Not logged inAmerican Welding Society Forum
I have a few questions. I have been over ridden in a rejection that and I would like to hear your input. Below is the link to the sample.
So the samples are supposed to be .025 vertical and .050 for the bottom of the plate on this T-joint fillet weld. Now the top .025 plate isnt the same thickness all the way to the root. I have listed the thickness along the the side of two different samples.
I had a few discrepancies here. If I was to just judge it based off the 10% lack of root fusion both samples are rejected. I was over ridden based on the idea that since there are different thickness toward the end of the sheet that I was to measure .025 at the thinnest part and add the rest of that "thickness" that isn't really there. When I add the thickness that isn't physically really there then it "consumes" the the rejected area and I was told that that Is how I have to look at it. I don't feel this is right. It went from being rejection for LOF in 6 of 8 samples to now be passable. The top photo is how i was told to "judge" a macro sample. I just cant seem to jump on board with that idea.
As for judging concavity. There needs to be an even right triangle to measure the largest size the weld can be with excessive concavity. What is hard to grasp is that when you measure the 10% and you have a concave surface the code only asks you measure to the leg in the direction of the LOF. Even though the weld itself might not even meet the weld size in due to that profile. I was hoping someone can clear this up for me. Not everything is a straight line when I comes to this. Especially if the thickness varies and that calls are made based of 1.5T for minimum weld size as well.
to see what D17.1 says Figure 5.12 B, as well as Figure 7.1 for profile
Hey, it's not like your building rocket ships.
Yea pretty much. Hard to see that and not have an issue. Especially when 6 of 8 looks like that.
I'm gonna stay away from interpretation on this one for now... You have already heard my rant.
But the welds look to me (I know it's hard with a closeup macro) that the welds themselves may be considerably oversized.
What is the exact requirement for weld size (leg length) for a flat or convex fillet, and what are you actually seeing?
Are the operators using a .040 tungsten?
What is the included angle and tip prep of the electrode?
Are they using a heat sink or backing piece on the thin member?
If this is a performance qualification test, do you have a fillet testing fixture?
My bet is that some stricter process controls on the actual GTAW operation would benefit quality and improve your pass rates.
Just my opinion from far on the outside.
There isn't a maximum to evaluate it only a minimum of 1.5T. Only if it called on the print. And yea this is a qualification for a welder. Concave or convex is based off the typical AWS diagram photos, a clean 45 degree weld.
I'm gona ask a really, really dumb question
...what is that loose little object next to the joint in the further down pictures?
I took a snip of it but can't figure out how to put it in a post.
- Figured out how to attach it... (171k)
I am going to say that this is "trapped slag" and likely resulted from incorrect technique/manipulation of the rod/gun (providing this is an SMAW/FCAW process). Al Moore is a pretty strong proponent of discussions and examples of how these occur. In short, if this was pushed/forehand progression welded this is a likely result. Have also seen these rather commonly with E7024 electrodes with incorrect amperage application and not enough drag angle. Best regards, Allan
No bad questions on this forum unless it's to do your honest for your class haha. That is a form of what I would call over lap or cold lap but internally. When the bridge widens in this case the weld itself won't actually fuse but freeze and move throughout that gap at the root. Note it has no fusion line near it. Happens a lot with aluminum fillets that don't focus on the root seal.
Okay folks, I "finally" read the body of this post, WOW, sorry for interjecting my $.02 here. Way off on the description of what the picture shows now that I have a bit more information of what it is. If I have a better handle on the process covered by this photo and go with GTAW on aluminum, that's certainly an interesting item at the root. Even if this is GMAW aluminum, it's still an interesting little gidget. I would be more interested to hear of it's composition compared to the parent material. Especially since it doesn't appear to share any fusion lines to the surround parts. Best regards, Allan
Thanks the more minds the better. It is a 718 Inconel. I agree its quite an odd look. We get sent some things that just baffle me. IT was welded GTAW, and when macro'd its to be cut every other .50" and we use every other sample inside the edge drops. It goes to PT first then to To get slices. Ive been just trying to get why I would be over ridden for it my view of that actually being LOF beyond 10% which we calculated on a microscope.
Did you see how the code says on Figure 5.12B S1 & S2 measured leg size in vertical and horizontal member. To me that sounds like measure the weld that allows it to be measure in the profile in Figure 7.1 THEN run your 10% distance which would very much fail this route.
As for running the straight line... To me its the "thinnest member" at the bottom "T" with that number being the .018" which wasn't fused or welded on.
Sorry if my typing is bad I just got out of the hospital. Had a headache from a severe sinus infection, it took away most of my motor skills two days ago. I couldn't even talk with out gibberish, and couldn't write. Today and last night I had 3 CT scans and today an MRI and they said my brain was not affected by anything. That my brain swelled so much that it began to push against my front lobe and making me weak. Almost the same way a stroke is. They confirmed it wasn't one. It was pressure but still a scary day. They have me on meds for the next week. They said it will take a few days to my sharpness back once all the swelling out.
Hello J, I am going to play the ignorant one here again. I have been led to believe that percentages of inspection related callouts are generally applied longitudinally and not in a cross sectional sense and are either encountered on a weld face/weld root visually, by way of UT imaging, or by way of X-ray.
Since the code that you are referring to is one that I have almost zero in-depth familiarity with nor do I have a copy of it, are you saying that it specifies discontinuity size based on cross-sectional percentages? So in other words you believe that you first have to develop a scale of the image based on dimensions, ie. the thickness of the sheet, and then make comparisons between discontinuities as they relate to material thickness? I am "assuming" that once you have determined an accurate scaling procedure you are then "measuring" the various components of the sample and coming up with determinations of "passes or fails" based on percentages, is this where you are at here?
Sorry if I am not helping with your questions, I am trying to educate myself at your expense just a bit, and likely others who have come across this topic. I'll be watching/reading to see where you end up. Thanks again for tolerating me and best regards, Allan
I attached a few sheets of the code to show you what I mean. There is a very tight internal inspection clause that has a Class1 through 3. Anything under .063" fillets have to be qualified through visual, PT and macro. Check the pages and ask more questions haha.http://imgur.com/a/5gbge
I think the stuff you and Allan are seeing in that lack of fusion hole (which is large enough to park a Volkswagen) is prolly just GAK from the wet saw and or polish compound from the macro processing.
Holes are like flat surfaces in a bachelor pad... Something will go in it or on it, given enough time.
And if you can't find "GAK" in A3.0 that's their fault, not mine.
Hello Lawrence, as you have just framed your response, I could certainly see something "after-the-fact" that could certainly match what you just said. Hadn't thought of it in that way but it certainly makes sense and likely the most sense, thank you. Best regards, Allan
We checked if that was metal from the cutting but its actually it is in the under the weld. We use a ultra fine wet saw made specifically for exotic metals. Polishing would remove anything that hung back, as well as we check them for that same reason. I think if you view it in the ways of dabbing the filler while welding, it looks to me as if the puddle was wet a bit more ahead and when the next puddle dropped it froze before it bond with the previous droplet. It is Inconel and Inco and quite dirty even when being welded, the scale almost floats at the surface.
D17.1 for aerospace fusion welding. It is based on the old military welding standards, MIL-STD-2219, and MIL-STD-1595.
I don't have a copy of D17.1 sitting in front of me, so cut me some slack if I'm not on the button with my observations. First, why is the test plate not a uniform thickness? Why make this process more difficult that it has to be?
I do recollect the aerospace welding standard does allow some incomplete fusion in the root of the fillet weld, but that applies to production welds. I don't recollect if it applies to welder qualification samples. I will have to find my copy of D17.1 to verify what the requirements are.
Larry made a good point of saying the process has to be closely controlled. The use of appropriate fixtures is helpful. The use of a backing gas is a help. The samples have to be a uniform thickness and the edges have to be straight, true, and no root opening when qualifying on thin sheet metal. Cleaning is important because there is little margin for error.
Luck, there is nothing better than a little luck in the welder's favor.
Best regards - Al
We are a testing lab so they send quals and we do what we do. Welder quals yes are judged for fillets based on this 10%. So welds are pretty perfect and some are bad. I think it's a bad difference I thickness in one sheet. Just can't see a logic that passes this. Is it something you would consider knowing the LOF is greater than the limit?
How did they get the vertical memo to thin out like that? Too much cleaning with power tools?
It very well could be I dont know. It was the most extreme I have seen.
So I looked at the pix a good bit and I can't buy the strategy indicated by the red line. I don't think you can draw lines that allow for invisible material so your welds pass. I can easily draw a line representing the original thickness on the opposite side from the weld, and show a slightly skewed tee joint that would fail any possible interpretation. I'd invite whoever is overriding you on this to put their name at the bottom of the report.
Yea thats my plan. Seeing as I also cant see that at all.
Thanks for your input.
Powered by mwForum 2.29.2 © 1999-2013 Markus Wichitill