A little exhaust design indecision.
A little exhaust design indecision.
(OP)
I'm trying to design a new exhaust system for my Miata, and in this venture I’ve research a multitude of books but they all lead me down the road of build it and test it. I find myself wondering if there are any exhaust companies out there that TRULY spend a large amount of money on R&D to make a muffler with both superior deadening/attenuating qualities coupled with superior flow at low pressure gradients.
It seems illogical that any muffler i design could compare to a company that actually makes mufflers and has a large body of engineers working with them. I'm particularly interested in Flow master muffler systems because of their claim to using systems of chambers that create excellent scavenging effects yet flow the lowest of most main stream muffler companies. Its interesting however when I email asking for any technical specifications on they’re product they never respond, it adds in an opposite effect and leaves me asking myself if any aftermarket muffler company really puts a lot of R&D into the mufflers they just put together something that half way works and sell that.
Some of the ideas I had for increasing muffler flow performance if I do find out its not worth while to go with an aftermarket muffler are, creating trumpets on the wave length tuned pipes in the muffler chamber systems to increase discharge coefficient, creating ideal taper angles into and out of the muffler for better transition back into laminar flow, and insulating the muffler from ambient temperatures so that heat loss is minimized in order to keep heat and velocity high. Can anyone tell me if my ideas are good leads into a better muffler design?
I guess responding to either question or both, I’m struggling with what I should do, go with someone else’s design or find that maybe my thoughts are original or not cost effective for the major manufacturer and I could see increased attenuating qualities and increased flow.
It seems illogical that any muffler i design could compare to a company that actually makes mufflers and has a large body of engineers working with them. I'm particularly interested in Flow master muffler systems because of their claim to using systems of chambers that create excellent scavenging effects yet flow the lowest of most main stream muffler companies. Its interesting however when I email asking for any technical specifications on they’re product they never respond, it adds in an opposite effect and leaves me asking myself if any aftermarket muffler company really puts a lot of R&D into the mufflers they just put together something that half way works and sell that.
Some of the ideas I had for increasing muffler flow performance if I do find out its not worth while to go with an aftermarket muffler are, creating trumpets on the wave length tuned pipes in the muffler chamber systems to increase discharge coefficient, creating ideal taper angles into and out of the muffler for better transition back into laminar flow, and insulating the muffler from ambient temperatures so that heat loss is minimized in order to keep heat and velocity high. Can anyone tell me if my ideas are good leads into a better muffler design?
I guess responding to either question or both, I’m struggling with what I should do, go with someone else’s design or find that maybe my thoughts are original or not cost effective for the major manufacturer and I could see increased attenuating qualities and increased flow.





RE: A little exhaust design indecision.
So what do you wish to gain, and what are you prepared to sacrifice ?
For a road car, perhaps try to copy the original layout, using all the original factory mounting points and pipe lengths, but with slightly increased pipe diameter and use mandrel bends. If the original had resonators or expansion chambers fitted in unusual locations do the same. I would just scale everything up in flow area, at least as a starting point.
With a bit of luck it may flow much more freely with only a slight increase in overall noise level.
RE: A little exhaust design indecision.
Your ideas for reducing back pressure seem to be on the right track, although I don't understand why you want to keep temperatures high, unless you have a tailpipe turbo.
Bear in mind that back pressure does not affect most of the driving range at all, and, typically, doesn't even hurt full power a great deal.
Cheers
Greg Locock
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.
RE: A little exhaust design indecision.
Now, I'm a bit astonished by the 12 psi number, but guessing that it is a bit higher than your standard system, and your engine probably is no more than 200 hp, it seems likely that the best you could do with the factory cat is a 10 hp improvement. I think there are easier ways to get that sort of performance boost.
However, there is an easy way. Build a straight through pipe, put a butterfly valve at the tailpipe, and dyno the thing. Mess about with the valve setting and produce a graph of back pressure vs power.
Then post the results here.
Cheers
Greg Locock
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.
RE: A little exhaust design indecision.
But surely it would not help a traditional "sports" engine with generous valve timing, and where some attempt at exhaust assisted scavenging is being attempted.
While it may all be horribly illegal and antisocial these days, some of us find the new ways difficult to adapt to after a miss spent youth.
I have had some success with a rear mounted exhaust butterfly. This is lightly spring loaded with a hole drilled in one side only to unbalance the pressure. Under normal road load it tends to maintain a fairly small constant back pressure, and keeps the noise in.
At above around 90% throttle a cable mechanically forces the butterfly to the fully open position. It may not go much faster, but the sound is glorious.
RE: A little exhaust design indecision.
In racing, my passion, the higher the ultimate hp the engine makes, the greater the exhaust system mods effect overall performance.
To be PC these days, most tracks mandate some form of silencing device to hold noise levels to some arbitrary limit. None of the builders/owners/drivers that I know, care much for the idea save the rare time when a silencer actually improves output. Rare, but it still happens on occasion.
In my case, I am currently using the Flowmaster Hushpower II on both a 1380cc Mini and a 1594cc DOHC Lotus...both have several dyno sheets and generally all results tend to support Flowmaster's claims of very little hp loss (~2%). However, I have seen a drop in midrange torque output of aroung 5%. My tests were done on a Clayton chassis dyno back to back with an open exhaust. I have dyno sheets available for the last three or four 1380 tests from June and Sept. of this year, if your interested for comparative value.
If your going to do your own R&D, Flowmaster is a great place to start. It beats the heck out of starting from scratch as I did many years ago.
Rod
RE: A little exhaust design indecision.
RE: A little exhaust design indecision.
RE: A little exhaust design indecision.
Rod
RE: A little exhaust design indecision.
"I've just found a page that says that for a standard WRX turbo the exhaust +cat gives a back pressure of 12 psi at 225 hp. removal of the exhaust+cat gives another 20 hp.
Now, I'm a bit astonished by the 12 psi number, but guessing that it is a bit higher than your standard system,"
There is no way that is a correct backpressure. 12" of water maybe but even that would be high. Turbo engines do not like backpressure plus the turbo does a good job of converting impulse energy to boost so they are inherently quiet. Typical turbo backpressure would be about 5" water.
RE: A little exhaust design indecision.
I have only had a couple of turbo anythings and the only one that was dynoed in the 80's did not have backpressure anywhere near that number.
Rod
RE: A little exhaust design indecision.
Exception: Some recent Volvo engines specify 10 min, 25 max kPa ( 1.4 min, 3.6 max psi). I have asked; nobody outside of Sweden knows why a minimum is specified, or why it's so high.
I'm also astonished by the 12 psi number. That would be more credible, upstream of a turbo.
Mike Halloran
Pembroke Pines, FL, USA
RE: A little exhaust design indecision.
There is no magic muffler that can even out violent explosive pressure pulsations and have zero back pressure, think about it.
Some high frequency absorption is the best that can be hoped for in a straight through muffler. That will take off some of the "nasty edge" of the sound, and mellow it's character, but in reality the measured sound pressure level will not always fall as much as subjective hearing tests may suggest. Meeting a decibel limit is much more difficult.
RE: A little exhaust design indecision.
Rod
RE: A little exhaust design indecision.
Now Donaldson give a range of recommended pipe sizes for
various rated engine horsepower to give approximately 2psi.
75kW 2"
120kW 2.5"
165Kw 3"
230Kw 3.5"
375kW 4"
550kW 5"
635kW 6"
Truck and car systems might be roughly the same overall length, so the pressure drops may be comparable. These pipe sizes are well above what production cars use, and that should tell you that production cars are going to have much more than 2psi pressure drop.
Noise attenuation is an entirely different matter.
RE: A little exhaust design indecision.
Seems there might be a trading point between acceptable frequency /magnitude of the waves and pipe size/different mufflers needed to attenuate those frequencies... anyone concur?
RE: A little exhaust design indecision.
If I had to pick a number, I'd guess most production exhausts, with cats, have 4-6 psi back pressure at max power.
Cheers
Greg Locock
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.
RE: A little exhaust design indecision.
One is a steep acoustic shock wave, the other a low frequency high amplitude pressure wave.
Anything like tuned pipes or extractors that are going to maintain gas velocity, is going to be more noisy.
A plain pipe is a wonderful transmitter of sound. The old traditional ships bridge/engine room speaking tube is a good example.
Absorption mufflers can only attenuate the higher frequency components which changes the character of the sound, but does not reduce the overall measured sound power by very much.
To reduce a violently pulsing high pressure gas into smooth constant steady flow, there must be some restriction somewhere, there is no other way.
Any sort of pipe resonance will most likely increase the amplitude significantly at certain Rpm. Preventing drone at particular speeds is yet another problem to overcome.
None of this is simple or easy because it is not just one problem.
RE: A little exhaust design indecision.
RE: A little exhaust design indecision.
Cheers
Greg Locock
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.
RE: A little exhaust design indecision.
RE: A little exhaust design indecision.
RE: A little exhaust design indecision.
Cheers
Greg Locock
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.
RE: A little exhaust design indecision.
RE: A little exhaust design indecision.
RE: A little exhaust design indecision.
Eh? The speed of sound in the exhaust changes during warming, which has a pretty noticeable effect on the resonant frequencies. It's a pretty important thing to capture when trying to predict exhaust noise.
RE: A little exhaust design indecision.
Not true. Both parts of a turbo (compressor and turbine) are very inefficient at transmitting pressure waves. Bizzarely (in my mind) Some OEMs are actually dabbling with "sound pipes" these days. The basic idea is to route noise from the high-pressure side of a turbo (either intake or exhaust) to a site where it can be heard inside the car. From memory, BMW seem to like this idea at the moment.
RE: A little exhaust design indecision.
Is it the pressure wave that makes the sound or the mixing of extremely hot gas with cool atmosphere... I'd imagine a combination of both seeing how one determines the other PV=NRT buttt... we should be able to break down the ratio of which causes more.
RE: A little exhaust design indecision.
That's why a turbo's exhaust can be basically a low pass filter, the low frequency resonators aren't needed.
Cheers
Greg Locock
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.
RE: A little exhaust design indecision.
RE: A little exhaust design indecision.
running a series of experiments on a 2 litre propane powered 8 valve astra engine,that was dynoed at 135 bhp at 5500.this i equated to approx 150 on 97 octane petrol eg minus the oft quoted figure of 10 percent loss due to the propane displacing air and its lower energy per cc ,i used a exhaust system that was 2.5 inches dia,one straight through absorbtion type silencer 20 inches long and 6inches wide ,then a hedman turbo muffler 2.5 inch inlet .this ws then tested for back pressure .and at 6500 gave a max of 1.5 inches water.this was within acceptable limits.in my mind having donew considerable computer modeling and design work ,oh ic engines ,i se the air flow through the engine as bieng like a stream ,it flows in the air filter ,and out the exhaust .any restriction or back pressure is bad .any usable pulse tuning in the ex whether organ pipe pulse effect or inertia effect ,is a good thing if it creates a better cyl fill .but the average back pressure should be as low as poss .i wonder wether these ex presures of 12psi etc ,are pre turbine ?
thanks
robert
RE: A little exhaust design indecision.
RE: A little exhaust design indecision.
Using Donaldson's figure of 2psi for 120Kw with a 2.5" pipe, and your engine being around 100Kw, are you sure you dont mean 1.5psi not 1.5 inches of water ??
RE: A little exhaust design indecision.
RE: A little exhaust design indecision.
On a side note after reading Four stroke performance tunning A. Graham Bell cited some imperical evidence that alot of flow loss comes from poor housing desing, entry and exit angles that are too high, does anyone else back this statement?
RE: A little exhaust design indecision.
RE: A little exhaust design indecision.
Mike Halloran
Pembroke Pines, FL, USA
RE: A little exhaust design indecision.
RE: A little exhaust design indecision.
RE: A little exhaust design indecision.
My understanding is that at low throttle the spring keep the baffle "full closed" and like any regular muffler, and as throttle is increased, the increased flow would push the baffle open to reduce back pressure and increase performance.
I have heard of this being used on a small two stroke, but would the effect be the same on a 4 stroke?
RE: A little exhaust design indecision.
The justification is somewhat flakey (in my view). Manufacturers state that it's a fuel economy device - lower back pressure at higher speeds/loads means less fuel used.
RE: A little exhaust design indecision.
Now, the bizarre bit is that the external noise test is at full throttle. There is also a stationary noise test, but in my experience it is a formality.
Cheers
Greg Locock
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.
RE: A little exhaust design indecision.
True, but the entry conditions and test length (at least for the European pass-by tests) are such that the valve stays shut during the test.
RE: A little exhaust design indecision.
First, about backpressure: as was pointed out above, bacckpressure has little effect on engine performance except at high rpm and WOT operation, and even then it is dependent on a number called the "valve overlap factor" (VOF) which is defined as the area under the valve lift vs crankshaft rotation plot in the region of exhaust valve closing/intake valve opening. That is, the greater the valve overlap factor, the greater the engine's backpressure sensitivity, or loss of power with increasing backpressure.
I have stashed away somewhere a fascinating study done at Ford in the late '80's where power loss was plotted vs VOF for a variety of passenger car and light truck engines, done as part of the MN-12 (Thunderbird) program. It showed clearly that 4-valve engines (case in point was the 5.0L Porsche 928 S4 rated at 300HP) typically have very low valve overlap factors, compared to 2-valve engines, and very very little backpressure sensitivity.
I later confirmed this in testing done on the Lotus-designed ZR-1 Corvette's LT-5 engine, for which I designed a variable-backpressure muffler.
One other quick point at this time is that few if any aftermarket mufflers show any evidence whatsoever of attention to reducing noise at the driver's ear. You're right to note that the OEM's have vastly greater resources to devote to this task.
More to follow.
- R
RE: A little exhaust design indecision.
I'm not sure about other countries, but there's no noise test in our annual vehicle checks. Police can stop you if your vehicle is too (subjectively??) loud though?
RE: A little exhaust design indecision.
The late model production four valve engines generally have zero (or absolutely minimal) valve overlap. But if an aftermarket sports camshaft is subsequently fitted to one of these engines, there can then be significant valve overlap.
The exhaust reversion problem will be far more extreme than with a two valve head with similar valve timing because of the very large valve curtain area.
The variable area muffler is a clever idea, I have only read about them and seen pictures, but never actually tested one.
On my own turbocharged road car I have had some considerable success at reducing noise with an automatic exhaust butterfly located near the rear of the exhaust system. I developed this idea myself and am very satisfied with the results.
This consists of a very simple centre pivoted homemade butterfly valve fitted into the exhaust pipe just before where the pipe goes up over the rear axle.
It is very lightly spring loaded, and held closed against a mechanical stop. A 12mm hole is drilled on one side of the butterfly plate to unbalance the pressure. This causes the butterfly to open by itself from upstream pressure, and maintain a low but fairly constant exhaust back pressure.
At constant road speeds, the reduced pipe area at the butterfly vastly reduces the noise and any tendency to drone. Spring tension alters both the noise level and back pressure. It can be made as quiet or as noisy as you wish. About 0.25psi to 0.5 psi will work wonders.
A further refinement is to use a very long throttle cable extension to mechanically force the butterfly to the wide open position beyond some set accelerator pedal position.
An easy way to construct this is to fit a pair of welded bolt flanges to the exhaust pipe, these can then either be bolted directly together, or the experimental butterfly plate sandwiched between with gaskets. That makes it very easy to modify or remove.
The drilled hole in the butterfly plate needs to be about half way between the edge and centre of the plate. Using a simple flat, cut onto the edge of the plate to unbalance the pressure (instead of the hole) will not work, it causes the butterfly to flutter, accentuating noise. But with a hole located away from the edge, flow friction around both sharp edges of the butterfly seem to damp out any tendency towards flutter.
Although only made from mild steel, it has been completely reliable and because it is in constant motion and runs quite hot, corrosion or exhaust deposits do not seem to have caused any problems over four years of daily running. It has started to rattle slightly (at idle only) due to wear in the shaft. Fabricating another butterfly plate using better construction and materials is a job I will get around to one day.
RE: A little exhaust design indecision.
We took all the internal pipes for a muffler and made very simple conical flares at the inlet and outlet ends. We also made all ports (holes) in baffle plates into extruded holes.
The net result of these simple (but not necessarily inexpensive) changes was a 20% reduction in flow resistance of the muffler.
When we assembled one of these into a production muffler shell (using the adjacent muffler manufacturing plant), and furnished it to an acquaintance of mine running "Showroom stock", it showed shall we say, a "worthwhile" improvement in performance.
-----------
To get to the point:
a significant amount of the backpressure found in a production performance car exhaust system is often due to mismatches and transitions; for example, when a large diameter system pipe goes into a smaller diameter muffler inlet bushing, it is often possible to reduce backpressure by using smaller diameter system piping, matching pipe diameter to muffler inlet diameter.
There's more.
RE: A little exhaust design indecision.
Variable-tuning mufflers have been in (and out) of production for a number of years, as far back as maybe the late '80's in for example the Nissan Skyline (in Japan only, AFAIK), and in the '90's in certain Mitsubishi and BMW (3-series) models.
The Nissan at least used a Boden cable with an actuator driven by an RPM sensor, opening at a specified RPM, over a period of about one second, and closing then at about 250 RPM lower to prevent unnecessary cycling. This operated a butterfly on an auxiliary pipe that short-circuited a portion of the muffler. I presented a variation on this system to GM for Corvette applications in about 1990. They wanted it at xero cost, of course...
Then Arvin or maybe it was Walker came up wtih a spring-loaded, backpressure operated valve internal to the muffler; this was much later, and probably was licensed to them by a Japanese manufacturer. I don't recall if this ever saw the light of production.
RE: A little exhaust design indecision.
The idea is certainly not original, just as you say, but the technique is not often known about or copied by the do it yourself at home hot rod people.
Just placing your hand over the open pipe end, and partially closing off an idling exhaust pipe will significantly drop noise levels. The trick is to have something that will open up and close off the exhaust cross sectional area rapidly, with with changing gas volume. Whatever is used, MUST ABSOLUTELY NOT VIBRATE OR RATTLE, it needs to be quite well damped somehow, or it will just add characteristic noises of its own making things worse.
Another idea I have though about, but not tried is the use of a fairly large turbocharger exhaust wastegate located in the exhaust pipe. If fitted with a suitable light spring, it would open against the spring holding back some small residual back pressure. Just a simple low pressure blow-off valve. I suspect the poppet valve would vibrate fairly violently and chatter on it's seat and probably not work too well. But the pneumatic diaphragm in the actuator could perhaps be filled with some type of light fluid and damp the whole thing hydraulicly ?? That could possibly eliminate all the rattles but still make for a fairly fast smooth operating device. Not a low cost solution unfortunately, but some of those cheap Chinese external wastegates should be well up to the job.
RE: A little exhaust design indecision.
Norm
RE: A little exhaust design indecision.
http://
these are some very informative articles from the naca server on muffler design. Also search for david vizard's magazine article on the web, it has some interesting info as well including a backpressure vs power graph.
RE: A little exhaust design indecision.
Some of the statements in that article seem a bit odd.
Cheers
Greg Locock
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.
RE: A little exhaust design indecision.
here is a link to the beginning.
h
Greg any particular parts of the article seem odder than others? I have had a hard time finding a lot of exhaust info and not all of it seems to agree.