Muffler shape/design, flow, and noise?

[techno550]

I am searching for information on mufflers and their designs. I've found a good bit of information scattered about, but still have a lot of questions.

Most of the muffler designs I am investigating are "straight through" designs, perforated cores with some sort of packing material around them. The first (and perhaps biggest) question I have is how does the length of the perf tube effect the attenuation rate. The next is how does the shape affect the attenuation as well. I *think* the can shape around the perf tube would effect the target frequency. (or would that be mostly the hole size in the tube??) So something like an oval shaped can would then theoretically have a broader spectrum it effects than a round can. am I on the right track here?

The next big catch is something like a borla XR-1, which is a straight through perf core in an oval tube, but also has a sheet perforated metal dividing the middle of the perf tube as well. This brings up the question of target frequencies again, as that effectively halfs the "pipe diameter" in that area. It also brings up a question of how much rotation is there in the flow of the exhaust gasses and would this kill that... and if it does/doesn't is that a bad/good thing?

TIA
-Michael McCoy

 

[NCShane]

Michael,
Check out this Naca report on muffler design. It's pretty old but there is lots of information.

http://naca.central.cranfield.ac.uk/reports/1954/naca-report-1192.pdf

Shane

 

[techno550]

Thanks for the link. I have read that before, but it had been a while. It does not seem to specifically address the above questions though. (unless I read it wrong, which is definately possible.) the study seems to hit on various chambered designs, and then hits on a few perf core type, but small perforated areas and limited number of holes (and varied the locations), but didn't do a continuously perforated core in different volumes. I *think* this could mean it would act like an open volume, in which case the answer to the above questions would be that varying the length of the muffler doesn't change the peak attenuation, only the frequency at which it occurs. (and also the frequency at which zero attenuation occurs.) I'm still uncertain as to if the shape of the muffler effects the frequency (distance to a given wall) or if it is the volume that matters. If we treat it as a chamber, it'd be volume, but we've also seen that distance to a given end of the chamber matters as well. If it is a given length and the volume (assuming it is just volume that we are changing?) can be varied, the increased volume increases the attenuation rate. Am I on the right track?

any more input?

Thanks

 

[MikeHalloran]

It would probably be educational to plug the naca equations for a few of their designs into Excel and screw around with the dimensions to get a better feel for what affects what, and what you might expect in the envelope you have.

I did one for muffler 19; it came out about 95k, including a graph.

The second one will be much easier.





Mike Halloran
Pembroke Pines, FL, USA

 

[NCShane]

Well, I am by no means an expert in this area, but I did alot of research during my years working on a Formula SAE car in school. Another good source that I used is Design and Simulation of Four Stroke Engines by Gordon Blair. He basically says that an absorption muffler (packed deisgn as you describe) acts like a diffusion type muffler, with the packing material helping to absorb noise in the pass-bands and the frequencies above 1 khz. This is true if the total cross-sectioal area of all the holes(perforations) is greater than 5 times the pipe area. I'm assuming that this would be true for a continuously perforated tube like you have. So assuming that you can treat it as a diffusing type muffler, Gordon says that the attenuation of a diffusing silencer is basically a function of the expansion ratio (area of the box (chamber) over area of the pipe) and the relationship between the wavelength of the sound and the length of the chamber. The wavelength of the sound is simply used to define the pass bands of zero attenuation frequencies.
I would say that the shape of the chamber (oval or round) has significantly less effect that the area and the length. There are some corrections,etc for shapes other than round tubes, but I don't have them in front of me. You can find LOTs of SAE papers on this stuff, but it sounds like you're on the right track for a basic analysis.

Shane

 

[SylvestreW]

Acoustics of Ducts and Mufflers by M.L. Munjal (ISBN 0471847380 - Wiley Interscience publishing) is a pretty useful resource if you really want to get into the nitty-gritty.

 

[Rob45]

For a straight-through silencer, the lengths of perf'd sections greatly affect the primary tuning frequencies.

In the case of a pass tube that is perf'd its entire length, the silencer then behaves pretty much like an expansion volume, tuning primarily at a frequency corresponding to quarter-wave theory. That is, it will tune at odd integer multiples of the frequency determined by 4 x (four times) the silencer length; i.e., for length L, it will tune at nc/4L where n = 1, 3, 5...

Shorter lengths of perforation modify this behavior substantially.
There's an SAE paper detailing this, probably 15 or so years old.