This question comes up a lot from new and old members alike. So I thought a thread dedicated to it would be a good idea.

Here is some explanation I borrowed on 12/3/03 from http://modernmusclecars.net/articles...tdynamics.html

by: ragtop
Contrary to common popular belief, a larger exhaust does not always result in better engine performance. There are situations where an engine could loose low-end torque from "too little" backpressure. I know it seems counter-intuitive to think that an engine actually needs a certain amount of backpressure but here's why. Most engines are set up from the factory for a certain level of backpressure. Changing the exhaust can create a situation where the cam has too much overlap for the RPM range it's being driven in. In that case, the incoming fuel/air will come in the intake valve, only to have part of it sucked straight out the exhaust valve without getting burned! Obviously, engine power will suffer if it doesn't get a full fuel charge to burn. Proper backpressure will prevent this. So will choosing a more appropriate cam for your RPM range, however.

There is another, more complex reason why 5" pipes on a normal sized naturally aspirated engine won't work. It isn't that the engine needs more backpressure, it's that static pressure is only half the equation. The equation for the "equivalent" pressure at the exhaust port is P-pv^2, the static pressure minus the exhaust density times the square of the exhaust velocity at the port. A 5" pipe may see a slight reduction in static pressure but will kill off the velocity making it harder to push the exhaust out. Some tuned header systems can make the exhaust velocity high enough that the engine effectively has a lower amount of backpressure than the atmospheric pressure! This is known as exhaust scavenging and is what separates good headers from bad ones.

In the case of a turbocharged car, everything is different. They can run a huge exhaust pipe, like the HKS 5" pipes and see a performance gain rather than loss. The reason is the turbine. The exhaust coming out of the cylinders only sees the velocity going into the turbine. The velocity drop across the turbine doesn’t effect flow. On a turbocharged engine, there is no need to worry about the exhaust velocity downstream of the turbine. The size of the header primary tubes and collector, or the exhaust manifold design, still plays a larger role in determining exhaust velocity, but the pipes downstream of the turbo are a lot less important.

As for the turbo itself, you want to maximize the pressure (and temperature) difference across the turbine for the highest efficiency. A low velocity of the gasses exiting the turbine won't make it any less efficient; in fact, it can theoretically improve the efficiency of the turbine. So the exhaust on a turbocharged car can be designed to minimize static pressure, without concern for the exhaust velocity.

So what does all of this mean? It means that while a 5" exhaust would be desirable for a 350 hp turbocharged Supra, it would be a very bad idea on a 350 hp naturally aspirated Camaro and an awful idea on a 1.8 liter naturally aspirated Honda.

-ragtop


Not my words... so ragtop at the address above are where the credit goes.

EDIT:
Some of what he wrote is a bit convoluted to understand thorougly. What he is really arguing for is that you need a high velocity of your exiting gases, that's the entire purpose of the exhaust system... to evacuate gases at a high velocity. Without that you lose power and low end torque. Static pressure is what he is referring to as back pressure, but that is kind of a mislable. What he is referring to is the labeling of pressures as they would be on a pitot tube when using the pitot tube for measuring velocity. In that case you have one pressure point that is considered to be "static" and that is what he is calling both the static pressure and the backpressure... but that pressure is entirely velocity dependent. The higher your velocity the higher static pressure you have on a pitot tube.

There is a new link further down in this thread with a clearer explanation of all this.

 

This is some additional information from one of my posts in the thread:
https://www.evolutionm.net/forums/sh...5&pagenumber=3

First of all understand a couple fun things.

Pipe flow doesn't behave like normal air flow... so there is more to be concerned with.

Turbulent flow actually results in a greater volumetric throughput than laminar flow even though it's coefficient of resistance from friction with the walls is higher. It has a greater "average" velocity profile. So... in theory you can actually make an argument that mandrel bent is bad. How so?? Because mandrel bent piping is smooth enough that laminar flow can result. This will theoretically reduce your overall flow rate compared to the exhaust shop "crinkle" bend approach. So why do all companies make everything mandrel bent. Can we say marketing?!? Also, there are some cases when the unit is properly tuned with the right diameter and so forth that yes indeed mandrel bent is better.

The key to power with engines is allowing for high rates of exhaust transfer while maintaining backpressures that work with the original tune of the engine. Therefore the "funnel" idea helps this in some ways. It makes it possible to have a high flow rate of exhaust while maintaining enough backpressure to keep the engine running well. A high torque engine is one with some seriously bottled exhaust, a high horsepower engine is a very high flowing exhaust. What we need is in between. Our engine's are pretty torquey for their size, so we can afford to dump torque in favor of hp, and still come out with more torque than most uh.. civic engines.

If you want to understand the funnel effect look up the concept of a C-D Nozzle (converging diverging nozzle). Essentially what happens is that the gases are compressed by the geometry of the "funnel" making for a greater density of air. The same amount of mass has to flow through the nozzle regardless, so velocities increase. Now when the nozzle expands again the speeds go right back to where they were before. Except in one special case, when mach 1 is achieved at the throat. then the gases actually accelerate further and exit the nozzle at very high speeds until the pressures collapse back to atmospheric creating a sonic boom. (But the boom isn't really a factor... as your air compressor nozzle does this every single time you use it to blow dust off the car... no broken windows though )

So although you can increase your immediate tip exit velocity of your exhaust gases, you still have not increased your mass flow. So that is where back pressure comes in. If you have no back pressure you achieve your highest mass flow possible. The air "falls" right out of the engine pretty much with nothing to stop it. This result in lower torque... not sure why at this point.. I'll poke around and see if I can figure it out.

So although "funnelling" your exhaust is all great and dandy for higher exit velocities, you've done nothing to improve your mass flow rate, the mass flow rate is going to be determined by the engine and the restrictions present. You have to always take things back to mass flow as volumetric flow is misleading. But also in this case if you take your higher exit velocity and multiply it by the exit area, you aren't going to be any better off than at your slower but larger area'd flow.

Make sense? Shoot with questions as need be.

Hope this helps people a little bit in their exhaust and header decisions!

EDIT:

Backpressure is a misnomer and incorrect. It is merely a termed used to help people "visualize" the concept of what is needed in an exhaust. To avoid the syphoning off of the incoming mixture, or leaving too many leftovers of the spent gases, you need to provide the proper high velocity flow of exhaust gas. This is accomplished with the proper size piping. If your piping is too big you actually reduce your exit velocity of your exhaust gases. Basically it works out to the fact that based on the CFM output of your engine, the pressure waves generated, etc you need to size your exhaust properly to create scavenging. If your exhaust is too big this effect does not take place, and you will lose low end torque. Hope that clears up some discrepancies.

 

This "You need back pressue" is a bunch of BS.. Let me prove your theorey right out of the myth it really is.

Reasons you don't want to restrict exhaust. Same car, dyno's were made with in minutes of each other just opened my dual cut-outs to run open headers.





Lets talk torque.. 10-24+ foot lbs of increase in a huge rpm band. Please show mw where I lost torque..





No loss in the HP area either.




My point is don't restrict the exhaust. Don't believe the whole "You need back pressure to create Torque." It is a myth if your exhaust is set up correctly. With a good long tube header set-up you will have great torque running open headers.


Prove me wrong with some Dyno's. I gave you proof!!

 

You are trying to compare apples and oragnes here. Small displacement engines often run a different cam grind that leads to different valve timing. Additionally, large displacement engines can sacrifice some mixture due to the shear volume involved. So while your SS will benefit from an open header, and may have a cam grind that does not have as much overlap, our little 4 bangers don't do well with open header configurations.. I'll poke around and see if I can find dynos to prove what I'm talking about. But it is a pretty commonly known thing for import small displacement engines that "backpressure" is necessary to avoid too much low end torque loss.

Additionally, if you LOOK CLOSELY at your dyno, in the 2000 to 2700 rpm range you LOST torque. Which is exactly what we are talking about. So your dyno charts prove our point as well since the entire time we have been talking about LOW END torque. And as I stated above, these effects are amplified due the size and possible timing effects present on something like our engine.

EDIT: I'm stupid.. I was looking at the wrong line. SS's graphs show the right thing.

Also there are two different meanings of "open" with a header. One refers to a header with no exhaust attached, and the other refers to a header with no extra components (aka crimps, catalytics etc) with an appropriate exhaust sized to match the engine. The latter is good.. the former is bad.

 

Would you mind posting the amount of overlap your cam grinds have in them. I'd be happy to share with you what mine has. And it is NOT a little. Also, when you post graphs proving me wrong, let me know the set-up. Intake/Cam (with profile, duration,lift LSA)/Head work/Header/Exhaust used in the test vrs open header, the whole 9 yards..

Look at the grapgh with both HP and TQ overlay on both pulls. There is no TQ loss any where.. The TQ by itself shows it only becausr the grapghs are'nt dead on RPM wise.

BTW, I changed the tuning on the last open header pull and there was a drop in timing from 29.5* to 27.5* as well as fatening up the AF from 12.8 to 12.2..

 

Well moving your AF to 12.2 from 12.8 is going to help you a reasonable amount... but that is beside the point.

I have not seen any cam grind numbers for the Lancer cams. We aren't lucky enough to have all that info just yet... seems like performance info and parts are slow to come by for lancers.

I now see looking at yours what i was posting before was under the impression that the blue line was the open and the red was the standard (that is the way it is in the very first graph)

So you are right in that your dynos are showing gains across the way.

Pull off your hood, pull off your header, then run and let me know how that dynos.

I'm at work.. so I have to let this go for a while. But I will post when I get home.

 

What good would that do?

Have you ever used a flow bench before? Flow a head with and with out a 3 inch pipe on the exhaust side, what flows better?

Collectors can make or break a header.
Using a stepped header can make a huge difference.
Tube length can make a large difference.

There are a lot of variables.

Your missing my point, A well set up exhaust open will provide more torque and HP across the board vrs closed. It has to be a properley set up header though.

 

That's exactly my point though.. you need an exhaust system to get better perfromance. I found more info on it... and also part of the reason I have put "backpressure" in quotes like that is because I knew it was not quite the right term. I didn't know what the proper one was though... now I do!

check out:
http://www.automotiveforums.com/vbulletin/t13199.html

And that will explain it.

Basically it boils down to the fact that on an engine as small as what is in a Lancer you need significant piping to generate the velocities necessary to get good performance. You can't go with just throwing on a header with no exhaust, or a header with 4" pipe attached to it. It'll never work because of the lack of "backpressure" or in all actuality a small enough area to exhaust momentum and velocity to help the engine performance. Read through the link, and it'll be more clear.

I've been under the impression the entire time that you are pushing the concept of a header with more or less no exhaust, aka short straight finishing pipe and that's it. Which won't work on a Lancer. A properly sized straight through open (aka no reversals, baffling, etc) exhaust is what is needed and is best. And what I am arguing for. But that requires smaller pipe than most people thing.. and bigger is not better.

 

I can agree with you that you need the proper length and size of pipe after the header, but that is not for back pressure. It is for exhaust scavenging. With out the correct exhaust scavanging, your exhaust system will not flow correctly and you will loose HP/TQ.

It's not Lancer specific either. It comes back to the general theory for all combustion engines. My point of posting what I did was to try and correct the many mis-informed people about backpressure vrs scavenging.

 

Then we are on the same page. I was calling it backpressure (which is improper, but not what I thought you were correcting) but it is actually scavenging. i also was under the impression you were advocating huge *** piping on a 2.0 L which would also be incorrect. Some cars need bigger piping than others etc, and it is dependent on cams and on displacement. So yes not Lancer specific. What works on your SS certainly wouldn't work on a Lancer other than the principle and physics that lead to it.

Now we should be seeing eye to eye I'd guess!

 

Correct.. I'm not saying to go put a 4" exhaust on a small displacement engine. Just saying you don't want back pressure. You want proper exhaust scavanging. This is 1 reason why I have some extra piping under my car after the cut outs.

 

I rather call for effective flow rate than back pressure.

When I think back pressure, I think bottleneck. In my case for my RIP LS coupe, Mitsu designed the stock exhaust this way. On the inlrt of the cat converter, it was 2 inch while the outlet was 1 7/8 inch. At the final muffler pipe, it was a whopping 1.75 inches!

Going WOT at stock, I peaked HP at 4400 rpm as expected, but I didn't realized power immediately dropped off at 5200 rpm. Part of it is conservative cam lobe from the factory, a bit from the stock header design, and the most came from the exhaust.

When I upgraded header w/ SRI, my HP peaked w/ stock exhaust at 5700-6200 rpm before immediately power drop off. This was now due to MAF, stock 210cc, and stock exhaust.

Then I upgraded to a 501 MAF off the 3g eclipse (RS/GS/GT - all share the same MAF) or previous gen or modern Galant, used a set of 240cc injectors from a 3g eclipse, and tuned them w/ AFC. Now, I can bounce fuel cut at 6750 rpm. My peaked HP was now 6500 rpm and didn't feel a drop off until 6600 rpm, beyond the 6500 redline.

My car was a monster off the line from 0-65 mph, but on the top end I still suffered from stock exhaust. Passing power from 65-100 mph was average at best. From 100 - 113 mph, starts to pulling strong again in the 4K rpm mark. From 115 - 120 mph, it was agonizingly slow.

That's why I don't believe in back pressure. One must find an effective flowrate. If they truly believe in bottleneck/back pressure, hack off the lancer stock exhaust and use the Mirage LS/DE exhaust (both have the same size). See your car start choking from not having effective flowrate. For an extreme case, weld a Civic DX or Geo Metro penny size muffler. Let's see TQ figures now.

As for TC, no DSM or Supra guys will put an extreme 5 inch diameter exhaust piping or muffler. They go at max at 3 inches. Otherwise, people are seeing the sleeve section of the muffler which can be huge.

 

to much back pressure obviously will kill tq numbers
but he is right
you do need some backpressure an a NA small displacment engine for the fact he stated above
if your cam overlaps closely your intake charge will partialy get sucked out of the exhaust
thus loosing the mix for combustion
there is another way of proving it
put on a huge exhaust and go try and pass inspection
chances are that your VOC's (volatile organic compounds) (unburned fuel) are off the chart

 

exactly, thats why you DONT see dual exhaust 4 cyl engines, only a FOOL would put dual exhaust on a 1.8-2.3L 4 cylinder engine.....

 

actually you can put a dual exhaust on a 4cylinder... as long as it is sized appropriately... it'd probably be like 2-1" pipes... but it coudl be done... just the size matters.