Way back when, I had to change a muffler on the car I owned at that time. I never did a scientific measurement, but it seemed, after I changed the muffler, that the car was getting better mileage. I asked the car nut at work about this, and he said that if the innards of the muffler had collapsed, it would block the exhaust gases more than a muffler normally does and this would lower the efficiency of the engine. Said that those noisy mufflers they put on sports cars are there at least partially to up the performance of the engine (besides getting that sexy sound out of them, of course) and that a byproduct of this was greater mpg than would otherwise be the case.
Finally, I remember, vaguely, hearing about an electronic muffler being tested on trucks that worked by cancelling the sound of the engine so that it was almost silent, and that this resulted in increased mileage on the trucks.
So, exactly how much of this is true? And would using a high performance (and noisy) muffler result in any real savings on gas mileage?
Removing the muffler can result in worse, not better mileage. The muffler is tuned to make the motor run as it should and can provide a 'sucking’action which can improve cylinder filling.
The electronic muffler question I fail to understand. If it is what i am thinking it has nothing to do with the engine or its efficiency. Electronic noise cancelling works like this:
Imagine you are in a noisy environment. Now you put some ear mufflers on to blok the sound but you still get some. Now you put tiny speakers in there and you feed them a signal which is the exact opposite of what coming in so they cancel each other out. The mike is on the outside of the headset so it is all integrated.
This technique is commonly used in factories and other noisy environments. You hear music as well but it is amazing how well the noise cancelling part works.
I cannot see how this would have any relation to exhaust mufflers though…
I’ve also gotten better mileage since replacing my muffler. Went from 13 mpg to 25 mpg. I replaced the muffler and the straight pipe.
Electronic noise-reducing mufflers shouldn’t affect mpg.
An engine’s an air pump. Up to a point, the more air you let out, the more you can get in. Less restrictive exhaust = more efficient engine.
real world example: When I replaced the exhaust on the Jeep (stock 15-year-old single 2" with cat and restrictive muffler) with straight-through duals with glasspacks, mileage went from 10 to 12 mpg! 
Hey, 10 to 12 is 20%! (does that thing need a tune-up, or is it just really really old?)
racinchikki, what do you own? That’s a pretty amazing difference, especially just from the muffler.
The electronic thing I heard during one of the two previous oil crises we had back in the seventies. I can’t remember if it was a radio show or what. I surfed the web to see if there was any info at all about that twist, but there wasn’t, so there probably wasn’t anything to it.
'84 Jeep Grand Wagoneer, 360 cid carburetted V8. Slow (self destructs at 5500rpm, but can cruise at 100mph with no problem), inefficient, low-end torque monster (245 hp @ 3000rpm; 300+ lb-ft @ 1800 rpm :eek: )
Of course, it went up to about 13.5 mpg (damn good for that engine, especially with 165,000 miles on it) after I replaced the leaky fuel pump :eek:
[answering for the SO 'cause I’m here and she’s in bed mode] '88 Chevy Cavalier Z24. [/answering]
Thanks dear.
I replaced the straight pipe, too – it was busted up near the manifold. It wasn’t just the muffler. It’s still an amazing leap of differentness, but I’m not gonna argue. With gas prices nearing $2.00/gal, I’m just gonna sit here being glad I’m not getting 13mpg anymore…
Repeat after me: Air, or any gas, flowing through a pipe or passage, has mass and momentum.
Modern exhaust systems are not just tubes that keep the engine compartment from filling with smoke and monoxides. Each time an exhaust valve opens and belches a blast of hot gas into the exhaust pipe, a well designed system will use the slight negative pressure wave from the previous pulse- from an adjacent cylinder- to improve what’s known as “scavenging”. Meaning it helps draw more spent exhaust gasses from the cylinder, and move them down the pipe a little easier.
This reduces the amount by which the fresh, incoming fuel/air mixture is diluted (which produces more power) and also makes it easier for the exhaust gas to flow. (Again, more power.)
In a modern car’s system, the muffler is a part of the system. The muffler controls “backpressure”, as well as attenuating noise. Backpressure helps keep the scavenging effect from drawing TOO much gas through, since there’s a brief “overlap” period where both the intake and exhaust valves are open. If the scavenging effect is too great- or not properly controlled anyway- a portion of the fresh fuel-air mixture can be drawn through and out the exhaust.
Obviously, if it’s not getting burned in the combustion chamber, it’s being wasted- IE, you get less power per gallon.
There’s also a “reversion” effect- during the valve overlap, the suction formed when the intake charge begins, can try to draw exhaust back into the combustion chamber. This would, again, dilute the fuel/air mixture and thus reduce efficiency. The scavenging effect tends to reduce this, as does an exhaust system designed to promote better flow one direction, but not another. Some “anti-reversion” aftermarket headers have inner directional baffles at the flange which greatly reduces reversed flow without restricting the exhaust.
All these factors rely on a complete, properly-designed exhaust system. If the muffler plugs up, efficiency goes down. If the muffler “blows out” (all the internal baffles rust out and are blown out by exhaust) so that it’s just an empty can, efficiency goes down. If you yank the muffler off and slap on a glasspack or Cherry Bomb, chances are you lost a little power and cost yourself some mileage. If you saw off the catalytic converter- besides the illegality of doing so- chances are you lost some mileage.
That is, of course, assuming an otherwise fairly stock car or light truck. If you start playing with camshafts and stroker kits, you’re entering a whole new ball game.
But if you just slap on a K&N air filter and some so-called “platinum gap” spark plugs and think you need to “open up” the exhaust to help free up all that extra horsepower, chances are very, very good you’re actually making things worse rather than better.
Wow! You’ll have to explain some of that post, Doc Nickel.
What’s a Cherry Bomb? I’m not a hot rodder (to give you some idea, my current ride is a '91 Crown Vic LTD station wagon, not exactly a babe-mobile, but then I’m over 40 and getting gray, so I think I’m over the fully reclining bucket seat crushed red velour phase) so I’m not up on this kind of thing.
Your explanation made some good sense as I can actually see something like what you’re talking about when I watch the wood stove we have in action. It’s got a catalytic converter, and if you let the air out too quickly this won’t work and you end up wasting wood. On the other hand, you have to keep it clean so that it doesn’t get plugged up. So obviously the idea here is to keep the air flowing at the optimal rate for the appliance, somewhat like the idea in a car engine, according to your explanation.
Does this mean the OEM muffler is going to be the best in most cases? And what about the straight pipe? How much of a diff does that piece make? (See Racinchikki’s post above.)
Gunslinger, I can see how replacing a leaky fuel pump could increase your gas mileage. 
Okay, a “Cherry Bomb” is a brand name for another glasspack-type muffler. Glasspacks are short, tubular “mufflers” that are wide open, straight through. The “muffling” comes from the inner pipe being heavily perforated, and wrapped with fiberglass batting encased in a larger outer sleeve. The pressure of the exhaust forces some gas into the 'glass weave which muffles it a bit.
Restriction is low, but noise suppression is near-nonexistent. They’re more like “resonators” which simply change the tone of the exhaust note- say, to make it a deeper rumble rather than a tinny blaaaat.
They’re also short-lived. The 'glass burns up and blows out, then it’s just a perforated tube.
Your woodstove analogy isn’t too far off, but you have to remember that the engine exhaust ‘pulses’. There’s a “high” pressure wave immediately followed by a “low” pressure wave. The timing and duration of the pulses are, of course, dependent on the number of cylinders, the RPM and the design of the exhaust.
One of the primary benefits of a “header”, a fabricated exhaust manifold consisting of a long steel tube for each cylinder (or exhaust port) as opposed to a short cast-iron thing with only mildly-seperated inner passages, is that each individual cylinder gets far more benefit from the scavenging effect.
In fact, there’s a version of a header called a “Tri-Y”, which, on a V-8 engine, crosses some of the left-bank exhaust over to the right bank, and vice versa. This, because of the firing order of a V-8, optimizes the scavenging further still- each cylinder gets a strong “low” pressure signal from the cylinder immediately preceeding it.
The only drawback is that the crossover tubes make installation and fabrication far more difficult, and the power increase usually doesn’t make the extra hassle worthwhile. They work best at adding low-end torque primarily, so they’re saved for applications that need it.
In many cases, yes, the best replacement for a worn-out muffler is a stock OEM piece. Aftermarket pieces do tend to “flow” better, but as I explained above, more flow is not necessarily a good thing. The only time you should alter the exhaust system- at least on anything after the mid-eighties or so- is if you’ve made some fairly major modifications to the engine.
And, as I also said above, “major mods” does NOT mean a new “high flow” air filter and some fancy $8 sparkplugs.
By “major” I mean a new, more agressive camshaft profile, cylinder head porting, stroker cranks or larger-bore pistons, a larger carburetor and/or freer-flowing intake manifold (or larger injectors/different manifold for EFI applications, which themselves will require reprogramming the engine-management computer.)
Which reminds me of another point. With newer computer-controlled cars, the computer monitors- and adjusts performance accordingly- several engine sensors which usually include a catalyst temperature sensor, an exhaust O2 sensor, and others, which are dependent on certain exhaust system parameters. If you saw the muffler off, yes, the exhaust can flow through faster, but it also won’t heat up the catalyst sufficiently, which means the computer tries to alter the fuel/air mixture to compensate. Or the O2 sensor gives a false reading since it’s sensing more exhaust by volume than the sensor and computer are calibated for, so again the computer thinks something’s wrong, and again alters the fuel map (or even the spark maps) to compensate for a problem that doesn’t exist.
Or worse, the O2 sensor is reading a high oxygen level in the exhaust, but the catalyst temp sensor says it’s still too “cold”. In this case- and each car is different, mind you- the computer may think that something’s seriously wrong, and defaults to “limp home” mode. (which tends to be a fixed fuel/spark map that’s overly generous, and designed so that you can get off the freeway and to a service station, but it tends to be very rich- which reduces the chances of overheating- and have a late spark, which reduces the liklihood of detonation.) None of which, of course, are good for mileage and power.
Thank you, Doc. That was a four-star explanation.
Just to clean up a few problems, Doc, don’t mind me. The majority was very sound. 
A tri-Y is when you have two cylinders exhausting into a tube each, joined in a Y. The other two cylinders are set up the same way. The two pipes are then joined again into a final, larger pipe which is joined to the cat and then the muffler. This is done on either side. 4 exhaust tubes > 2 exhaust tubes > 1 exhaust tube. What you refer to is known as an X pipe, or an H pipe, dependant on the design. X pipes are proven to promote further scavenging and increase horsepower and tq, H pipes show no gains. Headers are generally an aftermarket item. X pipe crossovers have shown gains as much as 14 hp and 10 tq on an SBC. I have the article if you’d like a quote.
OEM mufflers are generally too distant from the manifolds or headers to scavenge very well, and are tuned mostly for noise output. Poor gas mileage solved by replacing the muffler is generally caused because the muffler is clogged badly enough that it backs up the exhaust and pressurizes it in the tubing which kills scavenging.
Quality aftermarket mufflers (Flowmaster et al) are designed so that they will increase HP and TQ, as well as incrementally increase gas mileage, but again, are somewhat ineffective unless they are in close proximity to the engine. They do, however, often flow much better than stock and have a much better note. Lesser backpressure will increase HP because of increased air speed, but hurt TQ because of lowered density.
Starting in (IIRC) 1998, the majority of vehicles has it’s primary 02 sensor heated so that there is not an ‘open loop’ running period required until the sensor heats up. The secondary sensor is behind the catalytic converter generally only used to judge the efficiency of the cat. It is normally not directly involved in engine management. Only OBDII (post-92 or so) vehicles have two O2 sensors. The LS1 (among others) has four.
The O2 sensor judges the air/fuel mixture in the engine, not velocity, optimally 14.7 to 1 air to fuel, by a voltage reading registered by oxygen content (reaction to the… erm, platinum? in the sensor) in the air. The preferred voltage is about .75-.78. Too much oxygen in the exhaust is a sign of a lean condition which can lead to detonation. Detonation triggers the knock sensor, which pulls timing, which hurts gas mileage and power output. Too little is a sign of an overly-rich mixture which will hurt gas mileage and power output.
Air is metered into the engine in two primary ways. MAP (manifold absolute pressure) and MAF (mass air flow) readings. MAP sensors read the density of the air, and MAF sensors read the speed of the air. Both methods give accurate measurements of oxygen content. MAF sensors are more likely to fail.
Exhaust is approx. 1550-1600 degrees (depending on a/f mixture) and will heat a exhaust-manifold mounted O2 sensor even if no exhaust save manifolds or headers are used.
Valve overlap helps increase the VE (volumetric efficiency) of an engine, but is not recommended for positive-manifold-pressure engines (e.g. turbo-, comprex-, super-, or pro- charged) because the positive air pressure provides enough force to ‘shove’ the exhausting gas from the piston and overlap would allow the intake gases to be exhausted by their own pressure.
BTW, anyone running platinum plugs and a K&N and fears that their stock exhaust ‘can’t take the extra horses’ is a riceboy.
If I got any of this wrong, please feel free to dispute it.
–Tim
No dispute, most is spot-on.
I’ll gladly confess a limited knowledge to more modern vehicles running the Onboard Diagnostics, what with having been raised on Quadrajets and HEIs and all. I don’t “dislike” engine management computers and whatnot, I’m just not as familiar with 'em.
To clarify your clarification, however, the O2 sensor can be confused by a higher exhaust velocity, (compared to RPM) since more gasses of a fixed oxygen content can be read as “more oxygen” when the computer makes an oxygen-vs.-RPM calculation.
It’s less important on older cars with relatively unsophisticated controls (but still electronically controlled) but an OBD-2 or OBD-3 can get a little flummoxed. (Not a “limp home” problem, but definitely a slightly-off mixture under regular driving conditions.)
Of course, with any OBD system, replacing the muffler with anything BUT a stock unit is illegal.
On the “Tri-Y”, that’s true that conventional off-the-shelf headers merely pair up two sets of exhausts on each side of a V-8, but it’s done more for ease of manufacturing rather than the optimum setup.
The best system would be to pair up sequential cylinders- in a typical V-8, that means that 1 and 8, 4 and 3, 6 and 5, and 7 & 2 are paired. However, due to the crossover nature, the additional header tube length actually needs to “skip” a cylinder so the previous pulse is “timed” correctly. IE, pairing 1 and 4, etc.
I recall Smokey experimented with it a couple of times- it’s best for torque and Smokey liked having lots of out-of-the-corners torque- but eventually dropped it because it was a fabrication and engine-maintenence nightmare for the fairly small benefit. He got a similar effect with more conventional headers by running a “flat” crank, but I digress.
However, you’re exactly right, the “X” pipe crossover adds much of the “true” Tri-Y benefits but with far less hassle.
An X improves scavenging since there’s something of a venturi principle going on, whereas an older type “H” pipe merely connects both sides of a dual system, which really just drops backpressure a little.
Though we get back into the original argument, where adding something drastic, like an X-over, changes how the exhaust works with the car… and vice-versa.
Standard disclaimer: It all depends on the car, the computer, and how you drive it, too.
Not quite true; if you use electronic noise cancellation, you remove physical noise cancellation from the design criteria, so you can design one for maximum efficiency/performance only without having to worry about any unacceptable noise it might make.