silencer design info

I found this thesis on the internet yesterday. It's a bit long but has some useful info in it.

www.ohiolink.edu/etd/send-pdf.cgi?osu1117631229

The top graph shows the silencer with no packing (black and grey) and with packing (the other colours). It shows the affect of porosity (the number of holes) and different sized holes. It is not exhaustive (no pun intended) but smaller holes seem better)

It also demonstrates why silenced exhaust noise seems lower in pitch, silencers absorb higher sounds better.

The lower graph shows the difference in the silencer with and without a baffle fitted 1/3 the way along. As you can see, fitting a baffle in a straight through silencer makes a lot of difference.

Something to remember if you make your own can.
Rescued attachment silencer.JPG

Having to regularly pack a silencer for a race car (every race!) I've certainly noticed quite a differrence in how you pack the silencer.

Also I've just ordered a 6" x 21" straight through silencer for my kit car. I was going to have fins/discs welded at two places along the perforated tube. This was to break up the volume in the silencer and also stop the majority of packing eventually blowing down one end. So from the above there does seem to be some merit in doing this.

has anyone on here made their own silencer?

quote:

---

Originally posted by cadebytigerhas anyone on here made their own silencer?

---

is it hard to do? I mean i cannot weld myself but i know people who can.

I cannot see it being that tricky really

I bought all the bits, had a friend do some welding, then bolted it all together. If you can build a car, you can build a silencer.

Interesting link, thanks for posting.

Note how the unpacked silencer is essentialy becomes a resonant cavity, peaking at multiples of it's fundamental frequency. Explains why absorbtive silencers are so damn noisy when the packing has been blown out.

Vizzard did a test a while ago which you can add to this if you're making your own silencer.

As i recall .......

Add an empty chamber at the start of the silencer, this should be big enough to hold 4 cylinders worth of expelled gas. This then becomes the tuned part of the "1" in the 4-2-1 / 4-1 system and you can accurately measure it. If not the '1' ends up somewhere in the exhaust. For most of us this will give little benefit but for the racers searching for an ex hp .....

From what I've read by Vizard, what you want is to tune the length of the "1" in either a 4-2-1 or a 4-1 manifold to the desired RPM, then have a chamber in your exhaust that is 8 or more times the volume of 1 cylinder in your engine.

It is also a good idea for the inlet tube to protrude into the chamber a little, rather than being flush with the front wall.

The ideal can will then have 3 chambers. An empty one then two with perforated tubes and packing material, as above.



[Edited on 30-8-2007 by smart51]

Thats quite a big chamber. For a 2 litre 4 cylinder car 8 times one cylinder is 4 litres which is 244 cubic inches. A 6" diameter can has an area of 28 square inches so it would have to be 8.6" long.

So the empty chamber at the start of the silencer would be 8.6 inches long and presumably the rest would be pro rate i.e. 24" giving a total of 32" long.

Quite possible I suppose but a lot bigger than I'm used to.

Any comments on my sums?

quote:

---

Originally posted by v8kido the empty chamber at the start of the silencer would be 8.6 inches long and presumably the rest would be pro rate i.e. 24" giving a total of 32" long.

---

So; presumably these two chambers (the empty one and the packed one) could actually be one chamber but with a divider in the middle.

The first chamber would be completely empty and the second chamber would have a divider and perforated tube surrounded by exhaust packing.

Could be quite neat and effective - anyone fancy building something like that and letting us all know how you get on?

The added baffle 1/3rd of the way along the fully-packed silencer looks very interesting - simple to try and of little consequence if it doesn't work as expected.

I might give that a try at the next re-packing.

Its worth noting that you shouldn't expect a 20 dB change by adding a baffle. The exhaust noise fundamental frequency for a 4cyl engine at 4500 RPM is 150 Hz and is 300 Hz at 9000 RPM. The first few harmonics make most of the noise so most of the din is being made where the silencer is least effective.

From calulations using simulated square waves, 2 to 3 dB might be saved with an internal baffle. That's still well worth the effort.

Hi I/we had to start making our own silencers for the race cars. This was due to the fact that all the others would just not get some of the cars below 105db for circuit use.

The secrets lie in the material being used for packing and also the materiel used for the silencer. But basically you need to make it out of aluminum 2mm. And using stainless stell of any guage is usless.

Pics can be found of it in my arcjhive at bottom of page. ( to big to post ).

cheers matt

Matt,

Do you have an ally tube to clamp onto the exhaust pipe or have you fixed it with a stainless tube. If so, how is this connected to the silencer main box.
It looks stainless from the picture but I didn,t realise you could connect/weld/solder ally to SS.

Hi the alloy tube was heated and shrunk on to the stainless tube.

Had my doubts if it would last/work. But has not given any problem's atall.

cheers matt

that is a really nice job! do you or would you make them for anyone else

Is the pope catholic?

Hi no not made any for sale other than the few fitted to our own LA cars,

To sell them on the open market they would be quite expencive Ie roughly £400.

And i havent the time to mass produce them.

cheers matt

the comments on the expansion box to "terminate the no. 1 resonant length" may be fine in theory or on mr Vizard's dyno, but in real life forget it!!!

The primary and secondary resonant lengths are of much higher importance, and most people even cock these up, so best to concentrate on these not on the "ultimate"

Well, I've just repacked about a quarter of my silencer, and fitted a baffle at the same time.

Because I've replaced some packing then I can't say for certain that it is quieter, but subjectively I believe it is. I'll need to use a proper noise meter to be sure, comparing it against previous measurements.

Coincedentally we tried this on our Radical Clubsport today. The initial feedback has been posative. Usually the sound testers measure us at 108 or above. Today we were 106 and we did feel the sound was very slightly muted compared to previously as some of the resonance in the car was reduced. The engine is a Hyabusa directly bolted tot he chassis.

We will need to test further but I'll also be trying this on my Striker when the new can arrives.

Just because the pitch has changed doesnl;t nesseceraily mean that the overall noise has got quiter. Only true way to check is using a DB(sound) meter

As I said earlier - much earlier! - the exhaust note did change, but I'd need a noise meter to confirm (and I have no pre-modification reference data).

However, I can say that the exhaust noise is less intrusive and less objectionable. It is still quite loud though!

Mind you, compared to the induction roar from my new bike carbs...