Help with Aluminum Fuel Tank

admin

Attachment 493561Attachment 493541Attachment 493511Attachment 493521Attachment 493531Attachment 493491Attachment 493501Hey Guys,I am looking for help with this aluminum fuel tank I am building. I am having (had) difficulties welding in the (2) internal baffles. The material was acting like it was not clean/ had oxide layer on it. It would take forever to get a puddle started, if it even would start and as I added filler it didnt lay down nicely like it normally does with my practice pieces. The weld puddle would build up from underneath instead of stacking like beads. This led to a gap forming between the outer shell and internal baffles, which as you can see from the outside of the tank has created some ugly looking protrusions/deformations under the weld areas. The 2 last pictures show the gaps. (there were more)I originally cleaned the areas as follows: wipe down w/ acetone, angle grinder w/ scotchbrite type pad, then acetone again. Once I noticed that the puddle was not forming and not flowing I took 220 grit sandpaper and sanded the parts by hand and cleaned again with acetone, but this did not help at all. The material is practically brand new, its not like it was sitting outside for years, so it shouldn't require unorthodox amounts of cleaning, right?All my practice pieces are welding fine. The material I specified is 5052 H32 and it is .090" thick, which is what my practice material is. You can see what my machine settings were on the practice piece, I dont understand why its not welding similarly on the tank. The fact that I am welding inside the tank with limited room doesnt help either, but I dont think that is really the issue.I am debating if I should scrap this and start over (the protrustions are highly stresses points that I am not comfortable with) but I dont know what I would do differently, other than paying someone a few hundred bucks to do it for me.Thanks for any help. Attached Images
Reply:One thing that will help with the over penetration is to clamp the piece down to the table or put a heavy backer on the underside of the baffles. Welding doesn't look bad.
Reply:What you are describing sounds fairly similar to what I frequently see when newer tig students get to alum T joints.Issues I see in those picts. No prep. The reason it's acting like it's covered in alum oxide is because it IS covered in alum oxide. That forms naturally when bare alum is exposed to air. Sanding is bad. Most sand papers are made from alum oxide, the same stuff you want to get rid of. You end up embedding it in the surface. You want to use a hand stainless brush to remove the oxide layer. A power brush just smears the oxide.My guess on the puddle issue is you aren't hitting it hard enough, fast enough. With T joints newer students often have issues getting the puddle started, and started where they want it. Many times they'll melt back the upper piece and not get the puddle to flow to the lower one. Trying to go to slow because they are "scared" of melting back the edge, just makes it worse. In many cases I suggest they stick the filler in the corner and light off on the filler and get the puddle to bridge and GO. The trick here is to simply use the filler to help "connect the dots" as it were right at the start, and then add filler to the puddle. You need to be sure you get the puddle to form on all three, then work the filler normally, not melt filler with the arc and drip it on cold plate.I also notice that dark dust on the piece that shows you dipped the tungsten and didn't stop and grind. A fouled tungsten changes it's characteristics and tends to spread out your arc making it harder to weld well. I end up showing students how easy I ( or they) can weld with a clean fresh tungsten, but as soon as they foul it, they go right back to having issues again. I'm also looking at all those beads on your practice pieces. I frequently see students do this, then wonder why they can't make good beads on individual pieces when it comes time to "prove" their stuff. Many times it's because the practice pieces get welded over and over and over and never get cooled between beads. That fools the students. The piece acts like they have the amps cranked way up and puddles can form at what seems like lower amps. With cold material that doesn't work.I'm also seeing a fairly long back cap being used in a confined space. I'm guessing all your practice pieces are being done out in the open, vs inside with restrictions. That complicates things greatly and your actual welds show it in the poor consistency vs the practice pieces. Set up your practice pieces and stick them in the "box" to weld. It gets  a lot harder to do when you need to get both arms the torch and the filler down inside. What works just fine out in the open won't work the same inside where you keep running into things..No government ever voluntarily reduces itself in size. Government programs, once launched, never disappear. Actually, a government bureau is the nearest thing to eternal life we'll ever see on this earth! Ronald Reagan
Reply:What machine are you using and what settings?
Reply:One thing to remember is that many abrasives are made out of aluminum oxide. It is also easy to get abrasives imbedded into the soft sticky aluminum.  If you are using aluminum oxide abrasives you may be imbedding the exact same material you are trying to remove. Now I use aluminum oxide abrasives to clean aluminum from time to time and generally don't have issues but if you do have an issue it's one of the first things to look at. Hand sanding with sand paper really doesn't work well for prep either. The high points in the grit scrape little vees in the oxidized layer but it takes a lot of sanding to actually break through that layer down to fresh aluminum. for small runs on lightly oxidized aluminum a small stainless brush works very well. For heavily oxidized material, long runs or big surfaces a stainless cup brush works very well as long as the surface finish is acceptable. Aside from that I would tend to agree with DSW about the fact that it still looks oxidized in the pics. If I look really close I can see your grinding marks but it doesn't look like it breaks through the oxide layer which just looking at the rest of the surface looks like it may be quite thick. Once you get through the oxide layer you will see a very noticeable change in the appearance of the metal. It will also get a lot sticker. Your wire brush or grinding wheel will take more effort to push through the soft aluminum than it does skipping over the hard oxide layer.Lincoln precision TIG 275Millermatic 140 MIG
Reply:What did your practice t-joints looks like?  Did you have this same problem with them?(the pics seem to show only flat practice)  Dave J.EDIT:  I think my eyes are playing tricks...now they look like lap welds.  Same question then though - what did your practice lap welds look like?Last edited by MinnesotaDave; 09-14-2013 at 12:46 PM.Dave J.Beware of false knowledge; it is more dangerous than ignorance. ~George Bernard Shaw~ Syncro 350Invertec v250-sThermal Arc 161 and 300MM210DialarcTried being normal once, didn't take....I think it was a Tuesday.
Reply:Is that a gas tank?  hmmm.  So when's the maiden voyage?  lol.  Do you think. possibly. that this project is somewhat beyond your skill level?   No!  Well then, Bon Voyage!!!No!!  Bomb Voyage!!!Last edited by ncountyline; 09-14-2013 at 01:33 PM.
Reply:Thanks for all the suggestions guys. I think you are right, it probably wasn't clean enough. It wasn't obvious if I had gotten the inside of the tank clean enough due to lack of vision. Although I prepped the parts in the exact same fashion as the practice parts, the results were quite different. I did attempt to clamp the parts to keep them from separating, but that didnt help much. I think one thing I forgot was to clean underside surface of the baffles, that might explain it also. Sandpaper....ugh whoops wish I would have known that before. I am in the process of doing all the welding on the outside and that is going as smooth as butter now without issues (and I did clean it with sandpaper and acetone-go figure(my abrasive pad on the right angle grinder flew off into the abyss)). If it weren't for the baffle issues the tank is turning out good, oh well, its not too bad for the 1st aluminum project. I will throw some pictures up later when I am finished.
Reply:My 2 cents worth is it's the heat causing the flat plate to warp and leave a gap. A joint like that I have much better luck using a wirefeed as you're in, got it welded and done before it has a chance to expand away. It's a lot harder to heat the flat surface of the tank than it is the edge of the baffle. Like showdog said, clamp it down and like DSW said hit it hard, hot and fast....Mike
Reply:Many, many questions have I. What filler rod are you using? What gas type are you flowing at 15CFH? What machine and what frequency? Do you have foot control(I see craters at the end of the weld).What I am seeing is typical your type of tank. So you are not failing that bad. You could finish it up, which you may as well, and it will work fine after the hammer makes it's appearance. Then you could build it again from .120 5052, and use 5356 filler which is the correct filler for 5052.Had I done this from .090, I would have used cleco fasteners or screws to secure the baffles. The baffles need to be welded not just on the flange side, but from another side to prevent welds from side splitting. The cleco holes can be plug welded later. I personally would have acetoned the grease off in advance and upped my freq, and lowered my balance to etch the crap off better. I also would have left the helium off as it is not needed. Other than that, get the soft face hammer out and keep going. At the least, keep it for posterity. I hate it when people start a project and don't follow thru. Do you or don't you use a foot control?Weld like a "WELDOR", not a wel-"DERR" MillerDynasty700DX,Dynasty350DX4ea,Dynasty200DX,Li  ncolnSW200-2ea.,MillerMatic350P,MillerMatic200w/spoolgun,MKCobraMig260,Lincoln SP-170T,PlasmaCam/Hypertherm1250,HFProTig2ea,MigMax1ea.
Reply:On the internal pic with the black soot, it looks like you just contaminated the tungsten with aluminum.  The puddle definitely won't run normally when this happens, and I wouldn't be surprised if it happens a few times, welding in a confined area with awkward access and limited visibility.  The fix is to stop, clean up tungsten, and continue on.  The black powder will wipe (or wash) right off and is not in itself a problem.  However the oxides that you'll make if you weld with a contaminated tungsten can be problematic.It looks well-prepped to me.  Looks like you used a flap disc to gently sand the surface.  Assuming you sanded the back side of the lapping pieces as well (so no trapped oxides in the weld joint), then that's perfect and appropriate prep for this, IMO.  And your weld beads that didn't get aluminum on the tungsten look good.  Watch out for those ending craters, though!  Easy way to put a pinhole leak or crack initiation site, especially if it's a fully-penetrated weld that craters from both the front side and the back side at the exact same spot.  Fix by tapering off the amps (often by a very large amount, it's a delicate balance) - practice on scrap.I don't buy the whole bit about staying away from aluminum oxide abrasives for aluminum weld prep.  If you keep the metals cool while brushing or sanding on them, nothing's going to embed under the surface.  If you heat the material up so much that it starts oxidizing, therein lies the source of the oxidation - it's not coming from the abrasive material.  And this is why using a power wire brush is not a good idea (easy to heat up material to oxidizing temp without realizing it).  Flap disc on an angle grinder, with gentle pressure and kept moving (so will not overheat the materials) is about a good of a method for prepping aluminum for welding as it gets, IMO.  (Especially if followed up afterward with a wipe or two of acetone around the weld area, until rag comes back clean).If your puddle has a problem with skinning on the surface or crap floating in it while under the arc, and your tungsten is clean (not contaminated) increase your %EP a bit by appropriate adjustment of your your A/C balance setting.  (Assuming you're using the old miller in the background of one of your pics... an Aerowave is it, with individual EN and EP amp settings?  The A/C balance setting controls %EN, so reducing it the A/C balance on that machine would increase the EP%.)Last edited by jakeru; 09-14-2013 at 04:11 PM.
Reply:Something I've been doing lately is using aluminum brightener to etch the metal before welding.  Don't clean til your ready to weld. I don't like wire brush marks. Most un used metal is clean enough not to mess with cleaning. Inverters are amazing.
Reply:Backing up that long, mid tank butt weld with backup strip tacked would help.(A lap joint for the tank walls is appropriate.)What is the end use for this tank?Why is the fill tube end so close to the tank bottom? This limits the tank filling, ordinarily, unless you've designed some exquisiteventing system associated with that top cover.Blackbird
Reply:I weld quite a few alloy fuel tanks and really you need to go back to basics and think out a plan before you even start cutting material.For a start, your material is too thin you really should be using 3mm (.120") if possible.Also material type, I much prefer to use 5005 as I've experienced cracking when folding 5052.For the tanks I build I'll try to design them so the lid is welded on last and fold a shell then install baffles first, then endplates, then if possible and I'm using angle mounts as in the pics below then having no lid allows me to clamp the angles around the shell so no distortion is possible.You need something solid that you can clamp the folded edges of the baffles as what you are describing as the shell of the tank moves away from the baffle is always going to happen unless you can clamp the baffle folded edge with the tank shell in between to a hard surface preferably a heavy steel bench.I've got some thick wall 30x30 RHS I use to reach from the end of the tank to the baffles with some heavy G clamps to make sure they are held flat against the bench.The steel bench will also act as a heat sink and draw heat away minimising burn through (you will still get it to some degree though).Has your welder got enough amperage for the job especially when welding the baffles and using a bench as a heat sink?You need big amps and you need it FAST so you can get the heat into it and QUICKLY get the tacks you need in place.I only stitch weld my baffles and one thing you really need to learn is a sequence of tacking to minimise the way the tank shell wants to move away from the baffle edges.Being right handed I'll put a tack at furthest point on the LHS where I intend to stitch weld I'll then start my welding so that I'm always welding towards the tack, if you don't you'll see the tank shell open up below the baffle especially if it isn't clamped down.Don't weld past the tack unless you've got another tack in place that you're then welding towards or it will open up again.For tank endplates and lid I'll try to fabricate the tank in a way so I end up with lap welds so with the endplates it allows me to stitch weld along a folded edge just like the baffles before welding the tank shell from the outside.Sometimes this isn't possible but I'll try to achieve it if possible, I've even gone to the extent of welding flanges to odd shaped end plates to get the same thing.For me, I like to weld in "false floor" horizontal baffles as doing this controls how the fuel moves in the last 25mm-30mm of tank depth and this is where it's the most important as with a "false floor" the fuel can't move away from the fuel pickup/s.I'll also return fuel from the fuel rail directly through the false floor.It should give you some things to think about and the pics below show how I build the tanks I do (Mustangs in this case)Regards Andrew from Oz.
Reply:beautiful work. nice.
Reply:Very nice awill4x4!You using what filler on your 5000 series alum sheet?Weld like a "WELDOR", not a wel-"DERR" MillerDynasty700DX,Dynasty350DX4ea,Dynasty200DX,Li  ncolnSW200-2ea.,MillerMatic350P,MillerMatic200w/spoolgun,MKCobraMig260,Lincoln SP-170T,PlasmaCam/Hypertherm1250,HFProTig2ea,MigMax1ea.
Reply:I can't keep up with your guys responses. Thanks for the quick advice. The rest of the tank welded up just fine. Besides the bulging from welding the baffles, the tank didnt warp at all. I didnt get any photos of the tank by itself, but here are some pictures of the tank resting in place. This is for a vehicle I am building, it is a replica of the Ariel Atom powered by a LNF engine (GM turbocharged 2.0L direct injection) (the rollbar is not finished yet). You can see the end face of the tank in the 2nd image. Yes the tank has a vent to the filler neck and to a carbon canister with fill limit and rollover valves, the filler neck being located close to the bottom has no effect on filling rate in my previous testing. Yes I am using a foot pedal and using Argon gas, I cant recall the filler metal, but it was the recommended filler for this series of aluminum. I did turn down the AC balance to 65% (used 70% for the internal baffles) for more cleaning on the rest of the tank, whether that helped or not I dont know, probably didnt make much difference as the outside welds had no issues. The tank uses a OEM fuel pump and I had a pump "ring" machined so that it accepts the pump, o-ring and locking ring.Oz, very nice example, interested in building me a tank if I decide to redo this one? PM me if you wish!
Reply:Originally Posted by shovelonVery nice awill4x4!You using what filler on your 5000 series alum sheet?
Reply:Originally Posted by maccosOz, very nice example, interested in building me a tank if I decide to redo this one? PM me if you wish!
Reply:Originally Posted by awill4x4I generally use 5356 filler wire, I've been caught out a couple of times on jobs I've welded with 4043 only to find the customer wants the parts anodised and unfortunately 4043 anodises black.Luckily only small jobs so I tend to use 5356 as my "go to" filler except when welding castings where I'll use 4043 or 4047.Regards Andrew from Oz.
Reply:Yup, just confirmed that I used 5356 for the filler rod. 3/32 filler diameter with 3/32 2% thoriated tungsten I believe.