Aluminum Welding Issue

admin

I may be putting this in the wrong area, but here goes anyway.I am building a solar car chassis at school for my senior design project. We are using aluminum for the frame members. It is 1/8" wall, 1" diam. The joints welded well, or so they appeared. We did some destructive testing and found that the welds are failing pretty easily. We did a destructive test of 4130. This piece behaved as predicted. Weld held, material failed in the HAZ.If anybody has any input on the aluminum issue, I'd love to hear it. It looks like the base metals are melting, they just aren't joining. The base material also appears to not be mixing well with the filler. Material is supposedly 6061, filler rod is 4043. Welder is a Lincoln Precision Tig 275. Using straight argon. The aluminum weld shown was done at 100 amps, slow travel speed. Joint preparation consisted of cleaning the surfaces with a 3m clean and strip wheel and beveling the joint edges around 45 degrees. The bevel was taken all the way down to the i.d. of the tubes.Last edited by scottguehne; 01-25-2011 at 08:59 PM.
Reply:I would make the car frame with the 4130. To protect the driver!What filler did you select for the 4130?Did you slightly heat the chromoly first to get rid of the moisture? (250ºF) The aluminum weld did not fully penetrate the joint. It is clearly shown in the photo.The weld bead is laying right on top of the surface.One or two person vehicle?Depends on the lengths involved. Can we see it?It looks like the aluminum weld itself failed, but you did not state the details of the welding.The cleaning of the aluminum must be done with a stainless tooth brush.Did you use a tranformer welding machine or inverter machine.Was it set on AC?Was it pure tungsten with a ball formed on the tungsten?Was crater fill on?Did you keep the sheilding gas on and over the weld until the postflow shut off?How may seconds was the post flow? You said the base metal was "supposedly" 6061, was it? Was the aluminum anodized???Why did the aluminum weld break and the tubing did not even bend?We need a photo of that test.Last edited by Donald Branscom; 01-25-2011 at 09:45 PM.AWS certified welding inspectorAWS certified welder
Reply:It looks as though you are not getting 100% penetration on the aluminum weld.  Since you are beveling, maybe you're getting suckback?  Full penetration would likely help you quite a bit since the lack of penetration acts as a stress riser instigating the crack formation.Second, the outside of the tube is prepped, but what about the inside and the cut edge?  What is your preheat?
Reply:The penetration is 100%. The photos may not show it the best, but the two pieces being joined together show weld material on the inside of the tube. If you look you can see the bubbling on the i.d. The weld appears to have no penetration because the filler and base metals don't seem to be joining properly. That is why you see the two very defined layers there. Preheat- technically none. After tacking the pieces together, the material is too hot to handle, I figured that is enough of a preheat. I say the material is supposedly 6061 because we did a hardness test on it, and the hardness corresponds to a 5052 I believe. We did the hardness test to try to determine the actual alloy. We thought due to the poor weld quality, we may have gotten a different grade of aluminum. The i.d. of the tubes have no prep. Not sure how you could even get in there. Fegenbush, you mention suckback. That is a new term to me. If you could give an explanation of that, I'd like to read it.I'll try to take a minute to answer some of the posted questions:"The weld bead is laying right on top of the surface."This wasn't a question, but I know that. That's why I've come to this forum, trying to find an answer. "One or two person vehicle?"Single person, not sure what that matters."Depends on the lengths involved. Can we see it?""It looks like the aluminum weld itself failed, but you did not state the details of the welding."Not too sure what details you would be after. Tig welded. "The cleaning of the aluminum must be done with a stainless tooth brush."I fail to see what the stainless brush offers that a cleaning wheel doesn't. I am willing to try some different things in the search for success."Did you use a tranformer welding machine or inverter machine."Transformer."Was it set on AC?"Yes."Was it pure tungsten with a ball formed on the tungsten?"Thoriated. Balled."Was crater fill on?"N/A on the particular machine. Amperage was 'peetered off' with an additional dab of filler to try to eliminate any craters."Did you keep the sheilding gas on and over the weld until the postflow shut off?"Yes."How may seconds was the post flow?"Around 7 seconds. "You said the base metal was "supposedly" 6061, was it?"Yet to be determined for certain. "Was the aluminum anodized???"No."Why did the aluminum weld break and the tubing did not even bend?"If I knew I wouldn't have posted here."We need a photo of that test."N/A. Brief description: Piece was cooled and placed in a vise, with the weld just above the jaws of the vise. A larger tube was placed over the smaller aluminum tube and force was applied until the weld failed.Last edited by scottguehne; 01-25-2011 at 10:19 PM.
Reply:Looks like you could use a little more filler in addition to any other problems you're having.My name's not Jim....
Reply:Originally Posted by scottguehne"It looks like the aluminum weld itself failed, but you did not state the details of the welding."Not too sure what details you would be after. Tig welded. "The cleaning of the aluminum must be done with a stainless tooth brush."I fail to see what the stainless brush offers that a cleaning wheel doesn't. I am willing to try some different things in the search for success.
Reply:No need to bevel to the id, halfway will do just fine. On my machine, I would set the machine to 150 and control my heat with the pedal. Using a stainless steel wire brush is recommended for aluminum because carbon steel ones deposit carbon particles on the part being cleaned. I would use 2% thoriated 1/16 inch tungsten. My frequency would be somewhere in the 60-80Hrz.This guidance should assist you in accomplishing your goal. Do not be afraid to change the setting to improve your results. Just make sure you keep track on paper so that you don't make multiple attempts at the same settings.Bill LambertArgon WeldingABQ NMSic gorgiamus allos subjectatos nunc
Reply:Just out of curiosity, what University do you go to?
Reply:Not too sure what details you would be after. Tig welded.
Reply:from what i am seeing, a couple things that you might wanna try. first off as said before, mechanical cleaning with a grinder and abrasives is a no-no for aluminum. your heat isnt hot enough and you are basically "soldering" the pieces together instead of welding them... make for sure that the puddle is melting the filler and not the arc from the tungsten. when dipping the filler and you pull back, try hard to keep the end of the filler inside the shield gas blanket, aluminum reacts harshly to the environment when molten, and if you pull the molten tip of the filler out of the shield, you are exposing it to nasty, only to dip that nasty back into the puddle, therefore introducing contaminants into the weldalso, as mentioned before, there are several alloys of aluminum that are non-weldable, and there isnt really a sure fire way to tell what alloy you have without some metallurgical sciences and testing going on (you need more than a hardness test to figure that out) for example 7075 alloy will NOT weld. it will melt and it will look like it welds fine, but will actually become so brittle that a child can break the welds loose. as this is a transportation type project, great care must be taken to ensure that you are using proper alloys, proper methods just for safety sake alone...also, will you be post heat treating the welds? if not, then you should be using 5356 filler. 5356 will not post heat treat, but it is a stronger weld when not post heat treating...hope that i have helpedmigweld 250parcmate 205weldall 250piforcecut 80i plasma'07 pro 300 miller'08 trailblazer 302mm350pdynasty 700dynasty 200deltaweld 452xr-a 50ft push pull feeders and gooseneck
Reply:Scott,You answered your own question in your first post."Aluminum was done at 100A and slow travel speed"Not enough heat (amps).Aluminum likes to be welded hot and fast.You got enough heat to melt the coped tubing but not enough to penetrate the solid wall of the tubing you were joining to .  Arc should be directed more at the solid tube rather than the coped edge.Slow travel speed generally also means overheating the filler, making it brittle.  That's why the filler looks dull and grey.With the 4130 you overheated the base metal.  Too much bead.  Fitup here is critical.  Preheat will help.  Faster, more consistent travel speed will help also.Syncro 250 DX Dynasty 200 DXMM 251 w/30A SG XMT 304 w/714 Feeder & Optima PulserHH187Dialarc 250 AC/DCHypertherm PM 1250Smith, Harris, Victor O/ASmith and Thermco Gas MixersAccess to a full fab shop with CNC Plasma, Water Jet, etc.
Reply:Agree with all the above.Your welds look dirty/contaminated. Clean as described aboveAfter you tack it, clean it again. Need at least 125 amps. Remember, roughly 1 amp per 0.001"I honestly wouldn't bevel much for 1/8". Leaves more base metal to tie into. But that is a personal preference Try again and please post stress test pictures!
Reply:When you (the OP) mentioned that the penetration was 100% you made the mistake of confusing burn through and penetration.  It is easy to melt through a thin piece of aluminum, much more difficult to get the 2 pieces to fuse together on the back side of the weld.  Consider an Argon or Ar/He back purge to prevent oxides from forming and helping the backside wet together.You might have burnt through the piece, but your penetration was hardly half the material thickness which despite what you may think is very evident from the pictures.I personally wouldnt want to sit in a car made of 1" aluminum tubing, to many issues with cracking from repeated stress and vibration.  Nonferrous materials fatigue very poorly.Have we all gone mad?
Reply:Originally Posted by sn0border88When you (the OP) mentioned that the penetration was 100% you made the mistake of confusing burn through and penetration.  It is easy to melt through a thin piece of aluminum, much more difficult to get the 2 pieces to fuse together on the back side of the weld.  Consider an Argon or Ar/He back purge to prevent oxides from forming and helping the backside wet together.You might have burnt through the piece, but your penetration was hardly half the material thickness which despite what you may think is very evident from the pictures.I personally wouldnt want to sit in a car made of 1" aluminum tubing, to many issues with cracking from repeated stress and vibration.  Nonferrous materials fatigue very poorly.
Reply:Originally Posted by sn0border88You might have burnt through the piece, but your penetration was hardly half the material thickness which despite what you may think is very evident from the pictures.
Reply:I now understand the difference in terminology here. I can now agree with the lack of penetration. What I don't understand is how the base metal can be melting, but not joining properly with the filler. You can all see that the filler is just sitting on top of the tubes. I welded those tubes at settings recommended by miller's website. I did not tell you guys how many tests were done prior to the test shown in the photo. One of the tests did involve no beveling, and the machine was set to 200 amps. I gave the pedal full throttle on that weld. I had to weld extremely fast, which is how a couple of you have said aluminum likes to be welded, "hot and fast." The same thing happened. The base materials would melt, with the filler essentially sitting on top. So, with that said, the burn through was there, but not penetration. To be honest, I'm still a little fuzzy on the true meaning of penetration. How can one have burn through, but not penetration? Burn through shows that the base metal is melting, so how can I have burn through all the way to the i.d. of the tube, but not penetration to the i.d. of the tube? On to the cleaning subject. As stated, I have been using a 3m clean n strip wheel to clean the aluminum. The wheel has been used for nothing but the cleaning of the aluminum. If the wheel left any residue on the material, I would think it would show in the weld, which it never did. It has welded cleanly. In the pictures that I have posted, the welding had been done days prior to posting, and the pieces had not been kept clean in any manner. That is why they look so dirty. They were welded clean. Any impurities on the surface of the aluminum got there well after the welding took place. With that said, I am getting a stainless wire brush and probably some stainless steel wool to clean the material with. I have also gotten some aluminum cleaning solvent from the local welding shop. How would you all recommend cleaning the oxides from the i.d.'s of the tubes? There is no good way to do it. The i.d. is just big enough to get a finger in there, much less any type of abrasive anything.We did around 7 or 8 tests on similar joints, and they all failed at the weld. The welds all appeared the same on the inside. There was burn through, but no apparent tying together of the two tubes. I do not put filler right in front of the arc. I place the filler at the leading edge of the puddle and let it wick into the puddle. Would leaving a gap between the two pieces help anything? From what I have gathered, it is best not to bevel the edges. And can anybody answer why the filler metal just sits on top of the base metals? As I stated before, I had the machine set to 200 amps, and I used all of them. Same result. I will get more test pieces and document each piece, from cutting of the piece all the way to its demise. I will document each step, maybe even attempt to do a video. That may not be until Saturday though. I feel like some more helpful folks are chiming in here, and I hope it continues. I do think the argument of penetration vs. burn through may help me out the most. More explanation of this would be greatly appreciated.
Reply:The flap wheel will not leave it clean enough for welding. Does not matter if it is brand new. Use the stainless "toothbrush".Also, a stainless wire cup on a grinder or wire wheel on a pedestal grinder will not do it. Elbow grease and a stainless brush is your friend. Unless I missed it, have you confirmed the alloy you are using yet? I think the cleaning is your main problem. Get clean, get full penetration, and then start looking at metallurgy if it's still jacked up. Knowing what you have is 50% of the battle. Cleaning is about 40 of the remaining 50%
Reply:Your right in thinking that the material is clean enough, for those who keep on suggesting he needs to clean his material properly then you fail to see what is really going on here.  It might not be the 100% correct way to do it, but is sufficient for what the OP is trying to accomplish.The part of the puzzle your missing is the cleaning action provided by the welding arc.  In AC current the EP portion of the sine wave bombards the part with ions which help to remove the oxides that readily form on the surface of aluminum. These oxides have a melting temperature of 3600F compared to the aluminum which is somewhere between 900-1200F.  If you can imagine it would be like trying to weld water balloons together, underneath the material is molten and fluid but on the surface there is a skin that prevents the 2 pieces from truly flowing together. So to circumvent this problem you have 2 options.  Either running a backing strip of some kind (not going to work with this diameter tubing) or to use a backpurge as the shielding gas will help protect the molten metal and aide in wetting action, allowing the 2 pieces to flow together.Because of the difficulty of getting a consistent weld and the fact that in many joints you are not going to be able to get a good back purge I would seriously recommend that you look to steel for the chassis design.Have we all gone mad?
Reply:Scott,Get yourself some flat coupons of similar thickness to your tube wall.Do butt welds til you get your heat/travel speed down to where you're getting good penetration.  Much easier to see on this type weld.Then move on to t joints.  Here you'll see that more heat (direction of the arc) is directed at the horz. piece than the vert. piece.  Form your puddle on the horz. and wick the puddle up to the vertical piece as you add filler.  Practice these welds til they become second nature.THEN, move on to the tubing.You're trying to master too many things at one time, and not doing any of them well.Also, and as has been suggested, if you're not going to heat treat the frame (which is highly unlikely) you're better off switching to 5356 filler.Similar welds to what you're trying to do are done every day in the marine fab business.  I'm thinking weight is a major factor in your "design" and that this "vehicle" will be for design purposes only and not driven on the road.The welds you're looking for can be plenty strong, but not until they're done properly.Syncro 250 DX Dynasty 200 DXMM 251 w/30A SG XMT 304 w/714 Feeder & Optima PulserHH187Dialarc 250 AC/DCHypertherm PM 1250Smith, Harris, Victor O/ASmith and Thermco Gas MixersAccess to a full fab shop with CNC Plasma, Water Jet, etc.
Reply:My 2 cents.Contamination/excess surface oxide from the preparation technique is a major factor in the poor weld properties.You need to know what the tube material is for sure.  If it is 6061, and you are obviously not going to heat treat the frame, then 5356 will give better weld strength than 4043.  The 6061 weld must be diluted with filler, if it is welded without filler, or with too little filler, it is very crack sensitive.  If the tube is not 6061, maybe it is a non-weldable grade of aluminum.
Reply:First Scott sez:"Joint preparation consisted of cleaning the surfaces with a 3m clean and strip wheeland beveling the joint edges around 45 degrees.The bevel was taken all the way down to the i.d. of the tubes.But then later Scott sez:From what I have gathered, it is best not to bevel the edges"For butt joints--you'll need beveling.You state no prep on tube ID. Does that mean no wet wiping on the ID?   (acetone/lacquer thinner)and apparently no wet wiping on the OD prior to oxide cleaning? X2 on you really needing to focus on basics-prior to tubesBlackbird
Reply:If  you are considering chrome moly steel tubing, a great source is www.aircraftspruce.com..  If you shop aircraftspruce'sr tubing charts carefully you will often find one or two ID/Thickness combinations there are very cost effective.... (generally in the $2.50 / foot range...)  Hint: Most bicycle tubing is 0.035" and .049" thick. connect tubing with TIG welds or fillet brazing... and for some tube notching hints, see http://www.cobratorch.net/ttn_dxf/ orhttp://www.cycle-frames.com/bicycle-...e-Notcher.htmlGood luck on the project..zip.Last edited by zipzit; 01-27-2011 at 11:09 PM.
Reply:just thought of this today while building a sissy bar for the biker dood down the street.... you arent diluting the parent metals enough... 6061 must be diluted with filler, otherwise all its gonna do is crack out... so.... you arent adding enough filler to the puddle. again, make sure that you are letting the puddle melt the filler and not the arc. doing tig on aluminum tube takes a bit of practice, you really should get some exp. from flats before moving on to round stuff.migweld 250parcmate 205weldall 250piforcecut 80i plasma'07 pro 300 miller'08 trailblazer 302mm350pdynasty 700dynasty 200deltaweld 452xr-a 50ft push pull feeders and gooseneck