turbo manifold advice

[greek]

so,here is the story:
i have a friend with a turbo mazda, i want to build a manifold for him as a present
i am reading about manifold the last 2 weeks & i believe that i can built one,i'll forget stainless (expensive,lack of 2nd bottle & regulator for back purging wraps & cracks)so it will be mild steel flanges & steam pipe
I need advise at flange thickness is 10mm thick enough or should i go thicker?
can i weld it without back purging or i will have problems with the inside of pipe?
what best welding method will be?
can weld-els be shaped with a press? the exhaust ports are oval shaped,do i need bracing to hold the turbo to help the manifold not to crack?
Finally will it hold?keep in mind that the cast chinese one he is using is repaird twice and is cracket at the moment
i know that some people in this forum have tons of experience with this so please give me some help

 

[FranV8]

This is a job I'd like to tackle sometime, for a Mazda too.

Anyway, thoughts from an armchair engineer:

I don't understand why people insist on a second cylinder for back purging, why not just use a y splitter manifold and a pair of argon flowmeters?

Mild steel using steam pipe seems to be accepted as more durable, and perhaps better suited to turbo manifolds than stainless, yes you could do a super shiny long equal length manifold, but there seems to be an accepted wisdom that even the gash looking log type manifolds perform very adequately. They must do, many manufacturers do use this style from the factory.

There is an American site where a workshop guy advertises making manifolds for unpopular models, he gives quite a few tips away.

Process wise, in MS you've got a few options, although I suspect TIG remains the generally accepted way, perhaps because you can get into more nooks and crannies, and have good control to get good wells.

Might be good to mount the flanges to thick plate to minimise distortion whilst welding.

Have fun!

 

[greek]

i realy dont insist in 2nd cylinder but i also dont have 2 flowmeters i will tig weld it if someone answer my question if mild still need buck purging as for the tipe of manifold i've seen dynoruns of both types live(some are on u tube as well)& equal length manifolds perform better

 

[pedrobedro]

You don't need to back purge mild steel.

 

[FranV8]

Looks like in general it is acceptable to not purcge mild steel

http://www.ukwelder.com/forum/index.php?showtopic=9717&st=20

But it's better. Personally I think I'd be happy doing it without purge, but until I try it I won't know.

Given that the problems of not purging could be slag on the inside, that might loosen at some inconvenient time and go through the nice shiny turbine wheel, i think I'd be inclined to build up the tubes in such a way that I could clean up the last weld reaching in with a die grinder.

Remember that if you go for equal length, in steam pipe, your manifold is going to end up weighing a load more.

You don't say which Mazda it is, but I take it as it is a Chinese manifold it's an MX5 - how many horses are you aiming for and on what engine? If it is the MX5/Miata/Eunos, one of the key qualities is the handling and balance, which will get pushed further forward with all the extra components (manifold, turbo, intercooler, associated pipework).

Sounds like good fun!

 

[greek]

the car is not mine it a mx5 1800 with forget internals & a 2871 turbo that push 340 whp at the moment,its a frients car and the manifold will be a gift for him.
still no one gave me an advice about flange thickness
i have to start from there i beleive

 

[piman]

Hello Greek,

go by the flange thickness of the original cast manifold, if you match that it will be stronger than the original, I would guess you could go 20% thinner but as the studs are suited to the cast one, just match the thickness.

Alec

 

[Pete.]

Use a 10 or 12mm flange. They tend to bend when you weld the tubes all to one side.

 

[FranV8]

340 whp - so add 15% transmission losses then that's 391 bhp @ the flywheel. That's pretty serious out of an 1800.

I understand now why the optimum manifold design is so key to you. Is this car used for drag racing?

 

[NTwoO]

I don't understand why people insist on a second cylinder for back purging, why not just use a y splitter manifold and a pair of argon flowmeters?

I see the thread also mentions a TIG welder, but if he now has a MIG set, then it'll require a second bottle for stainless irrespective of the purge solution chosen.

 

[greek]

no it's used on road & some track events its not an everyday car by nature anyway,thanks for the flange recomedation pete so it will be 14mm and after the mill it will be 12 hopefully, ntw00i i have a tig & a small mig that i think is not capable to weld more than 3mm only 90 amps

 

[hotrodder]

Two part article covering the making of a mild steel turbo manifold here... http://autospeed.com/cms/search/index.html?keywords=manifolds&x=0&y=0. There's a fairly recent thread with links to a buch of vids on header construction somewhere on the forum too and a bunch of other relevant threads

The oval ports... Chances are the port circumference will be different to commerically available pipe diameters and bends don't squash too well 'cause the radius of the bend alters the way the metal wants to move. Various solutions from the soul destroying (die grinding the transition via the gasket flange) to the more elegant like shaping/piecutting a short section of straight tube or using a short squashed cone, example here

A manifold needs to be able to expand/contract as it's heat cycled and more often than not bracing the flanges to each other or gusseting bits of it do more harm than good. The way to properly support a big turbo is with something akin to a turbo damper (turbo era F1 used simlar sets ups but again talking massive turbo and/or long branch manifolds)

Gasket flange... obviously a thicker flange allows for a heavier cut to correct distortion but you may then need oversized bolt holes as they'll move with the flange (take a look at the cheapnese ****e). If the flange is 'preset' distortion will be less of an issue as it'll pull the flange back- for example can tack/bolt the flange to a chunk of RSJ etc with the centre shimmed up if butting the primaries upto it and using fillet welds. Another method is to make the flange so that the primaries pass through (stopping short of the gasket face) and weld them from the engine side. The welds are closer to the flanges neutral axis so it'll pull less, light tickle on a large belt sander should be all that's needed.

The purging thing... stainless needs a purge or the chromium oxidises heavily during welding leaving less available to keep it stainless. Talking light scaling with mild steel which brushes/falls off easily. Use whichever process you're most comfortable with, pros and cons to each of 'em, TIG usually ticks more boxes for this sort of job with thin wall and/or stainless but not so much with heavy wall mild steel.

 

[Big Mick]

You need this book!
http://www.amazon.co.uk/Maximum-Boost-Turbocharger-Engineering-Performance/dp/0837601606

Cheers, Mick.

 

[greek]

thanks very much guys the flanges will be ready on friday so i'll update with fotos

 

[greek]

after almost a month we have an update