The value of back pressure in exhaust

[mnichols]

I've read on several forums that removing back pressure in your exhaust (removing your cat converter for instance) can reduce torque or even horsepower. I really don't understand how pressure in the exhaust manifolds can boost torque or power.

Many years ago I rode 2 cycle motorcycles, and in this configuration the use of expansion chambers (that sent a pressure wave back to the cylinder) to push the over charge in the exhaust back into the cyclinder makes sense.

In a 4 cycle engine this analogy doesn't work. So what is this all about, why does removing your cat converters hurt torque or power?

[javadog]

Originally Posted by mnichols

I've read on several forums that removing back pressure in your exhaust (removing your cat converter for instance) can reduce torque or even horsepower. I really don't understand how pressure in the exhaust manifolds can boost torque or power.

Many years ago I rode 2 cycle motorcycles, and in this configuration the use of expansion chambers (that sent a pressure wave back to the cylinder) to push the over charge in the exhaust back into the cyclinder makes sense.

In a 4 cycle engine this analogy doesn't work. So what is this all about, why does removing your cat converters hurt torque or power?

Actually, the basic principles are the same regarding pressure waves in exhausts, whether it for a two stroke or four stroke. Get hold of a copy of "Scientific Design of Intake and Exhaust Systems" by Phillip Smith, if you want to dig into the details.

Back pressure isn't what you want. Timing the pressure waves is what you are after. Changes to exhaust systems can result in changes to the number and timing of the pressure waves in a system and you can lose power at certain rpm ranges. It's not a result of the loss of "back pressure." Any change in section of a pipe will introduce a reflected wave. It's a long story.

JR

[peppy]

I thought that it had to to with cylinder scavaging.

There is a loss of low end torque, but gains can be had in the upper RPMs.
I know that there is much more to it than this.

[mnichols]

Thanks, I found a copy of Scientific Design of Intake and Exhaust Systems on Amazon, I ordered it and will do some research. Interesting subject.

[Hester]

Cool. Let us know what you figure out.

[dickepoohp]

Only anecdotal but: After several different attempts fiddling with the exhaust on an early 80 Honda CRX-SI, we finally just ran a straight pipe..no nuthin between the manifold and the tip. Greatly improved both power and fuel mileage (and being somewhat deaf, I LIKED the racket.)

Thinking we had found a wondrous and simple improvement idea, I did the same thing on an early 90s Mazda Miata...fearsome loud, but totally gutless...hard to pull away from a stop.
I don't KNOW how it affected top end as summer was too short to wait for it to get out of it's own way.
I can't contribute to the learned part of the discussion, only losing back pressure or whatever happened sans muffler/Cat ain't always an improvement.

[CharlesJones]

My understanding is that each pressure wave of exhaust gases (which you get for each firing of a cylinder) is followed by a pulse of low pressure. Torque can be improved by backpressure if it helps to time these pressure waves so that each low pressure pulse coincides which the next pressure wave coming from the firing of the next cylinder, thus setting up the scavengering effect.

I think for higher revs you need less backpressure to get this timing right, hence why an increase in backpressure generally helps torque rather than top end power. So as javadog says, its more to do with the timing of the pressure pulses down the exhaust rather than just backpressure in it's own right - it's just that backpressure influences this timing.

Something like that anyway.

[javadog]

Actually, what you want is a negative pressure wave to arrive at the exhaust valve of a given cylinder, at the point at which both the intake and exhaust valves are open, thus helping to draw gases through the cylinder. A positive wave arriving at that point would inhibit the exhaust flow and reduce the flow of fresh air/fuel into the cylinder.

Cam timing has a great effect on this. Cams with little overlap (modern cams can be like this) are not affected as much as the older cams with lots of overlap. Given that the exhaust length is fixed, as is the speed of the pulses, at different engine speeds the pulses will arrived at either helpful times to the valve opening/closing or at times that hurt the flow of gases. Some parts of the torque curve will be helped and others hurt. The idea is to design the exhaust system to add torque in the places you need it and to have the negative effects occur in the part of the torque curve that don't matter. In a racing engine, the negative effects would hopefully be below the normal operating rpm range of the engine.


JR

[Bill Verburg]

2 & 4 stroke engine exhaust works in the same way. The physical structures are different in that a 4 stroke has mechanically actuated valves and a 2 only has holes in its cylinder wall the are covered and uncovered to serve that function. IN both acoustic waves are reflected back to the exhaust port so that in a limited rev range there is a low pressure signal at the exhaust port which helps to scavenge exhaust byproducts from the combustion chamber. If the intake port is also open at that time fresh mixture is similarly drawn into the cc( and often right out the exhaust pipe).

For this to work the pipes need to be tuned to the same or at least close to equal length appropriate to the rest of the engine design and the width of the pipe has to be appropriate to the engine size and rev range. It's like tuning an organ. In addition the cams and intake need to be matched to the exhaust. A high speed set up won't work so well at low speeds and vice versa.

the point is to promote rapid non turbulent flow through the pipes and to have the low pressure signals at the appropriate port at the appropriate time

The issue is that w/ a muffled system the reflected signal is smushed out and doesn't work so well.

So w/ a muffled system the point is rapid non turbulent low w/ low back pressure. Back pressure is never good, it only saps power and torque, however it is a necessary evil in a muffled system.

To get rapid non turbulent flow the pipes still need to be sized correctly in length and width and attention to transition and merge details is still necessary

[richde]

Only a turbo setup always benefits from a lack of backpressure. (after the turbo, naturally)

Otherwise, it's dependent on cam profile.

[Bill Verburg]

Originally Posted by richde

Only a turbo setup always benefits from a lack of backpressure. (after the turbo, naturally)

Otherwise, it's dependent on cam profile.

Not true at all, each part works w/ all the other parts to eke the maximum from a given system. Back pressure like friction is a parasitic loss, necessary. you can't get rid of it but you can and do want to minimise it.

as w/ Goldilocks and the 3 bears there is a just right solution for any engine configuration. But the just right solution is still a compromise for low rpm, high rpm and noise.

The key factor in exhaust design especially in a muffled system is exhaust gas speed, you want to keep that as high as possible for as long as possible. That doesn't mean bigger pipes either. Too big can stall exhaust speed as badly as too small.

[Bill Verburg]

Here you can see a couple of dyno runs comparing some different exhausts, you can see the effects on a V8 but a flat 6 will be similarly affected

[Pele911]

Wow I need to rethink those Megaphones I just bought!

[Bill Verburg]

Originally Posted by jvreed

Wow I need to rethink those Megaphones I just bought!

reverse cone magaphones are a great tool for focusing the acoustic pulses at the ports, the down side, depending on the angle of the cones, is that they will often do that over a narrow rev range and of course they are loud. Cones of course are part of an unmuffled sytsem. Cones on a muffled system do not function in the same way

[Zeke]

Originally Posted by Bill Verburg

2 IN both acoustic waves are reflected back to the exhaust port so that in a limited rev range there is a low pressure signal at the exhaust port which helps to scavenge exhaust byproducts from the combustion chamber. If the intake port is also open at that time fresh mixture is similarly drawn into the cc( and often right out the exhaust pipe).

For this to work the pipes need to be tuned to the same or at least close to equal length appropriate to the rest of the engine design and the width of the pipe has to be appropriate to the engine size and rev range. It's like tuning an organ. In addition the cams and intake need to be matched to the exhaust. A high speed set up won't work so well at low speeds and vice versa.

A couple of the best points you will ever read. An early equal length HE or SSI style header is long enough to dump into a muffler that is almost like a collector tube on a race car. One could be brought to believe the Porsche engineers calculated the volume of the banana muffler carefully. However, the first of the 2 highlighted statements tells more: a one design banana can't serve a 2.0 carbed T to a 2.4 MFI S engine to max efficiency for each motor. But, if you remove the muffler completely, you will definitely lose torque, gas mileage and maybe some HP in certain ranges. Some type of collector and exit system will bring back some of the loss.

Finally, I don't think I've ever heard of an equal length header/collector system designed for low RPM operation, but it's entirely possible to design a system for any needs, but not all needs (as Bill pointed out). In the early days of enduro kart racing, we had an adjustable length expansion chamber that we manually shortened with a cable pull attached to the butterfly steering wheel. You could feel the sweet spot and keep pulling as you gained RPM down th straight until it was at its shortest length which should have been pretuned to the highest RPM for that track and gearing.
Going into a corner, you let the exhaust pressure blow the thing long again for the torque and you started all over with the pulling.

I believe these were outlawed many years ago, but for whatever reason, they disappeared. Now, you have variable length exhaust and intakes used today, especially in F1.

[Bill Verburg]

the MFI cars are a special case

all of the intake and exhaust components have to work together. MFI is a very inflexible system, it works great when all of the pieces related to intake exhaust and ignition are nominal(w/i design spec) but falls on its face when any thing goes out of spec. The only feedback loops w/ MFI are temperature, throttle and rev related. The muffler is one of the design components of MFI.

CIS, carb and EFI are far more flexible, each in it's own way and each w/ it's own limitations

The beauty of motorcycles, back in the old days, is that you only had 1 or 2 cylinders to deal w/ and parts were cheap and simple.
when I raced 2 strokes I had 2 sets of p/c w/ different port profiles and 3 different exhausts. Use was determined by the track.

4 stroke thumpers had 2 exhausts one for short one for longer tracks

same w/ the 4 stroke Triumph twins.

when you deal w/ multicylinder enfgines it gets very complicated very fast.

[356RS]

Great timing for this thread. I just started an attempt to build an muffler system for my outlaw MFI engine and Bill Verburg's knowledge, as well as the others on this subject is making me slow down and re-think my muffler project. I have a unique problem to solve because the MFI engine is in a 356. No muffler system can be purchased off the shelve to solve my problem. Limited space and the fact that I have never built a muffler system before is my challenge. The book by Phillip Smith is on order as suggested by JR. I'm all ears.

[JFairman]

Supertrap mufflers are cool because you can adjust the backpressure and noise level by adding or removing diffuser discs at the exit end of the muffler.
I've got an old BMW twin cylinder motorcycle with 2 of them and no crossover pipe.
I've experimented with the discs, and stacking more of them on to lower backpressure has a very noticeable drop in midrange torque while making the exhaust louder.
There is no exhaust scavenging effect on a another cylinder because there is no collector, so each cylinder has it's own pipe and muffler.

I understand cam timing and variable cam timing and it's effects, but would like to accurately know the theory behind backpressure and the torque curve too.
I guess it must come down to the speed of the pressure waves bouncing back and forth in the pipe in relation to RPMs.

[Bill Verburg]

Originally Posted by JFairman

There is no exhaust scavenging effect on a another cylinder because there is no collector, so each cylinder has it's own pipe and muffler.

That's not true, while the acoustic signals for one cylinder can be used to reinforce(or cancel) the signals at another cylinder that is not a requirement.

A thumper responds very well to changes in exhaust length, diameter and megaphone design.

cross overs are not necessary or even desireble on all engine configs. They are often used for noise suppression on twins even though they may not be desireable for performance.

[JFairman]

What I meant was the left cylinder can not have any exhaust scavenging effect on the right cylinder or the other way around because there is no physical connection between the two cylinders with the CC products/supertrap muffler exhaust configuration I have.