A A A Sound.Wood.Austria - selected measurement results of building components for multi-storey timber construction in Austria Heinz Ferk 1 , Christopher Leh, Markus Mosing, Jan Kasim, Selina Vavrik-Kirchsteiger Graz University of Technology Inffeldgasse 24, 8010 Graz, Austria Bernd Nusser Holzforschung Austria Franz-Grill-Straße 7, 1030 Vienna, Austria ABSTRACT With the increasing use of wood also for multistory dwellings, a wide variety of timber construction systems have been developed in Austria, of which in particular timber post and beam construction and solid timber construction in cross-laminated timber have become established. During the Internoise 2005, the first session on sound insulation in timber construction was initiated by Jean- Luc Koiumji. At that time, I was able to present the CLT construction method and show, using the example of the first multistory CLT apartment houses in Austria, that even high sound insulation requirements can be well met. However, it also became apparent that there is still a great need for research, especially in the low-frequency range, in flanking transmission and optimization. In the last years, within the Sound.Wood.Austria research project, acoustic laboratory measurements of typical Austrian wooden building components were carried out in order to determine the effect of various design measures on sound insulation in a systematic approach and to identify possible optimization potential. Exterior walls, apartment partition walls and apartment partition ceilings were investigated, but also effects from flanking transmission. In this talk, selected measurement results are presented and discussed. However, the results pointed out, that further research work and developments, preferably internationally networked, are desirable in order to exploit the acoustic potential of timber buildings in the sense of "Green Deal" 1. INTRODUCTION The exterior wall, the apartment partition wall and the apartment partition slab are decisive components for the sound insulation of multistory buildings. An important criterion for the selection of the structural components is the building system itself. Frequently used construction types are systems with continuous vertical partition with CLT double walls, systems with continuous ceilings, which are usually built as house-in-house systems due to the longitudinal sound insulation, i.e. with single-shell CLT partition with facing shells, floating screed and suspended ceiling, as well as typical "modular construction methods", in which individual, compact cells are stacked vertically separated from each other by means of special bearings. In the following, such components made of CLT in particular will be considered from a building acoustics point of view, which are frequently used in Austria for the first variant. 1 ferk@tugraz.at tt inter-ncie’” SCOTTISH evENT CAMPUS 3) O , o iLASGOW Source: H. Ferk Source: KLH GmbH Source: Arch. Riess Continuous joint „house in house system Cell Construction Figure 1: Examples for construction systems for CLT based multifamily houses. Austrian sound reduction legal requirements for external walls start from R w ≥ 43 dB, depending on the outside noise level. Between common rooms of multistorey apartment houses the standardized sound level difference must be at least 55 dB, between terraced houses at least 60 dB. The same is valid for separating floors, whereas the maximum standardized impact sound level must be less then 48 dB, and 43 dB for terraced houses. In the following sections, some results on exterior walls, apartment partition walls and apartment partition ceilings are discussed. 2. LABORATORY MEASUREMENTS The measurement of the sound insulation of the individual CLT-based components was carried out on a wall and a slab test stand in the Laboratory for Building Physics at the Graz University of Technology. To investigate the effect of different installation variants, individual CLT panels were examined in different connections to the testing facility. Usually, the connection of such components consists of stuffed mineral wool and linseed oil mastic sealing on both sides. The effect of replacing the mineral wool with an impact sound insulation board, an elastomeric strip and further a mounting with wooden wedges, i.e. "clamped", was investigated. It was found that the lowest sound reduction index R w for a 100 mm thick 5-layer CLT board was 31.9 dB with the impact sound insulation board, while the highest value of 34.1 dB was achieved with the wedged-in variant. Subsequently, all further measurements on impact sound insulation board were measured "on the safe side" with the impact sound insulation board. CLTP 100 mm, 5s, with different mounting conditions 70 70 70 dB ¾® dB ¾® dB ¾® 60 60 60 R Sound reduction index , R Sound reduction index , R Sound reduction index , 50 50 50 40 40 40 30 30 30 20 20 20 TDPT TDPT TDPT gekeilt MW Elastomer 10 10 10 63 125 250 500 1000 2000 Hz4000 63 125 250 500 1000 2000 Hz4000 63 125 250 500 1000 2000 Hz4000 Frequency, f, Hz ¾® Frequency, f, Hz ¾® Frequency, f, Hz ¾® wedged („edge fixed“) Elastomeric layer TDPT Impact sound R w (C;C tr ) = 31,9 ( -1,4 ; -4,2 ) dB R w + C tr,50-5000 = 27,0 dB R w (C;C tr ) = 34,1 ( -1,4 ; -4,2 ) dB R w + C tr,50-5000 = 28,8 dB R w (C;C tr ) = 32,9 ( -1,7 ; -4,2 ) dB insulation layer Mineral wool R w + C tr,50-5000 = 28,1 dB R w (C;C tr ) = 32,2 ( -1,5 ; -4,3 ) dB R w + C tr,50-5000 = 27,4 dB Figure 2: Sound reduction index R w for a 100 mm thick 5-layer CLT board with different connection to the testing facility. 3. EXTERIOR WALLS Based on a 100 mm CLT board, a vertical batten made of KVH 58/158 was mounted directly on the CLT. The space between the battens was filled with 160 mm thick "mineral wool light", a wind brake was mounted on top of it. The measurement result of the sound insulation index R w was 46 dB. Naturally, this type of exterior wall without a protection against penetrating rain cannot be used in practice in this form. In the next step, a horizontal inverted formwork on supporting battens and rear ventilation was produced and an evaluated sound reduction index Rw of 38 dB was determined for this structure. To investigate the effect of counter-battening, the outer formwork, and the battens in the area of the rear ventilation were mounted on horizontal counter-battening (at a distance of approx. 65 cm) for the subsequent measurement. This resulted in a back-ventilation space between the windbreak and the exterior façade that was partially extended to 56 mm. The arrangement of the counter-battens improved the weighted sound reduction index Rw to 42 dB, but with a clearly noticeable drop in sound reduction at 80 Hz. The results clearly show that the weighted sound reduction index with a rear-ventilated façade is noticeably reduced if a rather rigid attachment of the battens to the BSP and a likewise rigid attachment of the outer formwork to the battens result in an increased input of sound energy from the formwork, which is excited to vibrate, into the BSP. The effect can be seen in the entire frequency range from as low as 100 Hz. The "B numbers" in the following diagrams are reference numbers of the measurements. Figure 3: Sound reduction index R w for a 100 mm thick 5-layer CLT board with different ventilated formwork. In many cases facing shells are used on the room side. Together with the CLT, these represent a mass-spring-mass system from an acoustic point of view, which, however, in the case of direct mounting on the BSP, partly enables the transfer of the vibration energy of the facing shell into the CLT panel and vice versa. Based on a CLT wall with a ventilated exterior façade without counter-battens, the weighted sound reduction index of 38 dB is improved by around 8 dB, and by a further 1.5 dB with double cladding. Here, too, the improvement in the weighted sound reduction index results in a reduction in sound insulation in the low-frequency range - a drop in sound insulation at 80 Hz is very noticeable. Nevertheless, the overall weighted sound reduction index with spectrum matching value Ct r50-5000 improves slightly from 33.3 dB to 35.8 dB and 36.4 dB, respectively. 70 B18.686.010.310b dB o R w (C;C tr ) = 38,1 ( -1,5 ; -4,3 ) dB R w + C tr,50-5000 = 33,2 dB 60 R Sound reduction index , B18.686.016.310b 50 R w (C;C tr ) = 46,3 ( -1,8 ; -5,0 ) dB R w + C tr,50-5000 = 35,8 dB 40 B18.686.017.310 30 R w (C;C tr ) = 47,8 ( -2,3 ; -5,7 ) dB R w + C tr,50-5000 = 36,4 dB 20 10 63 125 250 500 1000 2000 Hz 4000 Fig. 4: Effect of a facing shell inside, single- and double-boarded The use of resilient brackets can prevent some of the unfavorable vibration transmission, partly due to the springing effect of the resilient brackets, and partly due to the reduced restraining effect of the bolting of the resilient bracket. The following diagrams show the effect of installing 5 resilient clips per batten compared to installation without resilient clips. The sound reduction index R w improves by more than 2 dB with single cladding and by considerably more than 6 dB with double cladding. An improvement is also achieved in the range below 100 Hz, with a slight reduction in sound attenuation in the range from 100 Hz to 200 Hz. The increase in sound attenuation is favorably parallel to the reference curve up to 500 Hz. The observed dip around 400 Hz can also be eliminated when using resilient brackets. Taking these results into account, the use or development of a suitable resilient bracket for the ventilated facade seems to be promising! Frequency, f, Hz o 70 B18.686.016.310b dB o R w (C;C tr ) = 46,3 ( -1,8 ; -5,0 ) dB R w + C tr,50-5000 = 35,8 dB 60 R Sound reduction index , B18.686.017.310 50 R w (C;C tr ) = 47,8 ( -2,3 ; -5,7 ) dB R w + C tr,50-5000 = 36,4 dB 40 B18.686.018.310 R w (C;C tr ) = 49,7 ( -3,1 ; -9,1 ) dB R w + C tr,50-5000 = 37,8 dB 30 20 B18.686.019.310 R w (C;C tr ) = 54,6 ( -3,6 ; -10,2 ) dB R w + C tr,50-5000 = 39,8 dB 10 63 125 250 500 1000 2000 Hz 4000 Frequency, f, Hz o Fig. 5: Effect of mounting the facing shell (single and double planked) on resilient brackets in comparison with direct mounting of the battens. Furthermore, it was also investigated how the use of a plaster base board instead of a wooden formwork for the ventilated facade affects the sound reduction. A wood fiber board GF40 was used, which has a tongue-and-groove joint and was coated for the measurements with an approx. 3 mm thick-film adhesive filler including a textile glass grid. In any case, the acoustic effect of the lower stiffness and also the sound-absorbing effect of the wood fiber board seems to add a favorable effect in the frequency range from 250 Hz and above. 70 dB o B18.686.023.310 (auf TDPT) R w (C;C tr ) = 47,5 ( -2,4 ; -6,2 ) dB R w + C tr,50-5000 = 37,1 dB 60 R Sound reduction index , B18.686.024.310 (auf TDPT) 50 R w (C;C tr ) = 52,0 ( -2,6 ; -9,1 ) dB R w + C tr,50-5000 = 32,3 dB 40 B18.686.020.310 (auf TDPT) R w (C;C tr ) = 37,5 ( -1,7 ; -4,8 ) dB R w + C tr,50-5000 = 32,1 dB 30 20 B18.686.026.310 (auf TDPT) R w (C;C tr ) = 44,0 ( -1,8 ; -5,9 ) dB R w + C tr,50-5000 = 31,3 dB 10 63 125 250 500 1000 2000 Hz 4000 Frequency, f, Hz o Fig. 6: Effect of mounting a plaster base board with 3 mm plaster compared to a wooden formwork outside. 4. PARTITION WALLS In the case of CLT based double walls, 2 types of panels were measured in different variations: 80 mm 3-layer and 100 mm 5-layer CLT boards were used and were arranged both symmetrically and asymmetrically. Also varied were spacing, planking, and variants with facing layer(s) were also studied. It should be noted that these are also "on the safe side" test bench values without any additions/deductions. The installation was carried out within the testing facility in a concrete frame with structure-borne sound insulation of the connection joint to the test wall, with a test wall area of around 11.6 m2 and a frame closed on all sides. Fig. 7 shows, using different CLT based boards with a 60 mm gap between, filled with light insulation (partition wall felt), that the thickness of the CLT panels has much less influence on the result than an asymmetrical construction, which shows advantages over the entire frequency range. Results Rw | Rw + C50-3150 | Rw + Ctr,50-3150 in dB 70 dB ¾® 80 mm CLT / 60 mm Dämmung / 80 mm CLT Mineral wool 100 mm CLT / 60 mm Dämmung / 100 mm CLT Mineral wool 60 100 mm CLT / 60 mm Dämmung / 80 mm CLT Bewertetes Schalldämm-Maß R in dB mineral wool 50 Sound Reduction Index 40 2 3 1 30 3s Massivholz 3s Massivholz 20 80 mm 100 mm 60 mm 100 mm 100 mm 60 mm 80 mm 80 mm 60 mm 48,9 / 45 / 35 48,9 / 45,1 / 36 54,7 / 49,5 / 38,9 10 63 125 250 500 1000 2000 Hz4000 Frequenz, f, Hz !" Fig. 7: Building acoustic influence of different thickness modifications on double-shell solid wood partition walls. Partition joint 60 mm (fully insulated); As known, a resonant frequency can be estimated for these double-shell partition wall constructions. In the present case of the walls shown above, values for the resonant frequency for the wall with 2 x 80 mm CLT of around 51 Hz, for 2 x 100 mm of around 45 Hz and for the combination of the different panel thicknesses an estimation of 48 Hz is obtained. When the distance between the panels is doubled to 120 mm, these resonance frequencies drop significantly to 36/32/34 Hz, which indicates a correspondingly more favorable behavior in the low-frequency range, since it is known that the favorable effect of a double-skin construction only comes into play from around 1.4 times the resonance frequency but can have a significantly less favorable effect in the range of the resonance frequency, depending on the damping. Overall, however, the sound insulation also increases with greater distance, as can be seen from Fig. 8. 70 dB ¾® 1) 80 mm CLT / 80 mm Dämmung / 80 mm CLT mineral wool 2) 80 mm CLT / 60 mm Dämmung / 80 mm CLT mineral wool 60 3) 80 mm CLT / 20 mm Dämmung / 80 mm CLT Bewertetes Schalldämm-Maß R in dB mineral wool 50 Sound Reduction Index 40 1 ) 2 ) 3 ) 30 20 80 mm 20 mm 80 mm 80 mm 80 mm 80 mm 60 mm 80 mm 80 mm 44,1 / 41,1 48,9 / 45,1 51,2 / 47,4 10 63 125 250 500 1000 2000 Hz4000 Frequenz, f, Hz ¾® Fig. 8: Effect of varying the distance between the wall shells Results Rw | Rw + C50-3150 | Rw + Ctr,50-3150 in dB Visible wood surfaces are therefore possible for partition walls with the appropriate shell spacing. For planning purposes, however, flanking transmission of the selected building structure must always be considered. If visible CLT surfaces are dispensed with, additional cladding with heavier boards, such as gypsum plasterboard, gypsum fiber or the like, can further increase the sound insulation, as Fig. 9 shows. 70 1) dB ¾® 2) 3) 60 4) Bewertetes Schalldämm-Maß R in dB 50 Sound Reduction Index 40 1) 2) 3) 4) 30 20 60 mm 80 mm 80 mm 60 mm 80 mm 80 mm 80 mm 80 mm 60 mm 60 mm 80 mm 80 mm 48,9 /45,1 / 57,4 64,9 / 58,8 56,6/ 51,1 62,2 10 63 125 250 500 1000 2000 Hz4000 Fig. 9: Effect of additional planking. Indication of results Rw | Rw + C50-3150 | Rw + Ctr,50-3150 in dB The behavior of double CLT Based walls is completely different when facing shells are used. If the respective facing shells are screwed to the cross-laminated timber wall via battens, this has several effects. Firstly, this leads to a significant reduction in sound insulation in the low-frequency range. As mentioned above, if the facing shells were not rigidly connected, it would be a four-mass oscillator. If the facing shell is mounted over a batten that is rigidly screwed to the cross-laminated timber wall, there is also a direct exchange of vibration energy between the facing shell and the cross-laminated timber via this rigidly connected batten, and a superposition of the eigenmodes of the individual systems capable of vibration. The effect of the facing layer is noticeable from about 100 Hz, the significant increase in sound attenuation continues here with a total increase of up to more than 15 dB within a one-third octave band. third octave band. This means that a very high level of sound insulation is achieved for uses such as speech, but only a low level of insulation of low-frequency noise below the third-octave band of 100 Hz is present. For applications such as apartment partition walls, even if the required single value, the weighted sound reduction index Rw, indicates sufficient minimum sound insulation, care should be taken to avoid such massive differences in sound insulation between individual one-third octave bands. Otherwise, this will lead to a kind of "band-pass filter effect", in which low-frequency noise components between the usage units will stand out in particular. Asymmetrical arrangements, e.g. of the facing shells, but also higher weight (and thus lower tuning) can bring at least small improvements here, as can also be seen in Fig. 10. 70 60 Bewertetes Schalldämm-Maß R in dB 50 Sound Reduction Index 40 30 1) 2) 3) 4 5) 1) 2) 3) 20 60 mm 60 mm 60 mm 60 mm 60 mm 60 mm 60 mm 4) 80 mm 60 mm 80 mm 80 mm 60 mm 80 mm 80 mm 60 mm 80 mm 80 mm 80 mm 60 mm 80 mm 60 mm 80 mm 5) 10 62,7 / 46,1 64,9 /49,5 65,7 / 51,9 48,8 / 45,1 58,7 / 49,6 63 125 250 500 1000 2000 Hz4000 Frequenz, f, Hz ¾® Fig. 10: Cross laminated timber partition wall as a double wall with a variation of facing layers on 6 cm thick, directly screwed battens. Results Rw | Rw + C50-3150 | Rw + Ctr,50-3150 in dB With independent facing shells, the transmiss ion of structure-borne sound energy between the solid wood wall and the facing shell can be largely prevente d. If the facing shell, as shown in Fig. 11, has a distance of 100 mm, for example, a resonance frequency of around 70 Hz results from the simplified consideration as a two-mass oscillator. Resonance is also evident in the 63 Hz third octave band. The sound attenuation of the deeper tuned facing shells increases extremely from around 21 dB at low frequencies to over 60 dB within an octave for the facing shell on both sides, so that low-frequency noise would also be particularly prominent here, but medium- and high- frequency noise is extremely well attenuated. 70 dB ¾® 60 Bewertetes Schalldämm-Maß R in dB 50 Sound Reduction Index 40 30 4) 3) 2) 1) 1) 20 2) 60 mm 40 mm 60 mm 40 mm 60 mm 40 mm 40 mm 60 mm 3) 4) 80 mm 80 mm 60 mm 80 mm 80 mm 60 mm 80 mm 80 mm 60 mm 80 mm 80 mm 60 mm 10 63 125 250 500 1000 2000 Hz4000 48,8/ 44,9 67,5 / 51,4 70,0 / 55,2 78,1 /54,7 Fig. 11: Cross laminated timber double wall with variation of free-standing facing shells. Results Rw | Rw + C 50-3150 | Rw + Ctr,50-3150 in dB Frequenz, f, Hz ¾® 2) 5. PARTITION SLABS In the case of solid wood ceilings, based on 5-layer cross laminated timber ceilings with 160 and 200 mm, various fillings, screed constructions and suspended ceiling systems were investigated. Assuming a 200 mm thick BSH ceiling, this achieves a weighted standard impact sound level L n,w of 82 dB. A floor structure with 60 mm loose chippings, a 30 mm thick TDPS mineral wool impact sound insulation board and a 60 mm thick screed improves the result to 46.9 dB. Additional suspensions further improve the single-number value but lead to a better single-number value at the usual suspension height, but to poorer behavior below 100 Hz, which can often lead to a devaluation of th e acoustic quality of the ceiling by the user. 70 dB ¾® L n,w (C I ) = 85,2 (5,1) dB 200 mm Brettsperrholz Norm-Trittschallpegel L n,w 60 Zementestrich (150 kg/m²) 65 mm MW mit s‘ < 10 MN/m³ 30 mm Split- Schüttung, ungebunden 60 mm L n,w (C I ) = 46,9 (-0,5) dB C I,50-2500 = 4,4 dB 200 mm Brettsperrholz 50 L n,w (C I ) = 39,1 (2,9) dB C I,50-2500 = 13,9 dB 40 65 mm Zementestrich (150 kg/m²) MW mit s‘ < 10 MN/m³ 30 mm mm Split- Schüttung, ungebunden 60 L n,w (C I ) = 34,9 (3,0) dB C I,50-2500 = 19,5 dB 200 mm Brettsperrholz 70 mm Federschiene 30 GKF 12,5 mm 65 mm Zementestrich (150 kg/m²) MW mit s‘ < 10 MN/m³ 30 mm 60 mm Split- Schüttung, ungebunden 20 200 mm Massivholz 70 mm Federschiene GKF GKF 12,5 mm 12,5 mm 10 63 125 250 500 1000 2000 Hz 4000 Figure 12: Effect of a suspended ceiling. The range below 100 Hz is clearly prominent. To avoid this disadvantage, a lower tuning of the ceiling (e.g. by a deeper suspension) would be necessary. Frequency, f, Hz ¾® 4. CONCLUSIONS The measurements were done in a parametric way of changing layers and materials which give a very good overview about the effects of different measures. Especially the behavior in the low frequency range it would be necessary to do further research. For the moment, constructions without a facing shell or a suspended ceiling show a better performance at the low frequency range. Therefore, it would be necessary to find suitable solutions for the reduction of flanking transmission for such systems. 5. ACKNOWLEDGEMENTS The authors would like to thank the Federal Ministry for Digitalization and Economy, the Research Promotion Agency FFG, the Austrian Wood Industry Association and the industry partners ELK Fertighaus GmbH, Griffnerhaus GmbH, Haas Fertigbau, KLH Massivholz GmbH, Saint Gobain Rigips/Isover/Weber, Stora Enso, Theurl GmbH, Vario-Bau Fertighaus GesmbH, Vinzenz Harrer GmbH for funding the project "Sound.Wood.Austria".. Previous Paper 241 of 769 Next