grabercars.com

Posted: Sun Sep 03, 2006 1:11 pm Post subject: Turbo header design

When desgning a header for a turbocharger, which of these 2 aspects does anyone belive is more important:

1) Equal length runners.
2) Minimal header volume

I know it's imortant to keep header volume down to reduce lag as much as possible, but also it's helpful to have even backpressure and even wave pressure to the turbo...

As a project I'm doing before I actualy start building my own car from scratch, I'm twin turboing a chevy 350, but I can't decide on my header design weather to go out of my way to ensure equal length runners (which will have a larger volume since the center 2 runners will be longer than they would be if I just turn them straght into the turbo in a central location)... or to go with just a set of smooth bends directly into the turbo (center 2 would be equal and outter 2 would be equal, but not all equal to each other)

So, I can't decide... engine bay space is NOT a problem, I could almost put 2 of these engines were this one is going Smile

Any thoughts would be appreciated.

Thanks,
--Ash

Posted: Tue Sep 05, 2006 9:11 am Post subject: Re:Turbo header design

The ideal would be to have both, however I've seen both designs used. For example, my turbo diesel has a header that just dumps exhaust gasses into the turbine, with the longest runner being maybe 6 inches. However, having longer, equal length runners is also a way to go, but does not diminish driveability. Have a look at the location on the Audi R8 racecar.

Post edited by: andyk, at: 2006/09/05 11:11
_________________
1997 Jetta GT
1997 Jetta TD (donor car)
McGill University FSAE 2003-2005
Concordia University FSAE 2010-

Posted: Tue Sep 05, 2006 6:36 pm Post subject: Re:Turbo header design

Well, think of the dynamics of the situation. Low runner length promotes speed... thus greater spool up efficiency. The contrast, equal length runners, creates a scenerio where each cylinder has the same pulse seperation during the exhaust stroke. The pulse seperation would result in higher HP numbers due to over-all efficiency, but at a higher spool time.

One way to do both is to use a divided turbo flange...
http://www.ssautochrome.com/level.itml/icOid/images/13B-TB_001_c.jpg

You can keep the pulses seperated with a short un-equal length runner Evil or Very Mad

_________________
"In slow, out fast
In fast, out backwards"

Posted: Tue Sep 26, 2006 10:24 am Post subject: Re:Turbo header design

I ponder this with my V12 twin turbos header design, even though I have no experience in header design I am currently unconvinced if it is going to make such a huge difference on a car running under 15psi?
It seems that header design is much less critical with turbos than for NA.

I'll be very interested in what you decide and how you come about your decision on the design.

Posted: Tue Sep 26, 2006 10:27 am Post subject: Re:Turbo header design

One question, if you keep the runners as short as possible on a turbo manifold, can you get it horribly wrong?

Posted: Tue Sep 26, 2006 1:06 pm Post subject: Re:Turbo header design

I don't think so... again, using the example of the turbo diesel in my Jetta. Exhaust manifold dumps it right into the turbo, while fresh air is just dumped into the intake manifold.

There must be some good logic to keeping the runner short, though... VAG has done it with the engine in the Veyron as well. http://www.seriouswheels.com/pics-abc/Bugatti-EB-16-4-Veyron-engine-study-1280x960.jpg
_________________
1997 Jetta GT
1997 Jetta TD (donor car)
McGill University FSAE 2003-2005
Concordia University FSAE 2010-

Posted: Wed Sep 27, 2006 5:23 am Post subject: Re:Turbo header design

Almost all OEM turb headers are extremely short "log" type.
not equal length at all.
I'll be running right at 15-16psi On this setup, so perhaps equal length doesn't matter... Nelson engines offers several twin turbo crate engines and thier designs are non-equal length runners, but thier design is primarily aimed at saving space in camaros and etc... were as mine is going into a 1969 Impala, so space is a-plenty.

I've also seen several high HP 4cyl turbo setups use very LONG headers (around 24" runners). I'm not sure what the advantage of this is other than to have some torque while waiting for the massive lag that a 500hp turbo ~2.0L must have Smile

So, I've decided my design objectives will be: minimal piping curves, short runner length, and access to spark plugs (both for hands, and for plug wires... plug wires +2000F exhaust != good spark for long Razz

)

But, after looking at a plethora of designs I think the reality is that when your cramming 15-16psi into a 5.9L v8, it doesn't really matter how the manifold is designed as long as it fits into the car. Smile

Posted: Mon Oct 30, 2006 2:08 am Post subject:

OK well thats decided it for me then..... I'll just possition the Turbo as close to the engine as my chassis will allow and make the headers fit with the least amount of bends and length.

I plan to use thick mild steel rather than the usual stainless... anyone got any thought on the material to use?

Posted: Tue Oct 31, 2006 8:13 am Post subject:

I have a dumb question on this subject.
I’ve read articles on exhaust header design (especially relating to F1 & racing cars). It seems that ALOT of money, time, & effort is spent on handmade Inconel exhaust systems with a focus on smooth flow, matched runner lengths, and lightweight, etc.
In this age of CAD/CAM/CFD and 5/6 axis machine centers, I’m curious as to why these systems are not just machined from billet stock or castings? It seems a lightweight alloy could be used to create a 2 or 3 piece design that would allow for a very efficient, compact, specific tailored design. This method would even allow for performance improvements with more efficient cooling (with water jackets, etc.).

Posted: Wed Nov 01, 2006 10:40 am Post subject:

I don't see how you would machine something like this from a solid chunk?
http://www.stikiller.com/productimages/t_16697.jpg

Or etc.
The bends are what make the CNCing impratical I think.

Even a stock log manifold would be hard to CNC I think :/

--Ash