Welding A-36 to 4130

[FDS]

We've been having a problem in our shop lately with weld cracks forming and are stumped at how to fix the issue. I'm hoping someone here can help. We are welding some very large pieces here in the thousands of pounds. The main cores we weld to are 4130 while the pieces we weld on are A-36. It all goes out to heat treat afterwards and then gets machined.

We have been using 3 different machines. We have a Miller 456MP, Deltaweld 652, and a Dimension 452. All have given the same results. We are using 70T-1 wire to weld the 4130 to the A-36. We have been preheating everything to about 450*F or so the day prior to welding and then holding it there overnight to ensure even heating inside and out. The A-36 parts are bevelled on each side so we can get a full weld the thickness of the part. One side gets welded a couple of passes and then the whole assembly flipped over and a few passes on the other side. Repeat until finished. Each one is large enough that it generally takes the good part of a day to get it welded. The pieces are also post heated and allowed to cool very slowly.

We have seen some hairline cracks on the weld joints before heat treating and have had to fix these. These get annoying. We just had 2 pieces return from heat treating and they have some very large cracks that need to be addressed. We are stumped as to how to fix this. We've tried several different things and can't seem to get down to the root of the problem which is why I'm posting here. Is it a heating issue? A welding rod issue? Etc...

If anyone can suggest a cure for our issues, I'd appreciate it.

[zapster]

I think the problem lies in the 4130 aspect of it..

It should be "normalised" before welding..

Heat to 1700 degrees..
Anneal between 1450 and 1550 degrees..

Weld..
Re-harden @1600 degrees..
Quench with water or oil...


A3 is tool steel..

And it needs the basic same procedure...

Thats according to my boss and the "Machinists Handbook" pages 1650 to 1660..

...zap!

[FDS]

They are normalized when we get them. After they are welded up, we send the assembly out to get normalized, quenched, and tempered. Then they come back and go on the lathe.

[pulser]

Any photos of the cracks, or a description of were they are located?

I assume the cracks are in the 4130 heat affected zone next to the weld or are they actually in the weld metal? Centerline or transverse?

Sounds like the A36 is positioned perpendicular to the 4130, like a "T" joint, and the A35 is double beveled. Do you fill the bevels, and then fillet reinforcement as well?

[zapster]

Maybe we're missing something here..

Are both pieces normalised before welding??

...zap!

[jrfv]

I think you need to use wire with 1 or 2 percent molybdenum content. 4130 is one of the chrome/moly grades isn't it?

[FDS]

Any photos of the cracks, or a description of were they are located?

I assume the cracks are in the 4130 heat affected zone next to the weld or are they actually in the weld metal? Centerline or transverse?

We are getting cracks at the welds themselves. the strange thing is that we've been doing this for years with no issues until a couple of months ago. Nothing has changed with our technique so we've been wondering if our supplier has changed the grade of metal we are getting. No idea. They claim not to.

Sounds like the A36 is positioned perpendicular to the 4130, like a "T" joint, and the A35 is double beveled. Do you fill the bevels, and then fillet reinforcement as well?

That's it exactly.

[FDS]

Maybe we're missing something here..

Are both pieces normalised before welding??

...zap!
Only the 4130 is normalized.

After welding we sent one out for NQT and they only tempered it at 700*F since the heat treat place checked the hardness on the A-36 as opposed to the 4130. It came back way too hard and was very hard to machine not to mention too brittle for what we need. We sent it back to redo at 1100*F and all is fine. Another piece we sent out was tempered at 1100* and it cracked at the welds pretty bad. We had to gouge them out and repair them. These were some pretty large cracks. No idea why it happened to one but not the other. We've had 2 come back this week cracked and one was fine. The guys doing the welding aren't new to us. They've all been doing it at least 15 years with one guy having over 30 years welding. They're stumped.

4130 is a chrome-moly grade. We need it for the strength. We've been over the Machinist Handbook for any info we can get and are just stumped right now. I'm not one of the welders. I'm in the engineering department trying to help figure this out. It's a little bit embarassing for us as this has never happened until recently and we just don't know why. Hairline cracks have happened before but nothing serious. We are trying to get rid of those too since we've upped our standards and testing procedures.

Since we haven't been to sure that we're getting what we need materials wise, we've been checking hardness on everything as it comes in. Before someone even touches it. Unfortunately that doesn't tell us the grade of the material but at least we know something is wrong if the hardness is off.

We've been doing this for a long time and just need a fresh approach as to how to deal with this. There is obviously something that we are overlooking.

[B2N3 Welder]

Is there a possibility its the weld wire? Can you or have you tried stick welding the material with 8018 B2 or 9018 b3 10018-d2 or 7018-1 may work.The Dual shield i used from esab worked really well also is dual shield 8000-B2,dual shield 88-cm,E91T1-B3. These have worked for me depending apon the chrome and molly content.These wires all are different in contents.

[pulser]

We are getting cracks at the welds themselves. the strange thing is that we've been doing this for years with no issues until a couple of months ago. Nothing has changed with our technique so we've been wondering if our supplier has changed the grade of metal we are getting. No idea. They claim not to.
That's it exactly.

We still don't know enough about the cracks to say anything about why they occur or how to prevent them. You say the cracks are at the welds, what does that mean?
Are they centerline, right down the middle of the weld, easy to see with the naked eye on the surface of the weld?
Are they transverse to the weld centerline?
Are they located along the edge/toe of the weld, along the toe on the 4130 side of the weld.
Since these are multiple pass welds, what weld pass is cracked?
You know what I'm asking here? Some photos would really help!

[olddad]

Memory isn't what it used to be and like has been said, really need pics to troubleshoot. I would think the pre-heat needs to be lowered to 350 with a max interpass of 500 or close to it. Solid wire, a 90 or 100 with high moly content. Spray or pulse with 95 - 98% argon. I would use spray mode with 98% myself but only because I've not used pulse enough to be confident with it.

[MoonRise]

Everything used to be good and now it isn't? Then something changed.

Your procedures are the same, so that leads to a suspect material change.

Since the 4130 is a semi-special alloy grade (meaning it's not just "mild steel", but should have the alloy percentages controlled pretty well), that would lead to suspecting the A36.

Bad batch of A36? Too much sulfur or phosphorus in the steel could lead to bad welds and weld cracking.

Or your torch liners have gotten crudded up with oil, and the oil is transferring to the weld and you then have hydrogen assisted cracking.

And like folks above have been saying, pics or a better description of the crack location and orientation can be a good clue as to what/why the cracks are occurring.

[MicroZone]

Something changed - perhaps the weldors are quenching too fast? Rapid quenching of the metal will create problems such as cracking and lamellar tearing.

[pulser]

So, what happened? Any new info on the cracking?
Inquiring minds want to know.

[olddad]

LOL...was going to give him a week and ask the same thing myself. They can't throw a question like that out there and just run off...LOL

[slagmagnet]

I think the best approach to this issue is to send samples of the suspect materials to a lab and find out exactly what they're dealing with. Otherwise it just turns into a dart throwing game-resulting in a lot of wasted time.

Assuming that's ok, get the expertise from the many welding engineers on staff of the welding wire company they are using. I have always found those folks to be quite responsive and helpful.

Unless a metallurgist is looking this problem over the liability consequences of a failed structure could be a hard lesson in getting to the bottom of the problem.

The carbon content could be an issue as well in either material much the same way as welding cast and it's attendant issues with cracking.

I agree with all the previous comments that all these things could produce this outcome. The only caveat to that comment is quenching with water or oil, I think that is too uncontrolled in a shop environment.

BTW, what are these structures used for?

Hope it helps.