34

Yes it is autogenous. It is actually 430Ti on one side and 430 on the other side. Any welding on this material effects corrosion. However, we have some heats that are not effected as bad by welding. Post weld heat treat is not possible. We are in the process of getting actual detailed as welded chemistry reports. The huge staff of metallurgists helping on this project are looking for tramp elements not controlled by the existing spec. The fix will either have to be a switch to another material or a custom material spec. That is why I posted my other question on the new 409 stainless. Supposedly it is not effected by either sensitization or 885 embrittlement. The only problem is material form and finding a domestic supplier. China makes lots of claims on it, but I am wondering about quality issues....I will leave it at that.

One of my other proposals would be to use an austenitic stabilized filler too but the joint does not lend itself well to filler.

By Mikeqc1 (****)

Date 09-29-2009 23:32

Im going to audit a Matl testing Facility they do most anything.
If you are interested i could send you the name.
i have not read the rest of the thread so i dont know if you are still looking.
MDK

By OBEWAN

Date 09-14-2009 12:46

The test I have a procedure for is ASTM A763 Practice X. It says it is for sensitization but that it will also detect sigma phase. It is a boiling acid and weight loss test. The concern I have is that there are no limits stated reagarding brittle/acceptable. The only test is weight loss and the limits are determined by the design authority and customer and must be "agreed upon".

By js55

Date 09-15-2009 12:52

Obewan,
885 is not a sensitization problem.
885 is the decomposition of a metastable Fe/Cr ferrite into an Fe rich alpha and a Cr rich alpha prime, the reason it is also called alpha prime embrittlement. As Stephen said, it is time temp related.
The best type of testing I can think of for this, since it is a mechanical issue is ductility or toughness testing.

On the other hand, it can be eliminated with a full anneal. But if service is in the sensitive range ~ 700degF to 1200degF it will happen again.

By OBEWAN

Date 09-15-2009 13:24 Edited 09-15-2009 13:54

I had always separated sensitization and embrittlemnt, but now I am finding some in the research literature who say they go hand in hand. Like where there is smoke there is fire - that kind of reasoning. The data seems to support it too in that the compounds between the grain boundaries are brittle, and can be the same compounds that contribute to IGSCC. That is why I am trying to learn more.

The ASTM A763 Practice X IS intended to test for sensitization via acid induced weight loss. It also says it will detect "embrittlement". The problem with the spec is that there are no acceptance limits stated. The design authority and/or customer determine the length of time in acid and acceptable weight loss. I think if there is any co-relation to be made to brittle structures is to be done with impact testing and then co-related to the acceptable weight loss, but that is just my assumption from my first exposure to the ASTM spec.

We are having trouble even finding a lab that can do this test, and may have to do it ourselves. We still have not heard from Bodycote testing yet. I am not comfortable going for this data with so many unknowns either.

Please disregard my prior posts, I see now that sigma phase is a third distinct problem. I will have to revisit ASTM A763 now, to see if that is in fact the kind of embrittlement it is intended to test for.

I guess I need to rule out all three problems, but it may require multiple tests. The customer's metallugrists are all currently stomping down the sensitization path though.

By OBEWAN

Date 09-15-2009 14:08

I re-read the spec. It does in fact say sigma phase, but also it says chi phase. I think both of these are due to exposure around 1000F. I am not sure we are in that range after welding since this is only a .035" tube wall. But, I am learning something new. I don't think I learned about chi phase in any of my welding classes. And we mostly focused on the results of exposures to certain temp ranges and not all the fancy greek names.

Post Weld Heat Treat is impossible on this part BTW which really puts my back against the wall unless we can switch it to a 321 SS or something for one side of the weld.

By OBEWAN

Date 09-29-2009 18:30

I read a report today that said hardness is one test that can detect 885 embrittlement. If the material is brittle, hardness can go up by a factor of two. Since we have all the original mill run hardness results, I may test hardness on some of the failed welds and parent metals to see if hardness went up.