Would you brace a stainless manifold?

admin

Ok, here is a quesiton that has been rattling around in my brain for a month or two- Some of you might remember seeing some pics of some of the stainless turbo headers I have been building. There is a debate going on in my head- should the manifold have a brace that ties the turbo flange back to the engine block? There are 2 lines of reasoning here: 1) A brace is good. Support the mechanical weight or the turbo and exhaust, and brace it against vibration- major source of fatigue failure.Fair enough, sounds good right? But, her eis the other argument2)  the manifold has 4 runners that a 1.5" sched 10 pipe, it is mechanically strong. What it needs is the ability to move around a bit as it heats and cools- a brace actually fights the header's natural tendency to change dimensionally as is heats up, and puts a wierd load on the header, causing fatigue failure. Any thoughts? I can see either theory making sense. (At the moment I am installing a brace).Thought this might make for a good debate with the metallurgists...
Reply:It is a complicated question.  We make aircraft engine components which are subject to a lot of fatigue stress.  We had a person with a MS in Mechanical Engineering along with the best FEA computer software to do the stress modeling.  What did they do?  They laid off the theory guy and spent $500K on a high cycle fatigue tester.  We call it the shaker.  Every part we make gets tested as a prototype.  Sometimes the first guess is right, and sometimes that parts crack requiring a second iteration in the design.  At any rate, the build and test results are better than the theory guesses ever were. Since I am a hands on engineer, I find that amusing.  As for restrained parts, it is a simple calculation to determine the stress based upon the thermal expansion and operating temperature.  If it is greater than the shear stress of the material, it will crack.  The number of cycles however is a different animal. Sometimes thermal fatigue is best evaluated by a real world test.  In the automotive world I am certain that is what they would do.  It sounds like you should expect your first attempt to be a prototype.Last edited by obewan; 05-30-2008 at 01:02 PM.
Reply:Agreed. All of these loads are cyclic/way below the actual shear strength of the manifold, manifold failures are pretty much always a fatigue related failure of some kind. The added complication is that austenitic stainless likes to work harden and become more brittle....I currently am using braces because the factory did- but the factoy manifold has short runners and is cast iron, so it probably didn't change dimensionally very much with respect to temperature- so that logic may not be valid. So here is another question- assuming a brace is good- can it by a piece of metal that is welded from one part of the manifold to the other, essentially "triangulating" it?
Reply:I can make this real simple...             It will either work or it won't.                            Go for it!
Reply:I'd try NOT doing it.  The reasoning is that the stresses caused by the heating and cooling cycles will be very great, and too much bracing can constrict it.  ...that stress may cause a failure. Either way, it's not a big deal.  If you have the ability to make a manifold, you have the ability to fix a crack, should it happen.
Reply:F1 cars use to have their turbo's braced.....I had a bracket cnc'd for my turbo...and using heim joints to compensate for the movement due to flew/expansion is a must...otherwise you might develop cracks.
Reply:Holy crap that is quite the bracing system! I can see how that would be the best of both worlds becuase I really get what Engloid is saying and that is my primary concern with bracing.  But, between that pic and the fact that a lot of OEM stuff is braced, it makes me think something should be done. Some background:I have built a prototype header that is on a friend's car that is currently braced. It has about 2000 hard driven miles on it with no issues as yet- but I know it can take longer than that for cracks to develop. Since then I have sold 4 more of them and am currently building them- the buyers know these are a pilot/experimental project and the headers were priced accordingly (in other words, I am losing my *** on the price. ). But this is the cost of research. And if they crack, I'll fix 'em. Hopefully I will learn what works and what doesn't and some peopel will gert some good deals on headers.I am almost tempted to ship 2 with braces and 2 without...A heim jointed setup would surely work, but the increase in cost/complexity would likely be significant... hmm....Thanks for the input guys!
Reply:The F1 turbo era set ups are an extreme example but i would model any 'brace' along similar linesThe YB manifold i posted the other week will have a similar set up- the 2wd manifolds used 'turbo dampers' from the factory due to their size (long runners) but the primary reason AFAIK was to dampen the resonance that came with this designI'm no metallurgist but there's a fair few possabilities that could contribute to fatigue failure of turbo manifolds...I can see the argument that bracing the manifold 'too rigidly' can actually cause problems as expantion/contraction are restrained. Like the F1 engine it's an extreme example but Ford had issues with it's Focus WRC exhaust compents some time back. They were running insane temperatures and the near molten parts would take a 'set' when cooling down and have to binned. Like the F1 engine we're talking inconel rather than stainless here but the mechanism is similarIt wouldn't surprise me if a similar process were behind the failures of many mass produced manifolds- fabricated in jigs under high restraint thus 'locking in' residual stress. The poor quality welds that often accompany these manifolds isn't going to help matters.Speculation- a couple of sources i'm aware of state the origin of type 321 as a solution to cracking issues with piston engined aircraft exhausts. Carrol Smith was one, can't remember the other321 is a stabilised alloy- contains Ti to prevent intergranular corrosion (Ti has a greater affinity for C than Cr). Maximum working temps are the same as 304 and 316, it's in the (approx) 400-800C range 321 has the edge. Shame it's so hard to come by (in the UK at least), especially as preformed mandrel bends.The low carbon varients like 304L are also immune (or at least less suceptable) to carbide precipitation but have lower strengths at elevated temps compared to 321.
Reply:Good points hotrodder. Another thought I had was that especially in the case of the F1 manifold- it is inconel and probably 18 gauge tube. I would be scared of not suporting it too!I had considered engine vibration- but calling it resonance is way more accurate. Definitely a significant issue. I am working with 1.5 shed 10 304L, so 0.109" thick.Ultimate tensile strength is a non-issue - you could use these suckers for jack stands!The locked in stresses are probably overlooked in a lot of applications as you says as well- although it is easy enough to avoid if a logical "order of operations" is followed.I think I will ook into a bracing method that works on similar prinicbles as discussed- something in tension that damps and supports the weight, but can move a litt.e
Reply:Originally Posted by CapnbondoI am working with 1.5 shed 10 304L, so 0.109" thick.Ultimate tensile strength is a non-issue - you could use these suckers for jack stands!
Reply:Alright enough teory! let see some pics of those manifolds
Reply:Hi There,All the cheap thin units coming out of china are braced across the top of the tube as it comes off the manifold plate. I have made these from SCH10 Mild steel steam pipe and had them HPC coated and havn't had one go, but stainless can be strange, in saying that i've seen some BIG turbos hanging from those thin #16 SS tube chinese manifolds and only fixed a few of them. My guess is it would be fine, but hey i'm no expert
Reply:Kicking around somewhere i have one of those cheap imported manifolds (wasn't worth fixing, owner never collected it and it may make a useful pattern if i'm ever asked to fab a manifold for an S13). Can't remember if it's sch10 but it's certainly thicker wall than 16GIf i remember i'll DPI it and post the results when i get the time- should be ammusing if it's as bad as it appears to beWannaweld, pics have been in the photos forum for some time now- you may well have already seen them
Reply:Look and see what the pros (with the big R&D bucks) do. The F1 bracing seems to take into consideration some of the fears...and fits in with my theory.If you don't want to stand behind our Troops, feel free to stand in front of them.
Reply:I have seen springs integrated into both factory and aftermarket exhaust components.  Perhaps a spring device/s would be worth considering