Impact Sound Insulation of Thermally Insulated Balconies

Abstract

With the increasing urban densification, balconies are gaining in popularity as they improve the living quality in homes. From a technical point of view, the thermal insulation between balconies and the building’s façade is state of the art. In Germany, the most popular balcony construction is a reinforced concrete balcony, separated from the building by a thermal insulation element (TIE), which is meant to reduce the thermal energy loss and thus ensure the sustainability of intelligent buildings. The impact sound transmission from balconies, however, is a problem that has not been addressed enough to date. The paper is based on a project of the same name within the iCity research with the main goal of providing acoustic quantities, e.g. an impact sound pressure level difference, for a TIE that can be used to compare the acoustical quality of products and used to predict the impact sound pressure levels within the building using the standard EN ISO 12354-2. Experimental and numerical studies have been carried out on various ceiling-balcony mock-ups without and with TIEs, e.g. by means of experimental modal analysis and validated finite element models, respectively. These studies showed that even doubling the width of the ceiling-balcony mock-up does not change the results significantly, suggesting that the proposed test set-up is suitable for standard testing. The analysis method and results presented here are for only one test set-up with and without a TIE that underwent constructive modifications during the tests. The selected TIE shows an effective sound insulation above 400 Hz and achieves a single-number rated impact sound level difference of $\Delta {L}_{\mathrm{w}}^{\ast}\approx 10\ \mathrm{dB}$.

1 Introduction

The increase in airborne sound insulation against outdoor noise, achieved by the development of higher quality walls and windows, leads to an increased sensitivity of the inhabitants against noise that is generated by neighbours. This is because the noise levels from the neighbours now emerge above the therefore lower level noise floor from the outdoors. Furthermore, outside areas of flats, such as balconies, are gaining in popularity, leading to more impact sound transmission, which can cause disturbances. These two points were taken into account in 2018, when the German standard of requirements on sound insulation in buildings (DIN 4109:2018-01, 2018) was revised. This standard now contains requirements for balconies on the normalized impact sound pressure level as L′_n,w ≤ 58 dB. For loggias that are often difficult to differentiate from balconies in modern buildings, the requirement is L′_n,w ≤ 50 dB. As L′_n,w quantifies the sound pressure level measured in the room when the ceiling or balcony is excited by a standardized tapping machine (Figs. 23.1 and 23.3), a lower level of L′_n,w means better protection from impact noise (e.g. a typical reinforced concrete ceiling without and with a floating floor has levels L′_n,w of around 70 dB and 46 dB, respectively).

Fig. 23.1

Diagonal impact sound transmission of a thermally insulated balcony into a receiving room of a neighbouring unit

In Germany, the most popular balcony construction is a reinforced concrete balcony, separated from the building by a thermal insulation element (TIE) meant to reduce the thermal energy loss. The design of the TIE is primarily based on static requirements. The elements consist of reinforced bars and thrust bearings, sheeted by thermally insulating material like extruded polystyrene. The main goal of the iCity project that is the basis of this paper was to provide characteristic acoustic values for a TIE that can be used for product comparison and used to predict the sound transmission in buildings. A first step to achieve this is to, through measurement and numerical investigations, understand the structure-borne sound transmission through these TIEs.

A not yet fully validated method, suggested by (Blessing, 2018), is to predict the impact sound transmission of balconies in the same fashion as currently done for floors, namely, according to Part 2 of the German standard (DIN 4109:2018-01, 2018) that uses single-number values (in contrast to frequency-dependent values). Yet, currently no standardized laboratory test procedure exists, to determine the “input value”. In other words, the characteristic acoustic values of TIE still need to be defined. The testing procedure shall also provide values for a frequency-dependent prediction following the European standard for building acoustics (EN ISO 12354-2:2017-11, 2017). This chapter describes how this task was tackled and how an approach was developed to predict the impact sound transmission of balconies. Finally, it discusses results of measurements on one laboratory test set-up carried out within this project.

2 Structure-Borne Sound Transmission in Buildings

For balconies, the most relevant requirements for impact or structure-borne sound transmission are along the diagonal path into an adjacent room of a second unit as shown in Fig. 23.1. If the balcony is not separated from the building using a TIE, it can be treated as a ceiling. The prediction can then be done according to (DIN 4109:2018-01, 2018), Part 2, taking into account a K_T value that describes the vibration reduction by the junction formed by ceiling and walls, e.g. with two flanking transmission paths f₁ and f₂ according to (EN ISO 12354-2:2017-11, 2017) from the exited balcony into the receiving room. A prediction with single-number values according to the German standard can be done using Eq. (23.1).

$$ {L}_{n,w}^{\prime }={L}_{n, eq,0,w}-\varDelta {L}_w^{\ast }-{K}_T+{\mu}_{\mathrm{prog}}\ \mathrm{in}\ \mathrm{dB} $$

(23.1)

with

${\mathrm{L}}_{\mathrm{n},\mathrm{w}}^{\prime }$ Weighted normalized impact sound pressure level for diagonal transmission

L_{n, eq, 0, w} Equivalent weighted normalized impact sound pressure level of the balcony for vertical transmission without flanking elements

$\Delta {\mathrm{L}}_{\mathrm{w}}^{\ast }$ Weighted impact sound level difference of the TIE

K_T Correction value for diagonal transmission

μ_prog Safety coefficient; μ_prog = 3 dB for impact sound

The German standard (DIN 4109:2018-01, 2018) does not provide an explicit K_T value for a transfer situation involving a balcony as shown in Fig. 23.1. In (Blessing, 2018), K_T = 5 dB was used for diagonal transmission from a floor to a room, but if this value is also suitable for balconies is yet to be shown. A lower value would be expected for balconies, as often large window/door areas at balconies limit the amount of sound energy going into the wall with the windows, redirecting it to the ceiling and walls on the diagonal below. In other words, the large window/door thereby decreases the diagonal vibration reduction at the junction (described by K_T) compared to a full heavy wall without window/door that has an assumed K_T = 5 dB.

The quantity ΔL^∗, termed impact sound level difference, is chosen in analogy to the approach to describe isolating elements for staircases made of reinforced concrete in (DIN 7396:2016-06, 2016). Further information on the development of this method is given in (Maack, Möck, & Scheck, 2020) and (Fichtel & Scheck, 2013). ΔL^∗ quantifies the increase of the impact sound reduction through the insulation element with reference to a rigid connection which describes an insertion loss notated by an asterisk *. The challenge now is to devise a laboratory test procedure and evaluation that determines ΔL^∗ as close to the real-world situation as possible.

3 Laboratory Test Set-up

In order to define a suitable laboratory test set-up and procedure, the transmission system “thermally insulated balcony” has to be understood thoroughly. Therefore, a laboratory test set-up has been built for experimental studies, consisting of a small ceiling and a thermally insulated balcony similar to test set-ups used by (Schneider & Fischer, 2008). The dimensions of the test set-up and the realization are shown in Figs. 23.2 and 23.3. The larger reinforced concrete slab represents the ceiling in a building and is supported on elastomer strips on two masonry walls. The mass spring system formed by the elastomer strips and balcony and ceiling has a resonance frequency of 25 Hz (Kluth, 2016). The smaller concrete slab represents the balcony. A laboratory set-up was built with a thickness of 18 cm, termed set-up 1a without TIE and set-up 1b with TIE.

Fig. 23.2

Dimensions of the laboratory test set-ups; dark grey bar depicts the TIE for set-up 1b with components

Fig. 23.3

Laboratory test set-up 1b with ISO tapping machine on the reference excitation position and velocity level measurement positions for determination of the impact sound level difference of the TIE (only those on the ceiling required)

4 Laboratory Test Procedure

The impact sound level difference is determined from velocity level measurements on the ceiling (Figs. 23.3 and 23.4). By Eq. (23.2) the radiated sound pressure level from the ceiling into an (imaginary) receiving room below the ceiling can be calculated.

$$ {L}_{\mathrm{p}}={L}_{\mathrm{v}}+10{\log}_{10}\sigma +6+10{\log}_{10}\frac{S}{A}\ \mathrm{in}\ \mathrm{dB} $$

(23.2)

with

L_p:

Sound pressure level in the receiving room

L_v:

Spatially averaged velocity level on the ceiling (ref 5e⁻⁸ m/s)

σ:

Radiation efficiency; assumption σ = 1

S:

Area of the ceiling

A:

Equivalent sound absorption area in the receiving room

Fig. 23.4

Side view of the test set-up 1a) without TIE (top) and test set-up 1b) with TIE (bottom)

A normalization to the reference absorption area A₀ = 10 m² results in the normalized impact sound pressure level from velocity level measurements according to Eq. (23.3).

$$ {L}_{\mathrm{n},\mathrm{v}}={L}_{\mathrm{v}}+10{\log}_{10}\sigma +6+10{\log}_{10}\frac{S}{A_0}\ \mathrm{in}\ \mathrm{dB} $$

(23.3)

The determination of the impact sound level difference ΔL^∗ of the TIE requires measurements on set-up 1a without TIE and on set-up 1b with TIE (Fig. 23.4).

$$ \varDelta {L}^{\ast }={L}_{\mathrm{n}0,\mathrm{v}}-{L}_{\mathrm{n},\mathrm{v}}\ \mathrm{in}\ \mathrm{dB} $$

(23.4)

with

ΔL^∗:

Impact sound level difference of the TIE

L_{n0, v}:

Normalized impact sound pressure level without TIE

L_{n, v}:

Normalized impact sound pressure level with TIE

To determine the weighted impact sound level difference $\Delta {L}_{\mathrm{w}}^{\ast }$ as single-number rating, the procedure according to (DIN EN ISO 717-2:2013-06, 2013) can be used as it is already a standard for floor coverings and isolating elements for heavy stairs.

5 Experimental Modal Analysis

To analyse the vibration behaviour of the test set-up, an experimental modal analysis was carried out on both set-ups, with and without insulation elements. For the experimental modal analysis, the velocity at each point of interest is measured, while the structure is excited at a reference point with a controlled force signal. The ratio of velocity and force is termed mobility Y. The term input mobility Y_P denotes that the force and the velocity are measured at the same point. High mobility values mean that only a little force is necessary to cause a large velocity response and thus peaks in the mobility indicate a resonant behaviour.

The modal analysis can be carried out using the reciprocity principle, by mounting a reference accelerometer at a reference point while exciting every point of interest, e.g. with an impact hammer. This latter method was used here for measurement convenience, as this way only one instead of hundreds of accelerometers needs to be attached to the surface. When visualizing the vibration patterns, the reciprocity once again comes into play and the reference position of the accelerometer becomes the excitation position. The measurement grid with a grid spacing of 10 cm (Fig. 23.3) results in 819 excitation points with the impact hammer. The reference position of the accelerometer was in the corner of the balcony where the highest vibration amplitudes are expected. The input mobilities at the reference position are shown in Fig. 23.5 for set-up 1a (dotted) and set-up 1b (dashed). Examples of vibration shapes at the so-called eigenmodes or intrinsic modes are shown in Fig. 23.6. The eigenmodes describe the vibration patterns of a system that can vibrate freely, without forced excitation.

Fig. 23.5

Input mobilities for set-up 1a and set-up 1b at the reference position for the experimental modal analysis in the corner of the balcony

Fig. 23.6

Vibration shapes of set-up 1a (left) and 1b (right) at selected frequencies

The first eigenmode of set-up 1b where the balcony oscillates as a cantilever beam is at about 12 Hz and is determined by the torsion spring stiffness of the TIE and the mass of the balcony. Studies performed by (Kluth, 2016) showed that this vibration is well perceived by a person standing on the balcony and may result in discomfort. For set-up 1a, this problem is not observed as its first resonance is not so pronounced and the frequency is higher. Investigations based on the finite element method (FEM) also showed that the decoupling of the ceiling and balcony from the masonry walls by the elastomer strips is not yet effective in this low frequency region. This effect was anticipated in the technical design to ensure the following two goals: (1) to be able to measure this cantilever beam vibration as it occurs in buildings in order to get insight into low frequency vibration problems and (2) to be able to measure the structure-borne sound transmission from balcony to ceiling in the common building acoustics frequency range from 50 to 5 kHz without influence of the supporting wall structure.

The vibrations above 50 Hz are dominated by bending modes of the plate(s). Without the TIE, the velocity level amplitudes on the balcony and on the ceiling differ by less than 2 dB. With the TIE, the balcony and the ceiling are effectively coupled in the frequency range from 50 to 400 Hz. Above 400 Hz, the vibration amplitudes on the excited balcony are significantly higher than on the ceiling. Here the TIE partially decouples the balcony from the ceiling.

6 Impact Sound Level Difference

The impact sound level difference ΔL^∗ is determined from velocity level measurements at the same six positions on the ceiling for set-up 1a and set-up 1b. The ISO tapping machine is positioned diagonally with one hammer at a corner of the balcony (Fig. 23.3) to excite as many eigenmodes as possible and thus to simulate a worst case for the impact sound transmission from balcony to ceiling.

The normalized impact sound levels measured on the ceiling are shown in Fig. 23.7 in 1/3 octave bands from 50 to 5000 Hz. At lower frequencies, both levels with and without TIE follow the same trend with peaks and dips varying around 70 dB. Towards higher frequencies, they diverge and the levels with TIE sink to values below 60 dB. Figure 23.8 shows the impact sound level difference evaluated from the L′_n,v values shown in Figs. 23.7 and 23.4. As expected from the results of the modal analysis, an effective decoupling of the balcony by the TIE is only given above 400 Hz. Above 400 Hz, ΔL^∗ first increases with frequency as it is typical for isolating elements but then reduces again above 2500 Hz. This is probably due to resonances inside the steel components (Fig. 23.2). The single-number rating of the TIE is $\Delta {L}_{\mathrm{w}}^{\ast }=10.2$ dB.

Fig. 23.7

Normalized impact sound pressure level of set-up 1a and set-up 1b measured on the ceiling

Fig. 23.8

Impact sound level difference without and with modifications of the TIE

7 Modification of the TIE

The investigated TIEs consist of statically indispensable tension and shear force bars, thrust bearings, foamed material for thermal insulation and fire protection boards (Fig. 23.2). The influence of each of these components on the impact sound transmission was investigated by modifications after the initial measurements. The fire protection, thermal insulation and load-bearing parts were removed gradually, and the impact sound level was measured for each modification step. After the last modification step, the TIE was reduced to a statically affordable minimum only leaving a few draw force bars, shear force bars and thrust bearings. The exposed area between ceiling and balcony was afterwards filled with concrete to obtain set-up 1a as it is shown in Fig. 23.4 (top). The effect of the fire protection boards and thermal insulation on the sound transmission is negligible. Reducing the tension bars by 67%, shear force bars by 60% and thrust bearings by 38% results in a significant increase of the impact sound level difference.

8 Finite Element Simulations

The main goal of the finite element simulations was to reduce the measurement effort needed to develop an appropriate laboratory test set-up for TIEs, in particular by defining the dimensions of balcony and ceiling elements. In a first step, the test set-up 1b was modelled in FE. The comparison between measured and simulated input mobility at the reference position at the corner of the balcony of set-up 1b is shown in Fig. 23.9. The agreement is very good in the whole frequency range. The measured and simulated vibration shapes were also drawn upon to further validate the FE simulation model.

Fig. 23.9

Input mobility of set-up 1b with TIE measured and simulated

In the next step, the velocity levels that result from excitation with the ISO tapping machine were simulated, and from this, the impact sound level reduction was calculated and evaluated. In Fig. 23.10, the impact sound level reduction is shown for set-up 1 as measured and simulated. The agreement across the whole frequency range is within ±5 dB, which is similar to the variations in nominally the same buildings and therefore acceptable. Note that the single-number rating only varies by less than 1 dB. In addition, the neglectable influence of the dimensions of the balcony and ceiling elements can be seen in this figure, for which the width of the set-up (length of the TIE) was doubled in the simulation from 200 cm to 400 cm. Again, the agreement across the whole frequency range is within ±5 dB, which indicates that the currently proposed set-up (Fig. 23.2) delivers suitable values to characterization TIEs for product labelling and for the prediction of the sound transmission in buildings. Note that the alteration of the FE element dimensions changes the single-number rating by less than 1.5 dB.

Fig. 23.10

Impact sound level difference: Measurement and simulation of set-up 1 and simulation of a modified set-up 1 with doubled sizes of balcony, ceiling and TIE

9 Conclusion

For the acoustical characterization of thermal insulation elements of balconies, a laboratory test set-up and method are proposed that can be used for product labelling and to predict the impact sound transmission in buildings. The quantity suggested, the weighted impact sound level difference ($\Delta {L}_{\mathrm{w}}^{\ast }$), which for the tested TIEs is around $\Delta {L}_{\mathrm{w}}^{\ast }=10\ \mathrm{dB}$, can be determined with velocity level measurements on the laboratory test set-up. For the investigated TIEs, a significant sound insulation between the balcony and the ceiling is observed in the frequency range above 400 Hz. Much trust is placed in the finite element simulations as the measured results on the laboratory test set-ups are in very good agreement with the finite element simulations. Therefore, various studies could be carried out with the FEM such as modifying the size of the laboratory test set-up. It was shown that doubling the width of the set-up (length of the TIE) from 200 cm to 400 cm has no significant effect on the simulated results of $\Delta {L}_{\mathrm{w}}^{\ast }$—which is promising for the acceptance of the test set-up.

Further investigations, involving measurements and simulations on various TIEs and measurements in real building situations, will be carried out within the frame of the iCity project in order to further optimize the test set-up and measurement procedure regarding simplicity and accuracy. The building measurements will be used to ensure that the normative prediction models of structure-borne sound transmission in buildings deliver appropriate results with the here proposed “input data” $\Delta {L}_{\mathrm{w}}^{\ast }$. Finally, the developed methods will be applied to optimize TIE products regarding the acoustical insulation properties.