More Contaminated (bumps) in Aluminum

admin

There were two posts (one tig & one mig) concerning bumps and appearance of the bead.This will takes three posts, but I played around a bit after working with an intern at the neighbors shop some time back and I made a horrible looking fillet (that the intern thought looked pretty good).The first pic is the weld at the neighbors done with a PT225, and in my opinion it is much smoother with the AC than my Square Wave 275. The macro pic is not that good but looking through a 20X & 30X "BIG EYE" the bumps are 125-250um in size, are metallic (aluminum) and shaped like a dough ball dropped on top of the puddle.The second is a pic of 3M roloc pads that I tried on the aluminum, they took the coating off the surface with one swipe but working further smeared the material. So it was one swipe and brush for cleaning.The third is 4043 fillets using the SW 275 in my shop, the bumps are the same shape but are in the 32-64um range.  These were cleaned OK but not anal.The fourth is a 5356 fillet with no visible bumps (not sure why the dimes look square, maybe camera angle?). Cleaning the same as above.more... Attached Images
Reply:The first pic in this post is an outside corner with 5356, not working for full pen really just trying to keep a wet puddle and working slow.The second is a back bend showing the 5356 filler is pretty robust without the full penetration.The third shows a very cold 5356 butt weld and a cold fillet weld.The forth pic... I just had to do it for zap!More... Attached ImagesLast edited by Matt_Maguire; 01-19-2011 at 02:20 PM.
Reply:The first pic shows a practice coupon I took from the neighbors shop. The backside had not been welded so I thought it would be good for a test. I did not grind down to the other beads, I just cleaned and brushed the surface reasoning that any crap between the coupons would be good for testing the difference with 4043 and 5356.The second pic shows a 5356 butt weld with a very small un-cleaned gap.The last picture shows 4043 on  the same coupon with different balance settings. I don't recall seeing any of the black specs you see welding over junk but they don't look very good none the less,  the few balls on the top weren't there till the balance got to 10.I'm reasoning the bumps in the fillets are coming from the filler rod (which was not cleaned at all), as they looked very similar in size and shape to "mig dust". I also didn't think to try a 4043 fillet at 9:00 to see if it was just dust blown off the rod and falling back down. I've seen others talk about abrading the rod for special work but in the day to day work this wouldn't be reasonable. I'm also thinking about some of the other work I've seen from a fire truck builder, a trailer builder and a boat builder that you look at and just go wow. None of these places are going to let someone get silly playing artist (well maybe the fire truck folks can).So to go back, I'm thinking as the oxides are blown off the filler, they take some AL up with them on the softer rod, and it's just dust dropping back down. I just can't reason that the silicon coming out solution would get together in that large of blob and maintain shape on top of the puddle near the arc. What very little aluminum I do "for real" is larger chunks where something is built up, so and an area is masked and the part is cleaned with Aluma-Brite and rinsed before heating to 300-400 and welded. I don't recall anything as ugly as the first fillet in the first post...Anyway the 5356 doesn't seen to have any bump problems at all???Opinions, thanks in advance?MattPS...  I 'spect Snowboarder is gonna post that outside AL corner (that's OK as long as he doesn't call me William)Done! Attached ImagesLast edited by Matt_Maguire; 01-19-2011 at 02:20 PM.
Reply:"Bumps" or sandy look is from too much heat. Give enough amperage to start puddle in 1 or 2 seconds and move fast will give shiny bead- no sand. Pretty al welds are done done with speed, going slow is leading cause of dull weld.(Heat)Rol loc disc will leave plastic behind and can cause problems, I have better luck with wire brush in die grinder. Looks like your getting it, trial and error, all it takes is practice.PeterEquipment:2  old paws2  eyes (that don't look so good)1  bad back
Reply:Matt - you've presented pretty clear evidence here that the "bumps" are associated with the 4043 filler rod, much more than (if at all) with the 5356 filler rod.I think there is a metallurgical explanation as follows.  In a nutshell, the "texture" or bumps are large crystals of silicon & aluminum.  The bumps are neither oxidation or "contamination".4043 has about 5% Silicon as a primary alloying ingredient to the Aluminum.  As such, it is a "hypoeutectic" alloy (silicon above 0% but below ~12%)  From its material specs, it also has a solidus and liquidus temperature of 1065-1170 deg F:http://www.alcotec.com/us/en/support...d/a4043tds.pdfOne characteristic of a hypoeutectic alloy, is it has a wide range between "solidus" and "liquidus", and indeed 4043 has a ~105 deg F range.  When a mixture is above liquidus temperature, it is homogenous.  But when it is in between liquidus and solidus, crystals can form within the melt (making a "slurry" of sorts.) http://en.wikipedia.org/wiki/Eutectic_systemhttp://en.wikipedia.org/wiki/LiquidusThe longer a substance is held in between liquidus and solidus temperatures, the larger the crystals can grow.  This jibes with castweld's assertion that welding "hotter and faster" reduces the surface bumpiness of a 4043 weld (as the rapid cooling of the weld wouldn't allow as much time for the Silicon+Aluminum crystals or "grains" to grow very large.)  That could be an interesting experiment to try and see how it affects surface texture.Another way to reduce the size of the Aluminum+Silicon crystals should be to switch to a filler rod with an Aluminum+Silicon composition closer to eutectic mixture, and a narrower range between solidus and liquidus.  4047, at 12% Silicon content, having a solidus and liquidus of 1070-1080F, (so only a 10 deg F difference) fits the bill.  It would be interesting to compare the surface texture of 4043 to 4047 welded with similar parameters aside from filler rod selection.  I'll bet that 10 deg F range between liquidus and solidus of the 4047 filler gives the Al+Si crystals little time to grow large when TIG welding, and you get a much smoother texture.Of course 5356 is another way to go to eliminate the Al+Si crystals, because it has no Silicon.  Now granted it does have other alloying ingredients (such as Magnesium) and a similar wide solidus-liquidus range (1060-1175) as 4043, but the crystals formed by the other ingredients must have a different physical structure than the Al+Si crystals.http://www.alcotec.com/us/en/support...d/a5356tds.pdf
Reply:Thanks Peter & jakeru,I should have updated this a while back.I had a chance to watch the neighbor welding a bell housing with soft AL rod where he had cleaned the crack and the area around the repair very well. I watched from behind him, shielding the torch with my hand and payed attention to the cloud above and around the torch. He was not cleaning the rod, just grabbing it and shoving it in.When the arc was started to form a puddle, the air around the torch had little if any smoke or dust in it. When filler was being added, a definite plume of white smoke appeared with sparklers and dust coming up. So i would have to say cleaning the rod would help with this but I'm not certain.I have to defer to Castweld that better settings and technique will also knock this down because I did a reasonable job of cleaning the coupons (but not the rod). I had no real appearance problems with butt welds or outside corners using the soft wire. When doing overlay build up with warm parts and 1/8" rod I can run hard enough to make "almost arrowheads" instead of dimes, and having filler over fresh beads everything looks fine. I have never built up any inside corners so I can't say how it would turn out.To the point of 4043 being a hypoeutectic, I think this binary is going to produce an acicular or globular eutectic structure at room temp. For the silicon islands to move and get together needs time. For them to form large islands may take a good amount of time. Aluminum is cast with up to 25% silicon added for strength, and the silicon islands at the surface are too small to see, but they are uniform "hard" islands that increase wear properties and make Al-Si castings very "abrasive" to cutting tools.Considering the above, I think re-melt and repeated re-melts are going to ruin ductility and would be very prone to cracking, but still won't form a bunch of "warts" & "blobs" on the surface. If anything being liquid too long may produce a "lamellar" type internal structure which could be a very weak weld.The more I think about it the more I'm convinced it's cleaning and technique. I think the 5356 is just more sluggish and I've not looked at anybody else using it to see if the cloud around the arc is full of junk. And if it is full of stuff maybe the puddle being less "wet" doesn't allow the junk to stick?Matt
Reply:so.. far  a real good read guys     . tanksidealarc 250/250 ac-dc tigidealarc 250/250 ac-dc tig #2 used for sticklincoln sp100hh125dual arbor grinder polisher30 yrs of hand tools52 pitch blocks 6p-26prake gauge -pitch gaugeG&D prop repair 918-207-6938Hulbert,okla 74441