Acoustic properties of commercial thermal insulators Valtteri Hongisto, Pekka Saarinen, Jarkko Hakala, Reijo Alakoivu Turku University of Applied Sciences, Acoustics Laboratory Joukahaisenkatu 7, FI-20520 Turku, Finland. ABSTRACT

The purpose was to compare the acoustic performance of commercial thermal insulators. Thirteen insulator types were studied. Six acoustic quantities were determined: sound reduction index of bare insulator, sound reduction index of encapsulated insulator (insulator between two boards), normal incidence sound absorption coefficient, airflow resistivity, dynamic stiffness per unit area, and reduction of impact sound pressure level in a floating floor. The ranges were 10  27 dB, 33  52 dB, 0.20  0.78, 3.0  2700 kPa  s/m2, 1.5  730 MN/m3, and 15  36 dB, respectively, when the thick- ness of the insulator was 100 mm. Closed-pore insulator types usually carried worse acoustic prop- erties than open-pore insulator types. Lower thermal conductivity was associated with worse acoustic performance of two acoustic quantities. Insulator manufacturers should consider to better declare the acoustic properties of insulator products, because this knowledge is needed in the acoustic design of building constructions and other applications.

1. INTRODUCTION

Thermal insulators are used structural applications, such as walls, floors, doors, roofs, duct wrap- pings, machine and wall linings, and vehicle envelopes. There is very little scientific comparative research about the acoustic performance of thermal insulator materials. The purpose of our study was to compare various acoustic properties of insulator materials to improve comprehensive understand- ing. The full version of the study is published in Ref. [1].

2. MATERIALS AND METHODS

Thirteen insulator types from several manufacturers were chosen to this study (Table 1). The thick- ness of each insulator type was 100 mm. The following acoustic measurements were conducted:  Airflow resistivity,  [Pa  s  m -2 ], was determined according to ISO 9053. Small value usually predicts high sound absorption coefficient.  Dynamic stiffness per unit area, s’ [N  m -3 ], was determined according to ISO 9052-1. Small value usually predicts high value of  L .  Normal incidence sound absorption coefficient,  0 [], was determined according to imped- ance tube method ISO 10534-2 within 100–3150 Hz. Mean absorption coefficient,   , was determined from these values. Specimen was installed against rigid background. Sound ab- sorption class was determined according to ISO 11654 (Class).  Sound reduction index (SRI) of bare insulator, R 1 [dB], was determined according to ISO 10140-2 within 50–5000 Hz. The single-number value, R w1 , was determined by ISO 717-1.  SRI of encapsulated insulator, R 2 [dB], was determined according to ISO 10140-2 within 50–5000 Hz. The insulator was placed between 13 mm gypsum board and 9 mm veneer board separated by 100 mm. The single-number value, R w2 , was determined by ISO 717-1.  Reduction of impact sound pressure level in a floating floor ,  L [dB], was determined according to ISO 10140-3 in a floating floor setup. The insulator (1200x2100 mm) was in- stalled on top of 160 mm concrete slab and a 19 mm veneer was placed on top of the insulator

21-24 AUGUST SCOTTISH EVENT CAMPUS. ? O? ? GLASGOW

so that the system formed a floating floor. Reduction of impact sound pressure level,  L w , was determined according to ISO 717-2. Table 1. Material s, thermal conductivity, and density of the insulator ty pes 1–13.

Insulator type (material)  [W/mK]  [ k g/ m 3 ] 1 ultra low density stone wool slab 0.044 25 2 low density stone wool slab 0.036 25 3 medium density stone wool slab 0.033 75 4 high density stone wool slab 0.037 100 5 ultra low density glass wool roll 0.042 11 6 low density glass wool slab 0.035 16 7 medium density glass wool slab 0.033 70 8 cellulose slab 0.039 37 9 wood fiber slab 0.038 50 10 expanded polystyrene board 0.036 17.5 11 polyisocyanurate board 0.022 30 12 phenolic foam board 0.020 30 13 cellular glass board 0.036 1 0 0

 3. RESULTS AND DISCUSSION The single-number values of each quantity are shown in Table 2. Frequency-dependent values are shown in Figs. 1–2. Table 3 indicates the linear associations between the reported quantities.

1.0

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70

0.9

1_100 2_100 3_100 4_100 5_100 6_100 7_100 8_100 9_100 10_100 11_100 12_100 13_100

1_100 2_100 3_100 4_100 5_100 6_100 7_100 8_100 9_100 10_100 11_100 12_100 13_100

0.8

0.7

0.6

R 1 [dB]

0.5

  []

0.4

0.3

0.2

0.1

0.0

125

2000

1000

4000

250

500

2000

1000

4000

250

125

63

500

f [Hz]

f [Hz]

Fig. 1. Left) Normal incidence sound absorption coefficient,    as a function of frequency, f, for insulator types 1  13. Right) SRI of bare insulator, R 1 , as a function of frequency.

21-24 AUGUST SCOTTISH EVENT CAMPUS. ? O? ? GLASGOW

Table 2. Single-number values for the 13 insulator types.

ID m'   Class s '  L w R w1 R w2 

[kg/m 2 ] [MN/m 3 ] [dB] [dB] [dB] [kPa  s/m 2 ] 1_100 2.5 0.73 B 1.8 36 11 51 6.6 2_100 2.5 0.75 B 1.6 36 15 51 12 3_100 7.5 0.7 C 1.8 35 21 50 31 4_100 10 0.68 B 11 32 22 50 33 5_100 1.1 0.72 B 3.1 33 11 50 3 6_100 1.6 0.78 A 1.2 36 15 52 10 7_100 7 0.73 B 0.43 36 17 51 16 8_100 3.7 0.73 A 0.75 35 12 50 11 9_100 5 0.71 B 1.9 34 10 50 5.9 10_100 1.8 0.27 D 44 20 16 38 530 11_100 3 0.22 E 10 16 15 33 2600 12_100 3 0.2 E 35 15 20 34 2700 13_100 10 0.26 E 720 15 27 35 68

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75

-15 -10 -5 0 5 10 15 20 25 30 35 40 45 50 55 60 65

1_100

2_100

3_100

1_100 2_100 3_100 4_100 5_100 6_100 7_100 8_100 9_100 10_100 11_100 12_100 13_100

4_100

5_100

6_100

R 2 [dB]

 L [dB]

7_100

8_100

9_100

10_100

11_100

12_100

13_100

Air_100

125

4000

2000

1000

250

500

63

4000

1000

250

125

2000

500

63

Gypsum

Veneer

f [Hz]

f [Hz]

Fig. 2. Left) The SRI of encapsulated insulator, R 2 , as a function of frequency, f, for insulator types 1  13. Air_100 corresponds to empty 100 mm cavity. In addition, the SRIs of veneer and gypsum are shown. Right) The reduction of impact sound pressure level,  L, as a function of frequency for insulator types 1  13.

21-24 AUGUST SCOTTISH EVENT CAMPUS. ? O? ? GLASGOW

Table 3. Squared Pearson’s correlation coefficients, r P 2 , between the reported quantities among the 13 insulator types. Statistically significant relationships are denoted by * p<0.05 and ** p<0.01.

    s'  L w R W1 R W2   1 -0.00 -0.03 0.51 -0.07 0.74** -0.02 -0.24

 1 0.67* 0.02 0.64 * -0.36 0.68* -0.89**

  1 -0.45 0.99** -0.46 0.99** -0.76** s' 1 -0.52 0.66* -0.47 -0.08  L w 1 -0.49 0.99** -0.74**

R W1 1 -0.44 0.12

R W2 1 -0.78**  1

The following conclusions can be made from the results. Sound absorption coefficient,   and   :  The range of   values is very large: 0.20  0.78.  Values are larger for open-pore insulators 1–9 than for closed-pore insulators 10–13.  Values are usually larger if airflow resistivity is larger.  The peak at 500 Hz for insulator types 10  13 is not assumed to be real. A small airgap may exist behind the airtight specimen, and the specimen-airgap-background behaves as a panel resonator. Sound reduction index of bare insulator , R 1 ja R w1 :  Differences between R w1 are very large, even 17 dB.  Values increase almost monotonically with increasing frequency except for insulator types 10– 12, which show dips at middle frequencies. This may be caused by coincidence frequency.  Unexpectedly, airflow resistivity,  , was not associated with R w1 , although small  is usually expected to predict small R w1 and large   . Sound reduction index of encapsulated insulator, R 2 ja R w2 :  There are significant differences between insulator types, up to 19 dB in R w2 .  The frequency-dependent curves follow well the classical theories of double panels.  Closed-pore insulators 10–13 show much smaller values than open-pore insulators 1  9.  Large   predicts higher R 2 value above the mass-air-mass resonance frequency.  Closed-pore insulators do not provide significant improvement compared to empty cavity. Dynamic stiffness per unit area, s ’:  The range is very large: 1.5  730 MN/m 3 .  The values for the closed-pore insulators 10, 12 and 13 are multifold compared to open-pore insulators 1–9. Reduction of impact sound pressure level in a floating floor,  L and  L w :  The range of  L w values is very large: 15  36 dB.  The frequency-dependent  L curves follow the theory of floating floor well: there is a mass- spring-mass resonance, f 0 , at low frequencies above which the DL begins to increase.

21-24 AUGUST SCOTTISH EVENT CAMPUS. ? O? ? GLASGOW

 Insulators with high s ’ have clearly, up to 15 dB, smaller  L w than insulators with small s ’.  Closed pore insulators have larger f 0 (appr. 200 Hz) than open-pore insulators (50–100 Hz). Thermal conductivity,  :  Smaller  (better thermal insulation) predicted smaller values of most important acoustic properties (   ,  L w , and R w2 ). Density,  :  Density was not associated with   ,  L w , nor R w2 .

4. CONCLUSIONS

Different insulator types carry significantly different acoustic properties. Insulator manufacturers should consider to better declare the acoustic properties of insulator products, because this knowledge is needed in the acoustic design of building constructions and other applications.

5. ACKNOWLEDGEMENTS

The project was funded by Paroc Group Ltd. / Owens Corning.

REFERENCES

1. Hongisto, V., Saarinen, P., Alakoivu R., Hakala J. (2022). Acoustic properties of commer- cially available thermal insulators  An experimental study. Journal of Building Engineering (In press).

21-24 AUGUST SCOTTISH EVENT CAMPUS. ? O? ? GLASGOW