This sounds like a very interesting problem. Usually this material does not give major problems in welding. Just a couple of possible problems:
1)You may have a rougue material somewhere. Usually I would suggest that it is the filler, but if you say that you managed to weld the 3" square bar (presumably with the same filler) without problems, then it points to the plate being the problem. Possibly you can do an analysis of this plate, looking especially for S or P.
2)You may have some insidious contamination present. Is the material new? If not, then what service was it in? If new, did you possibly perform some kind of a surface crack detection that left a residue? If you suspect contamination, then you could try heating the material up to around 200°C or so, to try and drive off any such contamination that is volatile.
3)Due to the large sections, your cooling would be very rapid which could result in not enough ferrite forming. This would make your material more sensitive to hot cracking. It would suggest that a moderate pre-heat, as suggested by another respondent, might be the solution.
4)You are using the wrong shielding gas (or single pass wire) which is leading to a build-up of a certain element in the weld, rendering it crack sensitive. As the butt weld in the square bar is less restrained than the weld to the plate, it may be able to resist the cracking while the higher restraint on the plate results in cracking. To check this, perform hardness tests on the weld metal and also double check the supplier catalogues to ensure that you have the right consumables for the job at hand. (Multi pass wire using a shielding gas with less than 3%CO2 etc.)
Let us know if you have other details, or if you figured out the problem.
Regards
Niekie Jooste
Niekie has a good point about shielding gas. 100% CO2 would tend to cause carbon pickup in the weld and perhaps some oxides. We have had good success with 75% Argon/25% CO2 mixes on 300 series stainless using FCAW. Limiting CO2 to 3% is a bit restrictive.
Also, some fillers are designed for single-pass welds. If you use them on multi-pass welds, the elements added to produce good appearance in single-pass welds can buildup to unacceptable levels and cause cracking. Single-pass wires that use higher silicon are agood example. Check the wire classification and manufacturer's literature to make sure you are not using a wire intended for single-pass welds.
Hi Marty
You are obviously right about the CO2 of 3% being too restrictive. I was having a blonde moment! The 3% is applicable to GMAW and not FCAW. In FCAW, the appropriate gas depends on the flux. As such, the supplier's recommendations should be followed.
Regards
Niekie
The problem has been solved by changing to a metalcore wire with no preheat. We still don't know what was causing the cracking but changing wire solved it!
Thanks for everybodys suggestions.
Where were the cracks located? In the bead? Along the toes?
As far as I could tell, the cracks were located in the weld metal in the bead. There did not appear to be any pattern to the crack location with regards to depth or location in the weld metal.
I sent the mill test certs to a metalurgist with the wire supplier. His comments were that there was high sulfur (.026 %) and copper (.41 %) in the 3" x 3" bar, and this may have been responsible.
However, the cracks occurred in the welds between the 1 1/2" plate and the 3" x 3" bar. There were no cracks in the splices of the 3" x 3" bars.
The 1 1/2" plate had low sulfur (.0003 %) and no copper listed in the mill cert.
Things that I would look at: Said you are using FCAW with CO2 sheilding gas. Which means that you probally dont have to worry about to little penetration. I welded low carbon steel using FCAW and 100% cO2 and had to back the penetration off a bit. Was getting internal throat cracks due to improper weld nugget geometry. If you look at a cross section of the weld make sure that the bead width to bead depth ration does not exceed 1:1.5. When you exceed this ratio then weld might crack in center upon cooling. Also cracking might be to to joint restraint. Couple things to consider.
Mr. Niekie Jooste
Could you explain more about item No. 3?
It was interesting for me and I thought cooling rate doesn’t influence ferrite Number. Even I thought the more cooling rate, the better; because it prevents chromium carbide formation.
I am curious about this too. I had thought that fast cooling promoted ferrite formation, and for that reason, you should not preheat stainless.
The cooling rate does influence the ferrite number because in alloys that typically contain some ferrite, there is a solid state phase transformation that is diffusion related.
I must however appologise because I got my transformations mixed up. As mentioned by H.Dibben, the faster the cooling the more the ferrite. It works as follows:
From the liquid state, a primary solidification mode is LIQUID to Austenite. There is however a secondary solidification mode that is LIQUID to austenite and Ferrite. This ferrite can then further transform to Austenite if it is given enough time. If not, it remains (or part thereof) as ferrite.
I would thus have to retract my advice of the pre-heat. Rather cool as fast as possible.
Regards
Niekie