Noise Reduction
The following tests have been undertaken to investigate the noise reduction characteristics of Armaflex insulation.
1. Acoustic performance tests on insulation systems based on Armaflex insulation Technical Report No:- AT 3461/1 Rev 0 23rd November 1994
evaluated:
System |  | |
1 | 1 x 25 mm | outer steel wrap |
2 | 2 x 25 mm | outer steel wrap |
3 | 3 x 25 mm | outer steel wrap |
4 | 4 x 25 mm | outer steel wrap |
5 | 2 x 25 mm | Arma-Chek Insulation System |
6 | 4 x 25 mm | Arma-Chek Insulation System |
The overall ‘A’ weighted sound insertion loss has been calculated based on ‘pink’ noise inside the pipe to give guidelines on the reductions which may be anticipated for high frequency noise sources such as valve noise.
The results of the tests are summarised (‘A’ weighted insertion losses to the nearest 0.5 dB).
System | overall dB(A) | Octave band centre frequency, HZ
Insertion losses (dB) |
| pink noise | 125 | 250 | 500 | 1 k | 2k | 4k | 5k |
1 | 20.5 | 0 | 3 | 14 | 17 | 23 | 24 | 32 |
2 | 24.5 | 2 | 3 | 15 | 20 | 31 | 30 | 35 |
3 | 25.0 | 1 | 1 | 16 | 22 | 34 | 32 | 35 |
4 | 25.5 | 1 | 2 | 17 | 23 | 36 | 33 | 35 |
5 | 14.5 | 0 | 1 | 3 | 9 | 18 | 22 | 32 |
6 | 19.0 | 1 | 2 | 6 | 13 | 28 | 32 | 35 |
Principles of test
For a sound source in a reverberant room, the diffuse field sound pressure level produced is proportional to the sound power level. Thus the reduction in source sound power achieved due to the application of an insulation system can be measured directly from the reduction in diffuse sound pressure levels.
Tests were carried out in a reverberation chamber, with a steady source sound level generated by loud speakers inside the pipe.
| |
| Corrections | 1. Background noise
2. Source level
3. True ‘pink’ noise |
True ‘pink’ noise is an idealised flat octave band noise spectrum. This has the effect in certain case of reducing the overall dB(A) reduction but gives a clear indication of the overall acoustic performance of the insulation to a flat noise spectrum.
In practice overall ‘A’ weighted insertion losses should be calculated based on estimates of the true noise spectrum within the pipe.
The insertion losses were then calculated as the difference in the corrected sound pressure levels for the insulation thicknesses tested, compared with the bare pipe.
Amplified ‘pink’ noise was played through one treble horn and one bass speaker via a 2.2 kHz two way passive crossover.
The pipe was suspended in the centre of a large reverberant test chamber (Building Research Establishment, Watford) at approximately 1 metre above the floor.
The test results show an increase in overall dB(A) insertion loss with increasing insulation thickness for both the steel clad and scrim coat faced insulation systems.
The individual octave band results show a consistent increase in insertion loss with increasing thickness in the frequency range 500 Hz to 8 kHz. These are the frequency bands to which the ear (and therefore the ‘A’ - weighting scale) is most sensitive, and therefore generally control the overall ‘A’ weighted insertion losses.
Noise reduction at low frequencies is principally controlled by the mass of the outer wrap, the insulation thickness having minimal effect.
The test results show that the systems tested are suitable for acoustic insulation applications requiring the following approximate insertion loss performance, for mid to high frequency sound:
System | Insertion loss to pink noise dB(A) | Armaflex thickness |
1 | 20.5 | 1 x 25 mm |
2 | 24.5 | 2 x 25 mm |
3 | 25.0 | 3 x 25 mm |
4 | 25.5 | 4 x 25 mm |
5 | 14.5 | 2 x 25 mm |
6 | 19.5 | 4 x 25 mm |
The Armaflex and steel provided significant acoustic insulation principally in the frequency range 500Hz to 8kHz.
The Arma-Chek system provided significant acoustic insulation principally in the frequency range 1 kHz to 8kHz.
2. Armstrong Acoustic Laboratory
The test method conforms to ANSI/ASTM C423-81a and E795 20 mm thickness Armaflex was mounted directly on the reverberation room floor
1/3 Octave Band Centre Frequency Hz |
| 125 | 250 | 500 | 1000 | 2000 | 4000 | NRC |
| Absorption Coefficient | 0.04 | 0.07 | 0.16 | 0.62 | 0.25 | 0.23 | 0.30 |
| Sample Sabins | 2.6 | 5.0 | 11.0 | 43.9 | 17.3 | 16.2 | - |
| % uncertainty | 4.0 | 0.8 | 0.6 | 0.6 | 0.4 | 0.2 | - |
The noise reduction coefficient (NRC) is the average of the coefficients at 250, 500, 1000 and 2000 Hz, expressed to the nearest multiple of 0.05.
3. Institut Fur Bauphysik, Stuttgart
Test method according to DIN 52218. During this procedure the overall sound level in dB(A) and the frequency distribution of the noise were evaluated by using third-octave filters.
The attenuation effect was determined from the measured values
| | |
| DL = L0 - L1 | where | L0 = sound level without insulation
L1 = sound level with insulation |
| |
| The DL values versus frequency for the fixings with stud pipe clips and the buried installation are shown graphically: |
| |
 | Sound attenuation effect, DL, of a 1" pipe insulation of Armaflex D-34 when fixed to the partition wall via 11/2" stud pipe clips tightened to the stop. |
| Type of fixation of 1" piping | Sound level in dB(A) | Attenuation effect in dB(A) |
| without insulation in 1" stud pipe clips | 53 | - |
| with Armaflex D-34 in 5/4" stud pipe clips | 32 | 21 |
| with Armaflex D-34 in 1 1/2" stud pipe clips | 25 | 28 |
| with Armaflex F-34 in 1 1/2" stud pipe clips | 23 | 30 |
| without insulation in plaster | 53 | - |
| with Armaflex D-34 in plaster | 35 | 18 |
| |
 | Sound attenuation effect, DL, of a 1" pipe insulation of Armaflex D-34 when fixed to the partition wall via: a - 5/4" stud pipe clips, b - 11/2 " stud pipe clips, and each clip was fully tightened to the stop. |
| |
 | Sound attenuation effect, DL, of a 1" pipe insulation of Armaflex D-34 when fixed to the partition wall via 11/2 " stud pipe clips tightened to the stop. |
|
|
|