martensitic structures

Date 02-14-2001 14:42

Anyone who whishes to get a god grip of hydrogen cracking
should read WELDING STEELS WITHOUT HYDROGEN CRACKING
written by N Bailey. Published in North America by ASM international, The Materials Information Society, Materials Park, Ohio 44073, USA.

European standards for avoiding hydrogen cracking are largely
based upon this book.
As for your specific question you should supply the heat input, material composition and plate thickness. This would give the readers of this forum a possibility to give you some quick advice.

Regards

Happy

By kam

Date 02-14-2001 16:26

Thanks Happy!!

I will get a copy of the article you mentioned. Did find an article by N. Bailey titled " Fisheyes, Hydrogen Embrittlement, and Removal. Interesting but doesnt answer my question.

I can get rid of this questionable microstructure, that is why I did not supply weld schedule information, I'm just wondering if its worth the trouble.

AWS "Welding Handbook" states that a martensitic structure is susceptible to hydrogen cracking but does not tell me to avoid it if possible.

Just looking for a few opinions on the subject and possibily some brain food (Articles, books, ect) so I can make a decision myself.

Thanks Again Happy

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

Date 02-21-2001 01:25

Martensite is the metallographic structure formed by the quick cooling of austenite. Austenite is the metallographic structure of iron and carbon steel stable above the "transformation temperature", which in degrees Celsius is about 723 (you convert to degrees F). When austenite is allowed to cool slowly, ferrite will form. If it cools quickly, the result is martensite. This process is called "tempering".
Martensite is very hard and brittle. So, the main problem in a weld containing martensite is not that it picks up hydrogen, because it'll pick up the same amount of hydrogem that ferrite will. The main problem is the brittleness of the weld. That's why welds prone to tempering, such as carbon steels with a carbon content of over 0.30 percent and chrome moly steels are wrapped with insulation after the weld is complete: to make them cool slowly. In the case of chrome moly steels, this isn't sufficient and the weld must undergo stress relief. How can these welds cool quickly and form martensite? The usual manner is because of breeze (or at least air drafts) blowing on the weld. That's why it's recommended to make the welds in a place protected from wind, or at least to use suitable windshields if the former isn't possible.
In your case, you're welding SAE 1020 steel with a carbon content of 0.20 percent. How can martensite form? Are you making the welds in a windy place without any protection? You know better than I. In any case, don't bother too much.Sure, martensite will form if the weld is allowed to cool down rapidly, but with a carbon content of 0.20 percent this is not a problem to worry about. The amount of martensite is too low to make any harm. Martensite starts being a problem to get concerned about when the carbon content is over 0.30 percent, which isn't your case. As you yourself say, fatigue tests are OK and so is hardness.
SAE 1020 steel requires that before welding the material is warm (or at least not too cold) to the hand. After the weld is complete, it doesn't require stress relieving (PWHT) if the thickness is 3/4 inches or less.
Giovanni S. Crisi
Sao Paulo - Brazil

By Ian Date 02-22-2001 11:34

Giovanni

I was interested to read your comments on Cr-Mo and cooling rates.

I make a very limited number of Kart racing frames for my own use and have previously used mild steel CDS (cold drawn seamless) tube and Mig'd or brazed the joints. The frames 'sag' a little after a couple of years and loose their springiness so I was intending to making a new frame using Cr-Mo tube (2mm wall thickness) and intending to Mig weld (GMAW) the joints (this seems to be the most common method of manufacture of frames in the UK and Europe). I was not intending to pre-heat or post heat the joints - they would cool naturally in air (no draughts). Am I likely to encounter problems ?

I could, if it proves to be a significantly better option, braze the joints but as I had to sell my bottles some years ago I would have to either borrow or buy a new set which I don't really want to do for insurance reasons.

I was investigating the Mig brazing process (as in another posting) but have not been able to find much out about the process or the equipment needed.

Any advice would be gratefully recieved.

Regards
Ian

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

Date 02-22-2001 22:43

Ian,
as you say, you're using "mild steel", which contains 0,25 percent carbon or less. This material isn't the best for the application you want to us it for.
If I were you I'd try using SAE 1040 steel, i.e., a plain carbon steel with a carbon content of 0.40 percent. I don't see the need to use chrome moly steel. Chrome and moly alone don't increase the steel mechanical strength at room temperature too much. What does increase the mechanical strength at room temperature is vanadium, that's why tools are made of Cr Mo V steel.
Using SAE 1040 you'll notice the difference. A word of warning: 0.40 carbon steel is a little more difficult to weld than mild steel. You'll have to be careful.
Giovanni S. Crisi

By DGXL

Date 02-26-2001 00:57

Hey Kam,
I am currently using GMAW-P on similar materials (1018 and 1025 welded to A 500 tube or A 53 pipe, both grade B) with very good results. The HAZ and martensite are minimal compared to that of conventional GMAW arc transfer modes. I have macroetch samples with macrophotographs to show this as well. In fact, the metallurgical engineer (a P.E.) that performs my mechanical tesing was a little surprised when we first started R&D for this equipment 3 years ago. He was taken back by the relatively narrow HAZ and minimal martensitic structure.

The presence of martensite is unavoidable and is inherent to any arc welding process using carbon based materials. Our parts performed well under tensile and compression tests as well as the proof testing of the overall weldment (this included welds with known flaws).

By the way, what equipment are you using? I am using Miller Phoenix power supplies with Auto-M controllers for mechanized process control. Good Luck.

DGXL

By kam

Date 03-21-2001 16:42

We are using Miller Auto Invision. Now I'm getting cracks in the HAZ at the weld nugget interface. I thought this might happen. The weld size is very small (approx. 1.5mm) and is a highly restrained joint. Looks like a preheat is going to be required. Material is SAE 1026 insteal of 1020. There is no doubt that what I was seeing was martensite. Two outside labs verfied that. Base materials are acting as a heat sink which is causing rapid cooling of the weld pool. I understand the basic principles of metallurgy concerning the 3 T's (time, temperature, transformation) and the mechanical properties of martensite. What I was looking for was whether I could make a correlation between hardness of the martensite and suspectibility to cracking. Considering the joint design, can i monitor hardness in the HAZ and predict cracking. I need to do alittle more research.

Thanks for all your replies
Lots of very good comments....keep them coming.

By HappyWelder (*)

Date 03-22-2001 13:46

There is no question about that susceptibility to
cracking increases with hardness. However, it
might not be possible to quantify, since hydrogen levels
and restraint also are important factors.

If you find a nice correlation please post it. I've got problems
with martensite and hydrogen myself. Any input would be
highly appreciated.

/Happy

By DGXL

Date 03-27-2001 02:49

Hi again KAM,
This is getting interesting. Will IJP be a problem if you try reducing the average heat input? Try lowering the background amperage/voltage parameters (reduce heat) if it will not cause any non-fusion problems. I had only 1 or 2 welds out of thousands with cracks using this process as stated in my earlier post. Both cracks were in the HAZ which propagated into the weld metal. Are the discontinuities your experiencing soley in the HAZ? Hotter is not always better (see the post on the aluminum even though your materials are ferrous).

The problem with my application was IJP/IF discontinuities in the root and groove wall areas. I had wrestled with this for some time before I changed the default parameters on my Auto-M controllers. I obtained instant positive results when this change was made.

When I increased heat input by tweaking the parameters,that is when I first found the cracks. At that point I backed up a little and achieved what I was looking for. Send me an e-mail and let me know how you do on this, or, try contacting Miller, their pretty good with tech-talk.

By kam

Date 04-10-2001 19:21

Interesting. I wonder if the cracking is occuring during the heating phase of the process. My experience with induction heat treating found that cracks can happen either when that power or heat is being applyied (expansion cracks) or while the part is cooling (quench cracks). Most of the expansion cracks were caused by the heat opening up small inperfections in the surface.

The cracks that I am seeing start in open area such as a void due to incomplete joint penetration and stay right along the nugget and do not head towards the nugget or the base material. I believe that area is called the amixture zone. What I have read tells me that this area is one of the weakest parts of the weld due to chemical seggreation.

What article on aluminum are you refering to? I would like to read it. Last weekend I was playing around with my mig welder at home trying to get a good aluminum weld. The only way I could get a good bead was to travel hot and fast.

thanks for the info

kam

By - Date 03-31-2001 18:52

It is interesting to see things "develop", and I thought that I would give my opinion on the developments. I assume that when you say that you are welding GMAW-P, you mean that that you are using a pulsed spray transfer mode.

The weld size that you mentioned, (1.5mm) is indeed very small. Even with relatively thin base metals, you are liable to have very fast cooling rates. This will necessarily mean more martensite formation. Many people assume that heat input (VxA/Speed) is an absolute indicator of the cooling rates. This is not so, because your cooling rates is related to your temperature gradients. Even if you maintain the same heat input you can vary the temperature gradients.

As an example, if you double the Amps, and double the welding speed, the heat input remains the same. The temperature gradients will however be much bigger. (If you disagree with me on this, let me know, and we can debate it a bit.)

Another factor to consider, is the size of the HAZ. If you are obtaining a very narrow HAZ, it is an indication of high temperature gradients. A small HAZ is advantageous, because the "problem area" is then smaller. Even if a tiny crack develops here, it will be rapidly arrested from propagation by the more ductile weld metal and parent metal. The area of grain growth will also be very small. This area generally leads to a poor weld strength and impact properties.

We therefore have two factors that tend to "balance" each other. Low thermal gradients lead to less martensite, but wider HAZ's. High thermal gradients lead to more martensite, but a narrower HAZ. I believe your problem lies in achieving the "sweet spot" where you have a small enough HAZ, with relatively low martensite present.

Depending on your current welding parameters, it may be advantageous to either increase your temperature gradients, or decrease them. It is impossible for me to tell you which. Obviously, keeping everything else the same, an increase in pre-heat temperature would lead to lower thermal gradients.

Regarding your question of a correlation between hardness and succeptibility to cracking, the answer is yes, there certainly is such a correlation. The correlation is however not simple, because it relates to a number of parameters:
1)Joint design. (The higher the constraint, the lower the hardness to prevent cracking.)
2)Material microstructure.
3)Amount of hydrogen in weld and HAZ. (The less the hydrogen, the higher the hardness before cracking occurs.)

As a rule of thumb, in the industry in which I work, (Pressure vessel manufacture.) a hardness of below 240HB is considdered "safe" for the type of materials that you are talking about. In your industry, it may be different.

I hope this helps.

Niekie Jooste
Sasolburg
South Africa

By kam

Date 04-10-2001 19:35

Very well stated. I have noticed the same correlation between heat input parameters and HAZ patterns that you talked about. What i'm planning on doing is monitoring the hardness and size of the HAZ to see if I can find the point where cracks are lickly to occur. This should not be a problem but its going to take alot of time and work. The amount of joint restraint should remain the same (highly restrained - no give at all or I guess you fully restrained). The hardness I am seeing is around 45 Rc. Lower hardness should assure a crack free weld.

I am using a pulsed spray transfer.

Thanks for the comments

kam