Sheet metal inlet manifold cracking. Help needed.

admin

Hi all,  I have not done a lot of aluminium welding, however previous aluminium i have welded pipes and intercoolers have been successful. Hopefully someone can help me, i purchased a custom inlet manifold which is off a turbo engine, running 18+psi.  It is made of 3mm aluminium sheet (quite thin i know).  I had to modify it which required a cut and shut.   I linished back the weld along the top edge because some of the welding didn't look visually nice.  Anyway to cut a long story short, on the dyno, the manifold cracked along where i linished the weld back.  After reading some info, it would seem the weld filler material provides the strength and linishing it back was a mistake unless it was welded from the inside as well.  Anyway, i rewelded along the top where it cracked.  I thought i had pretty good penetration and a sound weld, which i cleaned up and prepared as best i could.  Only a few weeks of constant thrashing he manifold has cracked again under boost.  It has split right along the weld.  I used ER 5356 filler rod, 2.4mm rod and tungstun.  I waved over the weld after welding with 'pulse' which made it look a little more uniform and visually nicer, however not sure if this attributed to making the weld weaker.  I did not preheat the aluminium.I have a Kemppi AC/DC mastertig 200. Is this fixable with re-welding once again?  Is a different filler rod less prone to crack/fatigue with pressure?  Is 3mm to thin for strength? (I've been told you really need 4-5mm for a job like this that will see over 20-30psi regularly.Any help would be appreciated,Regards,Brendon. Attached Images
Reply:It's difficult to tell exactly what's going on in the pictures.Flat sided containers are not good for holding pressure.  I suspect the plenum is ballooning under pressure.  This places your weld in bending rather than just in tension.Can you add reinforcement to keep the plenum from ballooning?  A box with exterior ribs would be stronger than the box alone.Another option would be to replace the flat sided box with a circular tube.Dynasty200DX w/coolmate1MM210MM VintageESAB miniarc161ltsLincoln AC225Victor O/A, Smith AW1ACutmaster 81IR 2475N7.5FPRage3Jancy USA1019" SBAEAD-200LE
Reply:Hi Andy thanks for the response.  That was one thought going through my head, to cut off the 'box' plenum where the runners are and roll a 3mm sheet with a 3" die and use a rounded shape plenum instead of box, that would mean only two welds.  The other option was reinforcement like you mention, i.e maybe some angle aluminium over that edge to provide support, however this would leave it visually looking crap.  I'll try to get some better photos in a moment.
Reply:I would run the ribs around the box the short direction, not down the edge.  You want to support the middle of the flat area so it doesn't blow out.  If you have the sheet properly supported then the edge weld only carries tension and not bending.  Not having a solid bead on the inside it giving you a notch that starts a crack.An angle down the edge wouldn't really solve the ballooning problem.  It would probably just delay the failure.Dynasty200DX w/coolmate1MM210MM VintageESAB miniarc161ltsLincoln AC225Victor O/A, Smith AW1ACutmaster 81IR 2475N7.5FPRage3Jancy USA1019" SBAEAD-200LE
Reply:Hmmm, food for though, couple more close up shots of the crack.  I really liked the box shape and i know it's been done before without cracking, so wasn't sure if it was my weld penetration/filler rod or sheet metal too thin at fault.  I do understand what you mean about ballooning however i have seen this shape inlet/plenum work ok for some others. Attached Images
Reply:Unknown alloy aluminum sheet metal, sharp corner under pressure and vibration, possible incomplete weld penetration on the inside corner leaving a ready-made crack-initiation site, use of 5356 filler (which is more crack sensitive to begin with than 4043 filler), use of 5356 where conditions exceed 150F (probably not during use due to the hopefully cooler air charge coming in) but certainly after shutdown underhood temperatures will exceed 150F, linished (= burnished aka 'creased inward') into a concave surface profile, hmmm anything else wrong here?Sorry dude, once you cut and welded and cracked and rewelded that aluminum and it cracked again, I really-really think you have to:- get a 'better' design/materials/workmanship going on.  The best laid schemes ... Gang oft agley ...
Reply:Temps out of the turbo could be several hundred.  Depending on how well the intercooler (really aftercooler) works, then temps could still be above 150F.I'd call it a faulty design that other people may have gotten away with because they built with thicker material.Dynasty200DX w/coolmate1MM210MM VintageESAB miniarc161ltsLincoln AC225Victor O/A, Smith AW1ACutmaster 81IR 2475N7.5FPRage3Jancy USA1019" SBAEAD-200LE
Reply:5356 is actually more ductile that 4043, however more sensitive to "hot crack". The fact that the crack is straight down the middle, and knowing the application leads me to believe that it was incomplete fusion coupled with less that ideal weldment placement. The lack of multiple " spider" cracks would suggest that it was not a thermal stress issue from the mag content. Complete fusion without oxide inclusion and other "grunge" on the backside of the weld can only be done with a backpurge if using the tig process. 3mm is plenty thick for that manifold pressure and quite honestly it should take the abuse if welded properly. Move the weld further away from the bend, tig with backpurge 100%penetration  or weld with O/F method. Base material im guessing is either 5052 or 6061, in either case 5356 would be my filler of choice.
Reply:Another thing to consider with it being as thin as it is is if the weld and surrounding material is cooling to fast.  We had a weld crack on a 1/8 structural cage today due to rapid cooling from a fan blowing over the weld block though not directly.Welding Supervisor Department of Corrections.
Reply:Ahhh thanks for the all the input everyone.  Upon further inspection it appears the plenum had been ballooning under boost as you can see/feel the sheet metal is convex accross the top which used to be flat.  I hope goes with AndyA's theory of the weld bending and finally fatiguing?  I mean this did last quite a while at larger boost levels before it finally split.  I think i'm going to cut the 'box/plenum' section off and roll one sheet over on itself maybe using 3-4mm.  This should have more strength and resistance to ballooning not to mention less welds?  Can anyone tell me, is 5005 grade sheet aluminium and 5356 filler rod suitable for the task?  This is what materials i have available to me.  I figure while 3mm doesn't have much strength in a box configuration, it would be ok in a rolled configuration eliminating welds?Thanks.Brendon.
Reply:5005 would not be my first choice, I would rather see a 6xxx series or a 3xxx series material. 5005 tends to work harden pretty quickly and is not really happy at sustained elevated temperatures.
Reply:Welder killed when he pressurized an aluminum tank with air and it ripped open and cut him in half:http://dhbc.ky.gov/NR/rdonlyres/2D04...lExplosion.pdfAlso mentioned in :http://www.nationalboard.org/SiteDoc...ins/WI2011.pdfAnd it wasn't directly from -his- repair welds, but rather poor/inadequate ORIGINAL materials and design and workmanship (compounded by his use of air pressure to leak/pressure test something, instead of using an "incompressible working fluid" aka usually plain water to pressurize the tank).Just an FYI to make you aware of what can happen even at relatively 'low' pressures (I believe it was thought that the referenced aluminum tank failure occured at a pressure of around 40 psi or so, lower than the 'nameplate' rating on that item of 60 psi, but waaaay higher than the calculated after-incident analysis using the actual ASME PressureVessel Code requirements, which worked out to 2.5 psi 'safe' design pressure limits, per the consulting engineer's report.)[not saying or claiming that his inlet manifold will fail and rip him in half, the example was just an FYI about how pressure vessels are not so simple/easy as to be a DIY item.]See, it's not always so easy to design and make something that is safe and effective once pressure (and heat and vibration) are thrown in.  Impossible?  Certainly not.  But not quite so simple as 'weld up a box shape and pressurize it' either.Also note that aluminum is a certainly useful material and all, but it has the 'interesting' design characteristic that aluminum WILL fail from fatigue (if there are cyclical/vibration loadings going on), pretty much no matter what, eventually.  With steel, you can almost always design it so that it will last 'forever' from a fatigue standpoint, while with aluminum you pretty much can not no matter what you do.Cyclic loading and vibration will 'do in' aluminum.  Eventually.And as mentioned by others above, for 'clean' welds on a closed shape you really have to backpurge.  Cause you can't get in to backgrind and weld from the inside, you have to make SURE that the weld root is clean and complete with no 'crud'.  The best laid schemes ... Gang oft agley ...
Reply:Can you add rods across the inside of the box to act as stay bolts used in boilers.  This would tie the box together so it would not baloon out.Dan
Reply:Stay bolts will probably mess up his airflow.  While you could do that to solve the pressure problems, it's not going to make a good intake.Dynasty200DX w/coolmate1MM210MM VintageESAB miniarc161ltsLincoln AC225Victor O/A, Smith AW1ACutmaster 81IR 2475N7.5FPRage3Jancy USA1019" SBAEAD-200LE