I wonder if there is not a bit of dross from the plasma cuts that is lending some oxide that isn't getting cooked off..
It would only take a quick skim to remove oxides to count this out.
maybe you are at a critical point with the travel speed?...too fast to allow the outgassing before the weld freezes?
John,
That was my first thought when i saw the imp speed. I have used the same settings except the imp is about 16, not 21. Lawrence also gives good info.
I reduced the IPM to 16 for about 8' and ended up with as much if not more porosity. Preheating to 150F caused the tacks to break so I stopped with it. I could beef up the tacks but that would cause other issues...
I couldn't grind the faying surfaces this time because the parts were already fit. I will try this next time. Although I don't do that on any other thicknesses we weld (.3125 - .750) and I never see this type of defect. In your experience is thinner plate more susceptible to contamination than thicker plate? Seems like it would be...
The welds on this product amount to two 40' welds and one 8' weld per part. It is fit-up/welded on a floor plate. Some here believe that there is "dirt or air" being "sucked up into the joint". I'm fitting on top of a floor plate that is swept off prior but it is still possible that the bottom side of the joint is in contact with some dirt. The bottom side is not visible during welding. I wonder if I was able to put some heat resistent gasket under the joint before I fit if that would help. Copper backer would be nice but an 80' copper backer is not in the cards.
One thing that I noticed, there is some fine debris in the joint that is very hard to keep out (slag from tacks, tiny bits of ??) The joint is blown out but as tight as it is it may only serve to wedge the debris deeper.
A little more detail on the shape of the joint... It is not perfectly square because of the the kerf from the plasma torch puts about a 3 degree bevel on each part. So what I end up with is a 6 degree inclusive single vee so to speak (factor in chamfers too). I have always welded the first pass on the side where the kerf opens up but maybe I would have better luck with the other side where the opening is tightest.
i would bet the farm on its atmospheric gasses being introduced from the bottom of the puddle, we had a big investigation on a very similar problem and never had a repeat after taking the measures we did. for 1/4 inch you may even consider going to an automated flux core or hard wire or metal core depending on your end requirements. plasma dross generally ends up as many many surface inclusions but does not extend down to the root. were the pockets of porosity columnar in shape in line with the seem?
darren
The porosity was fisheye at the surface and extends to the root. And it is always in line with the seam. I need to stick with SAW process with L-61 but it wouldn't be a big deal to change fluxes if that would help. So what were the measures that you took pray tell.
sorry to take so long to get back to you was a busy weekend.
well we went away from saw for the thinner stuff such as the .250 if you need to adjust your wps then a consult with some of these learned folks on the forum would be a start but if you are using wholly one manufacturers product then i would assume that they would be very interested in providing you support.
the porosity you described is exactly what we had and found it to be atmospheric contamination from the underside of the puddle.
there is one suggestion of creating a bed of flux on the bottom of the plate is this possible as i think that would be the cheapest and fastest solution. could be implemented within a few minutes/hours.
darren
as an after thought here we had a problem with too much penetration with one job and we ended up extending our stick out so as to pre heat the wire and have more of a roll on "prell" effect instead of a high penetration effect. might be an easy fix if it works.
another question is, is the flux new or does it have a recycled portion?
ive run into this before. watch the plate from the bottom and see what colour the puddle gets to. the problem i had was it was getting too hot and atmospheric gases were being introduced from the bottom and developing columnar porosity along the centre of the seam. i went to two pass stick and it was cured. in my opinion it is because the weldment is getting to too high a temperature while being welded. if you need to do this with saw then either a piece of pipe cut in half lengthwise on the bottom of the first pass with inert gas, or eliminate the joint prep and just mark with lime stone marker for aid in sight/alignment of the weld would probably eliminate this problem.
darren
I AM A TANK BUILDER AND WE USE SAW ON OUR ROUND SEEMS WE HAVE PROBLEMS WITH POROSITY ALSO SOMETIMES THE FLUX HAS SOME MOISTURE, AND MOST OF OUR PROBLEMS ARE THE POWER SOURCES! I HAVE ALSO DROPED THE VOLTAGE AND INCREASED THE AMPS, SEEMS TO HELP THE POROSITY A BUNCH HOPE I WAS OF ANY HELP, MIKE
The voltage seems like it might be a bit high for 1/4" plate. You are penetrating clear through the joint but the plate is cooling too rapidly to allow the gases to escape. (just a guess)
If you're using a concentric cone lower it down a bit. It sounds like porosity related to flux burden. Happens occasionally when reducing wire sizes, current, puddle sizes, or increasing travel speed, on SAW. Its more often associated with switching from agglomerted fluxes to fused fluxes.
Its also related to moisture in the flux. Its just that with bigger wires and generally bigger puddles and generally higher current the gases have more time to escape before puddle solidification.
Is 780 or 781 flux an option? Is a flux back-up bed practical with this joint?
Found this old post of mine and figured I would post the solution I found a couple of months ago.
Starting from the beginning...
The procedure:
This is a B-L1a-S groove. Filler/Flux combination is F7A2-EM12K (L-61 with 860 neutral flux). Base Metal is A36. No preheat is required and we haven't been using any (material temp is well above 32F). We use crows feet on either side of the joint so that we can check after/during the welding for proper alignment. Both sides have a 1/32"-1/16" chamfer to aid tracking. All joints are machine cut with a plasma torch. DCEP. The root opening is 0. CTWD is 3/4" - 1".
The equipment:
I'm using a LT-7 Lincoln Tractor with a 1000A CV power source. CV is enabled on both the power source and the tractor. Ground is secured directly on the part being welded.
THE SETTINGS THAT WORK:
1/8" wire L-61 860 flux
1st Pass
300 A
30 V
25 IPM
2nd Pass
450 A
30 V
25 IPM
The problem as "darren" pointed out, was that atmospheric gases were being introduced from the bottom of the joint. Just to put things in perspective I am welding up to 80ft at a time of .250 plate. Trying to elevate the plate is not an option and a flux bed would be a lot of trouble for several reasons. The settings above work because the first pass is just a "backer pass" so to speak. It works to seal up the joint but not go too deep and introduce the gases that want to tunnel up through the puddle before it can freeze. Once I figured out that I could make the first pass porosity free by turning the Amps down to 300 (from 395), I just had to get the correct setting to hit the root from the other side. I verified CJP with a macroetch and we're off and running.
Thanks for all the replies and help guys.
C.
One more thing for you to check:
It is possible that the 1/4" plate is made from a different process than the other plates due to it's thickness. Such as the plate > 1/4" might be killed (i.e. fully deoxidized), and the 1/4" plate might not be killed, it could be rimmed, semi-killed, etc. A relatively easy way to check would be to compare the plate certs. It might make sense with how you slowed down your travel speed and got as much or more porosity. You were probably melting more base metal.
If it's SA36 it has to be in the killed condition according to section II part A. Or is there some fine print on that one?
According to the ASTM A36-05 that I have here, killed or semi-killed steel is only required over 1/2" thick in Section 6. It doesn't hurt to check anyway.
I'm using ASTM A36 / A36M-90. LOL! S91.1 at the end says "The steel shall be killed and have a fine austenitic grain size." Obviously I need the new code on that one. Thanks for the pointer. I'll keep that in mind when buying and welding.
Forum
