Sound absorption of LWS with rWPCs as supporting system

Fig. 32 showed the sound absorption of LWS with rWPCs as supporting system.

First of all, the sound absorption ability of Jacobaea maritima garden modules was compared with previous works. The substrate depth in this study was less than the studies showed in Table 5, therefore it was reasonable that the garden modules in this study had less αs in low-frequency range, since the overall weight and thickness of substrate layer would determine low-frequency absorption [9]. The number of plants per square meter in this study was between the work of Davis et al. [11] and Azkorra et al. [10], while the overall sound absorption of this work was less than [11] and exceeded [10].

Fig. 32 Sound absorption of A1 and B1 covered by 100 % garden modules compared to the effect with only garden modules.

The less absorption comparing to [11] should be mostly caused by a thinner substrate layer and less exposure of the substrate. The substrate in this study was inserted in pots,

0

plant bodies should also be accounted. Boston ferns in [11] had wide and dense crowns, which would create larger leaf area density comparing to Jacobaea maritima applied in this study.

On the other hand, helichrysum thianschanicum applied in [10] had relatively different morphology, which might create vibration or scatter mechanisms less efficient for noise attenuation than boston fern or Jacobaea maritima. The phenomenon could be confirmed by its similar absorption result with [8] applying also boston fern but with much fewer plants per square meter (Table 5).

Therefore, Jacobaea maritima were potential for noise reduction purpose, which presented comparable performance to boston fern. Secondly, the integration of rWPCs supporting system of A1 and B1 both improved sound absorption for garden modules, which could be observed in Fig. 32 and Table 11. Table 11 showed the average αs of settings with garden modules categorized by frequency range full, low, medium, high according to [11]. To further understand the functionality of rWPCs to LWS, various settings were examined in the following section.

Table 11 Average αs of different frequency ranges of LWS with various supporting systems (100 % coverage rate).

Frequency (Hz)

As for sound absorption coefficient exceeding unity, edge diffraction and the sealing of specimen sides might be two possible explanations. Edge diffraction is caused by non-planar sound wave reflection created by the edge, causing more absorption [56], which was also mentioned in [11]. However, sound absorption coefficient reaching almost 1.4 should be caused by the unsealed sides of specimen, which should be accounted as extra sound absorption area, especially for the case of [11] with exposing of high absorptive substrate. In this study, the difference between sealed sides and non-sealed sides of rWPCs was examined in advance, with results indicating the two cases had little influence to the measurement of rWPCs, and substrate was inserted in plastic pots. Therefore, the unsealed sides should not lead to critical influence on the results of this study. However, the area of plants was indeed a factor that need to consider for green wall measurements in reverberation room.

4.2.2.1. Acoustic mountings with LWS

The average αs of full, low, medium, high-frequency ranges for various supporting system settings and garden modules were listed in Table 11, with the overall trend in Fig.

33. It could be observed that A2 had the highest absorption averagely, and highest absorption at frequency around 1,000 Hz. A1 had the best sound absorption performance at medium-frequecny band, low-frequency band for A6, and high-frequency band for B3.

Settings utilizing Bw showed better performance at high-frequency range, which might be due to lower density comparing to Aw. Though A6 gave higher absorption at low frequencies, it had lower absorption for medium to high octave band comparing to A2, suggesting that even with coverage of garden modules, the existence of pores would still influence acoustic performance.

Fig. 33 Sound absorption of A and B with various air gap covered by 100 % garden modules.

Fig. 34 Sound absorption of A and B with various air gap covered by 100 % garden modules by 1/3 octave band.

Last but not least, though increasing air gap to an extreme 30 cm, there was no significant improvement on sound absorption observed except for octave band 100 and 125 Hz. In the study of [11], there was also little difference after the addition of 5 or 10

0.00

0 500 1000 1500 2000 2500 3000 3500 4000

Absorption coefficient

cm air gap. As a result, for better making use of space and effectively attenuating low-feqeuncy noise, an air gap of 2.5 cm was recommended in this study.

4.2.2.2. Coverage rate for LWS

Coverage rate of garden modules were studied with rWPCs setting A6 and B4 chosen for their better sound absorption at low-frequency range, considering the difficulty of improving the performance of such range by light substrates and plant bodies. Fig. 35 was done by subtracting 1/3 of total module gradually and for Fig. 36 subtracting 1/8 with modules. Fig. 35 and Fig. 36 both indicated that with the increase of coverage rate, the sound absorption of all octave bands were improved, but was more effective for mid-high frequencies.

Fig. 35 Sound absorption of various garden module coverage rate with setting A6

Fig. 36 Sound absorption of various garden module coverage rate with setting B4.

On the other hand, for coverage rate of less than 65%, A6 had better performance, while B4 showed a better effect for 100% coverage (Table 12). The result also suggested that the coverage rate of around 33% in this study had similar performance compared to work of Azkorra et al. [47]. As a consequence, it should be noticed that to reach certain sound absorption level, sometimes it would not be necessary to apply a high garden module coverage rate. Better performance of B4 than A6 at 100% coverage rate might also suggest that air gap of certain thickness would be necessary for improving sound absorption for higher octave bands, since 5 cm and 10 cm air gap of Davis et al. also showed less sound absorption decline at high-frequency range [23].

Table 12 Average αs of LWS coverage rate study.

Setting Coverage rate of garden modules

0% 30% 33% 65% 100%

A6 0.18 - 0.39 0.53 0.65

B4 0.14 0.33 - 0.51 0.67

Table 12 also suggested that the effectivity of increasing the coverage rate to improve

0.00

100 600 1100 1600 2100 2600 3100 3600

Absorption coefficient

sound absorption would decrease with the rise of the coverage rate. Consequently, it is possible to find a solution most efficient for the required acoustic purpose by choosing suitable coverage rate and supporting system, which might further decrease the overall cost and effort of maintenance for LWS. However, it should also be taken into account that the scattering pattern of modules should not lead to concentration of sound wave of a certain frequency, though such phenomenon was not obvious in this study, but was seen in work of Wong et al. [8].