SPROING!!!! Some header stuff

admin

Some of you may remember some of my posts of some of the headers I make. Thought I'd post this since there had been some discussion a while back about "locked in stresses" in headers- probably the #1 killer of turbo headers IMO. Here is what they look like when they're done: The technical term for this process is caled "de-sproinging". :rofl:Basically there are forces locked up in the runners when you use a 1 piece head flange. When you then vibrate and heat cycle the header, these stresses try to seek out the weakest points in the header (stress risers) and cause cracks. So, after they are done,  I cut in between the flanges. At which point they go "SPROING!" and try to eat your bandsaw blade. Wee!Here is a pic after cutting:SPROING!that 1/8" of deflection represents probly close to 100lbs (!!!) of force trying to create a crack. No wonder one piece stuff dies an early death, eh? Here is a picture of it realigned (aka desproinged :rofl- the rightmost runner may need just a touch more, but once I get to within 1mm or so I generally call it good as it is hard to straighten to within a tighter tolerance than that without overshooting, at least for me). Note the slight gaps are supposed to be there, I try to get these the width of the bandsaw blade:Closeup example of how one of the flanges that sproinged th most lined up after cutting:and a closeup of the same joint after straightening:The straightening process is tricky but it just takes a little practice. I easy fixed the 2 finished headers in a couple of hours. Also made one of the braces: Nothing spectactular but I think it helps damp some of the engine resonance and they take 20 minutes to make so might as well. ::wiggle: After this they will be machined flat at the machine shop. Fun stuff!Last edited by Capnbondo; 06-24-2008 at 04:08 AM.
Reply:looks like a lot of work to me! But you make it look nice and easy. And those welds look good to, for all that out of position you have to do. Great job from were I stand!Miller Thunderbolt 225Millermatic 130 XPLincoln HD 100 Forney C-5bt Arc welderPlasma Cutter Gianteach Cut40ACent Machinery Bandsaw Cent Machinery 16Speed Drill PressChicago Electric 130amp tig/90 ArcHobart 190 Mig spoolgun ready
Reply:Thanks!  Other than the odd one-off, this is the first "production run" I have done (making 5 of them). I'm still fine-tuning the TIG welding, doing all of this work in one go has really helped me improve- the next batch will be lightyears ahead of these. I kinda rushed these a bit- like I'm sorta nervous. But they've come along well and I have learned a lot. I have to drill it into my head that rushing is NOT faster in the long run. I also need to get some 0.40 filler, 1/16" is just too much I think.
Reply:pretty nifty... i never really paid that much attention to the locked in forces in the headers, but now that you brought it up, i understand and it makes sense to me. just out of curiosity, how do you actually straighten/realign everything anyway? just use a vice and a little tugging? seems too complicated for that.Later,Andy
Reply:Wouldn't the preferred method be post-heating in an oven and then cooling slowly?  Perhaps a re-cycled ceramics kiln might be adaptable?-MondoMember, AWSLincoln ProMIG 140Lincoln AC TombstoneCraftsman Lathe 12 x 24 c1935Atlas MFC Horizontal MillCraftsman Commercial Lathe 12 x 36 c1970- - - I'll just keep on keepin' on.
Reply:Some sort of post-heating/normalizing process would be cool, but especially with stainless I think you need a pretty serious setup.It only takes about an hour to cut and straighten so it's cheap insurance and not particularly labor intensive- and then I know it is good with relative certainty. Andy,I basically "flame straighten" them. I run the tig over a suitable weld joint at a suitable position on the weld, add a little filler to keep the alloying happy. A pass about 1-1.5" long is generally all it takes. You just have to pick the right spot. If you overdo it or do it in the wrong spot things can go downhill fast. The first couple I did, I got to cut the flange off and try again. Now it's easy.
Reply:Originally Posted by mondoWouldn't the preferred method be post-heating in an oven and then cooling slowly? Perhaps a re-cycled ceramics kiln might be adaptable?
Reply:I dont know if im missing it or what , but the reason you cut it is to re leave the stress , how does it get this stress? From the heat? How are you closing those gaps then , wouldn't it just put the stress back on? I'm lost...
Reply:I don't think you'd do much good to bother trying to heat treat it.  Think of how hot they get when being used.  They often glow red hot.  If that doesn't take out stresses (as well as put them in), I don't think anything will.
Reply:Originally Posted by EngloidI don't think you'd do much good to bother trying to heat treat it.  Think of how hot they get when being used.  They often glow red hot.  If that doesn't take out stresses (as well as put them in), I don't think anything will.
Reply:I believe that one of the most common problems with headers is their lack of penetration.  Here's the scenario I have in mind.  Most guys you see on the net the build turbos use little or no open gap.  It's obvious by the width at the top of their bevel.  Therefore, you know they likely got no or little pnetration.  So lets say you have fit up a weld joint with about 1/16" gap.  You get spotty penetration, and the gap ends up closing on you before you get all the way around it.  For example purposes, lets assume that alternating quarters got penetration vs didn't.  The quarters that didn't will be touching at the root (inside).  As this heats, you essentially have a thin layer of metal here that heats up faster, with it not being a solid piece all the way through.  Just imagine half your bevel penetrated and the inner half just touching.  As these parts expand, the inner parts without penetration are pressing against each other and the weld joint is tying to expand, in this area, at a fate faster than the parts that are fully penetrated.The problem is that most people welding manifolds simply aren't able to get good penetration all the way around.  Even the "top" manifold companies really struggle with it.  That said, nobody really has the science behind each one of the thousands of designs, to know how much penetration is needed, what angles you can get by with coping one pipe into another, etc.  It's a fine line, and nobody really has the money/time to find where it is.
Reply:I can totally see what you're saying- I actually feel the opposite though. The situation you described should result in a failure right at the weld- as the end of each pipe presses against the other, the pipes are literally trying to rip the weld apart. This is almost more of a yield failure than a fatigue failure (would probably not take many cycles to occur) and cause a failure right at the weld.While i can see the logic in this, I'm not sure the failure mode is consistent with what typically happens. Most of the failures I have seen occur in the HAZ, in other words probably just too much heat, or maybe 304 just embrittles somehow. More of these manifolds look like they've been overheated- possibly in the pursuit of penetration,The best scenairo might be full pen using a fairly agressive bevel, around 60 amps, and doing 2 passes on (.109) schedule 10. My thought though is that a single pass at 60amps or so with tight fitup has a minimal chance of contamination and at roughly 60-70% penetration is still as strong/stringer than the HAZ, and the HAZ has been minimized. But that is just my current theory.