Elsevier

Applied Acoustics

Volume 133, April 2018, Pages 58-63
Applied Acoustics

Airborne sound insulation between dwellings, from 50 vs. 100 Hz – A compilation of Swedish field surveys

https://doi.org/10.1016/j.apacoust.2017.12.017Get rights and content

Abstract

It has for long been debated whether 50 or 100 Hz is the proper lower frequency limit when evaluating airborne sound insulation between dwellings. Although 100 Hz is the lowest third-octave band within most regulations, there is an ongoing interest in paying more attention to lower frequencies. In Sweden, evaluation from 50 Hz became mandatory already in 1999 wherefore unique experiences are available by now.

In this paper, extensive data in terms of field measurements and questionnaire surveys from in total 46 building objects of various constructions have been compiled. A number of single number quantities, standardized by ISO as well as alternatives, are compared concerning their correlation against the subjectively rated annoyance responded by the residents. The statistical evidence for a 50 Hz limit was found to be small considering the total database but when the lightweight buildings were analyzed by their own, the importance of frequencies below 100 Hz became clearer.

The overall recommendation is to include frequencies from 50 Hz in order to achieve good sound protection against a broad variety of sound sources, including music and other possible items generating low frequencies.

Introduction

The first modern international standard regarding evaluation of airborne sound insulation between dwellings was launched in 1968 [1]. The evaluation procedure was based upon a reference curve originating from the sound reduction properties of a 25 cm brick wall [2]. The frequency range was set to 100–3150 Hz and the single number quantity (SNQ) Rw – the frequency weighted sound reduction index – was defined. Later on, complementary spectrum adaptation terms ranging from 50 Hz were introduced in 1996 [3] in order to take sound insulation at low frequencies into account. The term C50-3150 was adopted 1999 into the Swedish building code (BBR) in which requirements for R’w + C50-3150 were stipulated. This was a result of extensive complaints from residents, whereby Swedish authorities as well as the building industry realized that the legal regulations must be stricter.

Up to now, Sweden is still the only country that mandatorily requires sound insulation to be evaluated from 50 Hz and a lot of experience has been acquired over these almost twenty years. A reason for Sweden being a pioneer in this context is that lightweight constructions are more commonly used compared to most other markets and that lightweight buildings, especially at that time, were prone to suffer from poor sound insulation at low frequencies.

During the last few years, an increased interest in low frequency sound insulation has been noticed. As an example, the ISO standards regarding airborne and impact sound insulation nowadays include a specific low frequency measurement procedure to be applied under certain circumstances [4]. Other examples are to be found in the great amount of research activity in the field. Several studies based upon listening tests have been reported (2014–16) [5], [6], [7], [8] but also attempts to optimize evaluation parameters against subjective perception can be found (2016) [9], [10] as well as questionnaire surveys (2015) [11]. A common discussion in the given studies is whether airborne sound insulation between dwellings should be evaluated from 50 Hz or if it is good enough to keep 100 Hz as the lower frequency limit. A coarse general conclusion based upon these studies is that frequencies below 100 Hz are of no great importance when assessing sound insulation. This standpoint is however clearly rejected in the reflective paper by Rindel (2017) [12] who believes that the results of the studies in fact point in the opposite direction, i.e. that frequencies 50–80 Hz must not be kept out of the single number evaluation. Rindel also claimed that the authors severely have misinterpreted their own results and presented wrong conclusions. A key factor in the understanding of this problem is whether it is preferable to optimize the evaluation process in order to find the parameter that gives the highest possible correlation for a specific type of sound source (e.g. loud speech). Or is it better to apply a more protective perspective, from the residents point of view, leading to higher demand for sound insulation in terms of a broad banded SNQ that covers most sources in dwellings? Related to this, it is well known that various sound sources, as speech and music, put different demands on the frequency-weighting factor since many types of music contain proportionally more energy at low frequencies [13]. For the time being, it is obvious that no general consensus within the addressed question of 50 or 100 Hz exists.

In this paper, the Swedish experience of evaluating airborne sound insulation is given with the purpose to find out whether SNQ’s ranging from 50 Hz show higher correlation against the residents’ rated annoyance compared to D’nT,w ranging from 100 Hz. Also the potential difference of correlation in between lightweight and heavyweight constructions will be covered. The following presentation is the compiled results from field measurements and studies of residents’ rating of sound insulation during more than 30 years.

Section snippets

Method

Data from six Swedish research projects and consulting cases have been compiled to a common database. In total, 46 building objects of multi-storey residential houses are included. The objects cover a variety of building techniques and can be divided into lightweight (almost exclusively wooden based) constructions (24 of 46) and heavy concrete constructions (22 of 46). The projects/cases are Aku20 (awaiting publications), AkuLite (2013) [14], SBUF (2011) [15], Boverket Byggkostnadsforum (2007)

Measured sound insulation

The reported values of DnT,w show a spread from 49 to 66 dB whereas DnT,w,50 range from 47 to 64 dB. The overall average of DnT,w is 58.9 dB and the standard deviation is 3.8 dB. The overall average of DnT,w,50 is 56.3 dB and the standard deviation is 3.5 dB. The distribution according to the Swedish sound classification [22], based upon DnT,w,50, is shown in Fig. 2. Sound class C corresponds to the minimum requirement of the national building code, DnT,w,50 ≥ 52 dB, while sound class B and A

Discussion

Overall, the analyzed relations between SNQ’s and annoyance show low correlation. In most cases, the correlation is so weak that the statistic relation is insignificant. The reasons for this are important to discuss.

The overall average sound insulation was found to be 56.3 dB for DnT,w,50 which indicates that 70% of the residents should be “not at all annoyed” according to the large study carried out in Norwegian dwellings 2002–2015 [23]. The residents in the study of our paper are in general

Conclusions

The extensive data of measurements and questionnaire surveys from 46 objects of apartment houses suggest that the evaluation of airborne sound insulation from 50 Hz by DnT,w,50 is a slightly better choice than evaluation from 100 Hz by DnT,w. However, since most of the included buildings were originally designed to fulfill the requirements of the Swedish building code, based upon DnT,w,50, the result could possibly have been different if they were designed for a 100 Hz criterion. Airborne noise

Acknowledgement

The authors gratefully acknowledge the financial support by the Swedish research council Formas.

References (23)

  • P. Virjonen et al.

    Optimized single number quantity for rating the airborne sound insulation of constructions: living sounds

    J Acoust Soc Am

    (2016)
  • Cited by (6)

    • Airborne and impact sound performance of modern lightweight timber buildings in the Australian construction industry

      2021, Case Studies in Construction Materials
      Citation Excerpt :

      Acoustic response of a room below 100 Hz depends on boundary conditions (room modes, room volume and dimensions) [67] and methods in measuring, predicting and deducing airborne sound transmission in low frequency range is an active field of research [65,72]. In some countries like Sweden, the building code requires to fulfil weighted standardised level difference with frequency range extended to 50 Hz DnT,w,50 [73]. This aims at protection towards low frequency range and low annoyance to airborne sound transmission.

    • Assessment of the sound reduction index of building elements by near field excitation through an array of loudspeakers and structural response measurements by laser Doppler vibrometry

      2018, Applied Acoustics
      Citation Excerpt :

      Over the last years, an increased interest in sound insulation at low frequencies can be observed. Although in current standards the assessment of sound insulation in the frequency range down to 50 Hz is still optional [1,2], most studies recommend the inclusion of frequencies below 100 Hz [3,4]. However, it is also well known that measurents of the sound reduction index according to the standard ISO 10140:2010 [5] at frequencies below 100 Hz are affected by a large uncertainty [6–8].

    View full text