Mysterious Change in Weld

admin

This is a welding quiz.  Have you ever had good weld penetration on one part, and for no apparent reason, inadequate penetration on the next part?Attached are photos of two GTA weld cross sections that show this actual problem.  These autogeneous GTA welds (a) and (d) were made under identical conditions, one right after the other.  The welds were made on 2" diameter, 304L bars, identically machined and cleaned prior to welding.200 amps, 0.080 arc gap, 9.5 volts, 5 ipm, argon @ 10 scfh, 3/32" 2% thoriated tungsten with 20 degree included angle and 0.030" flat.Weld (a) = 1.85 mm deep x 9.80 mm wide.Weld (d) = 3.20 mm deep x 7.42 mm wide.What caused the difference in weld shape?  Any guesses?Sorry for the photo quality, had a heck of time getting a small enough file.Last edited by pulser; 08-28-2007 at 06:09 PM.
Reply:There aren't any pictures to look at...John -  fabricator extraordinaire, car nut!-  bleeding Miller blue! http://www.weldfabzone.com
Reply:I would guess that the time you spend with the arc over it, since if you left it more over the metal , youll apply more heat, and get more penetration. That´s just my guess, even though they are a lot of factors which could affect the welding penetration.
Reply:couldn't open your photo file, but by best guess would be arc blow (or magnetism)
Reply:contaminated tungsten?
Reply:Again, sorry about the photo, I'm going to try attaching it again, this time as a pdf.Good guess on contaminated tungsten, and the arc blow, both could definitely be a possible causes, but not in this case.  And regarding the time over the weld, etc., absolutely every welding variable you might think of was very carefully held constant in an automated welding setup, for these two welds.Hint:  1.  "good weld penetration on one part, and for no apparent reason, inadequate penetration on the next part".  2.  "The welds were made on 2" diameter, 304L bars, ............."The surprising answer soon to be revealed. Attached ImagesMystery Welds.pdf (36.0 KB, 295 views)
Reply:Hint:  3.  http://www.aws.org/wj/supplement/S-1999.html
Reply:okay... saw the photos, went to the link. man, there was gobs of info in that link! Big words and math and science and everything! You say all parameters and prep were identical. Looking at the photos, (maybe it's my screen), the grain structure seems to differ. Are the bars different from each other??? Or did you make one weld down the length of the bar and one across it?   Probably the answer will be super simple and then I will have to go chew on a ground clamp because I didn't notice something basic...
Reply:Position!!!!
Reply:Ground placement....you welded in the direction of the ground...you welded away from the ground....JUST SHOOT ME!
Reply:Bluestreak,thank you for your interest, some good guesses, you're getting warm.The grounding setup and direction of welding was the same.  Its not the weld setup or weld parameters.OK  , the welds were made on two different bars!  The bars were the same type of steel, 304 stainless steel, but not the same heat (batch).  Now, what could possibly be different about the bars that would affect weld penetration like this?  This phenomena first became an issue in precision GTAW of stainless steel in the 1970's, and the mechanism responsible for it was first clearly defined in the early 1980's.  Since the 80's, there have been probably a hundred research papers written on the subject, but I suspect there are many in the welding industry that are not aware of the problem, even though they may have experienced it.  It is most noticeable in precision mechanized welding, where you are trying to get exactly the same weld every time.  I suppose in manual welding, you may experience a batch of steel that "welds differently" but you compensate by adjusting your technique.
Reply:HA! I did get it... sort of... "Looking at the photos, (maybe it's my screen), the grain structure seems to differ. Are the bars different from each other??? "Heat would have caused the grain structure to vary.
Reply:Why are these welds different depth and width?  Because the chemistry of the 304L stainless steel bars are different!    The bar with the shallow/wide weld had a low sulfur content of about 10 parts per million (ppm), while the deeper/narrower weld had a high sulfur content of about 150 ppm.How can the sulfur content affect the shape of the weld?  Through a phenomena called Marangoni flow, or thermocapillary flow.  Also described as heat-to-heat variation, it is a phenomena changing penetration as a result of the chemistry of one batch, or heat, of material versus another.  In a pure steel, or one very low in surface active elements such as sulfur, selenium, and oxygen, the surface tension of the molten metal is low at the high temperature center of the weld, and high at the low temperature toe of the weld.  This difference in surface tension creates a pattern of fluid flow that circulates the hottest metal from the weld center, out to the edges of the weld.  As arc power is increased, this flow pattern is driven stronger, and the weld gets wider and wider, and increases in depth very little.  So if you've got a part with a joint that you've been welding for years, and now you've got inadequate penetration due to a low sulfur material, you can try cranking up the amperage, but it won't do much for you!  The opposite is true of steels containing higher amounts of elements such as sulfur.  In the high sulfur steel, surface tension is highest at the hot weld center, and lowest at the cold weld toe.  This then creates an inward flow which converges at the center and drives downward in the weld, creating a deep/narrow weld, with a high depth-to-width ratio.The cutoff point in sulfur content is roughly 20 ppm and below is low sulfur, poor penetration material, and sulfur above 60 ppm is likely to be high penetration material.  The prediction of dept/width tendency is very inaccurate however, because of other surface active elements in addition to sulfur, and elements such as calcium and titanium that might tend to combine the sulfur or oxygen and render them inactive with regard to Marangoni flow.The problem of heat-to-heat variation in stainless steels began to puzzle and frustrate welders during the 1970's, when "better" cleaner steels with less impurities such as sulfur were being introduced.  A well established weld procedure that had worked fine for years, suddenly might give great difficulty in getting enough penetration.  After many years of many people pulling out their hair over this, they came to realize the problem occurred when they got a new batch of steel.Attached is a chart that relates steel surface tension as a function of temperature across the weld to the fluid flow and weld shape. Attached ImagesMarangoni.pdf (9.8 KB, 76 views)
Reply:Originally Posted by pulserWhy are these welds different depth and width?  Because the chemistry of the 304L stainless steel bars are different!    The bar with the shallow/wide weld had a low sulfur content of about 10 parts per million (ppm), while the deeper/narrower weld had a high sulfur content of about 150 ppm.How can the sulfur content affect the shape of the weld?  Through a phenomena called Marangoni flow, or thermocapillary flow.  Also described as heat-to-heat variation, it is a phenomena changing penetration as a result of the chemistry of one batch, or heat, of material versus another.  In a pure steel, or one very low in surface active elements such as sulfur, selenium, and oxygen, the surface tension of the molten metal is low at the high temperature center of the weld, and high at the low temperature toe of the weld.  This difference in surface tension creates a pattern of fluid flow that circulates the hottest metal from the weld center, out to the edges of the weld.  As arc power is increased, this flow pattern is driven stronger, and the weld gets wider and wider, and increases in depth very little.  So if you've got a part with a joint that you've been welding for years, and now you've got inadequate penetration due to a low sulfur material, you can try cranking up the amperage, but it won't do much for you!  The opposite is true of steels containing higher amounts of elements such as sulfur.  In the high sulfur steel, surface tension is highest at the hot weld center, and lowest at the cold weld toe.  This then creates an inward flow which converges at the center and drives downward in the weld, creating a deep/narrow weld, with a high depth-to-width ratio.The cutoff point in sulfur content is roughly 20 ppm and below is low sulfur, poor penetration material, and sulfur above 60 ppm is likely to be high penetration material.  The prediction of dept/width tendency is very inaccurate however, because of other surface active elements in addition to sulfur, and elements such as calcium and titanium that might tend to combine the sulfur or oxygen and render them inactive with regard to Marangoni flow.The problem of heat-to-heat variation in stainless steels began to puzzle and frustrate welders during the 1970's, when "better" cleaner steels with less impurities such as sulfur were being introduced.  A well established weld procedure that had worked fine for years, suddenly might give great difficulty in getting enough penetration.  After many years of many people pulling out their hair over this, they came to realize the problem occurred when they got a new batch of steel.Attached is a chart that relates steel surface tension as a function of temperature across the weld to the fluid flow and weld shape.
Reply:Originally Posted by pulserThis is a welding quiz.  Have you ever had good weld penetration on one part, and for no apparent reason, inadequate penetration on the next part?Attached are photos of two GTA weld cross sections that show this actual problem.  These autogeneous GTA welds (a) and (d) were made under identical conditions, one right after the other.  The welds were made on 2" diameter, 304L bars, identically machined and cleaned prior to welding.200 amps, 0.080 arc gap, 9.5 volts, 5 ipm, argon @ 10 scfh, 3/32" 2% thoriated tungsten with 20 degree included angle and 0.030" flat.Weld (a) = 1.85 mm deep x 9.80 mm wide.Weld (d) = 3.20 mm deep x 7.42 mm wide.What caused the difference in weld shape?  Any guesses?Sorry for the photo quality, had a heck of time getting a small enough file.
Reply:2006 post there is something in the drinking water LOLVinnie
Reply:pulser,Thanks for the info (nice way to present to BTW).  Keep reminding us that we really need to know what we are welding.Next question, what is the best way to compensate?  High frequency pulsing? electrode shape? all the above?Con Fuse!Miller Dynasty 350Millermatic 350P-Spoolmatic 30AMiller Multimatic 200Hypertherm PowerMax 1000G3Miller Maxstar 200DX
Reply:I didn't realize that sulfur content ( which I assumed to be tightly controlled) would vary by enough to affect the weld deposit, especially considering that a lot of the sanitary fittings are brought in from Asia. Approved procedures will obviously have to be looked over again, although I have never witnessed the efect that you have shown us. I will start paying closer attention to the MTR's.   Originally Posted by Donald BranscomWhen you say "autogeneous" do you mean with no filler?
Reply:Originally Posted by Pro-FabI didn't realize that sulfur content ( which I assumed to be tightly controlled) would vary by enough to affect the weld deposit, especially considering that a lot of the sanitary fittings are brought in from Asia. Approved procedures will obviously have to be looked over again, although I have never witnessed the efect that you have shown us. I will start paying closer attention to the MTR's.  Being  'old school', I never use the term. But all of the young welders that I hire come out of trade school calling the 'fusion' welds autogenous. Years aago when trying to get weld specs on some TIG rods from the UK, I found the term to refer to oxy/fuel welding when the filler was bare wire and no flux was used.