Return-Path: Received: from rly-xl01.mx.aol.com (rly-xl01.mail.aol.com [172.20.83.70]) by air-xl01.mail.aol.com (v96.8) with ESMTP id MAILINXL14-5ad3f7b69fa1e2; Wed, 01 Oct 2003 19:58:36 2000 Received: from www.orcon.com (mail.orcon.com [12.152.176.212]) by rly-xl01.mx.aol.com (v96.8) with ESMTP id MAILRELAYINXL13-5ad3f7b69fa1e2; Wed, 01 Oct 2003 19:57:46 -0400 To: Lancair_ES@YahooGroups.com, lml@lancaironline.net Cc: newlan2dl@aol.com Subject: acoustic insulation MIME-Version: 1.0 Message-ID: From: "Dan Newland" Date: Wed, 1 Oct 2003 16:56:20 -0700 Content-Type: multipart/alternative; boundary="=_alternative 0083A2D088256DB2_=" X-AOL-IP: 12.152.176.212 X-Mailer: Unknown (No Version) --=_alternative 0083A2D088256DB2_= Content-Type: text/plain; charset="us-ascii" Hi Folks As I had said before, I have been working with Dave Morss on his lovely Legacy to try and make the plane quieter. We are about half way done and I thought I'd give a preliminary report on the findings. First, to describe the tests up to this point only with the goal being to isolate the variables by testing: TEST # CONDITION 1. An empty plane 2. An insulated plane less the firewall blankets, (blankets are 1/2" fiberglass/1/8" lightweight barrier/ 1/2" Orcofoam) 3. A heavy firewall blanket in addition to #2 4. A light weight firewall blanket replacing #3 5. Area around instruments fwd of glass shield 6. No sidewall blankets (firewall unchanged) 7. 1/2" Orcofoam 8. 1" Orcofoam Note too that readings are taken during takeoff, max power climb, cruise climb and cruise and that dB readings are done for dB A and dB C. The following results are in dB A which is weighted more closely to how we perceive sound. The same tests in dB C showed even greater differences in noise reduction. And I most often will read the cruise readings but since this is preliminary data, I'm just going to generalize. Naturally all comparisons are in the same flight regime. Still to follow, tests 9-18 will be similarly done by testing an area both with and without blankets and of different types of constructions that will thus tell us the relative amount of noise coming through and the effectiveness of the blanket constructions. The aircraft areas are divided up as: A. Firewall, (2 pieces) B. Fwd of glare shield and behind the firewall, around the instruments, (3 pieces) C. Sidewalls, (4 pieces) D. Baggage comp. floor and bulkhead, (2 pieces) E. Flooring (under carpet and over nose wheel well, 4 pieces) F. Glare shield, (1 piece) Also I should point out that all of these materials are FAR 25.853 (12 second vertical burn test) compliant. As an experimental, we are not required to comply with this but would you want anything less? An I should also point out that unlike many planes, Dave's Legacy doesn't have a trim liner so the outside of the blanket needs to be upholstered to really dress this up. That has a secondary effect for us in that the interior trim liners can be a big benefit for reducing noise into the plane IF DONE CORRECTLY! This type of installation will not be done on this aircraft, nor will constrained layer damping (the typical black foam attached to the fuselage skins) be tested. Both of these would probably net some improvements in reducing cabin noise. The tests will have to be repeated for #1 and 2 since the digital recorder decided to write over all of test 1 and most of test 2. However, the rest of the tests are otherwise completed. Anyway, we have removed over 4 dB from the plane just in the best vs. the worst of tests 3-8 so I would expect perhaps 6-7 dB to be removed from the basic plane which is a very considerable amount of noise. Dave has been very perceptive that way and his trained observations have been very close to what we've seen in the laboratory results in terms of overall noise. I also had tonal analysis done and there are a LOT of harmonics at low frequencies pretty much where you would expect in the 135 Hz range (2700 RPM/60 = 45 Hz with 3 cylinders firing per second = 45 X 3 = 135 Hz). So the initial tests show that the massive firewall blanket (test 3) made a considerable improvement in the overall noise. This resulted in an improvement of .5 dB to 1.9 dB based on the comparison against the light weight blanket (test 4) and seems to be about .5 dB or better than incomplete data from part of test 2. The bottom line is that you should spend a considerable amount of time and effort fitting a very good blanket to the firewall! And a few pounds here for a barrier are well spent! Tests number 6, 7 and 8 showed that insulation on the sidewalls also could eliminate over 4 dB of noise by insulating the firewall. The 1" Orcofoam was a very good performer for sidewalls (the best so far in a real life test) and separate laboratory acoustic tests have shown it to be an excellent absorber, (I haven't seen anything better in any FAR approved insulation). The other thing that was interesting is that the area around the instruments can affect the noise in the cabin, but not in the way I expected (I LOVE unexpected results!) As it turns out, the blankets with barriers were NOT as good as the plain un-insulated cavity (the spread was between 0 dB and .6 at the most so it isn't huge). This means that the blankets behind the firewall and around the instruments DO have an effect and thus can be optimized. So while I cannot show data that verifies this, I would speculate that an absorber blanket only would improve the acoustics and as with most acoustic treatments, more is better. To summarize it, the barrier didn't appear to help and actually made it slightly worse. OK, a brief aside, sound absorption of a material is very dependant on thickness and frequency. The most effective absorption is for the thickness of the absorber to be 1/4 the wavelength of the frequency. The formula for this is 1127/frequency = wavelength in feet where 1127 = the speed of sound in FPS at sea level. So as an example, we already know there is a spike in the 135 Hz frequency due to the 3 bladed prop and the 6 cylinder engine (firing every other stroke) at 2700 RPM. So how much absorber is needed to absorb this frequency? 1127/135 = wavelength of 8.35' so 1/4 of that is 2.1' needed of absorber...obviously we won't be planning on absorbers to get rid of this frequency! However it can help a lot on getting rid of harmonic frequencies from the engine and prop along with all the other noise generated separately. So inverting the formula, frequency = 1127/wavelength. Since we have 1" of insulation, we should see best absorption at a wavelength of 4" which works out to 3394 Hz. And since our hearing is typically good from between 20Hz and 20,000 Hz approximately, this will cover a lot of frequencies. Looking at our spectra, we definitely see the absorption improvement at these higher frequencies. Another bit of trivia, a HARMONIC is the doubling, tripling, quadrupling, etc of the fundamental frequency so the harmonic frequencies of 135 Hz are 270, 405, 540, 675, etc. Octaves are related but not the same, whereas an octave is a doubling of the frequency. The octaves of 135 Hz are 270, 540, 1080, 2160, etc. The way we humans (and even some sub-humans I dated back in High school) perceive sound are nonlinear. We perceive octaves easily but not necessarily a harmonic. Thus our music scale is based on octaves while music notes typically have many harmonics. OK, I'm stating to get out of my league. Anyone interested can read a much better definition of this in a book titled "The master handbook of acoustics" by F. Alton Everest. Very readable and helpful for those with an interest. Please keep in mind this is preliminary and we are still testing. More surprises may be in store but it is pointing us in a definite direction. Perhaps what will happen (and I hope it will) is that for the next series of tests, I will get a metal plane such as an RV to test further on and refine the data even more. Particularly as it relates to planes with trim liners which can make a huge difference. And you know I love you guys but the standard rules regarding copyrights apply here so please don't go and publish these results thus making millions of dollars for yourselves leaving me holding ashes... So very soon I will have these materials put together as kits so small plane builders can buy them. In the mean time, if you'd like, check out our website at www.orcon.com. Yes this is a small project in an otherwise big company but it's sure a lot of fun! Further results to follow. Dan Newland Lancair Super ES #061 Application engineer Orcon Corporation 1570 Atlantic Street Union City, California 94587 phone: (510) 476-2150 cell: (510) 290-6885 fax: (510) 489-6702 --=_alternative 0083A2D088256DB2_= Content-Type: text/html; charset="us-ascii"
Hi Folks

As I had said before, I have been working with Dave Morss on his lovely Legacy to try and make the plane quieter.  We are about half way done and I thought I'd give a preliminary report on the findings.  

First, to describe the tests up to this point only with the goal being to isolate the variables by testing:

TEST #          CONDITION
1.                An empty plane
2.                An insulated plane less the firewall blankets, (blankets are 1/2" fiberglass/1/8" lightweight barrier/ 1/2" Orcofoam)
3.                A heavy firewall blanket in addition to #2
4.                A light weight firewall blanket replacing #3
5.                Area around instruments fwd of glass shield
6.                No sidewall blankets (firewall unchanged)
7.                1/2" Orcofoam
8.                1" Orcofoam

Note too that readings are taken during takeoff, max power climb, cruise climb and cruise and that dB readings are done for dB A and dB C.  The following results are in dB A which is weighted more closely to how we perceive sound.  The same tests in dB C showed even greater differences in noise reduction.  And I most often will read the cruise readings but since this is preliminary data, I'm just going to generalize.  Naturally all comparisons are in the same flight regime.

Still to follow, tests 9-18 will be similarly done by testing an area both with and without  blankets and of different types of constructions that will thus tell us the relative amount of noise coming through and the effectiveness of the blanket constructions.  The aircraft areas are divided up as:

A.        Firewall, (2 pieces)
B.        Fwd of glare shield and behind the firewall, around the instruments, (3 pieces)
C.        Sidewalls, (4 pieces)
D.        Baggage comp. floor and bulkhead, (2 pieces)
E.        Flooring (under carpet and over nose wheel well, 4 pieces)
F.        Glare shield, (1 piece)

Also I should point out that all of these materials are FAR 25.853 (12 second vertical burn test) compliant.  As an experimental, we are not required to comply with this but would you want anything less?  An I should also point out that unlike many planes, Dave's Legacy doesn't have a trim liner so the outside of the blanket needs to be upholstered to really dress this up.  That has a secondary effect for us in that the interior trim liners can be a big benefit for reducing noise into the plane IF DONE CORRECTLY!  This type of installation will not be done on this aircraft, nor will constrained layer damping (the typical black foam attached to the fuselage skins) be tested.  Both of these would probably net some improvements in reducing cabin noise.

The tests will have to be repeated for #1 and 2 since the digital recorder decided to write over all of test 1 and most of test 2.  However, the rest of the tests are otherwise completed.  Anyway, we have removed over 4 dB from the plane just in the best vs. the worst of tests 3-8 so I would expect perhaps 6-7 dB to be removed from the basic plane which is a very considerable amount of noise.  Dave has been very perceptive that way and his trained observations have been very close to what we've seen in the laboratory results in terms of overall noise.  I also had tonal analysis done and there are a LOT of harmonics at low frequencies pretty much where you would expect in the 135 Hz range (2700 RPM/60 = 45 Hz with 3 cylinders firing per second = 45 X 3 = 135 Hz).

So the initial tests show that the massive firewall blanket (test 3) made a considerable improvement in the overall noise.  This resulted in an improvement of .5 dB to 1.9 dB based on the comparison against the light weight blanket (test 4) and seems to be about .5 dB or better than incomplete data from part of test 2.  

The bottom line is that you should spend a considerable amount of time and effort fitting a very good blanket to the firewall!  And a few pounds here for a barrier are well spent!  

Tests number 6, 7 and 8 showed that insulation on the sidewalls also could eliminate over 4 dB of noise by insulating the firewall.  The 1" Orcofoam was a very good performer for sidewalls (the best so far in a real life test)  and separate laboratory acoustic tests have shown it to be an excellent absorber, (I haven't seen anything better in any FAR approved insulation).

The other thing that was interesting is that the area around the instruments can affect the noise in the cabin, but not in the way I expected (I LOVE unexpected results!)  As it turns out, the blankets with barriers were NOT as good as the plain un-insulated cavity (the spread was between 0 dB and .6 at the most so it isn't huge).  This means that the blankets behind the firewall and around the instruments DO have an effect and thus can be optimized.  So while I cannot show data that verifies this, I would speculate that an absorber blanket only would improve the acoustics and as with most acoustic treatments, more is better.  To summarize it, the barrier didn't appear to help and actually made it slightly worse.

OK, a brief aside, sound absorption of a material is very dependant on thickness and frequency.  The most effective absorption is for the thickness of the absorber to be 1/4 the wavelength of the frequency.  The formula for this is 1127/frequency = wavelength in feet where 1127 = the speed of sound in FPS at sea level.  So as an example, we already know there is a spike in the 135 Hz frequency due to the 3 bladed prop and the 6 cylinder engine (firing every other stroke) at 2700 RPM.  So how much absorber is needed to absorb this frequency?  1127/135 = wavelength of 8.35' so 1/4 of that is 2.1' needed of absorber...obviously we won't be planning on absorbers to get rid of this frequency!  However it can help a lot on getting rid of harmonic frequencies from the engine and prop along with all the other noise generated separately.  So inverting the formula, frequency = 1127/wavelength.  Since we have 1" of insulation, we should see best absorption at a wavelength of 4" which works out to 3394 Hz.  And since our hearing is typically good from between 20Hz and 20,000 Hz approximately, this will cover a lot of frequencies.  Looking at our spectra, we definitely see the absorption improvement at these higher frequencies.

Another bit of trivia, a HARMONIC is the doubling, tripling, quadrupling, etc of the fundamental frequency so the harmonic frequencies of 135 Hz are 270, 405, 540, 675, etc. Octaves are related but not the same, whereas an octave is a doubling of the frequency.  The octaves of 135 Hz are 270, 540, 1080, 2160, etc.  The way we humans (and even some sub-humans I dated back in High school) perceive sound are nonlinear.  We perceive octaves easily but not necessarily a harmonic.  Thus our music scale is based on octaves while  music notes typically have many harmonics.  OK, I'm stating to get out of my league.  Anyone interested can read a much better definition of this in a book titled "The master handbook of acoustics" by F. Alton Everest.  Very readable and helpful for those with an interest.

Please keep in mind this is preliminary and we are still testing.  More surprises may be in store but it is pointing us in a definite direction.  Perhaps what will happen (and I hope it will) is that for the next series of tests, I will get a metal plane such as an RV to test further on and refine the data even more.  Particularly as it relates to planes with trim liners which can make a huge difference.  And you know I love you guys but the standard rules regarding copyrights apply here so please don't go and publish these results thus making millions of dollars for yourselves leaving me holding ashes...  

So very soon I will have these materials put together as kits so small plane builders can buy them.  In the mean time, if you'd like, check out our website at www.orcon.com.  Yes this is a small project in an otherwise big company but it's sure a lot of fun!

Further results to follow.


Dan Newland

Lancair Super ES #061

Application engineer
Orcon Corporation
1570 Atlantic Street
Union City, California 94587

phone: (510) 476-2150
cell:  (510) 290-6885
fax:   (510) 489-6702 --=_alternative 0083A2D088256DB2_=--