Juicy aluminum pipe project--some advice solicited

admin

Hello all,I am working on a project welding some 4" aluminum extruded pipe back together.  This stuff was cut for shipping purposes and needs to be re-welded for its final use.  I have to weld 8 sections back together to form 4 20' lengths.  They are cut into 10' pieces.I beveled one, fitted and welded it out using some plates and an angle iron clamped to it.  Looked true and a straight edge confirmed it.However after welding it out it has an ever so slight bow to it.  The finished product must be true.  Can anyone suggest how I might accomplish this?  I have welded tons of iron pipe with no issue as tolerances like this was not an issue.  This aluminum pipe will be turned by a motor and with the current bow it will turn with an oval pattern.Any ideas?  It occurred to me to weld it in opposite segments at a time.  Top, bottom, 9 o'clock, 3 o'clock, etc.  The pipe looks to be 5/16" walls.The pictures show the first one I welded out with the crown ground down.  That's how the finished job will look like.  If distortion is a fact of life, how can I straighten it?Thanks,Tony Attached Images
Reply:therrera, first that is not pipe.Pipe is a uniform round extrusion and you don't have pipe.  I'm not sure what it is? but it's not pipe.Second while I don't have a solid end all reply that deals with the weld, it is clear the extrusion with all the legs is not going to be joined 100%. There is no way to weld the interior legs fully without puddle sag, that you can't dress in a 10' length.Third, you may get some penetration of the square section recess (?) but you will have to re-machine or broach those extruded tracks.Fourth, you can probably set up a bending frame to deflect the resulting 'as welded' lengths using a standard three point frame.this image of a TIG butt weld in 3" sched 40, 6061-T6 shows a weld closer to the surface of the pipe wall; there is no 'crown' so the toe does not 'pull'. I suggest that you could draw an aluminum (pipe butt) weld by using a 'cap weld' instead of using just the second pass full to the top of the V bevel. The toe of the cap, as you show in your post, could be drawing the extrusion unevenly?Last, the two opposing channels in the 'pipe-like' extrusion will lead to different amounts of contraction and potentially an uneven final 'lay' of the butt welded extrusion.I don't know if this is any help?  If I were welding this extrusion; I'd make the top and bottom channels' bevel larger/deeper/wider and make those two segments  the first welds- just the opposing channels.  Then I'd do the remainder in quadrants and (re) bend the result if needed.  I'd keep the weld top surface flush to the extrusion.cheers,Kevin MorinKenai, AK
Reply:the first thing i would have done, if the money was there, is buy a stick of 3 x 3 x 1/4 / 3/8ths angle.i would have put it on top of at least five horses and trued it up.if the od, is four and those stiffeners are a quarter then the id would be about three.  you could pick up a few feet of 2 1/2 in sch forty al pipe or a tube with 3 in od.  then you take a skill saw and run the blade down the pipe long ways.in the case of the pipe it will pop open.  in the case of the tube it may provide you with enough removed material to squeeze it into the opening.   either way i would probably want a sleeve in there.then when i welded it out i would not let the weld come above the surface of the extrusion so there is no raising it while rolling it and welding in the angle iron.maybe that's overkill but you're the one with the customer.or go the big bucks and shop here.http://www.vebru.eu/Weld-Tech%20folder.pdfthe stiffeners are in there to prevent sag so i am assuming this is for rolling up something.  in arizona it could be a pool cover.  make sure everything is aligned;  in the photo of the channel one side looks offset to the other.i just don't understand why they didn't get 20 footers in the first place.Last edited by fdcmiami; 06-04-2012 at 09:19 PM.
Reply:Originally Posted by fdcmiamii just don't understand why they didn't get 20 footers in the first place.
Reply:Thanks,I appreciate the tips.  I will keep the weld flush on my next sets.  We stopped in order to get some ideas before continuing.  We will start again tomorrow at 8am.  I put some 1/2" "keys" in the slots shown so they aligned the extruded pipe evenly.  Then I used a 1/4"  x 2" angle iron (about 3 feet of it) to clamp down on it and finally I put in an alignment plate on the top to get one side perfectly aligned.  We checked with a straight edge after I tacked it up and it was true.  After tacking was done I tapped the "keys" out of the way by sliding them completely along the slot and out.What would a straightening jig look like and how would I apply pressure to the side that needs bending back?  I have a 10' steel table six horses, a concrete driveway and hammers, come along, hi-lift jack, clamps and so on.Thanks,Tony
Reply:I agree with what Kevin wrote. With what I can see you are working with, I would maybe use a couple of sticks of channel to use as a saddle and clamp them one on top and bottom to the pieces. For getting the ones welded straight? I would cut them apart and start over if you try straighting the pieces you may end up with two bends one at the joint and a new one where it bent to compinsate for the first?? if you dont use a jig. good luck
Reply:therrera,if you tacked up clean/fair/true (?) then increase your tacks to 8; quadrants first and then divide them. Make the tacks a full hot dime/ and back out with an even count. This should give you an even tackup that will not pull.  I'm assuming the butt fit/root gap is uniform?  To make a press frame you'd take and rigid frame/table or floor, and use the jack/hydraulics/screw in the middle and the outer points to hold- common base is floor/table/frame stronger than extruded 4".   Press on the high side away from the pressure to straighten.  Since this #1 aluminum # 2 heated past original temper #3 not a purely symmetrical extrusion; you'd do the press a little and measure the stroke. Repeat, go slow, do less-not more, take time to avoid cutting out and reheating HAZ in 6061 alloys.  HAZ yield of 6061-T^ (assumption of alloy?) goes down hill each time you get back in the same area with more heat.Use a saddle of ply or pipe segment to make the press foot, not the end of a jack.Cheers,Kevin MorinLast edited by Kevin Morin; 06-05-2012 at 12:26 AM.Reason: typos
Reply:Thanks for the feedback on this project.  I will incorporate these ideas tomorrow (later today).  I will make sure I level the pieces being welded and then true them up using angle iron braces and positioning tabs.I will put 8 tacks evenly spaced out, weld in opposite segments not allowing a build up beyond the pipe's surface.  I will make sure the gap is even all around.  In the first weld the gap was uneven as one side was almost touching (1/16" gap) and the rest had 1/8" gap.  Apparently the cuts made by the factory or distributor were not true.As far as a jig for straightening:  I will build a three point deal using a fixed point in the center and one end of my steel table.  The third point will have room for a porta power ram to use for straightening the deal.  I may have to use it to do them all if needed.  We'll see....By the way, my client said that he had to travel to China to talk with the factory reps about this extrusion to make sure his idea is going to work.  He said he has $6,000 invested in these pipes including aver $1500 or $2000 in shipping cost from China to here.Thanks again,Tony
Reply:therrera, that's why they were shipped in short lengths? the length or handling issues were judged easier in 10' over 20' lengths?pretty complex extrusion try to 'save' on international transport.Go to the trouble of squaring the ends, likely with a sander, before anything.  Its not rare to find extrusions line saws are 'off' and the ends not true 90's. That is where your original dog leg/kink/bend point most likely came from in the one length already welded.When you dress the weld off, first, use only very small diameter pads, sanding or grinding to keep the worked area narrow. There are 1-1/2  or 2" pads available for all the grits and the 3M Scotchbrite (tm) buffing pads too.Second try to keep the sanding tracks just on top of the weld bead until it barely touches/brushes onto both sides of the extrusion. This way the final dressed area will be less noticeable and more even looking.]Finally, if not a trade secret, what will the shape be used to do?Cheers,Kevin Morin
Reply:they must have gone over land on the back of a yak and then loaded onto a canoe for the trip across the pacific.  don't know what's up with that story. lol
Reply:Tony-I would expect, despite best efforts, that some bowing can be measured, especiallywhen making 2 checks for straightness, each one 90˚ from the other.-extrusions by their very nature are not necessarily dead straight, nor perfectly round-If that 20' assembly is only supported at either end, it can have some inherent sag-has the customer defined the amount of measured runout acceptablewith an agreed upon inspection method for both you and him to use to check?It's going to be something--.010, .125, ,250These are issues normally resolved before the work begins, unless thisis an R&D or 'Best Effort' repair.Blackbird
Reply:Hi everyone,The advice give out was right on the money.  I made sure even gap all around, tacked in 16 places, welded in small increments alternating opposite sides and welded the joint flush with the surface.The next one still pulled a great deal less than the first one, but still pulled.  While it was still hot we put the center over the horse and applied pressure on each end to bend down the in the direction to counter the bend, it worked.  Apparently while still hot, the metal was pliable enough to let itself be bent relatively easy.  In fact, we bent at first too much and was able to correct it by immediately turning the pipe over and bending again, but lighter.The third pipe was the same, a slight pull  but easily straightened. The last one came out perfect, with no distortion and required no straightening.  The only difference was that I welded it out in one inch increments alternating opposite sides so it took about 12 or so turns to get it all done.  It was the best of the lot and now I think we have the procedure down for how to do these should it come up again.  We cut the first pipe and welded it back using the above procedure.  It came out well, needing a slight straightening as described.  We used clamps and strong backs of channel and angle iron at 90 degrees from each other to hold true in both directions.  My welds were rough compared to the sample posted here earlier which was real nice, but the client was happy with them.My client knew that this was a trial and error situation and in fact he was told by the factory reps that it was impossible to weld them once cut.  He worked with during the whole project as a helper and giving his advice or observations along the way.  He was in the body repair business and has a good eye for metal work.For example he took a dollar bill and folded it until it was the thickness of the gap under the straight edge before we straightened the pipe.  After we would bend it he used it as an improvised "feeler gauge" to tell us if we were having the desired effect.  He said they used to do that when straightening and using bondo on bodies to tell when a surface was true. So his help was vital to getting this done and to his liking.I appreciate all your advice and this was a good learning experience for me as I have very limited experience welding aluminum pipe or tubing.Until next time,Tony Attached ImagesLast edited by therrera; 06-05-2012 at 08:29 PM.Reason: to add more detail.
Reply:the factory reps probably would argue that since the stiffeners had been cut and separated that the mechanical properties of the extrusion have been compromised. if your customer was happy then that's all the counts but personally i would have inserted a sleeve within the two tube sections and my root pass would have connected the sleeve and each tube.did they mention what it was for?either way, good luck.
Reply:Hi again,I think  you're right about the sleeve and we talked about that but my client decided that since the purpose for these tubes is to act as a roller for a scrolling sign, they won't be subject to much weight nor stress.The design of the extrusion matches a motor's drive gear that sleeves into the end of the pipe and will turn it. It will be hung vertically and mounted on bearings top and bottom on a frame being built by my client.Apparently this system is part of a niche market for applications like this.  That's all I know.Thanks,Tony
Reply:If I have something welded like that that's been distorted from the heat of welding by just a little bit, the distortion can often be "counter-acted" by re-heating (e.g., melting and running a puddle) along whatever portion of it that needs to be contracted to pull it back into alignment.  You can repeat, running the puddle across it more than once.However - with aluminum when doing this, if it's a crack sensitive alloy, there could be a concern of causing hot-cracking if the puddle is not diluted with additional filler rod.  If that's the case, you can dilute (add) additional filler, and then grind off when done.  Net result can still be "strategic" shrinkage.It sounds like you got a solution figured out already though (by bending), which is great, I just wanted to share the "re-melt a puddle to shrink" trick.
Reply:Tony, it is great news that you could take all the information, make an improved work plan and use the www.weldingweb.com to help your day-to-day work.If that is not the purpose of this site and other specialty sites I have no idea why they're here.Congrats, glad you won and the client was right there working along side you; that can be a real challenge (!) but it sounds like you and he were able to get the job done, and he can go on with his scrolling (like auto-furling sails on a boat?) bill boards and you have a new solution for your local market.Thanks for posting not only did your post(s) allow others to see what you're doing, they could offer their few cents, and you could learn (because you had the attitude of listening to others) and most of us have never met each other! this ain't your Dad's shop floor- well it's not MY Dad's shop floor.Cheers,Kevin MorinKenai, AK
Reply:Originally Posted by therrera...the purpose for these tubes is to act as a roller for a scrolling sign, they won't be subject to much weight nor stress.
Reply:Hello again,thanks for all the tips and tricks.  The device my client is building is in essence a prototype.  There are smaller scrolling signs but he is building a relatively large one.  Now that I think of it, these are probably going to be hung horizontally and not vertically as that would put the height of the device at over 20 feet and not very moveable as it would be sticking up where it could encounter power lines, etc.  I imagine that there will be issue that he will have to resolve besides the joining of these halves.  Such as the sag that will be inherent when hung horizontally and so on.On the first one that was warped too much, I took a torch to it to heat it up and make the aluminum soft.  Once heated we applied pressure on the ends with the pipe centered over a horse.  We bent it too much!!!  Then we flipped it over and bent it too much that way as well!! We finally straightened it pretty good but could not close enough for our taste.At that point we decided to simply cut the weld and join the unused ends together using the new game plan (short beads, skip welding on opposite sides with each bead, no weld crown, etc.).  So the idea of heating the joint with the TIG torch would work.  However the issue of how much heat is safe to put to aluminum comes to mind and I think there must be a thin line where structurally the material may be weakened and compromised.I am not versed well on the properties and issues of aluminum even though I have been welding it for 30 years now on and off.  I first learned back in 1982 when I got a hold of a craftsman high frequency unit and put in on my old lincoln buzz box 225 amp AC welder and rigged up an air cooled TIG torch to it.  No remote amp control or even an on and off switch.  Move close to the metal and the arc starts, bingo! But I practiced and practiced at every opportunity and began offering aluminum welding services along side the steel repair and fabrication.  Since it was such a low powered setup, all I could offer was aluminum fuel tank repairs on big trucks and light aluminum welding.I free-lanced for two years back then when a recession put a lot of welders out of work.  I was one of them but what separated me from my laid off peers was that when I was working I started gathering the tools needed to operate on my own if and when such a situation developed.  The initiative to learn aluminum welding was driven by the need to make work.  I learned the difference between being out of a job and having work.  They are two different things. You don't need a job to have work.  Although a job is one form of work.I also learned to have a great deal of respect for all those who operate a small business after having walked in those shoes.  When I punched a clock I could go home at the end of the day and forget about it. For the small business owner his (or her) second shift was just beginning,  taking care of all the little things that make the business be able to keep its doors open that as an employee I never saw, nor worried about.I just knew that the power was there when I turned the welding machine on, there was rod and gas to weld or cut with, the vehicle(s) ran and had gas in them, there was material to work with, etc. etc. etc.It was a good education, one that has stayed with me for all time.Thanks again,TonyLast edited by therrera; 06-09-2012 at 03:34 AM.Reason: to add clarity
Reply:Originally Posted by therrera...since the purpose for these tubes is to act as a roller for a scrolling sign, they won't be subject to much weight nor stress.The design of the extrusion matches a motor's drive gear that sleeves into the end of the pipe and will turn it...