1/16" aluminum 3d

admin

Theoretically, if you had to make a watertight 1:1 scale basketball out of aluminum by welding together longitudinal 1/16" thick strips no wider than 2", assuming you had the ability to cut the strips very accurately with alignment marks etc, what would be the preferred weld pattern? Fit up quarters, tack the insides, then tack quarters into place, then do short beads all over the outside until done?Warping is going to be a a big deal. Im thinking getting the 3d shape stitch tacked will prevent further distortion when running longer beads, but it may not be possible to run the longer beads at all. In fact, the entire project may not be possible. But that is what makes it fun!Thanks for any thoughts. Maybe this belongs in the project forum but it seemed like a fit up question.Karl
Reply:Its going to be pretty hard to dribble an aluminum basketball! Warpage would probably have it looking more like a football.Miller xmt304,  Miller S22 p12, Miier Maxstar SD, Miller 252 w 30A, Miller super32p12, Lincoln Ranger 9, Thermal Arc 181I with spoolgun, Hypertherm 10000 ,Smith torches. Esab 161lts miniarc.
Reply:FUTBOL! GOOOOOOOOHHHHL!Has anyone actually ever tried something like this? Or is this going to turn into a discussion about tacos and pringles?
Reply:If I was making a 9" aluminum sphere I would google commercial restaurant supplies and weld two mixing bowl together. Wagner has 8" and 10" alum spheres, but not 9". There are shops that can spin a disc of aluminum into a 9" hemisphere. That's old school, trying to find one will be hard.
Reply:Thanks for the advice. The basketball is only an example. What I am making cannot be spun, because there is no plane of radial symmetry. So, back to the original question. Seems like most people would run screaming from the job.
Reply:Oh, there is at least one spinning shop around here. Thre are a few places on youtube.
Reply:I guess we're not sure exactly what you are trying to make.  There are many 3D modeling programs like solid works or formZ that will take your shape and unfold it so you can take a flat pattern to use on a sheet, fold it, tack it, weld it.
Reply:Im not asking how to model it or how to unroll it or cut it. Im asking how to weld it after it is tacked, without destroying it from warping. If possible.
Reply:Doesnt seem like this is generating any traction. I will try doing long seams on opposite sides and skipping around a bit. The manifold guys do that but they seem to use a hammer to correct for warp.  This thing will not be free to warp however, so it may break tacks. If long welds dont work I will use short welds on the next one. I also think preheating may help a bit.
Reply:atg, if I were to build in the manner described I'd make sure the weld was in proportion to the parent metal and then weld in 1-3" stitches depending on which seam I was welding.I would say that beveling the edges or the gores uniformly and evenly was critical, tacks were super critical and I'd use very small tungsten and 0.023" welding wire. I don't hand feed so I'm not saying I know how to use that size wire with the two hand TIG method.it seems to me that if the pieces are tacked up from rounds that some forming has been done since the spherical surface is not the original flat surface of the outlined gores. It that were not done to a high degree of uniformity, then I'd guess  the resulting sphere could never be any better shaped than at cold forming and tack up.If the outlines were plus or minus 1-32" down toward 1/64" fair curves, and the forming was less than a thickness out of round, then tack up could be done on a flange edge for good part of the shape. Once the form was tacked (TIG @ <1/16th dia. tacks) then a 1/16 or smaller TIG bead could be put in by planning opposing contraction forces around the center of the sphere, and with context of the seams.Not sure this is any help, since I don't use two hand TIG but that's what I'd do.  It seems like a job for the English Wheel to me?Cheers,Kevin MorinKenai, AK
Reply:Thanks Kevin that is the input I was looking for. It is all laser cut to less than 0.005" so the fit is very good once each panel is bent and twisted. As I alluded it is not a sphere and the top of each "gore" has a fair bit of twist. I just bend them by hand. I know there arw better ways to make it from an efficiency standpoint but it is an art thing and in my head there are pretty welds all over the outside. Sort of cubist. Like Wilde said, All art is quite useless.
Reply:Oh the gores are not compound surfaces. They curve on one axis and they twist but are not curved on the second axis of the original sheet plane. So it is cold form and tack.
Reply:Hydro forming?  Weld two flat round discs together then fill with water?
Reply:Oh hell why not laser sinter the thing while we're at it. I'll just pump out an stl and fire it off to the nearest stereolithography place.
Reply:If I was trying to avoid warping on a project like this, I'd solder or braze, depending on your strength needs.  I would use Oxy-Acetylene (because that is what I have and what I know), an appropriate aluminum brazing filler and an appropriate aluminum brazing flux.  As long as the brazing is strong enough for your purpose, that is the way I would go.  If you were going to use flanges (NOT butt joints), you could easily solder that.  If you are using one of the anti-corrosive alloys of aluminum (I forget the series...?50xx?, I would make sure that was an etching flux.  Either soldering or brazing would pretty much eliminate the warping issue.  I tend to get my aluminum soldering/brazing materials from Kent White/Tin Man Tech (again, because that is where I was trained and what I know).  There are, however, plenty of suppliers online for these materials.Kev.Last edited by Machine_Punk; 06-27-2014 at 03:57 PM.Kevin / Machine_Punk from The Aerodrome Studio - Lincoln PowerMIG 210 MP - Meco N Midget w/custom welding station - Vintage Victor 100Current Projects: The Aerodrome Studio
Reply:atg, I'll admit that from the basketball example to the more recent posts I'm not able to visualize the form but... if the edges can be tacked with small enough tacks so they're intact not deformed making the eventual full welds odd at the tack location? THEN, as one of the most capable welders here has explained (referring to the welder who does the jewelery-TIG welded engine manifolds on other posts don't recall his name) the starts and stops are the main focus of welding this type of project to obtain high end results.What angle, roughly is between the surfaces?  I'd assumed they were compound as the spherical gores of ball are compound so your example in OP lead me off a bit.   If the edges are from 60 to 120 degrees, outside angle, where there is an edge to edge weld, then fine wire and a small electrode should give the best results with the least distortion.The greatest distortions will happen as the two surfaces come closer to 180, and at that point the weld distortion will be the greatest, I know you know that, but my point in repeating it is that the welds where the angle between shapes is closer to 90 could be larger, hotter, bigger and the two edges will obviously work to resist contraction more that a weld approaching the in-plane butt weld.If its not too much risk to the idea; a few images of the joints may be informative to others?cheers,Kevin MorinKenai, AK
Reply:Attachment 755231
Reply:As my buddy Steve says, I dont torture the aluminum, but I do ask it some difficult questions.
Reply:atg, that makes things clear.The Clecos are not my idea of a good application because they require the welds to fill those huge gaping holes for the clamps. I'd have suggested a series of 1/8 or 3/16" arc edges to use as back up to the seams for tacking.The tacks shown illustrate the proprotion problems mentioned above. The penetration of the tacks I see are about 6x to 8x times to big. If you are trying to do this with a transformer (60Hz) I'm not sure it will happen. An inverter can adjust the HiFreq to 250 or up so the bead is held narrow, that will allow the proportion to parent metal welds required.  The arc cone of a transformer is always wide even with small or tiny tungsten, the arc cone foot print is not controllable like it is with inverters.I'd tack this with 250 Hz hi freq.  125 to 200 PPS pulse and a 0.040 or 0.060" tungsten and drop of 0.023 wire.  I'd prefer to tack on a curved backing bar, and then move to the next tack by curling that gore down along the last gore's edges. Welding this out with a set of curved backing bars cut and sanded to the back face shapes, implies a series  of three or four curves like a drafting French Curve to adjust to the shape being supported.I'd say the welds will be longer to the more curved end, and shorter to the left or less curved, where the pieces are less self supporting and more butt joints.The Cleco holes will never allow uniform welds, and I'd estimate that at least 100 hours of starts and stops in this thickness and material would be needed before you'll approach the uniform end to end bead on one of these forms.is the inside important? or just the outside? will this be sanded fair (I know you'd mentioned the welds before)? There are other joints to make this shape that may be more 'weldable'?Cheers,Kevin MorinKenai, AK
Reply:Yes its a big ask but i wanted to try something ambitious even if it failed. It is basically . A learning exercise.Initially i planned to tack using the holes to poke 1/16" rod through to provide a better tack. That may yet work.For now i am tacking the back side btw holes, using clecos for alignment. I am working toward a tack setting similar to what you describe. FYI I am able to tack using the spot setting without penetrating at all. But i worried the small tacks would not hold the warping metal.The big penetrated tacks were just an experiment that did not work. Thanks for the pulse freq suggestion; i did not realize that high freq pulse would help. If i move back to poking rod thru the holes there will be no holes to weld over.I do not know what arc edges are. Will look it up.Fitup is harder than i thought but i am getting there now. Yes it is a lot of time but I only do this for fun anyway.Hadnt planned to sand but I can do whatever I like. It is my project.
Reply:Originally Posted by Kevin Morinatg, that makes things clear.The Clecos are not my idea of a good application because they require the welds to fill those huge gaping holes for the clamps. I'd have suggested a series of 1/8 or 3/16" arc edges to use as back up to the seams for tacking.The tacks shown illustrate the proprotion problems mentioned above. The penetration of the tacks I see are about 6x to 8x times to big. If you are trying to do this with a transformer (60Hz) I'm not sure it will happen. An inverter can adjust the HiFreq to 250 or up so the bead is held narrow, that will allow the proportion to parent metal welds required.  The arc cone of a transformer is always wide even with small or tiny tungsten, the arc cone foot print is not controllable like it is with inverters.I'd tack this with 250 Hz hi freq.  125 to 200 PPS pulse and a 0.040 or 0.060" tungsten and drop of 0.023 wire.  I'd prefer to tack on a curved backing bar, and then move to the next tack by curling that gore down along the last gore's edges. Welding this out with a set of curved backing bars cut and sanded to the back face shapes, implies a series  of three or four curves like a drafting French Curve to adjust to the shape being supported.I'd say the welds will be longer to the more curved end, and shorter to the left or less curved, where the pieces are less self supporting and more butt joints.The Cleco holes will never allow uniform welds, and I'd estimate that at least 100 hours of starts and stops in this thickness and material would be needed before you'll approach the uniform end to end bead on one of these forms.is the inside important? or just the outside? will this be sanded fair (I know you'd mentioned the welds before)? There are other joints to make this shape that may be more 'weldable'?Cheers,Kevin MorinKenai, AK
Reply:Fundamentally i realize the shape and joint type are not easily weldable. Poor design but i will muddle thru and learn what i can.
Reply:I was thinking dcen with low amps and helium doing a lay wire technique might be fast enough to stay ahead of the heat flow along the seam.
Reply:But i have not tried it yet.
Reply:Originally Posted by Kevin MorinThe Clecos are not my idea of a good application because they require the welds to fill those huge gaping holes for the clamps.What is it?
Reply:Im impressed that all the people who think its a fool's errand are keeping that thought to themselves.
Reply:Thanks for these suggestions they are clearly wiser than my plan. I dont know of alu braze that matches color but i will do some reading.Still on the weld program for now though.Karl Originally Posted by Machine_PunkIf I was trying to avoid warping on a project like this, I'd solder or braze, depending on your strength needs.  I would use Oxy-Acetylene (because that is what I have and what I know), an appropriate aluminum brazing filler and an appropriate aluminum brazing flux.  As long as the brazing is strong enough for your purpose, that is the way I would go.  If you were going to use flanges (NOT butt joints), you could easily solder that.  If you are using one of the anti-corrosive alloys of aluminum (I forget the series...?50xx?, I would make sure that was an etching flux.  Either soldering or brazing would pretty much eliminate the warping issue.  I tend to get my aluminum soldering/brazing materials from Kent White/Tin Man Tech (again, because that is where I was trained and what I know).  There are, however, plenty of suppliers online for these materials.Kev.
Reply:Ok well getting pretty good looking welds with lay wire and 1/8" rod is not terribly hard They look cold on top but are fully penetrated. Hardest part is staying on the seam while zooming along; it is easy to get off track. They are smooth instead of beaded.I hope i can use a regular technique with smaller wire but it is harder to stay ahead of the heat. Backing seems like good advice.Good tacks continue to elude me. I am getting smaller wire to dab into the tack.Last edited by atg; 07-02-2014 at 01:42 PM.