We have welded some A36 plate to A36 plate 1 1/4" thick to 1" thick, nothing complex, stuff we have done a million times before but usually with A572-50 not A36. Also we have some stiffiners of A572-50 to the A36 plate. Either the next day, or within several days some of the welds have cracked out.
The toe of the 1/4"-3/8" fillet weld let go and cracked the entire length of the weld. MT insp. confirmed this, a nice clear definate indication.
For the life of me I can't figure out why this keeps happening on this weldment. After two days of just sitting on a bench, we had two more develope in a area that three different inspectors looked at previously, so I know they were not their before, they are new. At first I thought it was lack of fusion and that maybe one individual welder didn't have his machine set high enough. But now that's its been a week and I have sat down and spoken with the welders, I feel comfortable saying their is no way its just one guy.
I had four different welders(people) with four different machines(just calibrated 2 months ago), all FCAW 1/16 E71T-1 75/25% shielding gas, no pre-heat because it wasn't thick enough to require it or cool enough in the shop to warrent it. Where the detail sits the rust was ground off to bright shiny metal. I will admit that they didn't grind through the mill scale, they cleaned it up(D1.1 tightly adhered mill scale not required to be removed). The welders feel that the mill scale was very thick on this one particular plate and that is the reason, that they couldn't burn through it.
I have checked the mill cert for one plate and I found that the level of Manganese is out of range by .05. ASTM A36 calls for plate over 3/4" through 1 1/2" thick to have between .80-1.20, and my cert says the value is 1.25. Can .05 more magnaese really matter? I am waiting for the cert on the other plate.
These plates did have heavy rust and oxidation on them and have some pitting. I didn't see them before they were welded. I am wondering if their is some contaminate on them that I can't see and thet may be affecting them as well.
Does anyone have any other thoughts or suggestions
Thanks Chris
eekpod,
What was the C at? If it was at the upper limit that will result in a pretty high CE. Also, did you look at note B for table 2 in A36 regarding Mn? You might just be having a restraint problem. Maybe try a preheat and slow cool.
The symptoms you describe sound like the affects of delayed cracking, other wise known as cold cracking, or most fitting; Hydrogen induced cracking.
This is based on the fact that you mentioned cracks forming hours or even days after the weld was performed, which is generally attributed to Hi cracking (assuming it's not fatigue related.) In fact several codes and hydrogen procedures call for the inspection of the part hours or days after the weld was performed to allow for creep and hydrogen induced cracking to form.
you make no mention of following Low hydrogen practices, if so this information is probably redudant.
In terms of cause, I cannot be certain but several things I would look at:
Lack of preheat: Although by code you may not be required to preheat. Under low hydrogen practices it would be advised to drive out moisture, and burn of contaminanents like oils etc. I'm surprised you aren't preheating 1.25" material, an old rule was passed down to me about preheating above .75" due to the quenching affect of thicker segments as well as uneven contraction.
Lack of PWHT: The time it takes for hydrogen to diffuse out of a weld is a function of temperature , joint geometry and weld thickness. Low hydrogen practice calls for hours at elevated temperatures (which I can't remember off top of my head) to allow hydrogen to diffuse. It won't diffuse readily at room temperature.
Filler comanitation: Rod with too much moisture will cause excessive hydrogen to enter the weld. likewise hydrocarbons from oils and paint etc can all cause contamination. It sounds like you cleaned the base material thoroughly but you'll be chasing demons forever if it's contaminated filler. Very rarely shielding gas an be contaminated or leaks can allow moisture to get in. I have limited FCAW experience so I'm not sure of the proper Low hydrogen procedure for FCAW filler.
you mention slightly higher manganese levels, my metallurgical knowledge isn't great but manganese and other alloys have an affect on the carbon equivalency which is used to determine the need for preheat. .05% manganese by itself would not be of huge metallurgical concern but it might just be the straw that broke the camel's back.
Hope that helps, hopefully one of the seasoned forum goers has some good information
A36 is normally bullit proof. Even with no preheat. But sometimes it just happens. A36 is a large grain relatively dirty (meaning generally higher S and P) plate (spec maximums dont do justice IMO to the difference between it and pressure vessel plate). This will also mean low Z Axis toughness and ductility. The delay in the cracking points to hydrogen, probably from contamination more than moisture. A36 is pretty forgiving with moisture. But the S could have led to small de-laminations, or at least large elongated sulfide inclusions, and when combined with contamination was just enough.
If you have high restraint this won't help.
Yes, we did have some high restraining going on to keep the parts in the odd oreintation that they called for. I thought that as well, but when I had other welds crack out that were not in the restraint area, I figured I had a different problem.
I just found out this moring that the cert the mill supplier gave me doesn't match the heat number written on the part, so I'll have to wait for the new cert and see the values and report back.
Normally I would recommend preheat on something like this, but with shop production schedules, and it not being "required" the call was made to not take the time to pre-heat. Believe me when I say I think we all wished we had done it now. Not realizing there was going to be a problem, like I said before we have done this on other weldments, using A572-50 plate, not A36.
I don't think its filler wire contamination, because I had four different machines welding on this at the same time, and I didn't have any issues with the welds that were made before anf after this particular part.
I think that the surface of the plate was contaminated sometime in the past, and I can't see it, but it may be affecting me now. I am wainting to hear from the supplier.
eekpod,
Just curious. How are you managing to be out of the short circuit transfer mode using 75% argon?
I understand that. I was thinking that the argon needed to be at least 80% in order to get beyond short circuit mode.
Differences in gases percentages, volts, and amps to arrive at short circuit, spray transfer, globular, etc. are really only variables for GMAW, not FCAW (I believe). Working on LeTourneau Log Stackers in OR we used FCAW 1/16 71-T1 all position, all the time on materials beyond 2" thick with 75-25. In some applications we ran, and I do now in my shop in AZ, 100% CO2. I use 80-20 on GMAW for easy change from short to spray.
Don't have proper books in front of me right now to confirm this but am reasonably sure this information is correct.
Have a great day, Brent
As mentioned, the Ar requirement is only for GMAW
I would suggest sending a small chuck in for chemical analysis over review of the mill certs. The mill certs may only look at the A-36 required elements.
The mills melt so much scrap metal these days. Don't be surprised to see enough significant enough quantities of chrome, nickle, etc.
Like Hogan said, add them up to maybe an already high end acceptable limit of carbon and you can raise the CE enough to cause concerns.
The problems already mention in conjunction with high restraint and intersecting welds can set the scene for cracking problems.
While FCAW is considered to be a low hydrogen process by many people, the selection and storage of the electrode can influence just how much diffusible hydrogen is introduced into the weld.
On the question of intersecting welds, some shops do not clip the corners of stiffeners and run the adjacent welds into the corners for the purpose of sealing the joints to prevent leakage or simply because they are not aware of the potential cracking problems. You can add the residual stresses as you would vectors to determine the magnitude of the resultant residual stress. Bare in mind that once a crack has initiated (maybe from an unfilled concave weld crater) the energy required for propagation is approximately 50% of the force required for initiation.
The following sketch shows how to add the tensile forces when the welds intersect (in a common plane). This isn't usually a problem with A36 where the ratio of yield strength to tensile strength is low, but as the ratio approaches 1 (as in the case of A992), the problems become more pronounced. Three intersection welds, each at 90 degrees to the others, has a very high potential for cracking.
Best regards - Al
eekpod,
We had the same problem last year. After all the analysis and testing we did, it turned out to be the thickness of the mill scale. Our welders, too, complained about the mill scale thickness during welding. The next morning the welds were cracked. We started grinding the mill scale off of all plate 3/4" and thicker, where it was to be welded, and have not had a cracked weld since. We use GMAW 95/5.
Hope this helps.
MICHAEL B
Do you think there may have been the chance the welds didn't fuse completely through the mill scale to begin with and it was simply a matter of the welds full cooling, contracting, and then becoming visually detectable?
I've had cases where MT would produce an indication and everyone was convinced they were false indications due to "geometry". Even when the toes of the welds were finished by grinding, the indications would be intermittent.
What parameters are the welders using, i.e., are they borderline spray or transitioning between spray and short circuiting. I've encounter many welders that like to reduce the voltage and wire feed speed because "it was too hot". Can't say that I blame them, spray mode can be uncomfortable on a hot day!
I like to listen to the welder's grumblings. There is usually something behind all of it. It can frustrating at times when they use their jargon and it isn't clear what the problem is. If you can sift through the noise and get to the heart of the matter, you can take corrective action and head off the problem before it gets out of hand.
Best regards - Al
I agree Al, and the last drawing you posted shows how we have it, a longitudinal weld, with a stiffiner perpendicular to it, with the corner clipped and the two welds don't touch.
Anyone recommend a lab I could send a piece to for some type of anaysis? In in the New England area.
There is an increase allowed for Mangananese with the reduction of carbon, check the footnotes on the table. its something like each reduction of .01 % c and increase of .06% manganese is allowed, but check it out.
Cold cracks, are caused by the combined effects of low ductility of the weld,residual stress and diffusible hydrogen in the weld. A weld's ductility may decrease with a high carbon equivalent and a high cooling speed after solidification. Residual stress in a weld can be larger than expected if it contains weld discontinuities such as incomplete fusion, incomplete joint penetration, overlap, undercut, slag inclusions, and porosity. The source of diffusible hydrogen in a weld is mainly moisture in the welding consumable and atmosphere.
Perhaps oil is contasminating the joint?
MDK
you should put a meter on the machine after a few hours of welding. if you increase electrode dia. make sure your machine can handle it. we helped a fab shop with the same trouble,the start of the shift most welds are good, but after extended welding the welds would not penatrate the base metal. hope this helps.
Hello cwi-cws, following your line of thinking and example, I have seen instances where the power supply grid for a shop would be seeing different load patterns and depending on these, the supply voltage to the welding machines could vary as much as 40 to 50 volts from the optimum range. These variences did in fact have a considerable effect on the machine performance and weld quality. Just a bit more for consideration. Best regars, aevald
You had a 40 to 50 volt swing in the input voltage?
There had to be a problem with the electrical grid to see that big a swing. Even while large electric motors are starting, I wouldn't expect the voltage to vary that much. The power company's switch board must have been swamped by angry callers complaining the fluorescent lights were out, incandescent lights were dimming, and the television pictures were shrinking to the size of a tea saucer.
Then again, with our aging infrastructure, it wouldn't be out of the question. Usually, problems like you mention are because the facility is on the outer fringes of the power grid and too far from a substation.
Best regards - Al
Hello Al, this was in an older industrial complex beneath the St. John's Bridge in Portland, Oregon. The line voltages were supposed to be 480V, but as I said they would vary immensely during the course of the day sometimes. This was also 20 or so years ago. I have not run into anything like that since either. Best regards, Allan
Allen, while I can't say I saw swings that large, I worked in shops in NW Portland from the late 70's through the early 90's and we could always tell when power usage was up. In the winter when heaters came on, in summer when air came on, especially if you worked night shift and people started getting up between 6 & 7 and our power voltage levels would drop. Again in the evening when they got home from work.
While you can't base all your work loads around these problems, we tried to save certain work to times when we knew power levels would be higher and more stable.
Have a great day, Brent
Hello Brent, from your timeline, it's possible we may have crossed working paths during the 80's possibly. I worked some manufacturing, fab shop, shipyard, and structural gigs during that time. Best regards, Allan
This steel didn't happen to come in from the Ukraine did it?
Forum
