Received-SPF: softfail
 receiver=logan.com; client-ip=216.240.18.37; envelope-from=echristley@nc.rr.com
Message-ID: <49651977.8060908@nc.rr.com>
Date: Wed, 07 Jan 2009 16:07:03 -0500
From: Ernest Christley <echristley@nc.rr.com>
Reply-To: echristley@nc.rr.com
User-Agent: Thunderbird 2.0.0.19 (X11/20090105)
MIME-Version: 1.0
To: Rotary motors in aircraft <flyrotary@lancaironline.net>
Subject: Re: [FlyRotary] Re: Muffler design (was 20B isssues . . )
References: <list-3414975@logan.com>
In-Reply-To: <list-3414975@logan.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

Tracy Crook wrote:
> Agreed, this is definitely a new concept and I hope it works.
>
> The only thing that bothers me is that, as you pointed out, the 
> bandwidth needed for attenuation is about 0 to 12 Khz.  That 
> represents MANY octaves even if we bump the lower limit to a realistic 
> number ( only a closed pipe will work at zero :-).  This bandwidth is 
> mutually exclusive with the term "tuned".  And if by tuned they mean 
> low pass, then it is not a new concept since that is covered by 
> conventional mufflers.  There may be something to this new concept but 
> until this contradiction is explained, I would hesitate to start 
> cutting parts.
The first parts I'm going to cut are wood sides and PVC pipe baffles.  
That'll be good enough to experiment with.  I've got a sound sample of a 
frequency sweep.  I can build a test muffler, play the frequency sweep 
in one end, and record it on the other.  Displaying both samples in a 
sound editor will tell me what, if any, frequencies are attenuated, and 
by how much.  Losses in the playback speakers and microphone can be 
documented by running a test without the intervening muffler.  I may 
have to spend money on good speakers and a good microphone (Oh, 
bummer!).  If that shows promise, I replace the PVC with stainless 
tubes.  If that works, I replace the wood with stainless sheet.

Which frequencies need to be attenuated?  Remember that Monty Roberts 
did some work a while back showing a sonagram of a rotary recording?  
The sound energies are much higher around 6 and 12kHz.  With that in 
mind, look at the graphs on the top of page 4 of this report
http://web.mit.edu/course/3/3.042/team1_08f/documents/complete-phononic-bandgap.pdf

It shows very significant attenuation in frequencies from 4 to 8kHz and 
then 10 to 14kHz.  This is with 13mm cylinders set in a 30mm square 
array.  Hollow cylinders showed similar efficacy.  Close enough to 1/8" 
pipe (mcmaster-carr part# 44635K422 ... cheap!) to make a prototype 
almost mandatory.  So, my current plan of action is to produce some 
sonagrams of rotary exhaust noise, and verify the target frequencies for 
attenuation.  With the experimental jig, optimize for bandgaps around 
those frequencies while using the smallest filling fraction possible.  
At that point, I will build a matching prototype.  Two sections of steal 
sheet, match drilled with 1/2" holes (or whatever size was found to be 
optimum).  Fill the holes with 3" sections of tube.  A bunch of welding 
later, a 8.5" wide muffler with a .500 filling fraction will render 
about the same flow area as a 4" diameter exhaust pipe.