Calculating Turbo header / manifold dimensions for best performance

Hi All,

I am now at the point of designing (or rather specifying) the dimensions of the exhaust manifold / headers for my project, a turbocharged K24 going into a lotus Exige.

I have done much research into this subject and I know how the wastegate take-off needs to be made for best boost control & that I will be aiming for an equal length design...however I am really really struggling to find good firm info / calculations on ideal runner diameter and length.

If I look at any fully developed race car (WRC, F1, turbocharged drag) engine, they all appear to utilise "tuned length" manifolds, so I cannot agree with the "internet experts" opinion that a turbo manifold doesn't matter, or that you just try to make it as short as possible - if the pros are not doing this, it's quite obviously NOT correct.

Does anyone have any formulas or comments on how to arrive at the best length?

I have tried programs like pipemax but this sadly does not cater for turbocharged engines (it does however come out with a 4-1 primary length of around 28" for my engine and cam spec), I know the exhaust cam profile must have an impact on these calculations as that will dictate the best possible pulse "tuning".

Very interested to hear anyone's views on the subject, I'm at the point where my fabricator is waiting to know the spec, so keen to work this out sooner rather than later.

Cheers,

Adam

Hi Adam,

Exhaust manifold design depends on vehicle application.

For example: Mercedes F1 in the first V6 Turbo season had a log-type manifold, which might seem odd but they did dominate from the get-go. This in comparison to Renault or Ferrari who chose to go with tuned length/equal length manifolds.

What isn't as common of knowledge is that Mere struggled with the thermal side of things, as the manifold type does allow one to present a lot of thermal energy as well as gas-kinetic energy to the turbine wheel, especially when used in conjunction with pulse converters (here-in lies the "secret" to using log-type manifolds, choices are not always packaging driven)

Now Merc F1 uses a tuned length design, which helps with the thermal management side of things.

In order to use PipeMax, as you're aware, you need quite some data of the cams, lobe separation angles, and so on... but people telling you it DOESN'T matter on a turbo car... I don't agree... the tuned length in order to tune the pulses which in turn aid VE matter LESS in comparison to a N/A engine. I would opt to reduce the effect of pulses on engine VE and would focus more on the reduction of boost-lag or boost-building side of things.

Things to consider are:

manifold volume

runner length (using 3rd harmonic, according to PipeMax's Larry Meax)

pulse converter type (Sulzer or Birmann types)

keeping sudden diameter changes to a minimum, as these create p-wave reflections

waste gate positioning (as you've stated) is very important

You might want to look at the old-spool F1 turbo-era for inspiration as well.

I re-emphasize: application is what should drive your choice.

style

street

drag

auto-X

time-attack

and so on.

I'm sorry I can't give you definitive numbers.

Just a representation what PipeMax did for a C32 Laurel with an L24

Header designed based on 3rd harmonics

same car (N/A)

no retune

same dyno

similar environment

+25 Nm (peak) difference between stock and PipeMax