Hydrogen-Induced Martensite Formation

By jon20013 (*****)

Date 05-16-2008 15:45

sounds like a great question for js55 :)

By Stephan (***)

Date 05-16-2008 15:50

Hi Jon!

Good to hear from you...

You are the first one who's enlightening me :-)

Thanks and best regards,
Stephan

By G.S.Crisi (****)

Date 05-16-2008 20:43

Never heard about it. I'm also waiting anxiously for an answer.
Giovanni S. Crisi
Sao Paulo - Brazil

By 803056

Date 05-17-2008 01:32

Hello Stephan;

This sounds interesting.

The formation of martensite has been described to me as: "The decomposition of austenite into martensite is known as "diffusionless transformation" because the quench doesn't allow excess carbon to precipitate." Hydrogen doesn't play a part in the process to my knowledge.

What you describe sounds similar to hydrogen blistering. Not knowing all the details or what paper you read, it is presumptuous for me to offer any opinion, but that's never stopped me before!

Let's say the substrate is carbon steel and the clad layer is an austenitic stainless steel. The interface between the two can have a mushy zone that is martensitic and subject to hydrogen cracking should atomic hydrogen migrate to the martensitic area.

One way that the presence of a martensitic zone in the HAZ can be confirmed is to use the modified WRC diagram developed by Kotecki and Lippold. A copy of the WRC - 1992 can be found in Appendix I of AWS D1.6, but a better diagram developed by J. Lippold and D. Kotecki can be found in their book;
Welding Metallurgy and Weldability of Stainless Steel
J. Lippold and D. Kotecki, Wiley Interscience 2005

If you were to plot the location of the carbon steel and the stainless steel using the chrome equivalency and the nickel equivalency of each, the tie line between the two would indicate the microstructures that could be anticipated in the HAZ.

The carbon from the carbon steel and the chrome from the stainless steel acting together would easily raise the carbon equivalency of the mushy zone of the HAZ to a point where martensite could form upon cooling.

Vessels in petro-refineries can experience hydrogen blistering as a result of being operated at high temperatures with a hydrogen rich environment. The hydrogen diffuses into the metal and collects to form a blister. I would think a base metal that has a martensitic region could experience delayed cracking as a result of the hydrogen at much lower levels than that required for hydrogen blistering.

I don't see how the hydrogen would be instrumental in promoting the formation of martensite. It would be interesting to know the particulars of the case you read about.

Best regards - Al

By Stephan (***)

Date 05-17-2008 16:36 Edited 05-17-2008 17:01

Hello Al,

thank you - once more - for your excellent reply!

Yes! This what you have described in the upper part of your response was - by now - my personal knowledge as well. Even that martensite can emerge, caused by the relatively well-known mechanisms. And that was the reason for that I was so surprised as I have read what the fellows have stated in their paper, even that - at least in my interpretation - hydrogen appears to be a cause for martensite formation as well!

And as you have so excellently stated by saying:

"The interface between the two can have a mushy zone that is martensitic and subject to hydrogen cracking should atomic hydrogen migrate to the martensitic area."

From the results of the paper I can confirm hereby, that even this is - amongst others - a supposed cause for the "disbonding" of the clad layer as a separation from the parent metal. At least as far the investigations treated within the paper have yield.

To be honest with you. I was glad to reading that neither the greatly appreciated fellows Jon (jon20013) nor Prof. Giovanni Crisi (thanks for replying Giovanni!) nor you my friend have ever heard of this. That naturally becalms me, but however it is nonetheless interesting what the gentlemen who have written the paper might have meant by writing so.

Thank you also Al for writing:

"It would be interesting to know the particulars of the case you read about."

If you don't mind, I would like to state hereinafter some of even the particulars you have mentioned. Since these are - at least from my humble point of view - extraordinary interesting.

Well, the paper I have mentioned has the title:

"Disbonding of austenitic stainless steel cladding following high temperature hydrogen service"

was written by M.F. GITTOS et al, as a year 2007 contribution to the IIW Commission IX (Behaviour of metals subjected to welding). The paper (IIW IX-H-653-07) deals with the investigations of a phenomenon which is particularly called "disbonding" what means a kind of removal of the deposited clad layer (different austenitic fillers were used) away from the substrate, as an effect from the component's shut down from service conditions. The components themselves which have been investigated are - just as you have correctly assumed already - high pressure vessels (hydrodesulphurisers, hydrocrackers, heat exchangers and vessels in coal conversion plant) made of 200(!) mm (~ 7,9 inch) thick 2.25Cr-1Mo steel grades. The inwards vessel surface was clad with different grades of austenitic filler materials

· 309L/347
· 309Nb
· 309LMo/308L

as well as a Ni-Base filler of a grade

· NiCr (65.6% Ni/17.9%Cr/3.58%Mn)

Different welding processes were used for the cladding operations as there were:

· Submerged Arc Welding (High- and Low Current Conditions)
· Manual Metal Arc Welding
· Electroslag Welding
· High Speed Submerged Arc-Strip Welding
· Submerged Arc Series-Arc Welding (not further explained but I assume it means SAW Multiple Wire Welding) and
· High Speed Submerged Arc Strip Welding

The service temperature was described as being up to ~ 450°C (~ 840°F) and the service hydrogen partial pressure is mentioned with up to 175(!) bar (~ 2,5 KSI). The vessels are being PWH-treated after manufacturing and it was found out that even this procedure has a considerable influence upon the later disbonding susceptibility (due to martensite formation, tempering etc.).

As you have already assumed Al, the major problem of disbonding is physically a kind of hydrogen embrittlement mechanism. This mechanism is agreed - in this particular application of components - as being based upon the diffusion of hydrogen through the clad-layer (and partially vessel wall) while the component is in service and a subsequent cracking "...on cool down to normal ambient temperature."

What finally has been found out by the extensively performed investigations, as recommendations to decrease the susceptibility for disbonding or increase the disbonding resistance respectively, was (concluded) as follows:

· NiCr Cladding
· Manual Metal Arc deposited cladding
· Finishing temperature for PWHT below 650°C (~ 1200°F)
· High ferrite in the first-layer cladding
· Increased cladding thickness
· Martensite in the first-layer cladding
· Use of vanadium-modified parent steel [according to (*)]

The authors state as well the following procedures to minimize disbonding susceptibility on already existing - in service - vessels as follows:

· Employing hydrogen release heat treatments before cooling
· Applying an additional low temperature PWHT

For the case of fabricating new vessels GITTOS and co-authors recommend the following measures as being most efficient:

· Avoiding low ferrite in the first-layer deposits
· Applying duplex (double) PWHT
· Use of V-modified steels

So far in a short overview the content of the very interesting paper. As you can see, the basis of the "disbonding" mechanism is even comparable with what you have already assumed (you are great Al!!) by saying:

"The interface between the two can have a mushy zone that is martensitic and subject to hydrogen cracking should atomic hydrogen migrate to the martensitic area."

Even this - beside my own humble interpretations of the results of the treated investigations - was the reason for that I was so surprised when I could read what the authors have stated what I have already posted in my original topic:

"A further possible mechanism for martensite formation could be hydrogen-induced transformation caused by the high hydrogen concentrations which develop at the interface (between parent metal and clad layer) during and subsequent to hydrogen charging. In this case, martensite formation would occur on cooling from hydrogen charging."

Since even this is something what I have never heard before. Hydrogen as possible "initiator" for martensite formation? Hmmm, strange...

Unfortunately even this particular statement was not further explained within the paper (at least as my English language knowledge has let me recognize this correctly), what was the reason for me to humble request help from my great fellows in the forums.

However, perhaps our outstandingly appreciated fellow Jeff (js55) may have mercy with me and can point me or us in the right direction by estimating what the authors of the paper might have meant by their statement. And I am absolutely sure, if Jeff has never heard about this, nobody may have ever heard about hydrogen as a martensite transformation initiator. :-)

Thank you once again and all the best to you and all the others,
Stephan

(*)
· SHIMOMURA, J. et al: "Improvement in resistance to disbonding of stainless steel-overlaid 2.25Cr-1Mo steels". ISIJ International,4, 1991 (pg. 379-386)
· BROUWER, R.C.: "Hydrogen distribution through the wall of pressure vessels made of conventional and V-modified steels" International Journal of Pressure Vessels and Piping, 56, 1993, (pg. 133-148)
· CAYWARD, M.S. et al: "Evaluation of hydrogen disbonding of stainless cladding for high temperature hydrogen service" Corrosion 94, Paper 518, NACE, 1994
· FUSARI, F. et al: Properties of strip surfaced overlays on the new generation of parent metals for the petrochemical industry" Welding in the World, 36, 1995, (pg. 173-180)

By DaveBoyer (*****)

Date 05-18-2008 04:16

Stephan, I would like to "zero In" on 1 line from the article: "martinsite would occur on cooling from hydrogen charging"

Could the problem be martinsite formation from retained austinite initated by the temperature change rather than the hydrogen itself ?

The suggestion of low temperature PWHT leads Me to believe that this mightbe an issue.

By Stephan (***)

Date 05-18-2008 08:43

Hi Dave,

thanks for your excellent input!

Yes, as usual you're right with your assumption.

As a first order of influence the martensite formation itself is seen as a mandatory factor for the later disbonding. I.e. - as Al has assumed already - martensite - or better the stresses caused by this constituent - in association with hydrogen are the cause for an exceed of the ductility and thus failure of the interface between substrate and clad-layer. Even these mechanisms are quite comparable with what we know when dealing with hydrogen-induced cold cracking seen as a particular kind of embrittlement.

The parts being treated within the investigations have been subjected to a (quote):

"...standardized PWHT of 30 hours at 690°C (1274°F),(...), for all cladding types. In addition, samples were heat treated at temperatures between 650 (1202°F) and 690°C (1274°F) for times of 1-30 hours. Data generated in the course of the investigation suggested that multiple PWHT could be of benefit in reducing the incidence of disbonding, and studies were therefore carried out with two stage tempering at temperatures of 690°C (1274°F) for the first stage and 565°C (1049°F) - 640°C (1184°F) for the second stage." (unquote)

As I have by now only stated one of those four factors which are seen by the authors of the paper as being relevant for disbonding initiation, please let me hereby state as well the other three of these (quote) "...sources..." (unquote) for martensite formation within the transitional zone between substrate and clad-layer (quote):

1. Martensite forms on cooling from welding close to the fusion boundary where the alloy content of the weld metal is reduced by dilution with parent metal.
2. High alloy dilution regions which just transformed to martensite on cooling from welding may re-austenitize during PWHT, reverting to virgin martensite on cooling from PWHT.
3. Carbon migration during PWHT from the parent metal into the cladding may destabilize austenite by causing chromium carbide precipitation with consequent elevation of the Ms temperature, again, resulting in martensite on cooling from PWHT. [acc. to(*)]

(unquote)

The authors pay very much attention to the PWHT procedures and the details of even those. So they state the very difficile relation between the tempering of original martensite (originated by cooling from welding) and the on the other hand emerging of "fresh" martensite acc. to item 3 above. Since the amount of - so-called in the paper - secondary martensite (generated by cooling down from PWHT) is of course a function of time and temperature of the PWHT procedure itself, they recommend a second PWHT subsequent to the first one. This however should be - of course - lowered in temperature to gain a positive effect in both tempering the first portion of PWHT martensite and reducing the effects as described under item 3 above. The authors conclude the coherences as follows (quote):

"This benefit(...) stems from two causes: firstly the tempering of the martensite present close to the fusion boundary is achieved and, secondly, the low temperature minimizes the formation of fresh martensite by either reversion or austenite destabilization."

(unquote)

They state a reference, see (**), for even these mechanisms which appears to confirm their results.

For me all this sounds very reasonable and understandable but nonetheless the fourth "source" as already several times stated above, namely the

"hydrogen-induced transformation caused by the high hydrogen concentrations which develop at the interface..."

is what has made me truly wondering on what might have been meant by that.

Hmmm, and by seeing now that not either you, Dave, as a very experienced expert in heat treatment issues and material metallurgical transformation topics with rather hard to weld steels and materials have heard of "hydrogen-induced martensite transformation" makes me believe that the authors might eventually have meant something towards what you have stated by saying:

"Could the problem be martensite formation from retained austenite initiated by the temperature change rather than the hydrogen itself?"

and the formulation within the paper has been chosen even just a bit "accidentally".

Once again a heartfelt "Thanks" for your precious input and my best regards,
Stephan

(*) GITTOS, M.F. and GOOCH, T.G.: "The interface below stainless steel and nickel alloy claddings" Welding Journal 71,(12) 1992, (pg. 461-472)
(**) VIGNES, A. et al: "Disbonding mechanism and its prevention" Interaction of steels with hydrogen in petroleum industry pressure vessel service, MPC, New York, 1993

By Joseph P. Kane (****)

Date 05-18-2008 14:21

This string should have been placed in the Metallurgy Topic category.

By Stephan (***)

Date 05-18-2008 14:31

Ouch!

Cheers Joe!

You're surely right!

Hmmm, I guess it's too late now...

Best regards,
Stephan

By DaveBoyer (*****)

Date 05-19-2008 05:03

Stephan, I appreciate Your complamentory opinion of My expertice, but I don't think I am actually an expert on these matters.

I know from my experience with some high chrome tool steels that not all of the austenite changes to martensite in the initial heat treatment, but may do so at a later time, namely after the material is tempered. This leaves untempered martensite. A second tempering is done in these materials to temper the newly formed martensite.

In some of the hardenable stainless alloys a cyrogenic cycle is needed after hardening and tempering to get all the retained austenite to convert to martensite. Then the item is tempered agin.

This is the basis for My asumption. Somewhere in the dilution zone there is a carbon/chrome mystery metal that is suceptable to retaining austenite.

By Stephan (***)

Date 05-19-2008 07:15

Hello Dave,

"I know from my experience with some high chrome tool steels that not all of the austenite changes to martensite in the initial heat treatment, but may do so at a later time, namely after the material is tempered. This leaves untempered martensite. A second tempering is done in these materials to temper the newly formed martensite.

In some of the hardenable stainless alloys a cyrogenic cycle is needed after hardening and tempering to get all the retained austenite to convert to martensite. Then the item is tempered again. (...) Somewhere in the dilution zone there is a carbon/chrome mystery metal that is suceptable to retaining austenite."

I truly hope you don't mind when I today the very first time - but with the very best intention - "disagree" with you! :-)

What, so my humble question, is experience if not what you have written, see above, as being a part of your kind reply?

You know, for me personally, a thread is not completed, as long as you haven't contributed to it!

Thank you once again and best regards,
Stephan

By js55

Date 05-19-2008 14:34

I have never heard of this before. Duh. Since this seems a seminal work. But a little brainstorming could be fun, though I'm sure the real solution will await much better heads than we.
But if I may, martensite can form from metastable austenite through deformation.
And you may have micro regions of metastable austenite at overlay interfaces. So, if a hydrogen build up is causing a deformation, a metastable austenite can transform.
We know that hydrogen can induce deformation. This is what leads to delayed cracking.
So, microzones of austenite with high volume percents of ferrite stabilizers, (metastable), hydrogen evolution increasing pressure and thereby causing a deformation, and deformation causing a transformation of metastable austenite into martensite.
Voila'. :)

By js55

Date 05-19-2008 14:46

And maybe this phenomena is more pervasive than we know its just that it happens on such a small scale we haven't noticed before.

Perhaps in order for it to be manifest you need a lot of hydrogen. Assuming a proper heat treat hydrogen will be predominantly absent from a pristine assembly. Charged in sevice hydrogen will become readily available. But its gonna have to be a lot since the solubility of hydrogen in austenite is high and the diffusivity of hydrogen in austenite is low. Meaning the austenite overlay is going to create a barrrier to the hydrogen (since the austenite is the surface against the medium). Higher temps, high hydrogen, more hydrogen evolution and some will find a way, I'm guessing. When it hits the ferritic microstructure (and the interface) the trouble begins.

By Stephan (***)

Date 05-19-2008 15:18

Jeff,

I humble bow my knees before you and to be honest I do not have any little problem with it!

"So, microzones of austenite with high volume percents of ferrite stabilizers, (metastable), hydrogen evolution increasing pressure and thereby causing a deformation, and deformation causing a transformation of metastable austenite into martensite."

Your analysis is absolutely... excellent!

And by being so reasonable it appears to be the clarification of all my questions.

I guess we all have expected so imploringly your contribution*.

Thanks so much,
Stephan

* You are truly one of the major reasons which making this the BEST welding forum in the world!

By js55

Date 05-19-2008 15:36

Stepehn,
I appreciate your confidence, truly. But, even though it may sound good don't mean its right.
It could be though. :)

By Stephan (***)

Date 05-19-2008 15:46

Jeff,

"But, even though it may sound good don't mean its right."

Normally... Yes, I would agree entirely!

But... in this case I know who has answered!

Hence:

"It could be though." :-):-)

My best regards to you again,
Stephan