View Full Version : What benefit do equal length runners and backpressure have on a turbocharged engine??

2nd May 2002, 13:21
Hello Techies,
in light of the numerous headers threads that are bouncing about, I thought I would ask two questions on a proper BBS. Here goes.....

1/ Is there any performance benefit from running equal length runners/headers?

To my knowledge, very short runners/headers etc will consume less exhaust energy than the long equal length runners (not good for turbine efficiency). Equal length runners are best suited to naturally aspirated engines where you utilise the runner lengths and inner diameters to induce ramming/scavenging, so therefore tuned to give torque benefits at a desired rpm and cam phasing. The general rule is, long runners for low rpm, short runners for high rpm, therefore there must be minimal benefit on the Impreza using equal length runners, may be even bad due the extra energy consumed??

2/ Is a restriction/backpressure post turbo good for torque?

As far as backpressure is concerned, once the gases have met the turbocharger, the remaining exhaust energy is consumed and the gases from the turbine outlet wish to find the quickest exit. Surely any backpressure will hinder this!
If this theory is correct then surely a 3" exhaust system will be better than a 2.5" system.
I always assumed that a 2.5" system would generate higher gas velocities Vs a 3.0" system, this would give better torque but not be able to flow as much gases at higher rpm. However on reflection, once again...this probably only applies to naturally aspirated engines not on Turbocharged.

Your thoughts would be appreciated.....


[This message has been edited by Jacko (edited 02 May 2002).]

Adam M
2nd May 2002, 15:24
the layout of the impreza engine, means that if you want equal length they are all going to be long asyou cant avoid the longest pipe being the length it is.

equal length headers are as you say designed to give a resonance at a specific rpm point proportional to their length, at this point you get fantastic extraction fo exhaust gases from the cylinder, which will mean much higher volumetric efficiency at this point, and hence more torque.

I am in agreement with you in that on a turbo charged engine any back pressure in the exhaust whill reduce the pressure ratio across the impeller or exhaust wheel. This means the gas will have to fight to get out on the other side. I cannot see how this is a good thing with regard to torque.

Again with gas speeds I can see what you are saying, but I still feel that when on boost the most efficient exhaust extraction is going to generate the most efficient boost.

Since the turbo is a much bigger contributor to the power of the engine, than the extraction of exhaust gases, I would have thought mods that improve turbo efficiency would outweight the inherent losses that this design can generate.

3rd May 2002, 07:06
To my thinking, the biggest benefit from aftermarket (or even ported OE) headers is due to keeping as much energy in the gases arriving at the turbo as possible, rather than any scavenging effect.

Logic says the turbo then has more energy as an output.

It's also interesting to note that porting the OE headers changes the tradmark burble considerably - it becomes harder but more civilised somehow.....like the gases aren't fighting to get out. And if you look at what needs to be trimmed, it makes sense, lol!

Everything I've read suggests that less backpressure post turbo is a good thing as well - certainly refitting my OE backbox dampens performance considerably!

Richard (just ported headers http://bbs.22b.com/ubb/smile.gif)

3rd May 2002, 11:07
I am still not convinced on the equal length runner front! Longer runners will consume more exhaust energy so reducing the turbochargers efficiency therefore induce more lag.

However, longer / separate runners should in theory have a ramming effect dependant upon their length/diameter and the Scoobs cam phasing (this would be different between UK Vs WRX Vs STi cams, as I believe they have different lift, duration and IVO/IVC & EVO/EVC timing).

IMHO, for a much improved bottom end torque (low thermal inertia/lowest energy consumption) and marginal increase in mid and top end (better matching of ports/better flow), it would seem that one should get the OEM headers ported and alleviate any restrictions in the OEM sections of the manifold/headers and uppipe. Better still, fabricate new headers and integral uppipe based upon the OEM original pipe routing and ensure good porting/tighter tolerance and try not to induce a Vena Contracta condition.

In theory, the long equal length runners should improve bottom end torque due to long runners (also good for low rpm ramming on a naturally aspirated engine). I cannot help thinking that the energy that would be consumed within this mangle of runners would outweigh the initial tuning benefit and may in fact cause a potential loss in bottom end torque/spool up. However once turbo up and spinning they would come on song and improved torque benefit should be visible at a certain rpm/cam timing but then tail off as they became out off song/out of tune Vs rpm. Equal length runners obviously improve the volumetric efficiency at these points so therefore flowing more air requiring more fuel to compensate.

Backpressure.......NOT GOOD! Biggest that can be packaged post Turbo the better. Gas velocity speed principle does not apply post turbo. Does any body wish to swap their complete 3" system with my tiddly little complete 2.5" system
All this b@ll@x about backpressure is good for turbo's...pah! Give the turbo something to work against...pah! If the poor thing doesn't have enough to do already...

Horses for courses !


Edited again to add contradictory thoughts...

[This message has been edited by Jacko (edited 03 May 2002).]

Adam M
3rd May 2002, 14:32

you are talking about longer runners meaning equal length?????


there are two issues here.

1. the length of the headers.
2. the relative length of the headers.

in the case of the turbo charged engine, I think the relative length of the headers is not important.

With NA cars, if you tune headers to have the same length, then at specific frequencies the exhaust is practically sucked out of the cycylinder, meaning there is less resistance to the pistons and so the engine has less work to do expelling wasted gases. This equates to higher ve and hence higher torque at that rpm point.

If this peak has a low q factor, you get a laregr spread of torque gain centred around this resonance point specific to the length of the headers.

I have read before that this is much less relevant on turbo charged engines due to the increased post ignition pressure meaning extraction is less of an issue.

I suppose the same resonance point could mean increased trbo spooling at the specific rpom, but I reckon the inertia of the core would act as a buffer so you would not notice the difference.

All this talk is great in theory, but the only way to know for sure is through practice.

As said before, the mrt tubular headers (not equal length), required Mark A to add 25% more fuel comapred with the standard car, and it went like a rocket even though it had three fat bastards and me in it.

Tim has had great results from his headers even though they were fitted at the same time as a lot of other stuff so it was difficult to know what was attributed to what.

It didnt stop him getting 344bhp @ 1.25 bar.

Alan mcrae uses tubular headers too, and he managed a healthy 58bhp at a recent rollign road day.

There is no question, that headers do make an improvement to the cars, but I would do all the other add ons first (except WI).

5th May 2002, 18:42
Does the impreza use 'sieamese' exhaust ports? (or what you call it in english)
Doesn't this induce a lot of overspooling between gas-ports, so adressing this would be a good thing. My experience says that one should atleast separate them another 15 cm before merging the gasflows.

6th May 2002, 06:12

OE, they're separate for a little more than 15cm already.


6th May 2002, 16:38
Ok, thanks for pointing that out, i haven't looked at a Subaru head since my old XT http://bbs.22b.com/ubb/wink.gif

But, this equal length issue, does it really matter, we are dealing with a pulse-driven turbo or am i wrong again?