High Performance Muffler Design 1

[Valdar]

Hello, I have been developing mufflers for racing applications for about a year now. The basic design found to be the most promising consists of a straight-thru cylindrical shape, with one or two internal baffles (2 or 3 tuned chambers) and a perforated tube core surrounded by fiberglass insulation; much like the classic 'cherry bomb' design.

I am wondering if anyone has experience in designing similar silencers and could share what they have learned in terms of what tuning techniques and geometry gives the highest attenuation? Stiffness is also important to consider..

Right now I am considering moving to an inline baffle (in the direct path of gas flow) and adding an annular open area around the inner tube for the gas to flow around the baffle without having to pass directly through insulation material.

The current designs are based on a Honda CBR600F4i motorcycle engine, and offer about 30 dB attenuation, down to about 110dB. The device is primarily 6061Al and weighs 750grams. Measurements were taken 45 deg from the outlet, 500mm away.

The target is 105dB and sub 750g.

These results have been achieved using standard textbook transmission loss equations and empirical tuning based on a 1/3 octave frequency map (fast weighting, 100 - 10000Hz)

There has to be more information out there... does anybody know where I can find it ? Thanks!

K.A

 

[MikeHalloran]

There exist roughly nine actual textbooks on related subjects. All are out of print and not expected to return.

There is in print one enthusiast book with an obvious title, better than nothing but probably of less use than whatever test apparatus you've rigged up.

See 'naca-report-1192.pdf', most reliably found with a Google search, for an incredibly dense (as in much information in not much text and a few pictures) and comprehensible (as in you don't need a PhD to read it) analysis of chambered (non absorptive) mufflers. You will still find it instructive for what you are trying to do. It may help you better place your dividing baffles, and the test rig in the pictures could save a lot of fuel.



Mike Halloran
NOT speaking for
DeAngelo Marine Exhaust Inc.
Ft. Lauderdale, FL, USA

 

[SomptingGuy]

One really good text is:

Acoustics of Ducts and Mufflers
Author: Munjal, M.L. (Department of Mechanical Engineering, Indian Institute of Science, Bangalore, India)
Hardback; Book; Bibliography, Index
346 pages
Published: June 1987
John Wiley and Sons Ltd
ISBN: 0471847380

It's not cheap, but it is still in print.

A question: Are you part of a "Formula Student" or "Formula SAE" team? I ask because their engines are based on the ubiquitous Honda CBR600. If so, why not investigate getting Ricardo WAVE under their sponsorship scheme?

 

[Valdar]

Hey, thank you very much for your prompt responses! haha, yes in fact I am part of a FSAE team (Waterloo), good call. Aside from that though I am very interested in continuing my research on mufflers and acoustics after university.

We already have Ricardo Wave software and have used it extensively in the past for engine tuning, however we have not looked into it's capabilities regarding mufflers.. I was unaware it had this feature and will look into it soon.

Thanks again !

K.A.

 

[Artsi]

"Theoretical and Experimental Investigation of Mufflers with Comment on Engine-Exhaust Muffler Design
Don D. Davis, Jr. George M. Stokes Dewey Moore George L. Stevens, Jr.
NACA Report 1192
1954"

http://naca.larc.nasa.gov/reports/1954/naca-report-1192/naca-report-1192.pdf

 

[SomptingGuy]

Valdar: We already have Ricardo Wave software and have used it extensively in the past for engine tuning, however we have not looked into it's capabilities regarding mufflers.. I was unaware it had this feature and will look into it soon.

It's used extensively in the tier-1 exhaust supplier community, because it allows you to tune for perfomance and noise (level and quality) at the same time.

 

[Rob45]

Valdar:
It's unlikely that you'll find a great deal of really useful information in writing, for fairly obvious reasons: the companies specializing in exhaust systems and mufflers are selling their technical expertise more than they are a few pounds of steel, and if they give the knowledge away, what do they have left to sell? Anyone with a few basic sheetmetal machines can build mufflers. That's why the muffler companies don't have training manuals or handbooks for their engineers - they don't want their competitors (nor their customers) getting hold of this information. (Ford a few years ago *thought* they could just produce their own mufflers in-house... )

The same holds true for those engineers with practical real-world muffler and exhaust system design experience: they've acquired this knowledge over a period of years, and it's their meal-ticket. If you meet one face-to-face, he'd probably help you out, but as far as giving you his favorite formulae for designing mufflers - that's highly unlikely.

And with the exception of the above-referenced NACA report, there are very very few texts with any really useful information. By the time you've built your second muffler, you will have learned more than most texts contain. That is, assuming that you've studied and understood the basics of fluid flow, thermodynamics, and heat transfer.\

With regard to your specific problem, reaching 105 dB(A): I assume this is a constant-speed test, and that you've considered incorporating a Helmholtz or other narrow-band tuner to attack the engine firing frequency? And that the outlet is configured to minimize jet noise? Beyond that, I can only suggest that greater volume gives greater attenuation, all else being equal, and that you might try to get some continuous-strand fibreglas roving, since it is somewhat more durable at elevated temperatures, while still being lighter than basalt-wool or stainless steel wool.

One final thought: I'd strongly recommend performing narrow-band analyses in addition to 1/3-octave, to more accurately locate any problem frequencies. This can help you find for example panel resonances of the muffler shell or of an internal part.
OK, one more final thought: you could search the archives on this site for things that I and others have written in the past on this and related topics. You might find something useful here.

Good luck,
- R

 

[magnograil]

Suggest you look into the NACA reports on muffler tests. Do not remember the number at the moment, but look at the NASA Glenn site and search on "muffler". They did theoretical calculations and acoustic tests on 77 configurations. Several were built and tested on a 7 cylinder radial engine.

 

[NormPeterson]

Mention of panel resonances brings something else to mind, that of integrity of the attachment of the baffles. Some years ago a whole series of aftermarket muffler from a well-known company was taken off the market, apparently the victim of internals coming adrift and rattling severely at various frequencies.

Norm