WPS for 1/2" fillet

Date 12-18-2006 17:11

You neeed to supply more information.

What is the application and what is the jurisdiction?
What Code are you welding to?
Are you following an approved WPQR?
What are your present parameters?
What position are you welding in?
What electrode/flux combination are you using?

By kard Date 12-18-2006 18:05

Welding to D1.1 Code. Yes I am following an approved PQR. My present parameters are 3/32 electrode, 450-50 amps, 35-37 volts and 15-21 ipm travel speed. Welding in the flat position. LA-75 with 860 or 960 flux. Not sure about what you mean by application or jurisdiction?

By HgTX

Date 12-18-2006 20:28

What are you planning to change in order to get to 2 passes?

Hg

By kard Date 12-18-2006 20:31

I am planning on increasing the amps

By Joseph P. Kane (****)

Date 12-19-2006 12:56

By "Jurisdiction", I meant State, City, County, WABO, a State DOT, Clark County, NYC Dept. of Buildings, Region ETC. By "Application" I meant is it for a building, or a tank or a ship, Exposed Weathering Steel Application, Etc. In short, do you need a WPQR, or are you "Prequalified". Are you stuck with a heat input increase limitation per D 1.1,Table 4.5 (17)? I am also confused as to how you get 860 and 960 flux qualified in the same WPQR. Are you using LA75 electrode? Are Charpy Impact Values specified?

Any armchair analysis of your question requires much more information than you gave initially.

So, the following is general observation and advice based on insufficient information;

If you are getting exactly a 1/2 inch Fillet weld size with three passes, raising the amperage by 10% will be the limit of change without making a new WPQR Test. A 45-50 Amp increase will not get you to 1/2 inch two pass weld on that basis alone. If you have to stick within the WPQR limits, and you are subject to the heat input increase limitation restriction, you will not be able to do what you propose. If you decrease the travel speed and increase the amperage, the closer you will get.

If your 3 pass nominal 1/2" Fillet Weld size is actually larger than a true 1/2" fillet weld, it may be possible to get the true 1/2" Fillet Weld deposited in two passes with a 10% increase in amperage.

Caution; A big increase in the deposited weld volume will change the deposited weld mechanical properties. While I have not done "Scientifically Controlled" tests, I have learned through testing that changes in heat input do not result in linear changes in the mechanical properties when using Submerged Arc Welding. So, be aware that you may affect the strength and elongation properties of the weld, and (If this is done in a groove weld), the Charpy Impact Values.

Why are you stuck with the WPQR limits? SAW is a prequalified processs under D1.1, as long as you comply with all the provisions of Section 3. If you are eligible to perform "Prequalified" welds, make a mock up, develop the parameters you need to make the weld in two passes and write a new WPS.

Sorry I can't offer you more. The only other experience I have is with multiple electrodes and lower voltage and travel speeds, and all for DOT, Military Combat Vehicles and Valve Body applications.

Joe Kane

By js55

Date 12-19-2006 19:40

I do not know that there is data to suggest that strength and elongation, as measured by codified tensile and bend tests, are effected by heat input. Certainly grain size, etc. can be effected, but any change in these mechanical properties will probably be well within a normal scatter band. I believe this is the justification for ASME considering heat input as a Supplementary Essential Variable only, and therefore applicable only to Impact Testing regimes.

By Joseph P. Kane (****)

Date 12-20-2006 12:24 Edited 12-20-2006 12:28

AS I said, I never did "Scientifically Controlled" tests. IN the early 90s, I did a series of SAW tests with different heat input levels and there was a noticable difference in the results. The most noticable thing was that the Charpy Impact values were not in the linear "Scatter Band". BY this I mean that in one series of tests, starting at 40 KJ/in and going to 72 KJ/in in fiver steps, the 50 KJ/in test failed Charpys with numbers like 13, 17, 22, 19, 16. The other four tests had numbers like 28, 56, 64, 108, 24. Multiple Charpy tests were done from each test coupon for the different AASHTO Zones. Elongation was not linear either. Tensiles were all within the limits, but there was no linear relationship in the % reduction in area / Elongation. I know the lab had properly calibrated equipment.

In another series for Flux Cored Electrode, I found that I was likely to fail to get 22% elongation when I welded with less than 26 KJ/in. There the Tensiles went up as the number of passes increased, but the tensile to Yield ratio was not linear. I had one side bend break and fail at approximately 70 degrees with no visible crack initiator and no visible microvoids at 5 X magnification,while the rest of the mechanicals including the other side bend test passed. Note all the bend specimens were polished with 120 grit.

Having said all this, I have been told by PHD welding metallurgists that the groove weld test results are not truly valid for Fillet Welds, partly because of the nature of the weld deposit shape and the cooling patterns. I have also been told that Charpy results are not really the best indicator of impact strength and cold weather performance. I am told that the Codes should be specifying CTOD tests! (Funy thing though, those people are/were on the AWS D 1 Committee and never managed to get that change to CTOD testing incorporated into the Bridge Code!) [All my tests were done I/A/W Bridge Codes.] I am not a trained metallurgist nor am I a graduate welding engineer. I did keep all the weld parameters and preheat/interpass strictly per the WPS. This was the limit of my "Scientific Control". The steel used for these tests was ASTM A588 Gr 50, all from the same 40 foot long, 1 inch thick plate.

I am not claiming to be able to write a scientific paper on this subject, but there was empirical experience.

By js55

Date 12-20-2006 20:47

Interesting results Joe. I stand corrected. Perhaps I should rephrase, I don't know that there are consistent results demonstrating codified tensiles and bends being effected 'deleteriously' below minimum qualification specifications by heat input variations. However, we always learn. My thought is that chemistry develops some 'bottom line' as far a mechanicals are concerned, that is without skill discontiuities (slag, fusion, cold lap, porosity) having an effect.
24 to 108 is quite a range. Was that all WM results or mixed WM and HAZ? Charpy's will certainly be effected with heat input. I'm curious why 22% elongation as a standard of concern when WPS bend geometery is oriented to 20%. 26 kj/in seems pretty cold for FCAW. And I truly suspect there was a problem with the bend that broke on you not associated with grain or micro/macrostructure. Perhaps not apparent at 5X mag. You are right about CTOD's and Charpy's. CTOD's, it is argued demonstrate 'real' material properties, or at least the best approximation available thus far, whereas Charpy's don't. Charpy's demonstrate only a general, yet consistent, result, for lack of a better way to put it. But CTOD's are a hard sell since they are time consuming and expensive. Can you imagine the costs associated with a file of, say, a 100 or more WPS's. And where are the failures justifying such an extreme code revision? Just because we can do something doesn't necessarily mean we should. Also, even though CTOD's are clearly more accurate (hence of favorite of metallurgical PHD's everywhere) I'm not sure they are necessary under 'minimum safety thinking' for code bodies, or given the great variance usable as WPS parameters, especially under ASME where the 10%/7% criteria isn't used. Perhaps the code bodies are demonstrating some minimalist wisdom in resisting the change, as would be the case with segregating fillets from grooves in WPS quals. I don't know, but thanks for the info.

By Joseph P. Kane (****)

Date 12-22-2006 02:55

OOOH! I really like the term "Minimalist wisdom"! However, I always attributed some of the incongruities of the codes to the general feeling that "we shouldn't change what we know works" Kinda like calling "old Farts" "Senior Flatulents". Much more "Diplomatic"! Very nice!

26 KJ/in WPSs are qualified because they want to deposit 3/16" Fillet Welds at high speeds. In the case of D 1.5 Weld Performance Qualification, they try to get the greatest parameter range qualified when they are performing Min-Max heat input WPQRs. In one test that I witnessed, they were actually trying to qualify at 22 KJ/in. They got a tensile that was in the mid high 90s and a yield that was in the mid low 90s, and a 16% Elongation! This was an E 71T1 electrode welded with straight Co2. They did not even achieve penetration to the root on both sides of the weld root. Classic "Wagon Tracks" showed on the radiograph. Oh Shit! - Oh Dear! as Mike Houle used to say!

In the case of D 1.5 and the NYSSCM, (Bridge Codes), the 505s are All Weld Metal. The Deposited WM chemistry comes from the broken end of one half the 505 specimen.

As to the Microvoids, I know that all tensile failures are the result of "Microvoid Coalessence". I have often seen this exhibited as rather bright "Bubble Holes", which are not really gas pockets after all. I do not know why this particular bend specimen failed, especially since the next specimen, taken less than 1 1/2 In. away, bent with no problem.

Who knows why! I didn't have the money to persue the mystery. On another test, I had a 1" plate coupon fail X-Ray, an indication showed an ill defined shadow toward one side of the joint for the full length of the plate approximately in the middle of the through thickenss. At First I thought it was a lamination, but how did it travel into the WM?!?! I confirmed it was there with Shear wave UT. Straight beam UT showed it to be 5/8 inch down from the top of the weld face in the WM and the BM. I sectioned it, and examined it at 10X, 50X, and 100X, and couldn't see it. I tried visible dye and UV PT, and couldn't see it. I tried yoke type MT and couldn't see it. I bent several of those specimens and nothing showed up! I "crease" bent it in the "Y" direction, and it didn't show up. (I forgot to try a hardness traverse.) A reduced section Tension Specimen broke on the other side of the joint and in the base metal. There was no deformation indicating a hard spot where we know the indication was located.

Your observation about the 3/4 inch ID radius of a 3/8 inch side bend specimen producing 20% elongation is right on. For some steels, the ASTM Standards specify less elongation for the parent metal, but the AWMT Specimen has to meet the higher elongation shown in Table 4.2 of AWS D 1.5, or Table 706.1 of the NYSSCM!

I do not know why. I don't want to belittle the code writers, because there is often good rationale for things at one time, even though the history of WHY is lost when successive generations of Code Committee Membership loose the older members with the historical knowledge. They are volunteers after all.

You and I ran up a lot of "war story", on this thread, Iwonder if we helped Mr. KARD?

By CWI555

Date 12-22-2006 03:31 Edited 12-22-2006 03:41

""Who knows why! I didn't have the money to persue the mystery. On another test, I had a 1" plate coupon fail X-Ray, an indication showed an ill defined shadow toward one side of the joint for the full length of the plate approximately in the middle of the through thickenss. At First I thought it was a lamination, but how did it travel into the WM?!?! I confirmed it was there with Shear wave UT. Straight beam UT showed it to be 5/8 inch down from the top of the weld face in the WM and the BM. I sectioned it, and examined it at 10X, 50X, and 100X, and couldn't see it. I tried visible dye and UV PT, and couldn't see it. I tried yoke type MT and couldn't see it. I bent several of those specimens and nothing showed up! I "crease" bent it in the "Y" direction, and it didn't show up. (I forgot to try a hardness traverse.) A reduced section Tension Specimen broke on the other side of the joint and in the base metal. There was no deformation indicating a hard spot where we know the indication was located. ""

If you seen it with UT, but did not see it physically or with any other NDE method, it may have been a segregation. I had a case of this not long back. The base material was A572 from Ukraine. I cannot speak for certain what your indication was, but the segregation noted here displayed dramatic shear wave results, and minimal response on longitudinal. The difference in behavior between the waveforms Longitudinal vs Transverse particle motion implied a similar condition as would be found in dissimilar welds (low nickel to high nickel for instance) being an area of segregation where the segregation line contained some larger grain structure. The plate was 41mm with the indication at 21mm.
The weld next to this area was problematic, with the apparent extension of the indication into the weld metal on an intermittent basis. The base metal I answered with some reasonable confidence, as I had UT and fluorescent MT (very faint but visible intermittent line) data to substantiate the indication. I also sectioned the base metal in this area found intermittent enlarged grain structure in the area of interest. What I did not understand and still do not understand is the apparent display in the WM. Sectioning and looking at the grain structure of the weld metal was inconclusive.

Any of you guru's out there that can explain this to me it would be appreciated.

By Joseph P. Kane (****)

Date 12-23-2006 01:31

I know this doesn't help Mr. Kard.

I am glad to find out that someone else out there has had a similar experience! I have had people speculate that alloy segredation was the culprit, Phase change was the culprit, a lamination that sealed itself due to eutectic chemical changes and dilution into the Weld metal dilution during the melting of the weld Puddle, IT wasn't there, and "you changed the stick-out during that pass"!

The UT was also performed by a third party, with a different machine than mine, so I believe that something caused a reflection. The shadow on the radiograph was approximattely 1/4 inch wide at one end and up to 1/2 inch wide in the middle of the plate, down to 3/8 inch wide at the opposite end. Multiple oblique shots were used to guess at the depth of the indication.

By CWI555

Date 12-24-2006 03:33

I can understand the phase change, and that may explain the weld metal. However; the segregation ran a signifigant distance into the base metal.
I'll have to do some further study on the samples with the theory you presented. Thanks for the lead.

By Joseph P. Kane (****)

Date 12-24-2006 22:06

In this test, the BM plate was 1" to 1 3/16" thick ASTM A588 Gr.50, Domestic Origin, and it was specially ordered to meet the chemistry requirements per AWS D1.5, Para. 5.4.2. (In the D 1.5 2002 edition. I don't remember what paragraph or what edition it was back in 1992) I rejected the plate for production, because it did not meet the ASTM thickness variation limits. Still, I didn't see anything wrong with using it for testing. We couldn't return it because we had it sent to a third party for cutting into WPQR test coupons 9" wide & 48" long. (I believe in making the tests longer than the code requires, so I can still do the mechanicals if part of the plate fails radiography, and I can perform Charpy Impact Tests for three different zones and also for FCM.)   The electrode was 1/16" ALLOY Rods Dual Shield II 71, from a "Bonded Lot" and CO2 Gas with Certified Dew Point.

As to the suggestion that I had a longer stickout in the pass that had the anomalous indication, I am sure that was not the case. I measured the ESO prior to the start of each pass with a gage. It was possible that the ESO got a little longer or shorter due to slight variations in the height of the previously deposited WM pass, but this welding was done using a travel Kat.

I do not know how far the anomally ran into the WM. I estimate 1/8 inch, based on the known joint geometry prior to welding. (Remember, I never saw it visually, and the width measurement estimate was based on Db drop during Straight Beam UT testing while the plate was still whole. This method of measurement is good for accept/reject measurements, but not good enough for forensic analysis.) The average visible depth of the HAZ was 3/16 inch, so there did not appear to be a lot of metal dilution compared to SAW welding.

I those days, I didn't have access to an STM or SEM, and did not know the capabilities of the STM. I did perform a lot of hardness testing with a PTC portable Hardness Tester, but that is too gross a test to be called a "Hardness Traverse".   I do not recall any hard spots with those tests.

I remember that someone else suggested that there might have been Hydrogen Embrittlement due to an un-noticed BM discontinuity. The reason It didn't cause a crack, was because successive passes tempered that pass. To me Hydrogen Embrittlement should not have been found in the undiluted WM zone.

The "Eutectic dilution formation" theory sounded plausible. It also explained the presence of the indication into the WM.   The rogue alloying elements in a lamination could have lowered the melting point of the BM on the top and bottom of the lamination surfaces enough to cause them to fuse together, and some of those rogue alloying elements diffused into the WM. Subsequent passes caused further diffusion and the visible indication would blend into the BM and the WM. (Much Like Diffusion Brazing) I have seen a 1 1/2" Dia. ball bearing rolled into the middle of an 8" armor plate, so it is not inconceivable that something could have been rolled into that plate. It had a thickness variation defect after all, and I never tested the other side of that cut edge on the neighboring plate. (I didn't think of it at the time.)

I find it significant that your experience was found in the middle of the through thickness! Mine was just below the middle. I have searched for some of the coupons that I saved from that plate, but cannot find them. I would like to have sent you my samples, but they would have only been 3" long and 3/8" thick. You couldn't even do a straight beam UT test per D 1.1 on a 3/8 inch wide coupon specimen.

By CWI555

Date 12-25-2006 22:21

( comments not meant to question Joe, they are meant for the those who haven't a clue what a STM, SEM, or segregation vs lamination is.)
Although I have used one, a scanning tunnel microscope would have been overkill, in my experience, as a euphemism, you would not be able to see the forrest for the tree. An electron scanning microscope was used (Micrion 9000 focused ion beam system at variable 100x to 250kx) which was overkill in itself. However it did make for some very informative imaging.

You have to be careful with the terminology, it goes in order of increasing severity, 1)segregation, 2)lamination, 3)de-lamination.

I've looked up my old reports and data, and Im leaning towards the dilution theory as I simply did not put two and two together at the time.(every now and then you have to kick yourself for not seeing the obvious) It would explain the WM coming up with indications. If the weld metallurgy became unbalanced due to dilution of the rogue elements into the WM, all sorts of strange things will ensue. That also explains my hv's being several points off in the haz/fusion interface area (2kg load) as compared to the known good weld samples. I don't think a portable hardness tester will get you to where you need to be to check for this problem, the indenter would be to big and end up giving you only a gross reading.

As for the sample size, they might not meet the spec for D1.1 but for UT, data could still be garnered from them for informational purposes using the right transducers. I would say based on finally re educating myself on how to add 2 + 2 you would probably find more info with micro hardness HV2 test and looking at it from 250x.

By Joseph P. Kane (****)

Date 12-26-2006 00:23

No offense taken from anyone. I'm only an amateur failure analyst. I probably did mix up terms here when I wrote of it. Sorry-I knew what I meant!!!!

Even If I found out the metallurgical etiology involved, I would still be at a loss to explain why it showed up on X-Ray. If it was a Delamination, with a less dense material in it, it would be at least 1% of the through thickness to show up like it did. At 100 power, it should have filled at least 25% of my field of view.

Bottom Line- I need a lot more education. Remember, none of the specimens from this coupon failed during the mechanical testing!

As for SEM / STM, My thought was to locate the area with SEM, then cut out the 3mm discs for the STM centered on the top boundary, in the middle and on the lower boundry, then try to analyze it. WHY, because it was available for free! Then again, I am a relative ignoramus as to what I would be looking at, but I have really knowlegable friends! I personally only have a 100x optical microscope. I could not see any decent microstructure variations at that power.

I already said that my hardness tests with the portable PTC tester was too gross for a proper hardness traverse. (However, I did have luck finding a hard spot on another fitness for service evaluation for TACOM, which was then cause to send the coupon to the very much higher priced experts at Lucius Pitkin Inc.! )
Again, it was amateur methodology, driven by lack of proper laboratory equipment and knowledge, but it potentially would have saved my client $9000.00. In the long run, TACOM accepted my recommendation despite the really top notch work by LPI. There were no known failures, and the contractor got paid.

As a consultant, I try to save my client money while finding a solution or compromise for the problem. Quite often, after I am unable to resolve the problem, I wind up referring the problem up to my PHD friends in my local ASM International section. In that type of situation, I often walk away without pay, unless I have follow-up work for the problem resolution.

I will continue to look for one of those 3/8 inch thick specimens for you.

By CWI555

Date 12-26-2006 03:50

X ray results only come with a density change, either more or less material by volume, or a change in density of material. If you can see it with x ray, and it passes mechanical, it's likely the later. To my mind, the problem with something like that is you have an unknown. Even if it passes mechanical on one day, there is no guarantee that the rogue elements won't have a problem in the future.

*Reference alloy 600 problems in the nuclear generation world, in that instance, the metallurgy was known, but un-accounted for environmental/thermal factors created a problem over time. There is no reasonable expectation that if you add an unknown variable (even one that passes mechanical) that the weldment or casting etc. will withstand the test of time, design stresses, environmental factors and so forth. To me it's an unacceptable risk. All the statistical modeling by the brightest minds, cannot account for an unknown like that. thats my personnel feelings on the matter, but whatever a client wishes to do is their business, however; I don't let it go un-reported either.

As for gear, not many people can afford the price tag on either sem/stm. I went looking today out of curiosity and found an sem like the one I used for 65K on the used market. That's a considerable amount of change for looking at something tiny. *warning* rant begin (and for those of you out there that grouse about the expense of that kind of work, that's one of the many costly tools required to do the job properly, tooling alone is not cheap, much less the education to have a clue what the results of the exams actually mean)*rant end.

Any inspector is continually educated if they are a good one. I've got 22 years in, and I consider myself to have just gotten my feet wet.

I think we've gotten way off the beaten path of the thread, so I'll let this be the last post on the matter. Hope I've helped in some way.

By js55

Date 01-04-2007 16:41

This one's gettin printed and filed. Thanks.

By johnnyh (***)

Date 03-22-2010 18:07

bump for JS55's first post