I'm going to start building a set of headers soon out of 321 stainless steel. Considering how expensive the materials are, I want to do as good a job the first time as possible. I'm starting with a box of 1 3/4" "U" bends.

What techniques have you guys used to work with this type of material?
For example whats the best way to cut and fit mandrel bends together?
Are there jigs available to ensure you cut the mandrel bend in the correct position?
What about a tool to hold two sections possibly bends together for fitting and welding?

I'd like to fit these in such a way as to minimize any gaps between the parts before I weld them in order to minimize the amount of rod I use.

 

Use a grinder with Zip cut wheels and extreme care for most of it to fit each piece of pipe. Rod??? I don't know ho much luck you will have if you pan on arc welding headers up... I would use a mig and tack spot weld them for a "fit" build first then tig finish weld them up when you have them fit as you want. There are tools similar to tail pipe expanders used in steamfitting for aligning fittings to pipe aptly named "internal alignment tool" but I do not know if they have smaller than two inch available and I have never seen one that was flexible for rounding bends. Its a lot of work and very fussy to make your own headers... Why not order a set of ceramic coated ones?

 

gaps are good as long as they aren't excessive. two differant ideas on fitting pipe, feather edges butted together and lands with a gap. lately i prefer feather edges butted together. there are all sorts of tools to hold pieces together so it can be tacked and welded.

 

Depends on the process you are using for prepping pipe for welding. A bevel is pretty much always a good idea but you leave a nickel thickness or so for stainless if you are sticking them and a knife edge if you are tigging them. Regardless header pipe wall thickness can be just butt welded with no problems cause its so thin. ede is right though there are tons of specialty apparatuses available for welding pipe and other items together acting kind of like a third hand but i do not know the name of any in particular that will help you out Dan. Perhaps a Google?

 

I'll be tig welding (GTAW) them together, the rod is for tig welding, its 347 filler rod if I recall correctly. These are for a turbo project thus the need for a custom set.

 

That sounds better. Turbos again eh? Seems to be all the rage these days but i guess I can understand why. In that case seeing as you do have a tig and its for a turbo and all ( high stress ) I would highly recomend that you purge weld them as this is going to give you the best quality. Please post up some pics here of your progress and completion if possible I would like to see how you make out. Thanks Good luck!

 

no need for any jigs to hold the pipe just take a sharpe and make a line across the 2 pieces and when your ready line them up and then just fuse them on 2 spots.... then proceed with welding....

 

Yeah, turbos again... been working this project for a couple years now, a little here and a little there but it looks like its finially gained the momentum it needs to get finished.

I will be be flowing argon to the back sides of the welds (back flowing) if thats what you mean by purge welding. I found a slick pair of specialty vice grips at a local welding store... they should do the trick for properly placing and holding the tubes while they get welded.

I will post pics as they get done... in the mean time check out these I built last year. I was going to use them on my f-body but changed my mind... I'll plan to use on some future project. They are made from schedule 40 weld els with 1/2" flanges.

 

another

 

10" tall from top of exhaust port

 

Yeah thats what i meant. Argon from both sides. Schedule 40 is pretty thick and needs to be bevelled for 100% penetration. Usually you want the two pieces about a nickels thickness where they meet and bevel up on a 45 degree angle or so. Oops I just realized you are in another country so a nickel thickness is about the thickness of a table knife. Do a root pass and a fill pass and it will be as good as solid pipe. I like your design of the diameter step up on the header. Very compact and nice looking. Only thing I would advise is to put a bevel on any material you use that is heavier than 16 gauge ( about 1/16" ) Thanks for posting up the pics. Even I am thinking about building a twin turbo set up now Its pretty hard to argue the power potential per dollar if you do your own fab work such as this. Very nice.

 

GN turbo's? Looks like what I have planned for a future project car... one that doesn't have so much rust.
Where did you get your stainless bends from? I've been looking for tight radius stainless but have come up with only 1 very expensive source... one that makes me thing Gail Banks is offering up a steal of a deal.
Did you get those flanges from an online store if so, mind sharing? One more question, last one I promise. If you got the flanges made, how much did it cost and was it water or laster cut (or EDM)?

 

Yep, weld els are designed to be welded together and come beveled... pretty slick stuff really. I made these manifolds over a year ago (before I had my welder). A friend welded them up with a MIG for me after I did all the fitting.

 

The turbos are aftermarket grand national turbos PT-51's from Precision Turbo the turbo pictured is actually new.

Not sure if you are talking about the weld el manifolds that are pictured or the stainless steel long tubes that I'm building so I'll answer both. On the mainifolds pictured, I puchased the flanges. They were unfinnished, machine torch cut. The metal was so hard near where the tourch had passed that it made them very difficult to work with. I got the header flanges from the same source and they arrived warped. I wouldn't recommend going the same route I went on the flanges... I would have had less time and money in them had I just started from scratch.

Building a set of stainless headers is expensive (at least $1000 buying tubing at whole sale). I really wanted to go this route on this car though so it was worth it to me. If you are concerned about the cost, you might consider building a set out of weld els ($100-$200). I have a friend who is building a set from old mild steel headers that he scrounged. I think he will have about $10 in his headers and they look great (now about 80% completed).

Stainless Steel Brakes now offers stainless tubing (but I dont think its advertised on their web page) well as http://burnsstainless.com/ and SPD.

As for the set of stainless headers I have in progress, the turbo flanges are laser cut 321 stainless from Precision Turbo.
I got my tubing, header flanges and merge collectors from SPD, the craftsmanship is top notch, expecially their flanges which are nothing less than a work of art. I'm a little upset by them though as they charged me a price other than the price we agreed to over the phone... because of this I wont do business with them again and my next order for down pipe material will be through someone else.

 

When I get around to it I plan on using some connections I have with the guys at the local waterjet to cut some flanges for me. I think this is the best route, go straight to the machiners, don't bother through the special turbo shops. This way you know what material, can specifiy the thickness, and the port profile.
I'm stock on waterjet, the lasterjet cutters are only good for things you're not going to weld because of the high localized heat. The waterjet stuff has a bit less tolerance but we're talking like 10thou for the 3/8 thick stuff.
Those are some beefy turbo's. What kind of power are you looking to get with them? I like the idea of them because of the built in wastegage and the 3 stud flange.

 

Dan W I am curious as to why you have decided to go for tubular headers rather than use those nice manifolds you already made up? I don't know much about turbos but I thought the exhaust manifold/headers are not too important in terms of power gains am I wrong?

 

I'm not Dan so I'll let him tell you why, just thought I'd give you my $.02 on the manifold vs header. First off, the purpose of headers on an engine isn't JUST flow, it's resonance tuning. Exhaust resonance tuning is very marginal with a n/a engine because of the high temperatures. I'm not going to bore you with the calculations or Helmholtz so just take my word on most of this. The idea behind headers is that they flow well, a turbo manifold can be designed to flow very well too, stock gen1 sbc manifolds are junk. Since with a turbo the resonance tuning is near nill (turbine eliminates the "plenum" to cause a pressure pulse). So you're left with flow and trying to "smooth" out the pulses at the turbine for efficiency. The same goes for turbo intakes. Forget about runner length and volume, make it for flow! Look at the GN indy intake manifold, it's a huge box with no runners. Well it has runners if you consider the thickness of the surface flange, lol. Idealy I believe the best turbo header/manifold would look like the lakester manifolds you see on old hot-rods, some flat head fords have em too. It has a tappered collector and the pipes flow into it nicely. If you loook at the LS1 exhaust manifolds they too are excelent designs for turbo's.
In the end they all really do the same thing but my point is that headers aren't necessary to make huge power. Don't forget the firing order conflict which means the exhaust pulsing is all out of sync too! This is where the turbo headers can be made to even out the pulses. In other words; uneven length pipes for a turbo are better than equal! This again is for the 90 degree cranks, a flat crank like ferarri is completely different, those would benifit from an equal length header no matter what form of induction. Have I confused anybody yet?

 

I think my combo will make about 800 at the crank on pump gas... 1000 or more on race gas. I chose the grand national turbo because I wanted a round entrance to the hot side... and the terry houston down pipe stubs really help elmininate a lot of the complexity from a packaging standpoint.

 

A couple reasons but I'll tell you right off power was not really one of them.

I think good headers could be worth 100 or more HP in this sort of application over a manifold. A turbo small block with good heads can easily make more power than is streetable no matter the sort of header or manifold... the difference between these two systems would probably be 700 vs. 800 hp... on the street this seems to me a moot point.

I'm building these headers in an attempt to smooth the exaust pulses to the collector... one exhaust pulse at a time making it to the turbos like this should add power (due to the decrease in backpressure between the head and turbo) and help the turbos spool better. Even though were talking turbos here, there are still exhaust pulses before the turbos so a longer tube header design can still help make more power... although probably no where near as important as in a naturally aspirated setup. The headers being made out of stainless should help mitigate some of the heat losses you would expect from a long tube.

The main reason I changed my mind is because of the intercooler setup I have planned. I'm going to incorporate a pair of air to water units into the air box above the throttle body. This intercooler setup was too wide to work well with the manifolds. Asside from that, I think the tuned header idea is cooler both looks wise and from an engineering standpoint.

 

Thanks guys, Good info I never considered the pulse effect on the turbine. I know of a 2500 horse twin turbo in Austrailia thats in a fourth gen and the headers look very radical with bends going all over the place and now my guess is that its a pulse/timing related issue.

Thats where I was coming from. I am new to the turbo game but had learned that heads/manifolds vs headers/ is not where to put your dollars for big horses initially at least.

 

JPrevost, I’m not sure what you’re saying there… from reading it 2x I got “I know why but there isn’t an advantage in using tubular headers over manifolds…”

Based on what I’ve seen, there is NO real world difference between manifolds and headers. 2 things determine output, heat retention and preventing reversion. If you’ve optimized both without restricting flow you will not see a difference between the styles. IMnHO, this is a lot easier to do with a manifold type design then it is with a header, but it can be done with both.

Harmonic tuning is a non issue in this case, since the pressure in the manifold will far exceed any harmonic effects, and the quickly changing pressure and temperature makes it next to impossible to tune for a harmonic anyway. For that matter, collector dimensions determine exhaust harmonics more then anything else and not only is the collector at the mercy of the turbocharger packaging, but the pressure increase due to the scroll located after it pretty much negates any effect that you’d get there anyway.

The biggest difference I’ve seen between the designs, and apparently, for the most part the determining factor in which is used appears to be how much $$$ was spent/did the owner expect to spend on their setup. Not flow or performance.

Dan, if you have some proof of a HP increase between both setups on comparable setups I’d love to see it.

 

After reading you're reply 4x I dont' know what you mean, lol.
There is an advantage to using tuned length headers over your typical "flow designed" manifolds. The difference is with tuned headers you can make more power by tuning the individual lengths to smooth out the pressure pulse at the scroll. You could do it with a manifold but it would be a rather complex casting and really heavy.
I too haven't seen numbers showing manifolds being any worse than tuned headers but at the same time I've never seen a test designed correctly to measure a difference. They usually have crappy small low flow manifolds vs large diameter tuned headers
The other reason manifolds are used more redily is packaging.
I'm a manifolds guy but there's something to be said about the smoother exhaust pulse, wonder how much of a difference it really makes.

 

Ok, if I’m following you’re saying that you’re going to get the pulses into the turbine at even intervals. How do you intend to do this on a V8? Unless you’re thinking of using a 180* crank or working in multiples of 3 cylinders/scroll it’s not possible, the firing order will not let you.

Not that it really makes a difference since most of the energy available in the exhaust is in the form of heat, not kinetic energy. The gasses leaving the turbine still have most of their kinetic energy but are cooler. On the other hand, with most designs, they loose much of that kinetic energy within a foot or so of the turbine as friction.



EXACTLY. On top of it typically they upgrade turbos, downpipes… in the process making it a no brainer what will make more power.

I’ve seen a few tests showing some exotic manifold designs producing way less backpressure at the port and way more power then a comparable header, but again, I don’t know that it is a fair comparison. The worst thing is that I could setup a test with the same turbos… that would still not be fair… all you’d have to do is put a “turbo cam” in the engine if you wanted to favor a design with more back pressure at the port and a more conventional cam if you wanted to favor a design with less pressure at the port.. Like with everything else automotive, it’s easy to stack the deck with “the combination.”

 

Hmmm... interesting views guys. I have been thinking of building up a twin turbo 406 what do you guys think I should do? Is max displacement as important on a turbo motor or is the stroke/bore relationship more vital? What about bottom end recomendations? How much boost before the insane factor sets in? Any other advice?

 

You're right about the tuned length headers with a standard chevy firing order. They are impossible to even out as is. One could argue though that a good tri-y design would help!
Another interesting thing is that if you have a single turbo there is no problem with this, maybe that's a reason singles do as well as they do on odd fire v8's . Now if you had $10,000 I bet I could come to a conclusion through some testing but what's the point right?
In other words, equal length, odd length, doesn't help but a tri-y might, just never been tested so who's to know.
You could also bring over the odd fire cylinder from the other side of the engine if you had the room . That would solve things. I'm actually suprised people don't do that on their dragsters with tubular k-members. It would only make sence.

 

LOL, I knew that was going to get controversial.

I dont know of any turbo small blocks running manifolds making the kind of power at the boost levels we're seeing from cars like Monty Williams corvette, John Meaneys 92 corvette or Troy Trepaniers 62 biscayne. These guys are all makeing 1200 hp or more on pump gas turbo small blocks. I'm no professional builder nor am I an engineer so I can't site a bunch of testing from personal experiance. I do a fair amount of reading though and Corky Bell sites improvements with well designed headers over manifolds.

I think it is possible to time the pulses one at a time to the collector by manipulating the length of each runner. I calculated cylinder volume and translated that into a length of tubing... I then plotted out the small blocks firing order for each bank in 90 degree intervals with where the exhaust pulses would be at each interval and lo and behold if the middle two header tubes on each side = the cylinder volume and the outer two are 50% longer, your exhaust pulses will arrive at the turbos one at a time... but maybe I'm wrong. Think about it and let me know.

The attached picture is a header from Monty Williams car... looks like he did exactly what I'm talking about.

 

I’ve thought about that quite a bit… I’m pretty sure that tri-y in itself won’t make that big a difference, but contouring the tubing and only having to deal with 2 pipes merging at a time would greatly simplify getting a good design.

You have to remember that with a turbo manifold, your “slugs” of exhaust gas have much less to do with what happens overall then the pressure in the system, and no matter what the shape is, straight tubes will pretty much have the same pressure in the whole system no matter what order you join them in.



Cubes + boost = win win situation. I don’t care if it’s optimized or not, it’s not that hard to make more power then what 99% of people know what to do with with that combination.

OTOH, I’m personally experimenting with this deal right now with the setup that I’m built… the whole setup is based around preventing reversion and controlling the direction of flow with as small a wall surface and manifold volume as possible, and I started with a cast LT1 manifold on one side (if you take a look at them without the heat shields on you’ll know why).

If it works the way that I expect I’ll probably end up ditching the setup and putting it on my truck since I actually expect it to spool a little too well for what I want.



I’d kill to be in a position to do that testing and get conclusive numbers. I see the point entirely.

that is recommended by Mcinnes and a few other big names in the field… I suspect that as long as you take heat loss control into account it would help. OTOH, if your car still has the manifold heat passages open this is a non issue since you’re accomplishing the same deal through the other side of the head.



I’m telling you, the earth is flat…



show me another build at that level that has tried manifolds?

The other reasonable argument is that is not really that outrageous a goal. Banks claims that kind of power is possible with their cast manifolds, and I don’t have any reason to doubt it. When it comes down to it it’s not that outrageous a number considering the parts going into those engine packages….


Corky Bell is full of ****.



I’m not sure how you came up with your numbers but it sounds like you over simplified. If you had it right you’d find that any particular volume would only work for a fairly narrow rpm and VE band. Basically you’d end up with a setup that only worked at X rpm with Y throttle position and load.

As far as I can tell his stuff is like it is for packaging reasons, nothing more, nothing less…

 

Thats my point, and thats what I was asking you to provide to support your claim.

I just checked their web page...they advertise up to 1100 hp... these cars make 100-200hp more in mild states of tune.

LOL, thaks kind of rude but ok everyone has a right to their opinion. I've actually talked to the guy btw and thats not the impression I got from him... He has also done some of the testing you were asking for regarding the header vs manifold thing.

Yes in an attempt to be brief I kept it simple and to the point but I dont think I over-simplified as far as explaining the basic idea goes.

I'm not talking about helmholtz or resonance tuning, I'm effectively talking about reducing back pressure... on most applications this will increase power across the board. One could try to maximize the lenght of the tubes beyond what I mentioned above as long as it was done equally to each tube and effectively get some resonance tuning but thats beside the point.

Maybe, why dont you ask him... Monty
I do know that he based his setup on cars like John Meaneys Vette and John talks about this briefly in his article to popular hot rodding... I don't have the article in front of me but he says something to the effect of 'are you paying attention here' when he talks about it. Perhaps someone reading this can quote from the article for us.

I think the theory is sound...
All internal combustion engines have exhaust pulses turbo or not.
Multiple exhaust pulses arriving at a collector at the same time makes back pressure.
Back pressure causes a loss of power because the engine has a harder time moving inert gases out on the exhaust stroke and left over inert gasses in the combustion chamber take up space that air/fuel could be using.
Reducing back pressure makes power in most cases
Timing the exhaust pulses so multiple pulses do not arrive at the collector at the same time will reduce back pressure
By adjusting the length of header tubes, you can time pulses to arrive at the collector one at a time rather than all at the same time.

Now this should help pretty much any motor make more power over a manifold but on a turbo motor, there is an added benefit...
When two pulses collide in the collector, they loose energy... by allowing the pulses to make it to the turbine without loosing this energy the turbo will spool faster.

 

Well not to stick my nose in but...........
On a normally asperated motor yes I agree 100%
On a turbo motor I disagree 75%
The reason being the exaust side on a turbo motor is ALWAYS experiancing a back pressure of 30 lbs or more(with wastegate closed part throttle). So from the second that exaust valve opens it is seeing back pressure, which voids any application of normally asperated theory.
Look at the gasses as small blocks getting shoved out one at a time. In a normally asperated motor these blocks need to be timed to avoid a "collision". But in a turbo motor these blocks are crashing into other blocks looking for a way out and coming back up the tubes (backpressure)

The reason to use "Headers" Over manifolds on ANY application is the increased volume of flow. Which leads to a increase of H.P.

The reason you don't see 1200 hp manifold motors, is because they can't flow 1200 H.P. Of spent exaust gasses.

I believe equal length turbo headers would be a great waste of time and money. Any difference in H.P. would only vary slightly on the matter of a few horses at most though this may even be lost by the cooling of the gas in the longer tubes.

But hey if you want them I'll make you a set.

 

Any car with a cat or muffler has backpressure in the exhaust... this is pressure above atmospheric, not above the pressure inside the combustion chamber when the exhaust valve opens. If what you say were true, exhaust gasses would never leave the combustion chamber. Its pressure differential that moves the gasses out. But you are sort of in the right direction about the it not being as helpful as in a naturally asperated setup because the higher the backpressure will limit gains. But this design is all about reducing backpressure.

Quote:

Look at the gasses as small blocks getting shoved out one at a time. In a normally asperated motor these blocks need to be timed to avoid a "collision". But in a turbo motor these blocks are crashing into other blocks looking for a way out and coming back up the tubes (backpressure)

What you just described is reversion.

Quote:

The reason to use "Headers" Over manifolds on ANY application is the increased volume of flow. Which leads to a increase of H.P.

Yep, certainly one of the key reasons but I'd argue that individual primary tubes rather than a log type setup is just as important a reason among other things.

Quote:

I believe equal length turbo headers would be a great waste of time and money. Any difference in H.P. would only vary slightly on the matter of a few horses at most though this may even be lost by the cooling of the gas in the longer tubes.

You realize were not talking about equal length headers here right?

Quote:

But hey if you want them I'll make you a set.

lol, thanks but no thanks. I saw an image of your work from another board... looks pretty good. For me this car hobby is about doing it myself... pay someone to work on my car???? Sacrilege

 

.