Welding Tips and Tricks

I talked earlier about posting failed tests to use as learning tools for others. Here are three that failed. I would like to discuss how and why they failed and how to fix it.

The first two pics are 3/8" 3G GMAW using pulse and both failed about the same way. It is obvious the HAZ was very brittle and they didn't get far at all in the bend. The second problem is serious lack of fusion. These MIG tests were done extremely close as far as settings. One thought is the sides were not held long enough not allowing the weld to melt in and penetrate. Why did two fail and one pass?

The third picture is a 1" 1G GMAW in spray transfer. This one failed because of lack of fusion. Here is a thought from us, when this test was done it was filled up to near flush and let cool over night so when we came in the next morning the plate would be cool, thus lowering the possibility of undercut brought on by a extremely hot plate. Because of this, is it possible that the weld did not penetrate because of the temp difference? Another thought is the travel speed picked up due again to a fear of undercut thus not allowing it to heat and penetrate?
-Jonathan

Here is a passed 3/8" 3G GMAW sent at the same time and using the same settings.
-Jonathan

Interestingly, even the "pass" shows flaws that make it borderline. The fusion/HAZ is very obvious, and there appear to be some tiny voids.

I also see, both in the "pass" and the "fails", that there seems to be a transition in heat input. There seems to be a "narrowing" (perhaps broadening? Hard to guess travel direction in the coupons) of the fusion lines.

I'll be interested in the analysis. My guess would be that the passes are too large (putting down too much metal too fast), and not getting enough heat into the base metal for full fusion, calling for a lower overall setting of wire and Voltage, with a higher relative wire:Voltage ratio, and a slower overall travel speed for a broader heating of the parent metal.

Did you have to adjust the WPS, or was it operator error?

Steve S

Steve,
Two more tests were sent up today and they tried to hang on the sides longer. If they fail again we might have to do some changing. However, the last picture was a pass using the same settings so I am leaning toward operator error. I have watched the passes before and noticed a faster overall travel speed. Technically, since one passed, the WPS is set but if we can't get these to pass I am not sure what to do. I am considering running another one as close to what I did before and see what happens a second time.

On a positive note, this was used to convince them to purchase a bender for test pieces. One of my selling points was if we had a bender in house we would have found these problems prier to sending them up. I still feel more time needs to be spent because even I had one that had obvious fusion problems but it passed. I will be putting up more as I have time.
-Jonathan

Edit; perhaps I should have included the amps, volts etc. I am not sure but I think the 3G was around 100-120 amps?? Will have to look tomorrow.

Good on you for selling 'em on the in-house bender.

What I see tells me you "may" have to adjust the WPS a bit. It may be set fine for a very- to above- average experienced welder, based on your ability, but some "tweaking" may allow you to get "passes" from average welders. The lower overall skill-level required to pass a test under the WPS (and still meet customer requirements), the lower your overall costs to produce the welds.

From what I see, these welds should not be that difficult. I'd suggest you have another welder (or two) who's experience you can trust and accurately gauge attempt the welds "blind", using just the WPS and your coupons, and bend them in-house. This will help you decide if the WPS is flawed in some way.

Steve S

Steve,
Care to give it a shot if I ship you materials and WPS?
-Jonathan

I'd be game, but I don't have pulse-MIG capability to fit the WPS. Unless your WPS has a provision for standard MIG...

That said, I'd give it a go with standard MIG.

Steve S

Otto Nobedder wrote:I'd be game, but I don't have pulse-MIG capability to fit the WPS. Unless your WPS has a provision for standard MIG...

That said, I'd give it a go with standard MIG.

Steve S

I don't have the AWS D1.1 standard in front of me to quote but globular, spray and pulse are pre-qualified processes therefore when a WPS is qualified in say pulse one can weld in globular or spray within that WPS' specifications. You could do it in short circuit transfer but since it is not a pre qualified process we would then need to send it out for a PQR/WPS. I would be willing to send you two sets anyway, would be interesting to see where you set everything.
-Jonathan

Minus the pulse, I can pull off a "globular-transfer" weld with the gasses I have access to. I can do "spray", also, but that's not a 3G process (at least within my ability...)

I'd also bet money I can pass this with short-circuit, but there's that "qualify" thingie...

Might be worth a go, just for a 3rd party opinion.

Steve S

Funny thing is, you can easily pass off a short circuit as a pulse I don't well enough, in my opinion. The two reasons we went with pulse is that is what we use the most and the cost if qualifying for short. Ok and the fact I like pulse better
Send me your address and I will send them out tomorrow.

Anyone else wanna have a go? I have around 50 extra tests
-Jonathan

Probably setting myself up, but, a PM coming your way...

Steve S

Superiorwelding wrote:
Anyone else wanna have a go? I have around 50 extra tests
-Jonathan

I'll take a couple, I don't have a mig, but I can stick them
Or were you joking?

~John

just to let you all know we are not inexperience welders between the two of us that have the failed tests there is over 40 years welding experience and as everyone knows if you have never failed a weld test you haven't taken very many at all. I suggest that before you rattle on about somebodies skill level learn something about them first.

AKweldshop wrote:

Superiorwelding wrote:
Anyone else wanna have a go? I have around 50 extra tests
-Jonathan

I'll take a couple, I don't have a mig, but I can stick them
Or were you joking?

~John

Sending some out, resend me your address.
-Jonathan

Superiorwelding wrote:

AKweldshop wrote:

Superiorwelding wrote:
Anyone else wanna have a go? I have around 50 extra tests
-Jonathan

I'll take a couple, I don't have a mig, but I can stick them
Or were you joking?

~John

Sending some out, resend me your address.
-Jonathan

Done.

In the mig certs we do,

The most common failure is in the 2f. Horizontal fillet 3 run. From lack of fusion. Too small a wire and not hitting the arc into the plate are the main reasons.

Mick

Superiorwelding wrote:Steve,
Care to give it a shot if I ship you materials and WPS?
-Jonathan

I would like to see the WPS.

hillbilly welder wrote:just to let you all know we are not inexperience welders between the two of us that have the failed tests there is over 40 years welding experience and as everyone knows if you have never failed a weld test you haven't taken very many at all. I suggest that before you rattle on about somebodies skill level learn something about them first.

I have apparently been misinterpreted, or communicated poorly, so I'll restate.

If experienced welders are having fails and "barely pass" welds (as you don't usually ask Joe Average to write the procedure), the WPS may need adjustment before it goes to a shop floor for everyone to test on, or there will be a lot of fails.

No offense was meant.

And, yeah, I've busted a few. Usually after a long expensive drive.

Steve S

Hey,

We did a test with 309 flux core in a single bevel butt joint. The nature of the wire led to an extreme failure rate in very experienced welders, upon xray. It was so high that the test was changed to a single V butt. Making it a lot easier.

Mick

I broke my first ever bend. Was done for fun.
Single bevel 2g, mig root, dual shield fill and cap.
Cap was good, root cracked. Was a little speck of something in there.
Didn't do any grinding during, just wire wheeled the flux off.

Sent from my SM-G900R4 using Tapatalk

Two more tests were sent out and both failed the exact same way as pictures before. This time my settings were duplicated, again they were not that far away anyway. I did find that I messed up the WPS, I put down a range of 90-120 amps. I ran two test plates today and my amperage was 80-90 amps for the entire run. No big deal since I passed with that amperage. The lab said this time one bend broke as to where you could still see the bevel in the plate. It is obvious that there is not enough amperage but I have a theory. I set up one plate and me am another guy ran beads together. I ran on the bottom and he on top. I immediately noticed a lack of dwell time on the sides. It was just a evenly times sweeping motion. I also noticed that in relation to the bevel, he might have overlapped the edge a very light 1/16", probably less. I was in the 1/16"-1/8" range on both sides. When I showed it to him he denied it and ran to grab another piecs he had done earlier. Handing it to me I turned it lookin down the profile and you could see clear as day that the weld was not penetrating the base material. The weld was stopped short of the end and you could see the edge if the bevel and thus how far out the weld did not go. So I think one problem is technique. As most have suggested, there I not enough dwell on the sides. We talked to the lab about switching to stringers and he laughed and said that he has not seen much success in 3/8" plate with stringers because of the void that can be created from a crowned weld. The mig weld won't penetrate into that void causing lack of fusion/cold lap.

Another theory I have. The pieces are still breaking because of extreme brittleness. When me and the other guy ran beads on the same plate I noticed that not only was his weld a smaller width and taller height but his HAZ was completely different than mine. Mine was what I would consider normal, even blue heat soaked out pretty far. His was probably 3/8" wide and deep blue/purple. So question is, is he concentrating his head so much in the bead that the heat stays right there causing the extremely brittle HAZ?

We also had a 6" sch 80 fail with the same symptoms as the plate, lack of fusion and brittle HAZ.

Guys I need help with this, any ideas??
-Jonathan

Here is my plate I sent up and passed. Don't know if it will help. In one you can see the coloring of the HAZ, the other was wire wheeled already, only pics I have though.
-Jonathan

(Post review says you've posted pics since I wrote this)

I don't suppose I've yet asked,

a) What alloy is the base metal? (nothing unusual, I assume, or you would have noted it...)

b) What alloy (and brand) Is the filler metal? (Also nothing unusual, I assume, for the same reason.)

... but asking these questions helps eliminate the obvious. If the parent metal, filler alloy, and associated sources are all familiar and proven, then the potential issues are reduced to process-related.

Steve S

You're not using an anti-spatter spray, are you?

I would think "not", but it's one more thing to check from the "obvious", list.

Steve S

What about pre heat and inter pass temp. If it is a special alloy, some cannot handle being to hot for too long.

Mick