Header Fabrication

Ron Earp

Admin
I'm interested in fabricating a set of headers from scratch. Team Stangweks has had a couple of different sets of headers made for our Mustangs but so far they don't exactly meet our rather demanding specifications. If you want a job done right you mostly have to do it yourself.

Does anyone on the forum have one of these header kits with the plastic bits that allow you to mock up headers? Like what is shown here:

[ame=http://www.youtube.com/watch?v=qZasjB1kZfE]Project: BOSS302 Equal-length stepped dyno headers - YouTube[/ame]

The ICE Engine Works kit?

http://icengineworks.com/Catalog2012Ed2FullVersionLR.pdf
 
Ron, I have doen this many times, It's really great fun. All you need is exhaust port flanges, collectors and a bunch of u-bends.Tack weld using a mig. finnish with tig. Position collector where you want it and start laying in tubes. Each tube must only be welded at flange and collector,each tube gets longer as it heats up, so no welding tubes together! U-bens come in different radia.
Have fun,
Randy:thumbsup:
 

Howard Jones

Supporter
Ron, Here's my advice. Build mild steel. Way easier than stainless. You don't need stainless on a race car. It's not like you will ever ware them out.

You need a TIG machine. 100-150 amps is enough, you will be working under 80-90 anyway.

I assume you know about supper clean material and wire brush everything spotless.

Buy a set of flanges. 3/8 thick. Stay away from 1/4 or even 5/16.

Buy a set of collectors. These are as expensive as you desire, from less than 50 bucks for both to mega money. You decide. For a 302 I would buy 1 3/4 ID to 3 inch. Use slip on, you will need this feature for mockup reasons.

The primary length will be about 32-36 inches. You can do a lot of math but that's what you will end up with. Try for equal length but +/- 5% is pretty good.

Cut a length of welding filler to the proper primary length. Lets say 34 inches. Play with this length and try and get an idea where the primary's must go.

I would do a stepped set 1 5/8 ID to 1 3/4 ID. Buy heavier gauge steel tubing, it's easier to weld. Start with the first down tube from the flange. I think it is easier to start at the front of the motor farthest from the collector first. This is usually a 90. Add to the longer down leg to be total about 1/3 the full primary length. Make a little long to begin with, then slip the next piece (now 1 3/4) over the first piece but don't weld yet just tape in place for now. This allows you to rotate as necessary. Continue to form the 1st pipe back to the collector.

The primary's will end up two closest to the block and two outboard of the inside two. The front port starts inboard as does the last port at the back of the block. The center two will end up out board. I would do the inside two first and then fit the center two last. Forward one of these two first and then the back center one last.

You are going to make mistakes but the tubing pieces, buy a lot of 90s and a few 120s along with a few feet of the bigger diameter, are really cheep compared to paying someone to build you a set. Don't weld to the flanges until you are done with the primary's and only slip on the collector last to tack it in place. Use the slip feature of step primary's to fine tune the fit of each primary in relation to each other. Lots of masking tape before you start welding, then tack .....check fit....tack another .....check bla...bla..bla. you get the idea.

To weld the primary's into the flanges form then into a square with a hammer. Practice on a waste piece first. Port the flanges to fit the square pipe, the pipe to fit the head port, and when you are ready weld the pipe to the flange inside the port side. Dress up with a porting abrasive and a air tool.

You can do this, Hell I did. It will take you a few days work but you can save thousands of dollars and when people ask you who did your header you get to say.............I did! Then tell them you will make them a pair for $5K. Be prepared to have people say OK without so much as a blink.
 
Last edited:
Ron

I did this sort of work for a living many years ago,and produced many SS exhausts.
If you choose SS back gas the pipes with argon.
I have a T piece in my gas bottle line to the tig,I then put a ball valve for on/off.
I then have a needle valve to regulate the gas, in the end if the hose I have a 6" length of brake pipe tube up the end of the hose.
This is so you don't melt the rubber gas hose,you cap the ends with old welding gloves,you have the brake bundy tube shoved in one end.
You have the needle valve preset for a small amount of flow so when you turn the on/off ball valve on it makes it consistent and you don't have to guess the flow.
It will fill the inside of the tube with argon ,wait 5-10 seconds then weld.
This will give you no slag on the inside of the pipe.
Just have to make sure you turn off the ball valve so you don't waist gas.
Its real hard trying to get it back into the bottle if you don't.
I would not build a set of SS pipe without back gassing.
The inside will look like the outside and no clean up.
If you have not welded SS before you are looking for a copper color at your welds.
If it goes black you have burnt the chrome out of the SS, In some areas it is difficult not to burn,usually on inside V's.
I put an old pic up to show you weld color
A linisher with a face plate is worth its weight in gold to get zero gaps.
Use .8 SS mig wire on the tubes and 1.6 on the flanges,most tube work you can fuse.

Jim
 

Attachments

  • DSCF0126.JPG
    DSCF0126.JPG
    259.4 KB · Views: 880
Last edited:

Ron Earp

Admin
This is for the Ford 3.8L V6 we use in our SCCA Mustangs.

We've been working with Burnstainless, an outfit in Charlotte, and PipeMax to calculate the proper primaries, diameter, and all the other parameters. We use Burns merge collectors and other high quality bits in the construction. We do have custom sets of headers that perform very well, but we're looking for single digit percent performance improvements.

I'll read through the suggestions here but we're confident we need 1.5" primaries for about 18", then 1 5/8" for about 18", merge collectors 3-1, fire ring, transitioning out to 1 3/4"", then to 2", then a y- all the latter in a very short distance. Then to a single 3" exit. Our biggest issue is we don't have this optimal setup precisely (close), and we don't have equal length pipes.

Would you guys try to use one of these kits, or just get a band saw and a box of bends?

We've got MIG and TIG, decent units, and finally have acquired enough skill to use them effectively. Well, sort of. We'll manage.
 

Terry Oxandale

Skinny Man
I've built the last 3 sets I've used, and found it to be quite rewarding, especially if you're using pre-fabbed collectors (Burn's are great). I use hose clamps to hold the ends together, with a hole in the middle of the clamps, 180 degrees apart to tack the tube in place once I've found the perfect position, then on to the next section, and so on until you reach the collector. The tacks are easy to cut and reposition as you get a final read on where the collector will be precisely positioned (I've always had to deal with very confined quarters). I'd also suggest you use the Hooker alignment sleeves ( Hooker Headers Alignment Weld Sleeves - Free Shipping on All Orders @ JEGS ), these will provide the "rod" material for the tack, plus provide great alignment of the tube ends. Aligning the tubes as needed, marking the alignment with a Sharpie saves the position, and allows you to set it down and then put it back together for a tack as well. Put all the tubes together, weld the tacks where you won't be able to reach them as an assembly, then weld them to the head flange last (with the collector slipped in place. There will be some shifting and re-adjustment. As said in the F5000 string, "nothing just goes together".
 
Last edited:
Hello Ron

To reply at your question ;
To have a good result using the kit you mention, you absolutly need to have similar dia and radius steel bended items so in your case it's really a waste of money to source such tooling ( this one is more usable for headers produscers who weld only one sort of dia bends etc etc )
By the past I had to built similar stuff you want to do and was using only saw and as much bendings and skill as you can found !
At the beginning we where used ( just to have a rought idea and take some pics ) to do paterns with that thincarboard tube aluminised ( air ducts) that you can bend as you want doing proper curves from the head to the 3/1 ; then using foam can spray we filled in , let dry and these where agood "guide" to start final steel tubes .
It is advisable to use stainless because as you will not weld with argon outside plus inside normal steell will give you a strange corrosion effect inside if penetration of weld is not correct.
And also because you will have maybe sometime to untackweld and reweld
to get better bend position ;so this is easier with stainless as this material accept better to reach melt temp twice , on contrary normal steel become brtitle once tackwelded many times .
On the future too it will be easier to repare or modify any weld on a stainless
header than on a old and burn steel one

For Le mans cars we where at the end using Inconel wich is an expensive and rare higth level stainless austenetic material!! ( and some others too in Titanium!)

I did on the past more than 50 different snakes and one important point we found is to be very carefull in doing nicely and properly the 3 in 1 junction ; this is 10/100 the success for and header .
You need really to obtain that nice welded "star " on the inside junction to help various gaz coming from head exhaust to increase venturi effect to the final big tube
Apologise for my poor english ; Hope this helps:worried:
 

Ron Earp

Admin
Thanks for the welding tips guys, please keep it coming.

Q: How come you wouldn't want to expand the pipes to be a slip fit for each segment? Or is that too much work and not worth it? Seems like that could make construction a bit more forgiving for leaks and welding.
 
Thanks for the welding tips guys, please keep it coming.

Q: How come you wouldn't want to expand the pipes to be a slip fit for each segment? Or is that too much work and not worth it? Seems like that could make construction a bit more forgiving for leaks and welding.

It's better to weld each segment carefully dia. to dia. coaxial and this is so easy to do with no slip ?
In case you want to change OD from a segment to another it's just a matter to machine a piece of aluminum cone and extand tube dia to match the bigger one, before tackwelding .
 

Kelly

Lifetime Supporter
Does anyone on the forum have one of these header kits with the plastic bits that allow you to mock up headers? Like what is shown here:....The ICE Engine Works kit?

FWIW Ron, here was my last effort, a set of equal length stainless 180s for my Pantera. I used the ICE building blocks and I don’t think I could have come as close on equal length with all the steps and different radii with other methods. I would tell you that the ICE blocks are a bit fragile (and pricey) though. If you look at the post below, the routing, and triple steps I think you’ll see why I used them.


It starts here under posts by Panterror:


http://pantera.infopop.cc/eve/forums/a/tpc/f/5650045562/m/4850059725/p/4

The info on my build picks up about the 7th page of the thread.


I really like the Burns stuff. It’s very good quality but it’s very expensive. I sourced my slip merge collectors from a fella on the West coast that built systems for cup cars and they were simply works of art at half the Burn’s price. If you want his name I can dig it out for you. On the stainless U-bends, I sourced from http://www.stainlessheaders.com/mandrelbends and same story only much less than half price of Burns with a couple caveats because the comparison isn’t exactly apples to apples.


I bought 16 ga 304 instead of 18 ga or lighter. If you want to build in lighter gauge or 321 SS their selection will be much less than Burns. Their stainless was very good quality American milled DOM. It will however arrive full of mandrel lube whereas I believe Burns would be clean. For me it was an easy trade because a $2 tube brush and $4 worth of degreasing agent in a 5 gallon bucket made them look like they had been bright dipped in about 2 minutes of brushing and the price difference was very large.


I went with 16 Ga for several other reasons.


It’s explained in the build thread above but even for as much primary as there is in that set of 180s, it’s only 6 lbs more for 16 ga vs 18 ga and they are much stronger, durable, and yes, a bit easier to weld too. It’s a street car but I don’t think it would have otherwise swayed me.


I was building triple step headers and with 16 gauge, each successive OD nested in the next ID snuggly with about ¼” half lap, with the step occurring downstream side. With 18 gauge or less, there would have been weld some gap to fill at each joint. This made many of the pieces self fixturing to a certain extent. The other thing I found on this build was the use of HVAC foil tape was very helpful during mock up. I had enough of those ice blocks to do almost an entire bank if I pieced together certain runs. That way I could experiment with routing a bit without generating (too much) scrap.


If you go through the build thread, you’ll see I ended up welding each primary step, then doing some fine tuning of the joints, putting all three segments together for the final weld for each primary. I was please with the final result and the primaries all strike bullseyes in the slip merge pockets.


One thing on slip merge collectors, they’re great for assembly and in most cases mandatory, but if this is an EFI car, they can be a source of leak and O2 sensor error. If it’s a carb car, no worries.


Good luck.
Kelly
 
Last edited:

Ron Earp

Admin
Ron, Here's my advice. Build mild steel. Way easier than stainless. You don't need stainless on a race car. It's not like you will ever ware them out.

Our current headers on both Mustangs are mild steel. But, we did have a problem with them - both sets "drooped". They started life all tucked up nice under the car but after a few race weekends the headers had sagged in the middle and had noticeable bends in the primaries that allowed the collector area to drop over 1.5". By the end of the season both cars had collectors below the oil pan and tranny mount.

Our thoughts were that while the headers were obviously supported at the front, they were only supported at the end of the system where the pipe terminates just in front of the rear axle. With no support in the middle they were able to droop and sag.

We took the pipes off the cars, put them on our spare motor, then heated them up and bent them back into position. And we fabricated supports for the system in the middle to avoid a repeat of this situation.

Question/observation - I don't think it would make any difference if we'd made these out of stainless since melting points of the two are similar. Is there any difference in "plasticity" of stainless steel when hot? I don't know what the proper word is for what I'm trying to express.
 

Kelly

Lifetime Supporter
Question/observation - I don't think it would make any difference if we'd made these out of stainless since melting points of the two are similar. Is there any difference in "plasticity" of stainless steel when hot? I don't know what the proper word is for what I'm trying to express.

It most likely will make a significant difference. In general, stainless steels will have much better high temp properties. It’s not just a melting point comparison. At some point, all metals that have elastic properties will behave more plastically as temperature rises and exhibit diminished mechanical properties. Creep is a typical issue for plastic behaving materials….gradual plastic deformation over time when subjected to stress, usually sheer for most creep susceptibility. When something becomes plastic it has no shape memory. For mild steels, at 900F and up properties start to fade and diminish rapidly from there. It’s not just ultimate strength but resistance to corrosion, fatigue, and general retention of properties for time at temperature and cycling and these factors can rapidly degrade lesser materials. Nickel is common alloying agent and in the extreme, Inconel and other high nickel alloys will immerge as the high temp metal of choice.

Intuitively it makes sense; by heating the steel, mechanical properties reduce, you can move the material easily and it stays put afterward. Hard to say whether it was creep or just plain old plastic deformation for your old exhaust. Always a good idea to support your exhaust system in a manner that supports it but doesn’t constrain movement in the direction of growth. For best practice look to the turbo guys. They want the exhaust entering the turbo to be as hot as possible and they come up with some ingenious ways to avoid hanging the weight of the turbos on the header.

Even 304 will likely only retain 1/3-1/4 its strength at your exhaust temperature but that may still be North of 25kpsi and perform much better than mild steel. 321 will be better but probably not needed. Inconel is commonly used by the turbo/pro-racing guys-very expensive. You’d be happy with 304. As mentioned earlier, you should use inert gas purge inside as it can make a big difference in the durability of the weld.

http://www.ssina.com/composition/temperature.html

Best,
K
 
One of the 'un-seen' advantages of stainless is that there should be virtually no corrosion buildup on the pipe interior over time, there have been many tests done where mild steel headers have shown large power drops due to corrosion, unleaded fuels have reduced this problem a little and you could get the system coated, but then you face the issue of difficulty to repair.
I think if you work out what it will cost to build & coat a Mild Steel setup you will be close to the SS price anyway.. and then dont forget the weight factor..its a race car, light is faster....although when you look at some drivers you wonder how!!:)
 

Ron Earp

Admin
We've heard about the issues with mild steel and buildup, as well as corrosion. However, over the last two years neither of these cars seem to have experienced any problems. Mine are ceramic coated inside and out, Jeffs are just painted black. Both are wrapped with header wrap. Maybe modern EFI and running a mixture that is more precisely controlled than days of old has something to do with it, I don't know.

Weight we're definitely worried about. So how thin a wall stainless could we use and still have good durability? And how thin a wall mild could we go?

We already use individual head flanges instead of one large head flange to save weight and help on install (VERY spatial challenged on the driver side). No leaking/sealing issues to date with over 120 track hours between the two cars.
 
Here is how I model an exhaust system prior to ordering bends and I have not come across an easier way. Please dont look at this as anything other than an alternative method and its not a guide to header design , the bench model is just an easy way of explaining the method. Cut out two manifold plates and two target plates for the collectors from 10mm ply. Drill the manifold plates to bolt to the head and dill a 3mm hole in the port center on one plate and 15mm od on the other, do the same to the target plates. Bolt the three mm holed plate to the head and bracket mount the target plate with the 3mm holes in place.It is no problem at all to achieve equal length headers, starting with the port that is furthest from the target plate fashion a piece of 3mm wire around whatever needs to be routed round bearing in mind all the bending needs to be done round a 7" diameter piece of ply, this will represent the 3.5" radius bends I use. Take that piece of wire and straighten it out and cut 7 more, fashion them in the holes until you are happy with positions but bear in mind that you may be using up to 1.75" od pipe so keep them well apart. Now if you believe all you read length of primaries is critical as is the position in the collector but I am not going there because its minefield of myths with little or no relevance to anything other than a 18000rpm F1 engine.




Once you have strangled the 3mm wire round everywhere you need just check that you have all the radius`s pulled correctly as you now need to swap port plates and transfer all the wire dimensions onto 15mm copper pipe, just happens the good old plumbers bender has a 3.5mm rad former. Its really easy to work and will give you a solid model with the correct radius bends.It really makes no difference how many joins you use because they very cheap.



If you need a better idea of exactly what it is going to look like just slip some foam pipe lagging on it. When you are all done you can just throw the whole lot at a decent tube bender who will just copy the models or dissect and measure your work then order. Either way its a very cheap and simple way of getting a precision exhaust system made up without struggling with steel tube.

Re collectors. These are the ones a friend of mine fabbed up for me.




 
Last edited:
If you use stainless you can purge the inside of the pipes with Argon and make it as pretty as the outside.
 
Back
Top